Attachment A
Request for Distillates Made from Cereals, Which Are Used before Distillation, to Have a Labelling Exemption
Dossier Finalised – August 1st 2006
I.C. Goodall, K.A. Campbell, C.D. Owen, J.M. Brosnan, G.M. Steele, J. W. Walker The Scotch Whisky Research Institute, Riccarton, Edinburgh, UK
Dossier Produced for the Allergens Working Group of CEPS, The European Spirits Organisation
P. Brunerie (Chair), R. Aylott, M. Feys, P. Liddle, L. Kragelund, N. Soper, M. Sobral Viqueira, J. Spouge
Table of Contents
Data or information which have been submitted previously in the dossier requesting a provisional exemption are indicated by the use of prefix and suffix asterisks in the Section Headings of the Table of Contents and the main text, as well as by the use of a dark blue text font.
Executive Summary...... 6
1. *Scope of Exemption*...... 7
2. Rationale ...... 9 2.1 *Rationale for Exemption as Presented in the Application for Provisional Exemption* ...... 9 2.2 Rationale – Additional Information ...... 10
3. Response to the Opinion of the EFSA Scientific Panel on Dietetic Products, Nutrition and Allergies on Commission Request N° EFSA-Q-2004-120 ...... 11 3.1 Introduction ...... 11 3.2 The Physico-Chemical Argument ...... 11 3.2.1 Summary of Work Presented in the Application for Provisional Exemption to Support the Physico-Chemical Argument ...... 11 3.2.2 Opinion of the EFSA NDA Panel on the Physico-Chemical Argument in the Provisional Exemption Application ...... 12 3.2.3 Revised Programme of Work to Support the Physico-Chemical Argument 12 3.3 The Analytical Argument ...... 12 3.3.1 Summary of Work Presented in the Application for Provisional Exemption to Support the Analytical Argument ...... 12 3.3.2 Opinion of EFSA NDA Panel on the Analytical Argument in the Provisional Exemption Application ...... 13 3.3.3 Revised Programme of Work to Support the Analytical Argument ...... 13
Page 1 of 173 3.4 The History of Safe Use Argument ...... 14 3.4.1 Summary of Work Presented in Application for Provisional Exemption to Support the History of Safe Use Argument ...... 14 3.4.2 Opinion of EFSA NDA Panel on the History of Safe Use Argument in the Provisional Exemption Application ...... 14 3.4.3 Revised Programme of Work to Support the History of Safe Use Argument ...... 15 3.5 Conclusions ...... 15
4. *Characterisation of the Substances for which a Labelling Exemption is Requested, the Methods of Production and the Route of Exposure* ...... 16 4.1 *Characterisation of the Exemption Requested* ...... 16 4.2 *Characterisation of Annex IIIa Raw Materials Used in the Manufacturing Process* ...... 16 4.3 *Characterisation of Methods of Production* ...... 17 4.3.1 *Production of Distillates* ...... 17 4.3.2 *Cereals Processing and Fermentation* ...... 18 4.3.3 *Distillation*...... 18 4.3.3.1 *Batch (Pot Still) Distillation* ...... 19 4.3.3.2 *Continuous Column (Grain Still) Distillation* ...... 21 4.3.3.3 *Neutral Spirit and other Grain Spirits* ...... 23 4.3.4 *Distillation System Comments* ...... 25 4.4 *Characterisation of Routes of Exposure to Consumer* ...... 25
5. Arguments and Data to Support Exemption – Physico-Chemical Evidence ...... 26 5.1 *Physico-Chemical Evidence as Presented in the Application for Provisional Exemption* ...... 26 5.2 Physico-Chemical Evidence - Additional Information ...... 27 5.3 Physico-Chemical Evidence - Opinion of a Physico-Chemical Expert ...... 27 5.4 Physico-Chemical Evidence as Presented by the Analysis of Model Distillations ...... 28 5.5 Conclusions ...... 29
6. Arguments and Data to Support Exemption – Total Protein Based Analytical Evidence ...... 30 6.1 Total Protein Based Analytical Evidence as Presented in the Application for Provisional Exemption ...... 30 6.2 Development and Validation Work to Improve the Sensitivity of the Total Protein Method ...... 31 6.3 Sampling Protocol ...... 32 6.4 Protocol for Total Protein Analysis Using the Bradford Analysis Microassay and AAA-DirectTM Analysis ...... 33 6.5 Results of Total Protein Analysis ...... 33 6.5.1 Results for Total Protein Analysis Using the Bradford Microassay Method33 6.5.2 Results for Total Protein Analysis Using the AAA-DirectTM Method ...... 37 6.6 Total Protein Based Analytical Evidence – Conclusions ...... 38
Page 2 of 173
7. Arguments and Data to Support Exemption – ELISA Based Analytical Evidence ...... 39 7.1 ELISA Based Analytical Evidence as Presented in the Application for Provisional Exemption ...... 39 7.2 Development and Validation Work to Improve the Sensitivity of the ELISA Based Gluten Method ...... 40 7.3 Sampling Protocol ...... 41 7.4 Summary of Protocol for ELISA Based Gluten Method and Results ...... 41 7.5 ELISA Based Analytical Evidence - Conclusions ...... 44
8. Arguments and Data to Support Exemption – History of Safe Use ...... 45 8.1 *History of Safe Use of Ingredients Requested for Exemption as Presented in the Provisional Application* ...... 45 8.2 Update on History of Safe Use of Ingredients Requested for Exemption ....45
9. Summary of Arguments and Data to Support Exemption for Labelling Exemption ...... 47
10. References ...... 49
Appendix 1 *Commissioned Physico-Chemical Expert – Dr. Eva Sørensen* .55
Appendix 2 Dr. Eva Sørensen’s Report on the Fundamentals of Distillation Theory and its Application to the Likelihood of Allergen Transfer during Distillation ...... 57
Appendix 3 Detail of Experiments to Highlight the Non-Volatility of Proteins ...... 66 Appendix 3.1 Introduction ...... 66 Appendix 3.2 Distillation of BSA Model Solutions ...... 67 Appendix 3.3 Distillation of Gluten Model Solutions ...... 68 Appendix 3.4 Distillation of Model Solutions – Conclusion ...... 68
Appendix 4 A Brief Review of the Scientific Literature Relating to “Gluten” Content in Cereals ...... 69 Appendix 4.1 Introduction ...... 69 Appendix 4.2 Wheat ...... 69 Appendix 4.3 Barley ...... 70 Appendix 4.4 Corn (Maize) ...... 71
Appendix 5 *Details of Samples for which Analytical Data were Submitted as Part of the Request for Provisional Exemption – Bottled Products and Production Distillate Samples* ...... 73
Appendix 6a List of Samples Analysed Using Improved Analytical Methodology – Bottled Products ...... 77
Page 3 of 173 Appendix 6b List of Samples Analysed Using Improved Analytical Methodology – Production Distillate Samples ...... 80
Appendix 7 Extended Analytical Survey - Sample Collection and Traceability ...... 85
Appendix 8 *Total Protein Based Analytical Evidence as Presented in the Application for Provisional Exemption* ...... 88 Appendix 8.1 *Introduction* ...... 88 Appendix 8.2 *Sampling Protocol* ...... 88 Appendix 8.3 *Materials and Method – Total Protein, Provisional Exemption* .....88 Appendix 8.4 *Results and Discussion – Total Protein, Provision Exemption* ...... 90 Appendix 8.5 *Analytical Conclusions – Total Protein, Provisional Exemption* ...92
Appendix 9 Development and Validation Work to Improve the Sensitivity of the Total Protein Method ...... 93 Appendix 9.1 Introduction ...... 93 Appendix 9.2 Method 1 – Total Protein Analysis Using NanoOrange® ...... 93 Appendix 9.3 Method 2 – Total Protein Analysis Using CBQCATM ...... 97 Appendix 9.4 Method 3 – Total Protein Analysis Using Bradford MicroAssay Protocol ...... 98 Appendix 9.5 Method 4 - Total Protein Analysis Using Mass Spectrometry ...... 102 Appendix 9.6 Method 5 – Total Protein Analysis Using AAA-DirectTM ...... 103
Appendix 10 Analytical Protocol for Total Protein Analysis of Alcohol Distillate Samples Using the Bradford Analysis Microassay ...... 120
Appendix 11 Analytical Protocol and Results for the Total Protein Analysis of Alcohol Distillate Samples Using the AAA-DirectTM Method 123
Appendix 12 *ELISA Based Gluten Analytical Evidence as Presented in the Application for Provisional Exemption* ...... 130 Appendix 12.1 *Introduction* ...... 130 Appendix 12.2 *Sampling Protocol* ...... 130 Appendix 12.3 *Materials and Method* ...... 130 Appendix 12.4 *Results* ...... 131 Appendix 12.5 *Analytical Conclusions* ...... 133 Appendix 12.6 *RSSL Reports on the Determination of Gluten in Cereal Distillates and Spirits Produced from Cereal Distillates Presented for the Application for Provisional Exemption* ...... 133
Appendix 13 RSSL Report Detailing Development Work to Increase Sensitivity of ELISA Based Methods ...... 150
Appendix 14 RSSL Reports on the Determination of Gluten in Cereal Alcohol Distillates using Improved Methodology ...... 161 Appendix 14.1 Introduction ...... 161 Appendix 14.2 Materials and Method ...... 161 Appendix 14.3 Results ...... 161
Page 4 of 173 Appendix 15 Independent Literature Review to Identify Occurrence of Allergic Reactions to Distillates as a Result of Specific Raw Materials Used Pre-Distillation ...... 172
Page 5 of 173 Request for Distillates Made from Cereals, Which Are Used before Distillation, to Have a Labelling Exemption
Executive Summary
A request is being made for the following ingredients to be permanently exempted from Annex IIIa of European Directive 2003/89/EC: distillates made from alcohol derived from the following raw materials employed as sugar sources prior to the distillation process: cereals containing gluten (i.e. wheat, rye, barley, oats, spelt, kamut or their hybridised strains) and products thereof.
The distillation step that all distilled spirit drinks and associated distillates must undergo both by law and definition is fundamentally a purification and concentration technique for alcohol and volatile compounds. No evidence has been found that the potential allergens contained in cereals can distil into the resultant distillates and be present in the alcoholic beverages available to the consumer. This is based on three areas of evidence:
1. Potentially allergenic proteins and polypeptides are non-volatile and hence do not distil.
2. Results of surveys of bottled spirit drinks and production samples, made from cereals used prior to distillation, demonstrate the absence of gluten and protein.
3. There is no clinical evidence or advice that allergens which originate from cereals containing gluten are a threat to consumers of spirit drinks, where these cereals are used before the distillation stage.
Page 6 of 173
1. *Scope of Exemption*
Labelling exemption is sought for distillates where cereals are used as raw materials in their manufacture before distillation. The scope covers distillates and spirit drinks, as specified in EEC Spirit Drinks Council Regulation 1576/89, and all other alcoholic beverages made from ethyl alcohol of agricultural origin. Exemption is not sought for alcoholic beverages that use cereals or any other of the Annex IIIa substances defined in directive 2003/89/EC after distillation.
All spirit drinks in the EU whether made locally or imported are regulated as to their manufacture by the EEC Spirit Drinks Council Regulation (1576/89). This states that spirit drinks in the EU may only be made from agricultural raw materials. Cereals are used to make distillates which are, or form the basis of, spirit drinks and other alcoholic beverages. Table 1 lists the categories of distillates and spirit drinks as defined in 1576/89 which are, or can be, made from cereals employed prior to the distillation step.
Table 1: Distillates and Spirit Drinks which are, or can be, Made from Cereals Employed Prior to Distillation Distillate Category Definition in 1576/89 Ethyl alcohol of Article 1) 3) h) “ethyl alcohol which … has been obtained by agricultural origin the distillation, after alcoholic fermentation, of agricultural products listed in … the Treaty. Where reference is made to the raw material used, the alcohol must be obtained solely from that raw material” Distillate of Article 1) 3) i) “obtained by distillation … of agricultural agricultural origin products listed in … the Treaty but which does not have the properties of ethyl alcohol as defined in (h) or of a spirit drink but still retains the aroma and taste of the raw material used. Where reference is made to the raw material used, the distillate must be obtained solely from that raw material” Spirit Drink Definition in 1576/89 Category Whisky or whiskey Article 1) 4) b) “produced from the distillation of a mash of cereals” Grain Spirit, Korn , Article 1) 4) c) “produced from the distillation of a fermented Kornbrand, grain mash of cereals” brandy Fruit spirit Article 1) 4) i) 2) “The name 'spirit' preceded by the name of the fruit may also be used for spirit drinks produced by macerating … certain berries and other fruit … in ethyl alcohol of agricultural origin or in spirit or distillate as defined in this Regulation, followed by distillation” Fruit ‘geist’ Article 1) 4) i) 3) “The spirit drinks obtained by macerating unfermented whole fruit … in ethyl alcohol of agricultural origin, followed by distillation, may be called 'geist', with the name of the fruit.”
Page 7 of 173
Table 1: Distillates and Spirit Drinks which are, or can be, Made from Cereals Employed Prior to Distillation (cont.) Spirit Drink Definition in 1576/89 Category Gentian Article 1) 4) k) “A spirit drink produced from a distillate of gentian, itself obtained by the fermentation of gentian roots with or without the addition of ethyl alcohol of agricultural origin” Fruit spirit drinks, Article 1) 4) l) “obtained by macerating fruit in ethyl alcohol of Pacharán agricultural origin and/or in distillate of agricultural origin and/or in spirit as defined in this Regulation” Juniper-flavoured Article 1) 4) m) “produced by flavouring ethyl alcohol of spirit drinks, gin, agricultural origin and/or grain spirit and/or grain distillate with Wacholder, juniper (Juniperus communis) berries” ginebra, genebra, genièvre, jenever, genever, peket Caraway-flavoured Article 1) 4) n) “produced by flavouring ethyl alcohol of agricultural spirit drinks, origin with caraway” akvavit or aquavit Aniseed-flavoured Article 1) 4) o) “produced by flavouring ethyl alcohol of agricultural spirit drinks, pastis, origin with natural extracts of star anise (Illicium verum), anise ouzo, anis (Pimpinella anisum), fennel (Foeniculum vulgare), or any other plant which contains the same principal aromatic constituent” Bitter-tasting spirit Article 1) 4) p) “produced by flavouring ethyl alcohol of agricultural drinks, bitter, amer origin with natural and/or nature-identical flavouring substances” Vodka Article 1) 4) q) “produced by either rectifying ethyl alcohol of agricultural origin or filtering it through activated charcoal, possibly followed by straightforward distillation or an equivalent treatment” Liqueur Article 1) 4) r) “produced by flavouring ethyl alcohol of agricultural origin or a distillate of agricultural origin or one or more spirit drinks as defined in this Regulation or a mixture of the above, sweetened and possibly with the addition of products of agricultural origin such as cream, milk or other milk products, fruit, wine or flavoured wine” Egg liqueur, Article 1) 4 ) s) “obtained from ethyl alcohol of agricultural origin, advocaat, Advokat the ingredients of which are quality egg yolk, egg white and sugar or honey” Liqueur with egg Article 1) 4) t) “obtained from ethyl alcohol of agricultural origin, the ingredients of which are quality egg yolk, egg white and sugar or honey” Väkevä Article 1) 4) u) “produced by flavouring ethyl alcohol of glögi/Spritglögg agricultural origin with natural or nature identical aroma of cloves and/or cinnamon using one of the following processes: maceration and/or distillation, redistillation of the alcohol in the presence of parts of the plants specified above, addition of natural or nature identical flavour of cloves or cinnamon or a combination of these methods” Rum-Verschnitt Article 9) 2) “manufactured in Germany and obtained by mixing rum and alcohol” Slivovice Article 9) 3) “obtained by the addition to the plum distillate, before the final distillation, of a maximum of 30% by volume of ethyl alcohol of agricultural origin”
Page 8 of 173
2. Rationale
2.1 *Rationale for Exemption as Presented in the Application for Provisional Exemption*
There is no evidence and no rational basis for considering that a consumer is at risk of an allergic reaction from cereals (an Annex IIIa substance) when used as raw materials, before distillation, in the manufacture of distillates.
The EU spirits industry supports the principles underlying the legislation in Directive 2003/89/EC – to protect consumer health and give information to sensitive consumers so they can protect themselves from adverse allergic reactions. These principles would not be aided by labelling for cereals when used before distillation in the manufacture of distilled spirit drinks, for two reasons. Firstly, distillation does not allow the non-volatile allergens which are proteins and polypeptides from the cereal raw materials to transfer into the final products. Secondly, such labelling would run counter to medical advice, thus leading to consumer confusion, diminution of choice and, potentially, a dangerous complacency with other products so labelled.
Although commercial distillation systems can vary, the physical principles they are based on are the same. All spirit drinks, by definition and law, undergo various distillation and rectification steps and due to the non-volatility of the cereal derived allergenic or food intolerance causing proteins and polypeptides, they will not distil over into the resultant spirit. For example, it is well known that proteins, polypeptides and amino acids cannot be directly examined by gas chromatography (GC) techniques without derivatisation or pyrolysis because of their non-volatility and research has been unable to detect gluten traces in cereal distillates used in vodka, gin, bourbon, whisky and other products.
A review article “Celiac Sprue” in the New England Journal of Medicine (Farrell and Kelly 2002) states under dietary guidelines for patients with coeliac sprue that “Wine, liqueurs, most ciders and other spirits including whiskey and brandy are allowed.” This advice has been disseminated through various charities and information providers (for example the Coeliac Society, www.coeliac.co.uk or the British Broadcasting Corporation website www.bbc.co.uk). An article in the Coeliac Magazine, “Crossed Grain”, specifically states that whisky can be consumed as part of a gluten-free diet. (McGough 2005). Thus, labelling the distilled spirit drinks produced from cereals under Directive 2003/89/EC would lead to confusion for sensitive individuals, with the medical advice indicating spirit drinks are safe, but the allergen labelling suggesting a threat. The Anaphylaxis Campaign, a UK based charity that offers advice and help to allergy sufferers, believes there is a case that labelling for the presence of things that are not actually there may cause complacency in those at risk and lead them to ignore ingredient and contamination labels on other products which may contain the allergen (e-mail, Gowland, 2004).
Page 9 of 173
2.2 Rationale – Additional Information
Certain precedents with regards to labelling alcoholic distillates have already been set in other countries. Standard 1.2.3 of the Australian Food Standard Code (2006) provides a formal exemption of allergen labelling for spirits. In Japan, alcohol is currently not subject to mandatory allergen labelling, although it is recognized that the subject requires further study. This information is contained in the Ministerial Ordinance amending (in part) the Food Sanitation Law Enforcement Regulations (Ministry of Health, Labour and Welfare Ordinance No. 23 of 2001). It states that for the specified allergens which should be labelled it is “difficult to judge whether the reaction if any is caused by antigenicity of the specified ingredient or the effects of alcohol. Therefore mandatory labelling is not required. For whisky made from wheat, labelling to the effect that it contains such ingredient is not mandatory under the Food Sanitation Law.”
Elsewhere the absence of potentially allergenic proteins is recognised as a means of exempting a product from allergen labelling obligations. In Canada, the proposed regulatory amendments to enhance the labelling of priority allergens in foods (Health Canada website, 2006) indicates that labelling of tree nuts, milk and cereals inter alia is required only if the ingredient, or a protein-containing derivative, is a component of the food product. Similarly, the Food Allergen Labeling and Consumer Protection Act (FALCPA) of 2004 (Title II of Public Law 108-282) in the U.S. permits a food ingredient that is defined as a major food allergen under the FALCPA to be exempted from FALCPA's labelling requirements if it does not contain allergenic protein.
Page 10 of 173
3. Response to the Opinion of the EFSA Scientific Panel on Dietetic Products, Nutrition and Allergies on Commission Request N° EFSA- Q-2004-120
3.1 Introduction
On October 19th 2004, the EFSA Scientific Panel on Dietetic Products, Nutrition and Allergies (NDA Panel) adopted its opinion on a request from the Commission related to a notification from CEPS, The European Spirits Organisation, on distillates made from cereals pursuant to Article 6 paragraph 11 of Directive 2000/13/EC (EFSA, 2004b). As a result of this opinion, cereals used (prior to distillation) for the production of distillates were granted a temporary exemption from the allergen labelling requirements introduced in Directive 2003/89/EC. This exemption was contained in Commission Directive 2005/26/EC.
This section of the dossier summarises the work presented in the request for the provisional exemption of distillates made from cereals, reviews the opinion of the NDA Panel regarding this request and presents a revised scheme of future studies established in response to the NDA Panel’s opinion.
Three principal arguments were presented in the request for provisional exemption of distillates made from cereals. These were:
1) The Physico-Chemical Argument. The size of the allergenic proteins and polypeptides and their lack of volatility prevent them from transferring to a distillate.
2) The Analytical Argument. Surveys of bottled products and the alcoholic distillates they have been made from have all been found negative for either gluten or protein.
3) The History of Safe Use Argument. Clinical experience and therefore advice show no real concern for distillates that originate from cereals containing gluten being a threat to consumers of spirit beverages where these cereals are used before distillation.
3.2 The Physico-Chemical Argument
3.2.1 Summary of Work Presented in the Application for Provisional Exemption to Support the Physico-Chemical Argument
Theoretical arguments, with evidence from a scientific literature review, were presented to support the position that the size of the allergenic proteins and polypeptides and their lack of volatility prevents them from transferring to a distillate.
Planned future work included commissioning an independent expert, with an in-depth knowledge of the distillation process from a chemical engineering background, to explain why proteins cannot carry over into a distillate. This explanation would
Page 11 of 173
include a detailed description of the distillation process as a means of process separation and tests to support the theoretical evidence supplied.
In addition, an expert on distilled beverages would be commissioned to provide a clear explanation of all the different processes involved in the production of distillates from cereals containing gluten.
To ensure that no evidence contrary to the physico-chemical argument had been published since the initial dossier was written, the review of the scientific literature would be updated.
3.2.2 Opinion of the EFSA NDA Panel on the Physico-Chemical Argument in the Provisional Exemption Application
The NDA Panel concluded that the “nature of the manufacturing process makes it unlikely that significant levels of allergenic proteins, peptides or fragments will be carried over into the distillate.”
3.2.3 Revised Programme of Work to Support the Physico-Chemical Argument
The NDA Panel raised no issues with the detail of the manufacturing process, providing that any industrial distillation is properly controlled. It was therefore decided that there would be no advantage in obtaining the opinion of a distilled beverage expert for submission in the completed application.
Whilst the NDA Panel accepted “that it is generally acknowledged that proteins, peptides or fragments will not be carried over into the distillate during a properly controlled distillation process”, it was decided that evaluation of this fact by an independent expert would add further weight to this central argument for exemption. To support this contention, the Scotch Whisky Research Institute would conduct laboratory simulations of the industrial distillation process using model solutions of relevant proteins.
3.3 The Analytical Argument
3.3.1 Summary of Work Presented in the Application for Provisional Exemption to Support the Analytical Argument
In an analytical survey of alcoholic distillates produced using cereals prior to distillation, and bottled spirit drinks made using these distillates, no protein was detected. Two analytical techniques were employed. An ELISA determination detected no gluten in a range of 46 distillate samples. The method was validated, with a limit of detection of 10 mg/l. A total protein determination, manufactured by Randox, detected no protein in a range of 39 distillate samples. The method was validated with a limit of detection of 5 mg/l, excepting gin distillates, which had a higher limit of detection of 10 mg/l.
Planned future work was aimed at extending the range of alcoholic distillates and spirit drinks surveyed, and, if possible, improving upon the sensitivity of the
Page 12 of 173
analytical techniques. 110 further samples would be collected from a wider spread of distillery sources with an approximate split of 70 bottled spirit drink products and 40 alcoholic distillates. All of these samples would be analysed for gluten using the ELISA based method.
Since most total protein methods surveyed were highly influenced by colour, it was predicted that only the non-matured samples would be measured for this analytical parameter. A method to increase the sensitivity of the total protein method had been proposed by Heriot-Watt University. This technique, or an alternative, once developed, would be used to sample production distillate samples from commercial stills.
3.3.2 Opinion of EFSA NDA Panel on the Analytical Argument in the Provisional Exemption Application
“The analytical data presented support the assertion that cereal proteins are not transported into the products of distillation during spirit manufacture at a level above 10 mg/kg. However, the analytical methodology did not address the allergenic activity of residual protein levels in the final product by using appropriate human sera. The further work proposed by the applicant to enhance sensitivity of the analytical methods and analyse further samples would provide further evidence.”
The analytical methods employed were described as being “appropriate for their intended purpose but they lack sensitivity”, and it was noted that for the total protein analysis “the use of a cereal protein standard would have been preferable” compared to the Human Serum Albumin standard used instead.
A comment made regarding the samples which were collected in the survey was that “full sample details were not available”.
3.3.3 Revised Programme of Work to Support the Analytical Argument
Limits of detection in the provisional application were 10 mg/l for gluten and 5 mg/l for total protein. Given the current methodology widely employed in the food industry for the analysis of proteins and the difficult nature of the spirit matrices, it was felt that increasing sensitivity would prove challenging. However, responding to the opinion of the NDA Panel, significant development work would be undertaken to improve sensitivity.
The planned survey of approximately 110 samples would be undertaken using the most sensitive validated techniques available. One alteration, which was not suggested by the EFSA panel opinion, but which was felt to be justified, was to bias the collection from finished bottled products towards production distillate samples. Firstly, it is the raw alcoholic distillate for which an exemption is actually being requested. Secondly, final products can be matured (whiskies) or have sugar or other components added to them (liqueurs) which can interfere with analysis techniques. To this end, the split of 110 samples was altered from 40 production samples and 70 finished bottled products to 75 production samples and 35 finished bottled products.
Page 13 of 173
Responding to the other points raised by the EFSA panel, a cereal protein standard would be used to calibrate any total protein method, if possible. It was also written into the revised programme of work that full details regarding the sampling location and production method would be collected with each survey sample. Some details regarding the initial samples would also be included in the final application. Whilst the EFSA panel noted that the analytical methodology did not address the allergenic activity of residual protein levels in the final product, the opinion indicated that increasing the breadth and sensitivity of the analytical survey would provide suitable additional evidence as to the non-allergenicity of distillates made from cereals.
3.4 The History of Safe Use Argument
3.4.1 Summary of Work Presented in Application for Provisional Exemption to Support the History of Safe Use Argument
An extensive literature review using the PubMed database of the US National Library of Medicine identified no reports linking coeliac symptoms or cereal allergy with the consumption of distilled beverages. Most sensitivities to alcoholic drinks do not appear to be immune mediated, but are more frequently pharmacological intolerances to specific compounds in these drinks. Where allergic and asthmatic reactions to specific non-alcohol components have been reported, these are wholly concerned with non-distilled drinks. Advice given by advisory groups and charities to Coeliac sufferers is that spirits and liqueurs can form part of the diet for a patient with Coeliac Sprue.
Planned future work in this area included two studies to be undertaken by relevant independent experts. The first study would provide further epidemiological evidence on whether any link exists between distillates produced from cereals containing gluten and allergic response. The second study would provide a clear statement of current medical knowledge and advice on whether individuals with coeliac disease could drink alcoholic beverages containing distillates produced from cereals containing gluten.
3.4.2 Opinion of EFSA NDA Panel on the History of Safe Use Argument in the Provisional Exemption Application
Whilst noting that the literature search did not reveal allergic reactions after consumption of distillates made from cereals, the EFSA NDA panel pointed out that adverse effects due to drinking of distillates made from cereals may not be perceived as due to wheat allergens or gluten but rather be attributed to alcohol, and under- reporting may thus have occurred.
The NDA panel inferred that the epidemiological expert would be responsible for carrying out epidemiological studies in order to confirm the absence of adverse reactions to distilled spirit drinks due to cereal allergens or gluten. This had not been planned; the epidemiological expert had been requested to provide an independent assessment on the existence of a link between cereal distillates and allergic responses,
Page 14 of 173
based on the medical literature available. The EFSA NDA panel concluded, in any case, that “clinical trials should be considered only if potentially significant amounts of proteins are found to be present in distillates using analytical methods with appropriate analytical sensitivity.”
3.4.3 Revised Programme of Work to Support the History of Safe Use Argument
The EFSA panel concluded that the critical justification to the argument that distillates made from cereal are safe for those sensitive to cereal allergies was the claim that proteins cannot be detected in the alcoholic distillates. The opinion that it would be unnecessary to carry out any clinical work, provided this claim could be further justified, meant it was unnecessary to retain the proposal for an independent epidemiological expert. Similarly, it was unnecessary to obtain an independent opinion from a medical expert.
Since an epidemiological expert would not be employed, the review of the medical literature provided in the initial dossier would simply be extended to encompass the literature published after its submission.
3.5 Conclusions
A revised programme of work was established, based on the studies and proposed programme of work presented in the application for provisional exemption, and the response of the EFSA NDA Panel to this application. This revised programme of work was as follows:
Physico-Chemical Argument Update the review of the scientific literature to confirm the absence of any evidence that cereal proteins are volatile. Commission an independent expert to provide an explanation, based on distillation theory, as to why cereal proteins will not transfer to the distillate. Perform laboratory distillations using appropriate model matrices to demonstrate that cereal proteins will not distil.
Analytical Argument Improve sensitivity of the analytical techniques used to measure gluten and total protein in alcoholic matrices. Improve applicability of the analytical techniques (use gluten for calibration of the total protein method, if possible). Extend analytical survey of production distillate samples and bottled spirit drinks to include another 115 samples.
History of Safe Use Update the review of the medical literature to confirm that there is no concern that potential allergens that originate from cereals are a threat to consumers of spirit drinks, where the cereals are used prior to distillation.
Page 15 of 173
4. *Characterisation of the Substances for which a Labelling Exemption is Requested, the Methods of Production and the Route of Exposure*
4.1 *Characterisation of the Exemption Requested*
Distillates used in the production of alcoholic beverages are based on the distillation of alcohol produced from the fermentation of sugars by yeast. There are a number of sugar sources from which this alcohol can be produced. Cereals represent one of these sugar sources. This request applies to an exemption from Annex IIIa of European Directive 2003/89/EC of “distillates made from alcohol derived from the following raw materials employed as sugar sources prior to the distillation process: cereals containing gluten (i.e. wheat, rye, barley, oats, spelt, kamut or their hybridised strains) and products thereof”.
This request for an exemption does not extend to any alcoholic beverages which contain substances produced from any of the Annex IIIa ingredients defined in Directive 2003/89/EC which are added after the distillation step.
4.2 *Characterisation of Annex IIIa Raw Materials Used in the Manufacturing Process*
All the cereals listed in Annex IIIa of Directive 2003/89/EC share the fact that they contain gluten. “Gluten” is the general term for a complex mixture of grain proteins within cereal grains which cause the food intolerance known as coeliac disease or sprue (EFSA, 2004a). Of those cereals listed in Annex IIIa, the most commonly used in the manufacture of distillates are wheat, barley and rye. The proteins within these cereals specifically identified as responsible for coeliac disease are the gliadins (wheat), hordeins (barley) and secalins (rye) (Wieser, 1996). The range of molecular weights of gliadin proteins are 28,000-39,000 Daltons and they are constructed from repetitive peptide units common to all gliadin subtypes (Stern et al, 2001). The hordeins and secalins are of similar size and peptide sequences to the gliadins and show immuno cross-reactivity (Palosuo et al, 2001). Although these intact proteins are of significant size it is clear that peptide fragments, for example resulting from food processing, still retain some toxicity with respect to coeliac disease (Stern et al, 2001). However, the peptide fragments which appear to retain their toxicity, such as the 33 amino acid peptide identified recently (Shan et al, 2002), are large molecules, which, like their parent proteins, will be non-volatile in an ethanol-water distillation system.
In addition to coeliac disease there are true allergens found in cereals, including the gliadin-like proteins discussed above (EFSA, 2004a). Those allergenic proteins identified so far are diverse, with wide ranges of characteristics such as molecular weights and thermostability (Simonato et al 2001). The examples of known wheat allergens in the EFSA opinion (EFSA, 2004a) have molecular weights between 15- 65,000 Daltons. Again these are all large molecules and even if a cereal peptide of 14-15 amino acids were to retain its allergenic properties, as has been reported for some other non-cereal food allergens (EFSA, 2004a), it would still have a high molecular weight in the region of 1200-1500 Daltons.
Page 16 of 173
4.3 *Characterisation of Methods of Production*
4.3.1 *Production of Distillates*
The distillates included in this dossier are produced using processes that differ in detail but follow very similar principles. These principles are summarized in Figure 1.
1. Cereal Processing Extraction of starch from cereal and conversion to smaller, fermentable, sugars
2. Fermentation Conversion of sugar to alcohol by the action of yeast
3. Distillation Concentration of alcohol, removal of non-volatile material, control of flavour
Distillate
4. Post-Distillation Processes These may include flavouring, maturation, sweetening and the addition of other ingredients prior to bottling
Figure 1 Flow diagram showing the principle steps in the creation of alcoholic distillates using cereals as the raw material
These common principles have been well documented for a wide number of products, such as Scotch malt whisky and its relatives (Simpson 1968, Lyons & Rose 1977, Watson 1983, Lyons 2003, Piggott 2003), grain spirits such as Scotch grain whisky (Pyke 1965, Panek and Boucher 1989, Campbell 2003), bourbon (Lioutas 2004, Ralph 2003), Canadian whisky (Piggott 2003) and rye (Ralph 2003), as well as products deriving from neutral spirits (Wilkin et al 1983), such as vodka (Madson 2003, Piggot 2003) and gin (Murtagh 1995, Piggott 2003). Neutral spirit is a common term for the high-strength, high-purity distillate defined as “ethyl alcohol of agricultural origin” in EEC Spirit Drinks Regulation 1576/89. It also provides the basis for a large number of other alcoholic beverages such as aniseed spirits, liqueurs and cordials (Murtagh 1995), as well as ready to drink products. Madson (2003) describes the main features of the distillation processes which are typical of the production of spirit beverage products deriving from the fermentation of grains and molasses, and provides a concise but reasonably detailed account of the principles of continuous distillation.
Page 17 of 173
In the production processes for cereal-based distillates, there are two main stages. These are cereal processing and fermentation, in which the components of the cereal are degraded and converted to a suitable substrate and fermented by yeast to produce alcohol, and distillation, where the alcohol produced during fermentation is collected and refined, prior to any maturation, blending or packaging processes.
4.3.2 *Cereals Processing and Fermentation*
Cereals processing is the first stage in the production of cereal based distillates. In cereals processing, which has many features in common with beer brewing, cereals such as barley, barley malt, maize, wheat and rye are heat treated in mashing and/or cooking processes to liquefy and gelatinise the main substrate (starch) and convert it into fermentable material (saccharification). These processes range from simple mash infusions (65 – 95°C) which are used for the production of malt whisky and similar spirits (Lyons 2003, Simpson 1968, Rutherford 1981), to high temperature atmospheric and pressure cooking systems (95 - 150°C) (Pyke 1965, Bringhurst et al 2003, Lyons 2003) which are used for grain and neutral spirits (Murtagh 1995).
In the case of Scotch malt and grain whisky, saccharification must take place solely as a result of the action of indigenous starch degrading enzymes, deriving wholly from barley malt, on the substrate. By law, no other enzymes or process aids are allowed in the production of Scotch whisky. For other distilled beverages saccharification by externally added enzymes may be permitted. The main biochemical changes which take place during processing are hydrolysis of starch, protein and other biopolymers to provide water soluble, low molecular weight compounds which form a fermentable sugar substrate (called wort), which is fermented to alcohol by the yeast, Saccharomyces cerevisiae (Lyons 2003).
After cooking and mashing, the wort pH would be around 5.5 (Rutherford 1981), falling to about 4.0 on completion of fermentation (Dolan 1976). The conversion of sugars to alcohol by yeast also results in the production of carbon dioxide. In small- scale production, the vessels used for fermentation are typically closed vessels of wooden construction, with no means of temperature control. Since it is uneconomical to collect the carbon dioxide, it is simply vented. In larger scale production, the fermenters are stainless steel, will have cooling facilities and in some cases the ability to collect the carbon dioxide. A typical fermentation will run for 40-48 hours (Piggott and Conner, 2003). After fermentation, the fermented liquor, called ‘wash’ or ‘beer’, is distilled.
4.3.3 *Distillation*
The principles of distillation have been well established for a long time now, and there are a number of textbook references describing the process in some detail (Piggott et al 1989, Russell 2003, Jacques et al 2003). Madson (2003) and Piggott (2003) give reasonably detailed, concise descriptions of the processes which are involved in the application of modern distillation techniques to ethanol (alcohol) production. Nicol (1989), Panek and Boucher (1989) and Campbell (2003), provide information which has been specifically applied to the production of Scotch whisky, as well as other products.
Page 18 of 173
The principal aims of distillation in the context of the production of alcoholic beverages are to separate the volatile from the non-volatile components (Considine and Considine 1984) and to increase the concentration of alcohol. The alcohol and volatile flavour congeners are fully separated from the grain solids and other non- volatile compounds (Panek and Boucher 1989), and partially separated from the less volatile water component with consequent increases in alcohol concentration in the distillate. All distillation systems operate on the same fundamental principles, obeying physical laws which state that different materials boil at different temperatures (Madson 2003).
Under certain circumstances, non-volatile and high boiling point components can also be entrained (carried through the process as liquid or solid particles within the vapour phase) during part of the wash (‘beer’) distillation (Ohtake et al 1995); however, proper distillation procedures prevent non-volatile material being transferred into the distilled product, since this process is undesirable in terms of flavour and product quality.
The separation of volatile components is determined by their relative vapour pressures and can be achieved in two ways, by simple (or batch) distillation or by continuous (fractional) distillation (Considine and Considine 1984). The former is achieved by pot still distillation, and is mainly used for traditional distilled products such as Scotch and Irish malt whisky, and certain forms of bourbon (Nicol 1989, Lyons 2003, Piggott 2003 and Lioutas 2004).
Continuous distillation is universally used for grain and neutral spirits and provides a well rectified product which can be used in a wide range of applications, such as grain whisky (Panek & Boucher 1989, Campbell 2003), bourbon, corn, rye and Tennessee spirits (Ralph 2003), as well as neutral spirits (Murtagh 1995, Piggott 2003). In this process steam rises through the distillation column, stripping alcohol from the fermented mash. Alcohol vaporises at 78.5°C but gluten or gluten fragments do not vaporise and rise through the still. In some cases the base spirit is redistilled through a multicolumn distillation system, consisting of three to five columns, to remove fusel oil and specific volatile substances. These additional distillation steps provide further barriers to protein carry over into the final product (Campbell 1988).
Efficient systems for the commercial production of neutral spirit from grain and molasses have been in operation for many years (Madson 2003). Modern distillation systems are multi-stage, continuous countercurrent vapour liquid contacting systems which are effective in separating, from a given feedstock, a purified product (Madson 2003).
4.3.3.1 *Batch (Pot Still) Distillation*
Nicol (1989) describes the design and operation of the simple batch distillation which is characteristic for the production of Scotch malt whisky. Spirits deriving from pot stills are normally prepared from a batch distillation process which comprises at least two, sometimes three discrete distillation steps.
Page 19 of 173
Figure 2 shows the main features of pot still distillation. In the first stage, which is known as the wash distillation, the fermented liquor (wash), which has an alcohol strength of around 8 - 10 % alcohol by volume (ABV), is distilled into a colourless product (which is known as low wines). This normally has an alcoholic strength in the region of 21% ABV, and is collected in a low wines receiver, where it is mixed with foreshots and feints, from the previous spirit distillation. Wash distillation is carefully controlled to avoid foaming (Nicol 1989) and entrainment of high molecular weight substances (Ohtake et al 1995) into the intermediate product (low wines). This is a quality control measure, since such entrainment could adversely impact on flavour.
Spirit cut (ca 75 – 64 % ABV) Low wines (ca 21% ABV)
Wash still Spirit still
LWF Residue Residue to still Wash (Beer) to still Low Wines Foreshots (ca 8-10% ABV) /Feints (LWF) Feints Spirit Receiver
(ca 25-30% ABV) Spirit Receiver
(ca 70-75% ABV)
Figure 2 Diagram showing the main features of a two stage pot still distillation process, typical of Scotch malt whisky and similar products
In the next stage, the low wines and feints mixture, which has an alcohol strength of approximately 25 – 30% ABV, is transferred to a second (spirit) still and further distilled into a final new make spirit at relatively high alcohol strength (70 - 75 % ABV). In the spirit distillation, the distillate is collected in three fractions, foreshots (first runnings), spirit (middle cut) (approximately 70-75% ABV) and feints. The foreshots and feints are returned to the low wines receiver and mixed with the low wines. The middle cut or new make spirit is then collected in a separate vessel and passed forward to the maturation process. Spirit distillation is not subject to the possible entrainment problems associated with wash distillation (Nicol, 1989).
For some products, such as Irish whiskey and certain Scotch malt whiskies, a third distillation is used to further refine the product. Lioutas (2004) describes a similar system for the production of a traditionally produced bourbon product.
Page 20 of 173
4.3.3.2 *Continuous Column (Grain Still) Distillation*
Continuous distillation can be used to produce whiskies such as Irish and Scotch grain whisky, bourbon and Canadian whisky, as well as other spirits.
Ethanol and volatile congeners Condenser Low boiling Reflux Reflux Lower product
Rectifying/Concentration Volatiles concentrated as Section they rise up the still
vapour Distillation Plates
Temperature gradient Feed to Still (Wash)(Wash) Feed Plate
Stripping Alcohol vapourised and stripped from (Analyser) feedstock as it falls through the still Section
Higher Boiling liquid
High boiling product, non volatile residues Steam (‘Bottoms’, Stillage, Spent Wash)
Figure 3 Simplified diagram showing the main features of a single column continuous still (based on Madson, 2003)
Continuous distillation systems are based on a typical distillation tower arrangement, whereby the feed enters close to the middle of the tower (Figure 3). In this type of system, the part of the column above the feed is known as the rectifying section, and the part below the feed is termed the stripping section. The column is designed so that there is a temperature gradient, with the hottest part of the column at the bottom (Madson 2003). In some distillation systems, the distillation tower can be replaced by separate stripping and rectifying columns.
Single Column Systems
Single column distillation systems are used for the distillation of products such as Canadian whisky and bourbon (Piggott, 2003). Bourbon distillers can also use a second distillation step in a doubler or thumper, a heated pot still-like chamber to increase the alcohol strength of the distillate (Ralph, 2003).
Page 21 of 173
Distillation columns have a large number of individual trays or plates, known as stages or contactors, which allow intimate contact between the rising vapours and the descending liquids to facilitate the separation and enrichment of the volatile components of a mixture (Madson, 2003). As the feed enters the column, it falls to the bottom and the heat vaporises the volatile material, which rises towards the top of the column. The volatile fractions are then further enriched as they rise through the rectifying section. The rectifier effectively removes any remaining excess water, resulting in a concentrated, purified product. The non-volatile and high boiling point material falls to the foot of the column and is discharged at the bottom (Madson, 2003). These are referred to as ‘bottoms’, stillage or spent wash.
Coffey Still Type Systems
The continuous grain distilling processes used in the production of Scotch grain spirit and neutral spirit have evolved over the years, from the early type of Coffey still to more modern fractionation columns (Piggott 2003). However, the principles of operation are still the same.
Coffey stills are still used for the production of grain spirit with few changes from when they were first introduced (Lyons, 2003; Campbell, 2003). Variations on this design are still in use in the Scotch whisky industry (Piggott, 2003). More modern design features are present in continuous stills for the production of North American spirits (Campbell, 2003). They are also used in some form for a wide range of potable spirits including bourbon, corn, rye and Tennessee whiskies (Ralph, 2003). Organoleptic qualities can be modified by manipulating the outlet strength of the spirit collected at the still (Piggott, 2003).
In its simplest form, a Coffey still comprises two main parts, an analyser (stripping column) and a rectifier. The beer or wash is preheated (“hot feed”), added at the top of the analyser and is further heated as it falls down the column. This process strips off the volatile materials from the wash and the alcohol rich vapour rises to the top of the column, where it is passed to the bottom of the second column (rectifier). The main function of the rectifier is to remove water from the product distillate (Panek and Boucher 1989). As the vapour rises through the rectifier, different components are separated according to their boiling points and their volatility (Campbell 2003). The volatile components are concentrated, while less volatile components are removed from the distillate (Panek and Boucher 1989). Normally a proportion of the distillate is recycled to the top of the still as reflux and this has the function of maintaining the liquid level on each plate and is critical in providing the equilibrium which results in the mass transfer which takes place throughout the column (Panek and Boucher 1989).
Figure 4 gives a general description of the principal features of a typical Coffey Still.
Page 22 of 173
Heads recycle from fusel oil still
Hot Spirit Vapour * Hot Feed Vent ~8-10% ABV
Condenser Spirit Draw Point
* Feed preheated to ca 90°C by passing through rectifier Reflux Spirit Cooler (not shown for clarity)
~94% ABV
Vapour Heads recycle
Analyser Rectifier
Vapour
Fusel oil still
Fusel oil Spent wash, stillage out out Steam Hot Feints Recycle ~10-15% ABV
Figure 4 Simplified diagram showing the main features of a Coffey Still (based on Campbell 2003, Piggott 2003)
Both the analyser and rectifier are subdivided by a series of stages by horizontal plates (usually 35 – 60 (Campbell 2003)). These are normally made of copper in Coffey stills, although steel can also be used. At each stage, vapour from the previous stage rises through the perforations in the plate, and is condensed, liberating its latent heat (Panek and Boucher 1989). This establishes a dynamic equilibrium which causes the slightly more volatile material to vaporise and continue to rise through the column, gradually becoming more enriched as it reaches the draw point, where the product is removed from the column (Campbell 2003). The flow through the columns is carefully controlled to prevent entrainment of the liquid phase. Additionally the plates in the analyser are carefully spaced, to avoid problems with frothing and subsequent potential contamination of the lower regions of the rectifier by the feed stock. In beer stills (analyser) there may be a demister, which removes any particles which might be entrained in the vapour (Panek and Boucher 1989).
4.3.3.3 *Neutral Spirit and other Grain Spirits*
Neutral spirits (which are essentially highly purified distillates) can be produced from almost any feedstock using a continuous still (Murtagh 1995). Neutral spirits produced from cereals have very low taste and odour, and are used for the production of white spirits such as gin and vodka (Murtagh, 1995; Piggott, 2003). Components with odour or taste characteristics are reduced to thresholds defined by the desired end product use of the spirit.
Page 23 of 173
High degrees of purity are achieved by increasing the rectification capacity of the system. This is done by adding additional ‘rectifying’ columns to the basic two column distillation system (sometimes up to four additional stages). The degree of rectification depends on the product which is required.
In the production of neutral spirit, rectification removes the last traces of undesired congeners such as fusel oils, which are a mixture of higher alcohols such as propyl, butyl and amyl alcohols and heads, which includes the acetaldehyde fraction (Madson 2003), and raises the alcohol strength up to the azeotropic point (97.2% ABV). The general layout of a continuous neutral spirit distillation system is similar to that of a grain still. However, the rectification stage is usually modified to include an intermediate hydrofining or extractive distillation column (Piggott 2003). This additional column is designed to remove the bulk of the fusel oils by injecting water into the spirit. This has the effect of separating the water soluble components such as ethanol and methanol from the less soluble fusel oils components, before the product is passed to the rectifier.
The rectifying column contains a large number of plates, which promotes a large amount of reflux in the column. This concentrates any remaining congeners so that they can be easily removed. However, in practice, the final removal of these materials may require additional fusel oils and heads columns and a stage for methanol removal, if this is not removed at the start of the process (Piggott 2003).
Wilkin et al (1983), describe some essential features of the distillation process for neutral alcohol production. While the terminology and names given to some of the process components may be slightly different for different producers, the overall principles are the same. A simplified diagram summarising the neutral spirit distillation system design described in this reference is shown in Figure 5. Foul heads to Condensers Condensers waste stream Heads Concentrating Column Heads stripping Column Heads draw
Vapour
Recycle to Beer column Feints Recovery Heads draw Heads draw Condensers Condensers
Wash (Beer) feed RectifyingColumn Fusel Oil Column Fusel
BeerColumn Product
Feints recycle
Fusel oil
Spent wash Steam to waste stream
Figure 5 Simplified diagram showing the main features of the distillation of neutral spirit (based on Wilkin et al (1983))
Page 24 of 173
4.3.4 *Distillation System Comments*
It is important to note from the preceding descriptions that all the various distillation systems that are used to produce distillates from cereal based substrates will prevent the transfer of non-volatile material from the feedstock into the final product. The principles of distillation are to concentrate the alcohol, to remove all the unwanted non-volatile material and select the desired volatiles to give a defined character.
4.4 *Characterisation of Routes of Exposure to Consumer*
The principle route of consumer exposure to distillates where cereals are used in their manufacture before distillation is through the consumption of the spirit drinks and alcoholic beverages produced using ethyl alcohol of agricultural origin (see Table 1). Such exposure does not pose a threat to those individuals sensitive to cereal allergens since the allergens are non-volatile and do not pass into the distillate.
Page 25 of 173
5. Arguments and Data to Support Exemption – Physico-Chemical Evidence
5.1 *Physico-Chemical Evidence as Presented in the Application for Provisional Exemption*
A dictionary definition of distillation is “the process of converting a liquid into a vapour, condensing the vapour, and collecting liquid or distillate. It is used for separating mixtures of liquids of different boiling points or for separating a pure liquid from a non-volatile constituent” (Penguin Dictionary of Science, 1986). In the production of distilled spirits, the conversion of liquid to vapour will occur in the temperature range of 78 to 100 °C.
Proteins and peptides do not distil. There are a number of scientific literature references which provide evidence of this fact. According to Lehotay and Hajšlová (2002), proteins and other high molecular weight substances cannot be analysed by normal gas chromatography. Gas chromatography is effective due to the volatile nature of the compounds under analysis at temperatures which include those involved in the distillation process. It is not suitable for the analysis of proteins because they are non-volatile. The non-volatile nature of proteins is employed in protein extraction techniques. Dickey et al (2002) describe how maize proteins can be extracted from milled maize by vigorous mixing in heated ethanol solutions. This process re- concentrates the ethanol for future extractions by employing a distillation step to separate the distilled ethanol from the residual protein fractions.
The non-volatile nature of proteins is addressed by Dr. Donald Kasarda in an article published on the website of the U.S. Department of Agriculture (2003), http://wheat.pw.usda.gov/ggpages/topics/Celiac.vs.grains.html. He addresses the issue of wheat proteins in grain alcohol by stating that “the toxic peptides (in fact all peptides) have low volatility, whereas alcohol produced by grain fermentation has a high volatility” and therefore that “properly distilled alcohol derived from a base of grain alcohol should be safe”. This argument extends to the distillation of alcohol derived from other protein-containing sources. This argument is supported by Campbell (1988). With regards to gluten proteins, Campbell concludes that “distilled beverages … do not contain measurable amounts of gluten or gluten residues” because “it is known that proteins are not distillable”. He supports his argument by demonstrating that two alcohol distillates contained no measurable solid residue.
Oldani et al (2001) were unable to detect gluten traces in vodka, gin, bourbon, scotch and grain spirit used as an ingredient in other products despite a tenfold concentration and an estimated detection limit of 0.0032 mg/L. They calculated a theoretical worst possible case of less than 0.006 mg of gluten per litre of spirit (although it is stressed they found no positive results). The theoretical limit of detection was estimated by applying a recovery factor determined at 10 mg/l to the lowest gluten calibration standard.
An earlier study by Vandenheede (1993) detected gluten in one sample of whisky. However, the assay used also gave positive results for port and cognac which are not produced from cereals. The authors themselves conclude that cereal distillates should
Page 26 of 173
not contain cereal proteins since distillation will eliminate any proteins. Positive results in the distilled spirits were attributed to the addition of cereal derived substances after distillation, or a lack of specificity in the polyclonal antibody assay employed at that time.
A recent study into the production of neutral spirits from cereals (de Laat, 2004) demonstrates that no gluten protein residues are detected in the final spirit (limits of detection quoted for methods used were 2.4 and 3.1 mg/l). Although each of the distillation columns in the neutral spirit production system would act as a barrier to the transfer of gluten protein into the final product, it was demonstrated that the first distillation column was sufficient to remove gluten protein residues.
The absence of any scientific papers considering proteins and peptides in conjunction with their volatility and/or the distillation process provides evidence as to the scientific confidence that proteins and peptides do not distil. A study of the literature over eleven years (1993 – June 2004) using the Thomson ISI Current Contents Physical, Chemical & Earth Sciences and Agriculture, Biology & Environmental journal databases was conducted. This study identified papers which matched the following search criteria: the text words PROTEIN or PEPTIDES or AMINO ACIDS in conjunction with the text words DISTILLATION or VOLATILITY or VAPOUR PRESSURE. At the time the study was conducted the Physical, Chemical & Earth Sciences database covered 1160 scientific journals and the Agriculture, Biology & Environmental Database covered 1099 scientific journals. (There will be some overlap in coverage). The only papers reported relating to proteins or peptides and distillation/evaporation were those considering protein residues remaining after distillation. Two abstracts from the Agriculture, Biology and Environmental Database of Journals by Mustafa et al (2000) are given by way of examples. Both refer to protein analysis of the products which remain in the original liquid following distillation and make no reference to protein presence in the distilled alcoholic sample.
5.2 Physico-Chemical Evidence - Additional Information
The literature review conducted in Section 5.1 was extended to April 2006 to determine whether any newly published data could be found on the subject of proteinaceous material and distillation (coverage had extended to 1193 journals in the Physical, Chemical & Earth Science database and 1136 journals in the Agriculture, Biology & Environmental database). No papers were found referring to the volatility of proteins, peptides or amino acids or their detection in any distillates. The search did identify a paper (Asano et al 2004) which refers to the use of continuous distillation to extract undesirable odours from milk protein hydrolysates. As with the abstracts from the papers by Mustafa et al, separation is achieved because the protein constituents are non-volatile.
5.3 Physico-Chemical Evidence - Opinion of a Physico-Chemical Expert
As indicated in the application for provisional exemption, an independent opinion from a physico-chemical expert would be sought to highlight the effects of distillation on the potentially allergenic proteins of interest. The commissioned expert was Dr. Eva Sørensen, Senior Lecturer at the Department of Chemical Engineering,
Page 27 of 173
University College London, London, UK (see Appendix 1 for further details). Dr. Sørensen’s remit was to provide a fundamental understanding of the parameters of distillation and produce an opinion on the likelihood that allergenic substances from the raw materials used to make distilled spirit drinks could transfer to the distillate. The final report studies the factors affecting the likelihood of any specific substance passing from an ethanol/water matrix into the resultant distillate and describes the physico-chemical conditions experienced during distillation.
Dr. Sørensen’s report is reproduced in Appendix 2 It confirms the contention, based on the concepts of distillation theory, that allergenic proteins would not be expected to transfer from the starting material to the final distillate. Specifically it states that:
“Compounds which typically do not have a vapour pressure are normally larger, or more complex, molecules. In relation to alcoholic distillate production, this would include proteins, peptides, amino acids, saccharides etc. To the best of my knowledge, no studies presented in the open literature have provided evidence that these do have a vapour pressure and there is general agreement in the scientific community that they do not.”
In summary, it confirms that for:
“Standard operation of batch and continuous distillation, as it is generally performed in alcoholic distillate production …. compounds which exert no vapour pressure, such as proteins, peptides, amino acids, saccharides… can therefore not be present in the distilled product.”
5.4 Physico-Chemical Evidence as Presented by the Analysis of Model Distillations
To support the distillation concepts presented in Section 5.3, model distillations of proteins were conducted at the Scotch Whisky Research Institute. To this end, two sets of protein distillations were performed. The first set of distillations used Bovine Serum Albumin (BSA) in an alcohol/water matrix; the second set of distillations used gluten. A distillation protocol was established, based on a single pot still distillation (see Figure 6). This is the simplest type of distillation regime employed in the alcohol beverage industry and the most likely to result in protein detection in the distillate, if proteins exhibited any volatility. Full experimental detail regarding the design of these distillations can be found in Appendix 3.
Both sets of model distillations, using BSA and gluten, confirmed the principle that proteins do not distil, as stated by Dr. Eva Sørensen in Section 5.3. No BSA protein or gluten protein were detected in the distillates produced. Duplicate distillations of both the BSA and the gluten containing matrices were performed and the level of gluten employed in the model distillation was an accurate estimate of the maximum concentration found in a cereal distillery wash.
In the case of the BSA protein distillation, even assuming there was protein in the distillate at a level below the limit of detection, a minimum reduction in concentration of 2,000 was achieved. Similarly, in the case of the gluten protein distillation, the minimum reduction factor would be over 45,000. All cereal based industrial
Page 28 of 173
distillations employ more than one distillation step (whether batch based as in this laboratory exercise or with a number of distillation plates in a continuous column based still). To determine the theoretical minimum reduction for an industrial process, even assuming proteins could actually distil, repeated application of the above calculated reduction factors would be required for each distillation step.
Condenser
Neck
Pot containing model solution Isomantle
Figure 6 Experimental set-up for model protein distillations
5.5 Conclusions
This section of the dossier has considered the physico-chemical evidence which supports an exemption for cereals used in the production of distillates prior to distillation. In summary:
1. A review of the scientific literature has confirmed the absence of any evidence that cereal proteins/peptides are volatile. 2. An independent expert has provided an explanation, based on distillation theory, as to why cereal proteins will not transfer through to the distillate. 3. Laboratory distillations using appropriate model matrices have confirmed that proteins, including the cereal protein gluten in particular, do not distil.
Page 29 of 173
6. Arguments and Data to Support Exemption – Total Protein Based Analytical Evidence
6.1 Total Protein Based Analytical Evidence as Presented in the Application for Provisional Exemption
In the application for provisional exemption, 39 samples of new make whisky and white spirits were analysed for total protein levels using a Total Protein Assay by Randox. Complete details regarding this set of analyses can be found in Appendix 8. The results of recovery experiments established a limit of detection of 5 mg/l for protein in ethanol, vodka and whisky samples. The limit of detection established for protein in gin was 10 mg/l. Recovery experiments were based on the use of the protein human serum albumin (HSA).
The range of alcohol distillates analysed and the levels of measured protein are reproduced in Table 2. The results show that no protein was detected in any of the samples tested.
Table 2: Protein in Samples of Ethyl Alcohol of Agricultural Origin, New Make Whisky, Gin and Vodka.
Sample Type and Number. Description Protein Concentration (mg/l) Vodka (cereal derived) Bottled Products S04-0710 Altai Vodka ND S04-0711 Wodka Wyborowa Pure Rye Grain ND S04-0717 Smirnoff Triple Distilled Vodka ND Altia Corporation – Rajamaki Plant – Dry S04-1065 ND Vodka
Gin (cereal derived) Bottled Products S04-0716 Tanqueray Special Dry Gin ND S04-0727 Bombay Sapphire Dry Gin ND S04-0829 Grants Gin ND S04-0870 Greenalls Gin Finished Product ND S04-1047 Hendricks Gin ND
Bourbon/ Canadian Whisky/Other Bottled Products Whisk(e)y Products S04-0973 Echter Nordhaufer Doppelkorn ND
Neutral Spirit Production Distillate Samples S04-0827 Girvan Distillery Grain Neutral Spirit ND S04-0872 Greenalls Grain Spirit ND S04-0900 Grain Neutral Spirit ex Velva (French) ND S04-0901 Grain Neutral Spirit ex Velva (French) ND S04-1034 Sedalcol Grain Neutral Alcohol 23/03/04 ND S04-1035 Sedalcol Grain Neutral Alcohol 07/06/04 ND S04-1036 Sedalcol Grain Neutral Alcohol 11/05/04 ND S04-1038 Sedalcol Grain Neutral Alcohol 14/04/04 ND Altia Corporation – Koskenkorva Plant – S04-1064 ND Grain Ethanol Cameronbridge Scotch Grain Neutral S04-1066 ND Spirit
Page 30 of 173
Sample Type and Number. Description Protein Concentration (mg/l) Grain New Make Spirit Production Distillate Samples (Scotch/Irish) S04-0826 Girvan Scotch Grain New Make Spirit ND S04-0874 Middleton Irish Grain New Make Spirit ND Cameronbridge Scotch Grain New Make S04-1067 ND Spirit S04-1076 Strathclyde Scotch Grain New Make Spirit ND
Malt New Make Spirit Production Distillate Samples (Scotch/Irish) S04-0869 Aultmore Scotch Malt New Make Spirit ND S04-0873 Middleton Irish Pot Still New Make Spirit ND S04-0903 Dalwhinnie Scotch Malt New Make Spirit ND S04-0904 Lagavulin Scotch Malt New Make Spirit ND S04-0906 Talisker Scotch Malt New Make Spirit ND S04-0907 Caol Ila Scotch Malt New Make Spirit ND S04-1071 Tormore Scotch Malt New Make Spirit ND S04-1072 Laphroaig Scotch Malt New Make Spirit ND Glentauchers Scotch Malt New Make S04-1073 ND Spirit S04-1074 Miltonduff Scotch Malt New Make Spirit ND S04-1075 Glenburgie Scotch Malt New Make Spirit ND
Gin (cereal derived) Production Distillate Samples S04-0712 Compounded Gin 193 ND S04-0713 Gin Distillate (LGS) ND S04-0871 Greenalls Gin Concentrate ND S04-0893 Distilled London Dry Gin Concentrate ND
ND = not detected. Limit of Detection (LOD) = 5 mg/l for all samples except the gin samples, for which the limit of detection is 10 mg/l.
6.2 Development and Validation Work to Improve the Sensitivity of the Total Protein Method
To provide additional evidence that protein does not transfer from the raw materials through to the alcoholic distillate, it was stated in the application for provisional exemption that a wider range of samples would be tested for total protein levels. Responding to the EFSA NDA panel opinion, a more sensitive method for total protein detection was sought. A range of analytical methods was tested in order to provide the most sensitive analysis for the measurement of total proteins in an alcoholic matrix. Full details can be found in Appendix 9. The discussion of method development demonstrates fully the problems experienced with the methods available and the specific difficulties of applying them to the alcoholic matrices under consideration.
Initial experimentation concentrated on the use of reagents which exhibit fluorescence on interaction with proteins. Sensitivity was shown to be good with the best of these methods, (a linear working range of 0.1-3 mg/l was found to be readily repeatable for one method). However, measurement of alcoholic matrices after evaporation and reconstitution in water highlighted issues with inter alia positive results from blank solutions and variability of results. It was apparent that the nature of the alcoholic
Page 31 of 173
samples made them unsuitable for analysis with the fluorescent techniques under consideration.
The Randox assay used to provide data for the measurement of total protein in the application for provisional exemption was a colorimetric method. An improved colorimetric method was sought and a modified Bradford analysis selected as being most suitable. By the use of a concentration step, repeatable recoveries at low concentrations were noted for BSA protein spiked into a number of alcohol distillate matrices. These validated detection levels, 0.5 mg/l for whisky/vodka/neutral spirit and 1.0 mg/l for gin, corresponded to a 10-fold increase in sensitivity compared to the previously reported Randox method.
One drawback to the Bradford Microassay method was noted. Some cereal based distillates, principally Scotch new make, contain long chain fatty acid ethyl esters which are only sparingly soluble in aqueous solution. These constituents, possibly exacerbated by the presence of copper and phenolic constituents, could interfere with the measurement technique. For those instances where the Bradford method provided a positive result, an alternative method carried out by a contract laboratory, AAA-DirectTM was shown to be effective in demonstrating whether this was due to protein or not. AAA-Direct™ is an LC method with integrated pulsed amperometry which analyses the amino acid content of a sample, following hydrolysis, to evaluate total protein content. Individual amino acids could be detected down to concentrations of less than 0.05 mg/l. Despite not measuring protein directly, the presence of cereal proteins at levels down to at least 0.5 mg/l could be detected by this method.
The AAA-DirectTM complimented the simpler in-house Bradford method. The Bradford method could act as a screening technique. Those samples for which a positive result was determined from the Bradford method could be analysed by the AAA-DirectTM method if it was suspected that the positive result was due to the presence of interferences. Neither method was able to analyse samples which have undergone a maturation process.
6.3 Sampling Protocol
Having developed more sensitive methodology for the detection of total protein in alcoholic distillates, this was now applied to an extended range of alcoholic distillates to provide further evidence that cereal proteins are not transferred into the distillate. These distillates are listed in Appendices 6a and 6b. Samples were requested from a wide range of distilleries and production facilities across the world. For full details of the sampling protocol and the implementation of traceability during storage and analysis, refer to Appendix 7.
Coverage of spirits produced from cereals was widespread in terms of both geography and distillation system types. Samples were obtained from Scotland, Ireland, France, Germany, The Netherlands, Denmark, Italy, Russia, Finland, Sweden, Japan, The United States and Canada. A number of bottles of popular brands were requested, as well as the alcoholic distillate that forms the basis of these spirit drinks. In terms of distillation systems, samples were supplied from distillation systems that use two pot stills, three pots stills, one column still, two column stills, multi-column stills and
Page 32 of 173
various combinations of pot and column stills. Where possible, additional differences were sourced. Some Scotch malt whisky distilleries use worms (a pipe passing through a water trough) in order to condense distillate vapours, whereas others use the more efficient shell and tube condenser arrangement, and examples of spirit using both these arrangements were collected.
In total, 39 bottled products and 76 production distillate samples were collected. Of these, 86 samples were non-matured and could thus be analysed for total protein. In some cases, repeat samples were requested from a distillery. In such an instance, repeat samples were taken from either separate still systems or separate batches of alcoholic distillate.
In addition, for those manufacturers producing ethyl alcohol of an agricultural origin derived from cereals, information regarding the markets to which they sell their product was requested. The following industries were identified: spirits, ready-to- drink alcoholic beverages, vinegar, flavour, food, pharmaceuticals and cosmetics. The principal outlet for neutral spirit for the distilleries that supplied samples for the project was the alcohol spirits industry. However, neutral spirit is also used in other alcoholic beverages, as well as food applications such as flavourings and vinegar.
6.4 Protocol for Total Protein Analysis Using the Bradford Analysis Microassay and AAA-DirectTM Analysis
Full details for the protocol used to determine total protein using the Bradford Analysis Microassay can be found in Appendix 10. Appropriate calibration curves were constructed between 0.2 mg/l and 10 mg/l using BSA as a protein standard. The limit of detection calculated from analysis of a range of blanks was 0.22 mg/l although recoveries close to this limit were more variable than at 0.5 mg/l where the method was validated and the average recovery was approximately 75%. Incorporating a conversion factor of 2.2 to account for an estimated difference in response between BSA and wheat proteins a limit of detection of 1 mg/l or less was achieved.
Those samples for which a positive response was detected using the Bradford Microassay were submitted for analysis using the AAA-DirectTM Analysis to determine whether this was due to the presence of protein or an interference. The full report for the analysis of sample using the AAA-DirectTM method can be found in Appendix 11.
6.5 Results of Total Protein Analysis
6.5.1 Results for Total Protein Analysis Using the Bradford Microassay Method
The results for the Bradford analysis of the 86 cereal derived distillate samples to which it could be applied (and 2 laboratory distillations from Section 5.4) can be seen in Tables 3 and 4. Where an absorbance outside the range covering the variation of the blank has been detected, the appropriate equation (see Appendix 10) has been employed to calculate perceived protein content (in terms of BSA). 100% recoveries have been assumed.
Page 33 of 173
Table 3: Results of Application of Bradford Microassay Method to 10 Bottled Products of Cereal Derived Origin
BRADFORD Sample Type and Number Sample Description ASSAY total protein conc. mg/l Vodka (cereal derived) Bottled Products ND S05-1978 Wyborowa Wodka, 40% vol. S05-2457 Grey Goose vodka, 40% vol. ND S05-3065 Stolichnaya Russian Vodka, 37.5% vol. ND S05-3125 Absolut Vodka, 40% vol. ND S05-3287 Smirnoff, 37.5% vol. ND
Gin (cereal derived) Bottled Products S05-2015 Plymouth Gin, 41.2% vol. ND S05-2461 Bombay Sapphire gin, 47% vol. ND S05-3285 Gordons Gin, 37.5% vol. ND S05-3063 (being analysed for ND almond as well) Beefeater London Dry Gin, 40% vol.
Bourbon/ Canadian Whisky/Other Bottled Products Whisk(e)y Products S06-0061 Genever Distillate, 56.9% vol ND ND = Not detected. Protein values of 0.22 mg/l can be detected, assuming 100% recoveries. Validation has shown that protein values of 0.5 mg/l will be consistently detected.
In summary, Tables 3 and 4 demonstrate that:
No total protein is detected in 66 of 86 analysed samples. Out of the 20 samples which give small positive results, 15 are Scotch new make malt samples. It was predicted that such samples would probably give positive results, given the presence of haze forming interferences. Two Scotch new make malt samples did not provide a positive response. It is assumed that the levels of interference are reduced in these samples to a level which does not interfere with the analysis. Five additional samples demonstrate a positive result for the Bradford analysis method. Four of these, given the distillation systems employed, might be expected to contain residual fats which may partially interfere with the method and hence give positive results. The exception to this is Sample S05-2194 which given the distillation system would not be expected to contain any residual fats, although the presence of a small white precipitate was noted upon drying. All absorbances for this sample were just within the upper limit of the blanks. However, four repeat analyses were all between the upper 2nd and 3rd standard deviation limits. Consequently, this sample merited further analysis.
Of 86 samples, no protein was detected in 66 samples using the Bradford Microassay screening method. A further twenty demonstrated small responses, but the nature of the samples and the observation of cloudy precipitates following ethanol evaporation indicate that the positive responses are probably not protein related. These twenty samples were submitted for analysis using the AAA-DirectTM Method.
Page 34 of 173
Table 4: Results of Application of Bradford Microassay Method to 78 Production Distillates of Cereal Derived Origin (including 2 laboratory distillations from Section 5.4)
BRADFORD Sample Type and Number Sample Description ASSAY total protein conc. mg/l Neutral Spirit Production Distillate Samples S05-2208 Grain Neutral Spirit, 96.6% vol. ND S05-2243 Grain Neutral Spirit, 96.3% vol. ND S05-2280 Ethyl Alcohol of an Agricultural Origin (Neutral Spirit), 96.3% vol. ND S05-2455 Ethyl Alcohol used for production of vodka, 96.2% vol. ND Ethyl Alcohol of French Origin used for production of Grey Goose Vodka, ND S05-2456 96.2 % vol. S05-2458/S05-2459 Grain Neutral Spirit, 96% vol. ND S05-3122 Finsprit - Cistern, 96.0% vol. ND S05-3123/3124 Grain alcohol, 96.2% vol. ND S05-3234 Grain Neutral Spirit, 96.5% vol. ND S05-3310 Grain Neutral Spirit, 96.5% vol. ND S05-3312 Grain Neutral Spirit, 96.5% vol. ND S05-3279/S05-3281 Grain Neutral Spirit, 96.3% vol. ND S05-3323 Grain Neutral Spirit, 96.4% vol. ND S05-3248 Ethyl Alcohol of Agricultural Origin, 96.3% vol. ND S05-4337 Ethyl Alcohol of Agricultural Origin, 96.2% vol. ND S06-0070/S06-0071/S06- Grain Neutral Spirit, 96.1% vol. ND 0072/S06-0073 S05-4395/S05-4396/S05- Ethyl Alcohol of Agricultural Origin - 96 % vol. minimum ND 4397 S06-0057 Grain Neutral Spirit, Sas van Gent, 96.3 % vol ND S06-0058 Grain Neutral Spirit, Delfzyl, ca 96% vol. ND S06-0059 Grain Neutral Spirit, Heibronn, ca 96% vol. ND
Grain New Make Spirit Production Distillate Samples (Scotch/Irish) S05-2203 Middleton New Make Grain Distillate, 94.47% vol. ND S05-2204 Middleton New Make Grain Distillate, 94.45% vol. ND S05-2206 NB New Make Wheat Whisky Spirit, 94.5% vol. ND S05-2207 NB New Make Maize Whisky Spirit, 94.6% vol. ND S05-2212 Loch Lomond Grain, 94.2% vol. ND S05-2244 Invergordon Scotch Grain Spirit, 94.4% vol. ND S05-2245 Invergordon Scotch Grain Spirit, 94.3% vol. ND S05-3233 Girvan New Make Grain Spirit, 94.5% vol. ND S05-3235 Girvan New Make Grain Spirit, 94.5% vol. ND S05-3313 Strathclyde New Make Grain Spirit, 94.4% vol. ND S05-3311 Strathclyde New Make Grain Spirit, 94.4% vol. ND S05-3282/S05-3284 Cameron Bridge New Make Grain Spirit, 94.4% vol. ND S05-3994/S05-3995 Port Dundas New Make Grain Spirit, 94.3% vol. ND S05-3996/S05-3997 Port Dundas New Make Grain Spirit, 94.3% vol. ND S05-4352 Cooleys Grain Spirit, 94.6% vol. ND S06-0150 Cameron Bridge New Make Grain Spirit, 68.5% vol. ND
Malt New Make Spirit Production Distillate Samples (Scotch/Irish) S05-0166 Macallan New Make Spirit, 71% vol. ND S05-2061 Glenmorangie New Make Spirit, 63.5% vol. 0.3 mg/l S05-2107 Aberlour New Make Distillate, 68.8% vol. 1.0 mg/l S05-2108 Glen Grant New Make Distillate, 68.7% vol. 3.0 mg/l S05-2109 The Glenlivet New Make Distillate, 68.8% vol. 2.5 mg/l S05-2110 Longmorn New Make Distillate, 68.3% vol. ND S05-2197 Middleton New make pot still distillate, 85.13% vol. ND S05-2198 Middleton New make pot still distillate, 86.12% vol. ND S05-2199 Bushmills New make single malt distillate, 84.0% vol. ND
Page 35 of 173
BRADFORD Sample Type and Number Sample Description ASSAY total protein conc. mg/l S05-2213 Loch Lomond Malt, 71.3% vol. 1.2 mg/l S05-2214 Loch Lomond Malt, 86.0% vol. 0.6 mg/l S05-2276 Laphroaig New Make Spirit, 68% vol. 0.9 mg/l S05-2442 Miltonduff New Make Spirit, 65.8% vol. 1.7 mg/l S05-2897/S05-2898/S05- 1.6 mg/l Craigellachie new-make spirit, 69.9% vol. 2899 S05-3073/3074 Talisker New Make Spirit, 70.2% vol. 1.0 mg/l S05-3075/3076 Dalwhinnie New Make Spirit, 67.6 % vol. 2.3 mg/l S05-3077/3078 Mortlach New Make Spirit, 71.9% vol. 1.6 mg/l S05-3321 Glenfiddich Distillery New Make Spirit, 74.50% vol. 2.6 mg/l S05-3486 Auchentoshan New Make Spirit, 79.9% vol. 0.4 mg/l S05-4353 Cooleys New Make Malt Spirit, 64.8% vol. 0.4 mg/l S06-0149 Cardhu New Make Spirit, 63.4% vol. ND
Bourbon/ Canadian/ Production Distillate Samples Other Whisk(e)y Distillates S05-2013 Maker’s Mark New Make Distillate, 65.0% vol. 0.3 mg/l S05-2062 Canadian Club Distillate 1, 67.4% vol. ND S05-2063 Canadian Club Distillate 2, 80.9% vol. ND S05-2076 Brown Forman Bourbon Distillate, 62.5% vol. ND S05-2077 Jack Daniels Distillate, 70% vol. ND S05-2150 Berentzen Traditionskorn (Korn Distillate), 32% vol. ND S05-2194 Oldesloer Doppelkorn (Korn Distillate), 38% vol. 0.2 mg/l S05-2279 Roggen-Kornfeindestillate (Korn Distillate), 96.1% vol. ND S05-2993/S05-2994 Jim Beam Distillate, 64.02% vol. ND S05-3032 Spanish Grain Whisky Distillate, 94.7% vol, ND S05-3034 Spanish Malt Whisky Distillate DYC, New make spirit, 68.5% vol. ND Crown Royal - Four Roses Bourbon Composite Flavouring ND S05-3269/S05-3070 Whisky(Crown Royal Distillate 1) 69.9% vol, Crown Royal - New Make Bourbon Whisky (Gimli) (Crown Royal ND S05-3254 Distillate 2), 69.0% vol. Crown Royal - New Make Rye Whisky (Gimli) (Crown Royal Distillate 3) ND S05-3276 69.9% vol. S05-4342 Wild Turkey Bourbon Whiskey Distillate, 55% vol. 0.3 mg/l S05-4344 Wild Turkey Rye Whiskey Distillate, 55% vol. 0.3 mg/l S06-0060 Genever Distillate, 51.39% vol. 0.9 mg/l S06-0151/S06-0152 Mackmyra Swedish Whisky, 68.9% vol. ND S05-2278 Weizen-Kornfeindestillate, 96.1% vol. ND SWRI Laboratory Distillates Production Distillate Samples Sample collected from distillation of a 10% ethanol solution containing ND S06-1283 1000 mg/l of Bovine Serum Albumin Sample collected from distillation of a 10% ethanol solution containing ND S06-1473 1000 mg/l of Bovine Serum Albumin (repeat) ND = Not detected. Protein values of 0.22 mg/l can be detected, assuming 100% recoveries. Validation has shown that protein values of 0.5 mg/l will be consistently detected. Estimated protein levels are based on 100% recoveries.
Page 36 of 173
6.5.2 Results for Total Protein Analysis Using the AAA-DirectTM Method
The results for the 20 samples analysed using the AAA-DirectTM method can be found in Appendix 11. The total protein content which could be assigned to a cereal based protein, based on these results, is presented in Table 5. Since amino acids were detectable down to levels of 0.02 mg/l in these analyses and the major amino acid in cereal proteins is present at levels >10%, a limit of detection for cereal proteins was calculated as 0.2 mg/l.
Table 5: Results of Application of AAA-DirectTM Method to 20 Production Distillates of Cereal Derived Origin, subsequent to Bradford Microassay
Malt New Make Total Protein Production Distillate Samples Spirit (Scotch) (mg/l) S05-2061 Glenmorangie New Make Spirit, 63.5% vol. ND S05-2107 Aberlour New Make Distillate, 68.8% vol. ND S05-2108 Glen Grant New Make Distillate, 68.7% vol. ND S05-2109 The Glenlivet New Make Distillate, 68.8% vol. ND S05-2213 Loch Lomond Malt, 71.3% vol. ND S05-2214 Loch Lomond Malt, 86.0% vol. ND S05-2276 Laphroaig New Make Spirit, 68% vol. ND S05-2442 Miltonduff New Make Spirit, 65.8% vol. ND S05-2897/S05- ND Craigellachie new-make spirit, 69.9% vol. 2898/S05-2899 S05-3073/3074 Talisker New Make Spirit, 70.2% vol. ND S05-3075/3076 Dalwhinnie New Make Spirit, 67.6 % vol. ND S05-3077/3078 Mortlach New Make Spirit, 71.9% vol. ND S05-3321 Glenfiddich Distillery New Make Spirit, 74.50% vol. ND S05-3486 Auchentoshan New Make Spirit, 79.9% vol. ND S05-4353 Cooleys New Make Malt Spirit, 64.8% vol. ND
Bourbon/ Canadian/ Production Distillate Samples Other Whisk(e)y Distillates S05-2013 Maker’s Mark New Make Distillate, 65.0% vol. ND S05-2194 Oldesloer Doppelkorn (Korn Distillate), 38% vol. ND S05-4342 Wild Turkey Bourbon Whiskey Distillate, 55% vol. ND S05-4344 Wild Turkey Rye Whiskey Distillate, 55% vol. ND S06-0060 Genever Distillate, 51.39% vol. 1.3 ND = Not detected. It has been calculated that protein values of 0.2 mg/l can be detected.
None of the 15 positive responses for the Bradford Microassay for the Scotch new make whisky can be associated with any proteinaceous material. It has thus been demonstrated that the Bradford assay can provide false positives, (possibly due to low concentrations of long chain fatty acid ethyl esters in Scotch new make whisky). This analysis applies to all the samples including S05-2194. Only one sample out of the 86 analysed contained protein, at a level of around 1 mg/l.
The specific source of the positive finding for protein in Sample S06-0060 Genever Distillate could not be identified but the analysis of this sample for gluten was negative (see Section 7 Table 8). Considering that the sample might have been contaminated with protein from a source other than the distillate raw materials, this sample was compared with another taken from the same distillery. Although not taken at the same stage in production, this comparison sample S06-0061 was actually sampled at the stage at which the product is bottled. This sample was shown to be negative for both protein (Section 6, Table 3) and gluten (Section 7, Table 8).
Page 37 of 173
6.6 Total Protein Based Analytical Evidence – Conclusions
No protein was detected in any of the initial 39 cereal derived alcoholic distillates, the limit of detection being 5 mg/l. No protein was detected in 85 out of the 86 cereal derived alcoholic distillates analysed in the extended sample survey. The Bradford Microassay method confirmed the absence of protein in 66 samples; the AAA- DirectTM analysis method confirmed absence of protein in a further 19 samples. The Bradford Microassay was validated at protein levels down to 0.5 mg/l. The AAA- DirectTM method was calculated as having a limit of detection of approximately 0.2 mg/l. One sample was shown to contain protein at a level of approximately 1 mg/l. A comparison sample, taken at a later stage in the production process, was shown to be free from protein. Both of these samples were negative for gluten (Section 7, Table 8).
The conclusion drawn from these analyses was that by using a total protein method with an increased and appropriate sensitivity, no potentially significant amounts of protein had been found in any distillates analysed in the extended sample survey. This provided further analytical evidence that cereal proteins do not transfer into the products of distillation.
Page 38 of 173
7. Arguments and Data to Support Exemption – ELISA Based Analytical Evidence
7.1 ELISA Based Analytical Evidence as Presented in the Application for Provisional Exemption
In the application for provisional exemption 46 cereal derived production distillate samples, matured spirit samples and liqueurs were analysed for gluten levels using an ELISA based analytical method. Complete details regarding this set of analyses can be found in Appendix 12. The results of recovery experiments established a limit of detection of 10 mg/l for gluten in alcoholic distillate matrices. The range of alcohol distillates analysed and the levels of measured gluten are reproduced in Table 6. The results show that no gluten was detected in any of the samples tested.
Table 6: Gluten in Cereal Distillates and Spirit Drinks produced from Cereal Distillates (Bottled Spirits and Production Samples)
Sample Type and Number. Description Protein Concentration (mg/l) Scotch Malt Whisky Bottled Products S04-0820 Glenfiddich Scotch Malt Whisky ND S04-0821 Glenfiddich Liqueur ND S04-0868 Aultmore Scotch Malt Whisky 12 year old ND Dalwhinnie Scotch Malt Whisky 15 year S04-0902 ND old S04-0905 Lagavulin Scotch Malt Whisky 16 year old ND
Scotch Blended Whisky Bottled Products S04-0707 Teachers Highland Cream Scotch Whisky ND Stewarts Cream of the Barley Blended S04-0708 ND Scotch Whisky S04-0709 Ballantine’s Finest Scotch Whisky ND Johnnie Walker Black Label Old Scotch S04-0714 ND Whisky S04-0715 Bells Scotch Whisky ND S04-0719 Cutty Sark Scotch Whisky ND S04-0720 The Famous Grouse ND S04-0721 The Famous Grouse Liqueur ND S04-0824 Grant’s Family Reserve ND
Irish Whiskey Bottled Products S04-0875 Jameson Irish Whisky ND
Vodka (cereal derived) Bottled Products S04-0710 Altai Vodka ND S04-0711 Wodka Wyborowa Pure Rye Grain ND S04-0717 Smirnoff Triple Distilled Vodka ND S04-0825 Taboo ND
Gin (cereal derived) Bottled Products S04-0716 Tanqueray Special Dry Gin ND S04-0727 Bombay Sapphire Dry Gin ND S04-0829 Grants Gin ND
Page 39 of 173
Sample Type and Number. Description Protein Concentration (mg/l) S04-0870 Greenalls Gin Finished Product ND S04-1047 Hendricks Gin ND
Bourbon/ Canadian Whisky/Other Bottled Products Whisk(e)y Products S04-0973 Echter Nordhaufer Doppelkorn ND
Neutral Spirit Production Distillate Samples S04-0827 Girvan Distillery Grain Neutral Spirit ND S04-0872 Greenalls Grain Spirit ND S04-0900 Grain Neutral Spirit ex Velva (French) ND S04-0901 Grain Neutral Spirit ex Velva (French) ND S04-1034 Sedalcol Grain Neutral Alcohol 23/03/04 ND S04-1035 Sedalcol Grain Neutral Alcohol 07/06/04 ND S04-1036 Sedalcol Grain Neutral Alcohol 11/05/04 ND S04-1038 Sedalcol Grain Neutral Alcohol 14/04/04 ND
Grain New Make Spirit Production Distillate Samples (Scotch/Irish) S04-0826 Girvan Scotch Grain New Make Spirit ND S04-0874 Middleton Irish Grain New Make Spirit ND
Malt New Make Spirit Production Distillate Samples (Scotch/Irish) S04-0869 Aultmore Scotch Malt New Make Spirit ND S04-0873 Middleton Irish Pot Still New Make Spirit ND S04-0903 Dalwhinnie Scotch Malt New Make Spirit ND S04-0904 Lagavulin Scotch Malt New Make Spirit ND S04-0906 Talisker Scotch Malt New Make Spirit ND S04-0907 Caol Ila Scotch Malt New Make Spirit ND
Gin (cereal derived) Production Distillate Samples S04-0712 Compounded Gin 193 ND S04-0713 Gin Distillate (LGS) ND S04-0871 Greenalls Gin Concentrate ND S04-0893 Distilled London Dry Gin Concentrate ND
Bourbon/ Canadian Whisky/Other Production Distillate Samples Whisk(e)y Products S04-0726 Jack Daniels Whiskey ND ND= not detected Reporting limit/Limit of Detection = 10 mg/kg (approximately equivalent to 10 mg/l)
7.2 Development and Validation Work to Improve the Sensitivity of the ELISA Based Gluten Method
To provide additional evidence that the cereal protein gluten does not transfer from the raw materials to the alcoholic distillate, it was stated in the application for provisional exemption that a wider range of analytical samples would be tested for gluten presence. Responding to the EFSA NDA panel opinion on the gluten analysis, a more sensitive method was required. The methodology as reported in Section 7.1 was sensitive to levels of gluten in spirit matrices at 10 mg/l. The contract laboratory, Reading Scientific Services Limited (RSSL), was commissioned to develop the
Page 40 of 173
sensitivity of the method. The approach taken was the use of a pre-concentration step, prior to the analysis as documented above.
The full report detailing this developmental work can be found in Appendix 13. Two methods of concentration were found to be effective, lyophilisation and the use of Vivaspin protein concentrators. Using the Vivaspin filters, a reporting limit of 0.4 mg/l was achieved (assuming 100% recovery). The method was validated at 1 mg/l to demonstrate that gluten can be quantitatively recovered from spirit matrices at this level.
7.3 Sampling Protocol
Having developed more sensitive methodology for the detection of gluten in alcoholic distillates, this was now applied to an extended range of alcoholic distillates. These distillates are listed in Appendices 6a and 6b. A fuller description of the sampling protocol via which the 115 cereal derived alcoholic distillates and spirit products thereof were collected can be found in Section 6.3.
7.4 Summary of Protocol for ELISA Based Gluten Method and Results
The protocol employed for the analysis of all 115 alcoholic distillates and spirit samples can be found in Appendix 14. A summary table of the results can be found in Tables 7 (bottled products) and 8 (production distillate samples including laboratory distillations from Section 5.4).
Table 7: Results of Application of ELISA Based Gluten Analysis Method to 39 Bottled Products of Cereal Derived Origin - pre-concentration step applied Sample Type and Number Sample Description ELISA ASSAY gluten conc. mg/l Scotch Grain Whisky Bottled Products S05-3232 Black Barrel Single Grain Whisky, 40% vol ND S05-3236 Cameron Brig Single Grain Whisky, 40% vol/ ND S05-3439 The North British Cask Strength Single Grain 1980, 60.3% vol. ND
Scotch Malt Whisky Bottled Products S05-1977 Glen Grant 5 Year Old Single Malt Scotch Whisky, 40% vol. ND S05-2443 Miltonduff, 40% vol. ND S05-3069 Laphroaig Single Islay Malt Whisky, 40% vol. ND S05-3286 Dalwhinnie, 15yr old, 43% vol. ND S05-3485 Auchestoshan 10 yr old, 40% vol. ND
Scotch Blended Whisky Bottled Products S05-1976 Chivas Regal 12 Year Old Blended Scotch Whisky, 40% vol. ND S05-2426 Dewars' White Label, 40% vol. ND S05-3061 Ballantine's Finest Scotch Whisky, 40% vol. ND S05-3993 Johnnie Walker Red Label, 40% vol. ND ND Irish Whiskey Bottled Products S05-2196 Powers Gold Label, 40% vol. ND S05-2200 Jameson, 40% vol. ND
Page 41 of 173
Sample Type and Number Sample Description ELISA ASSAY gluten conc. mg/l S05-2201 Bushmills Single Malt 10 Year Old, 40% vol. ND S05-2202 Redbreast 12 Year Old, 40% vol. ND
Vodka (cereal derived) Bottled Products S05-1978 Wyborowa Wodka, 40% vol. ND S05-2457 Grey Goose vodka, 40% vol. ND S05-3065 Stolichnaya Russian Vodka, 37.5% vol. ND S05-3125 Absolut Vodka, 40% vol. ND S05-3287 Smirnoff, 37.5% vol. ND
Gin (cereal derived) Bottled Products S05-2015 Plymouth Gin, 41.2% vol. ND S05-2461 Bombay Sapphire gin, 47% vol. ND S05-3285 Gordons Gin, 37.5% vol. ND S05-3063 (being analysed ND for almond as well) Beefeeter London Dry Gin, 40% vol.
Bourbon/ Canadian Whisky/Other Bottled Products Whisk(e)y Products S05-2014 Maker's Mark, 45% vol. ND S05-2078 Jack Daniels, 40% vol ND S05-2151 Edelkorn Vom alten Fass, 38.5% vol. ND S05-2991 Jim Beam Kentucky Straight Whiskey, .40% vol ND S05-3036 Blend Whisky, 40% vol. ND S05-3066 Canadian Club Barrel Blended Whisky, 40% vol. ND S05-3121 Breton Whisky, 40% vol. ND S05-4026 Crown Royal De Luxe Canadian Whisky, 40% vol. ND S05-4343 Wild Turkey Bourbon Whiskey, 50.5% vol. ND S05-4345 Wild Turkey Rye Whiskey, 50.5% vol. ND S06-0061 Genever Distillate, 56.9% vol. ND S06-0304 Hakashu 12 year old Japanese whisky, 43.5% vol. ND S06-0305 Yamazaki 10 year old Japanese whisky, 40% vol. ND S06-0077 Korenwijn 10 jaar vatgelagerd, 38% vol. ND ND = Not detected. Reporting limit (limit of quantification) is 0.4 mg/kg or 0.4 mg/l. Validation has shown that protein values of 1.0 mg/l will be consistently detected at a quantifiable level.
Table 8: Results of Application of ELISA Based Gluten Analysis Method to 78 Production Distillates of Cereal Derived Origin (including 2 laboratory distillations from Section 5.4) - pre-concentration step applied
Sample Type and Number Sample Description ELISA ASSAY gluten conc. mg/l Neutral Spirit Production Distillate Samples S05-2208 Grain Neutral Spirit, 96.6% vol. ND S05-2243 Grain Neutral Spirit, 96.3% vol. ND S05-2280 Ethyl Alcohol of an Agricultural Origin (Neutral Spirit), 96.3% vol. ND S05-2455 Ethyl Alcohol used for production of vodka, 96.2% vol. ND Ethyl Alcohol of French Origin used for production of Grey Goose Vodka, ND S05-2456 96.2 % vol. S05-2458/S05-2459 Grain Neutral Spirit, 96% vol. ND S05-3122 Finsprit - Cistern, 96.0% vol. ND S05-3123/S05-3124 Grain alcohol, 96.2% vol. ND S05-3234 Grain Neutral Spirit, 96.5% vol. ND S05-3310 Grain Neutral Spirit, 96.5% vol. ND S05-3312 Grain Neutral Spirit, 96.5% vol. ND
Page 42 of 173
Sample Type and Number Sample Description ELISA ASSAY gluten conc. mg/l S05-3279/S05-3281 Grain Neutral Spirit, 96.3% vol. ND S05-3323 Grain Neutral Spirit, 96.4% vol. ND S05-3248 Ethyl Alcohol of Agricultural Origin, 96.3% vol. ND S05-4337 Ethyl Alcohol of Agricultural Origin, 96.2% vol. ND S06-0070 to S06-0073 Grain Neutral Spirit, 96.1% vol. ND S05-4395 to S05-4397 Ethyl Alcohol of Agricultural Origin - 96 % vol. minimum ND S06-0057 Grain Neutral Spirit, Sas van Gent, 96.3 % vol ND S06-0058 Grain Neutral Spirit, Delfzyl, ca 96% vol. ND S06-0059 Grain Neutral Spirit, Heibronn, ca 96% vol. ND
Grain New Make Spirit Production Distillate Samples (Scotch/Irish) S05-2203 Middleton New Make Grain Distillate, 94.47% vol. ND S05-2204 Middleton New Make Grain Distillate, 94.45% vol. ND S05-2206 NB New Make Wheat Whisky Spirit, 94.5% vol. ND S05-2207 NB New Make Maize Whisky Spirit, 94.6% vol. ND S05-2212 Loch Lomond Grain, 94.2% vol. ND S05-2244 Invergordon Scotch Grain Spirit, 94.4% vol. ND S05-2245 Invergordon Scotch Grain Spirit, 94.3% vol. ND S05-3233 Girvan New Make Grain Spirit, 94.5% vol. ND S05-3235 Girvan New Make Grain Spirit, 94.5% vol. ND S05-3313 Strathclyde New Make Grain Spirit, 94.4% vol. ND S05-3311 Strathclyde New Make Grain Spirit, 94.4% vol. ND S05-3282/S05-3284 Cameron Bridge New Make Grain Spirit, 94.4% vol. ND S05-3994/S05-3995 Port Dundas New Make Grain Spirit, 94.3% vol. ND S05-3996/S05-3997 Port Dundas New Make Grain Spirit, 94.3% vol. ND S05-4352 Cooleys Grain Spirit, 94.6% vol. ND S06-0150 Cameron Bridge New Make Grain Spirit, 68.5% vol. ND
Malt New Make Spirit Production Distillate Samples (Scotch/Irish) S05-0166 Macallan New Make Spirit, 71% vol. ND S05-2061 Glenmorangie New Make Spirit, 63.5% vol. ND S05-2107 Aberlour New Make Distillate, 68.8% vol. ND S05-2108 Glen Grant New Make Distillate, 68.7% vol. ND S05-2109 The Glenlivet New Make Distillate, 68.8% vol. ND S05-2110 Longmorn New Make Distillate, 68.3% vol. ND S05-2197 Middleton New make pot still distillate, 85.13% vol. ND S05-2198 Middleton New make pot still distillate, 86.12% vol. ND S05-2199 Bushmills New make single malt distillate, 84.0% vol. ND S05-2213 Loch Lomond Malt, 71.3% vol. ND S05-2214 Loch Lomond Malt, 86.0% vol. ND S05-2276 Laphroaig New Make Spirit, 68% vol. ND S05-2442 Miltonduff New Make Spirit, 65.8% vol. ND S05-2897 to-S05-2899 Craigellachie new-make spirit, 69.9% vol. ND S05-3073/S05-3074 Talisker New Make Spirit, 70.2% vol. ND S05-3075/S05-3076 Dalwhinnie New Make Spirit, 67.6 % vol. ND S05-3077/S05-3078 Mortlach New Make Spirit, 71.9% vol. ND S05-3321 Glenfiddich Distillery New Make Spirit, 74.50% vol. ND S05-3486 Auchentoshan New Make Spirit, 79.9% vol. ND S05-4353 Cooleys New Make Malt Spirit, 64.8% vol. ND S06-0149 Cardhu New Make Spirit, 63.4% vol. ND
Bourbon/ Canadian/ Production Distillate Samples Other Whisk(e)y Distillates S05-2013 Maker’s Mark New Make Distillate, 65.0% vol. ND S05-2062 Canadian Club Distillate 1, 67.4% vol. ND
Page 43 of 173
Sample Type and Number Sample Description ELISA ASSAY gluten conc. mg/l S05-2063 Canadian Club Distillate 2, 80.9% vol. ND S05-2076 Brown Forman Bourbon Distillate, 62.5% vol. ND S05-2077 Jack Daniels Distillate, 70% vol. ND S05-2150 Berentzen Traditionskorn (Korn Distillate), 32% vol. ND S05-2194 Oldesloer Doppelkorn (Korn Distillate), 38% vol. ND S05-2279 Roggen-Kornfeindestillate (Korn Distillate), 96.1% vol. ND S05-2993/S05-2994 Jim Beam Distillate, 64.02% vol. ND S05-3032 Spanish Grain Whisky Distillate, 94.7% vol, ND S05-3034 Spanish Malt Whisky Distillate DYC, New make spirit, 68.5% vol. ND Crown Royal - Four Roses Bourbon Composite Flavouring Whisky(Crown ND S05-3269/S05-3070 Royal Distillate 1) 69.9% vol, Crown Royal - New Make Bourbon Whisky (Gimli) (Crown Royal Distillate ND S05-3254 2), 69.0% vol. Crown Royal - New Make Rye Whisky (Gimli) (Crown Royal Distillate 3) ND S05-3276 69.9% vol. S05-4342 Wild Turkey Bourbon Whiskey Distillate, 55% vol. ND S05-4344 Wild Turkey Rye Whiskey Distillate, 55% vol. ND S06-0060 Genever Distillate, 51.39% vol. ND S06-0151/S06-0152 Mackmyra Swedish Whisky, 68.9% vol. ND S05-2278 Weizen-Kornfeindestillate, 96.1% vol. ND SWRI Laboratory Production Distillate Samples Distillates Sample collected from distillation of a 10% ethanol solution containing 18 ND S06-1283 g/l of gluten Sample collected from distillation of a 10% ethanol solution containing 18 ND S06-1473 g/l of gluten ND = Not detected. Reporting limit is 0.4 mg/kg or 0.4 mg/l, assuming 100% recoveries. Validation has shown that protein values of 1.0 mg/l will be consistently detected.
7.5 ELISA Based Analytical Evidence - Conclusions
No gluten was detected in any of the initial 46 cereal derived alcoholic distillates, the limit of detection being 10 mg/l. No gluten protein was detected in any of the 115 cereal derived alcoholic distillates analysed in the extended survey. The method has been validated at 1 mg/l and can quantify gluten levels at 0.4 mg/l. The sensitivity of the method has increased by over a factor of 10 and confirms the absence of potentially allergenic material present in distillates derived from cereals.
Page 44 of 173
8. Arguments and Data to Support Exemption – History of Safe Use
8.1 *History of Safe Use of Ingredients Requested for Exemption as Presented in the Provisional Application*
Dietary guidelines for patients with Coeliac Sprue state that “wine, liqueurs, most ciders and other spirits including whiskey and brandy are allowed” (Farrell and Kelly, 2002). This advice has been disseminated through a number of websites by various charities and information providers, for example the Coeliac Society, www.coeliac.co.uk and the British Broadcasting Corporation website www.bbc.co.uk. This advice is supported by the general paucity of references to either coeliac symptoms or wheat allergy caused by the consumption of distilled beverages. As members of the Coeliac society can be expected to be particularly concerned to report any possible adverse reaction from the consumption of a specific food or drink the absence of any adverse reports is very significant.
A literature review using the PubMed database of the US National Library of Medicine failed to identify any reports linking coeliac symptoms or wheat allergy with the consumption of distilled beverages. The PubMed database includes over 14 million citations for biomedical articles from MEDLINE and additional life science journals dating back to the 1950s. Titles and abstracts were searched for COELIAC or CELIAC and WHEAT ALLERGY, which gave large numbers of literature reports (11,301 and 236 respectively). Linking these search terms with specific alcoholic beverages (WHISKY or WHISKEY, GIN, VODKA) and with DISTILLED SPIRITS gave no reported literature. Use of the more general term ALCOHOL with COELIAC or CELIAC returned a number of literature reports, but of those referring to coeliac disease, none reported distilled beverage alcohol causing coeliac symptoms but referred to the alcohol solubility of gliadins. Combination of ALCOHOL with WHEAT ALLERGY gave no reported literature.
Combination of ALLERG* and DISTILLED SPIRITS gave one reference to a review of allergic and asthmatic reactions to alcoholic drinks (Vally and Thompson 2003). From this review most sensitivities to alcoholic drinks do not appear to be immune mediated, but are more frequently pharmacological intolerances to specific chemicals in these drinks. Where allergic and asthmatic reactions to specific non-alcohol components have been reported, these are wholly concerned with non-distilled drinks. Only one reported investigation of spirit consumption triggering asthmatic attacks could be found, however the cause of the reaction was not identified and skin prick tests for routine common allergens gave no reaction to wheat (Breslin et al, 1973). Patients reacted to specific drinks but not the equivalent amount of ethyl alcohol. Figueredo et al (1999) report a case of a Type 1 allergy to barley malt and corn triggered by beer consumption, but the patient showed no reaction in oral provocation tests with a distilled spirit (10 ml – 50 ml).
8.2 Update on History of Safe Use of Ingredients Requested for Exemption The literature review conducted in Section 8.1 was extended to April 2006 to determine whether any newly published data could be found linking allergic reactions from consumption of alcoholic distillates to cereal raw materials used prior to
Page 45 of 173
distillation. Whilst the search for COELIAC or CELIAC and WHEAT ALLERGY gave an increased number of literature reports (12,725 and 270 respectively), linking these terms with specific alcoholic beverages, DISTILLED SPIRITS or ALCOHOL still gave no relevant reported literature. Similarly, no additional allergic and asthmatic reactions to specific non-alcohol components have been reported in distillates, according to the data returned by a combination of ALLERG* and DISTILLED SPIRITS.
The above findings have been confirmed by an independent literature review which is reproduced in Appendix 15.
The scale of spirit drinks production in the EU adds further weight to the history of safe use argument. The spirit drinks industry is a major outlet for EU agriculture. Each year the industry's raw materials include: 16 million hectolitres of wine, 2 million tonnes of cereals, 2.5 million tonnes of sugar beet and 0.3 million tonnes of fruit. Sales of spirit drinks in the EU in 2004 totalled over 260 million cases (of 8.4 litres, i.e. 12 bottles of 70 cl per case). In other words, over 8.5 million bottles of spirit drink were sold every day. Tens of millions of EU citizens thus regularly consume spirit drinks, safely and responsibly, without any evidence that these trigger allergenic reactions. This argument equally applies to the consumption of spirit drinks outside the European Union. The EU is the leading exporter worldwide of spirit drinks; in 2004 these were worth €5.2 billion.
Page 46 of 173
9. Summary of Arguments and Data to Support Exemption for Labelling Exemption
This submission requests that the following ingredients to be permanently exempted from Annex IIIa of European Directive 2003/89/EC: distillates made from alcohol derived from the following raw materials employed as sugar sources prior to the distillation process: cereals containing gluten (i.e. wheat, rye, barley, oats, spelt, kamut or their hybridised strains) and products thereof.
The physico-chemical evidence demonstrates that cereal proteins and peptides do not distil. This is supported by analysis of distillates made from cereals as reported in the literature and in the range of samples analysed by RSSL, SWRI and Intertek ASG (for information on these laboratories refer to Appendix 7). All distillates have tested negative for specific cereal proteins and for all other proteinaceous material (where this was carried out). Medical information and advice also supports the argument that such distillates are safe for consumption by individuals sensitive to cereal allergens. The named Annex IIIa cereals used as raw materials in the production of commercially available distillates are wheat, barley and rye. However, the arguments supporting the non-volatility of cereal proteins and peptides apply to all the varieties listed.
The distillation step which all alcoholic distillates and distilled spirit drinks must undergo, both by law and definition, is fundamentally a purification and concentration technique for alcohol and volatile compounds. No evidence has been found that the allergens contained in cereals can distil into the resultant distillates and be present in the alcoholic beverages available to the consumer. This is based on three areas of evidence:
1. Potentially allergenic proteins and polypeptides are non-volatile and hence do not distil.
Information from the scientific literature confirms that proteins and peptides do not distil.
An independent expert has provided an opinion on the process of distillation and its application to proteins and peptides. This opinion reiterates the contention, based on the fundamental concepts of distillation theory, that allergenic proteins would not be expected to transfer from the starting material to the final distillate.
Distillation experiments have been performed in the laboratory using model alcoholic matrices, containing both the potentially allergenic protein gluten and an alternative protein, bovine serum albumin. In each case, no transfer of proteinaceous material was detected in the final distillate.
Page 47 of 173
2. Results of surveys of bottled spirit drinks and production samples, made from cereals used prior to distillation, demonstrate the absence of gluten and protein.
An initial survey looking at bottled products and production samples failed to detect gluten in 46 samples with a detection limit of 10 mg/l. A subsequent survey covering worldwide production of cereal based distillates failed to detect gluten in 115 samples with a detection limit of 0.4 mg/l.
An initial survey looking at bottled products and production samples failed to detect proteinaceous material in 39 samples with a detection limit of 5 mg/l. A subsequent survey covering worldwide production of 86 cereal based distillates failed to detect proteinaceous material in 85 samples with detection limits of either 0.2 or 0.5 mg/l. One sample of genever was shown to have approximately 1 mg/l of protein present. The origin of this protein was unknown, but the sample proved negative for gluten and an alternative sample taken for the same product was negative for protein content.
3. There is no clinical evidence or advice that allergens which originate from cereals containing gluten are a threat to consumers of spirit drinks, where these cereals are used before distillation.
Alcoholic distillates are not associated with cereal allergies and by labelling them under 2003/89/EC, in spirits drinks or other food products, consumers would perceive a threat which is not there, leading to confusion and complacency with similar labels or warnings. Obviously, where spirit drinks are manufactured with any 2003/89/EC (Annex IIIa) substance after distillation, then labelling will have to take place, unless exemption has been granted for these ingredients.
In conclusion, this dossier demonstrates that cereal allergy sufferers are not at risk from cereal allergens in the raw materials used to make distillates. On this basis, CEPS requests a permanent labelling exemption.
Page 48 of 173
10. References
Asano, Y., Iiyama, Y., Kubota, T., Ochi, H., Mizota, T., and Kokubo, S. (2004) ABSTRACT: Removal of odors from milk protein hydrolysates using continuous distillation column system. Journal of the Japanese Society for Food Science and Technology-Nippon Shokuhin Kagaku Kogaku Kaishi, 51(1): 7-12
Breslin, A.B.X, Hendrick, D.J. and Pepys, J. (1973) Effect of disodium cromoglycate on asthmatic reactions to alcoholic beverages. Clinical Allergy 3:71-82
Bringhurst, T.A., Fotheringham, A.L. and Brosnan, J. (2003) Grain whisky: raw materials and processing. In: Whisky: Technology, Production and Marketing. Edited by Russell, I., Academic Press, London, UK, pp. 77-115
British Broadcasting Corporation Website (2006). Making the most of gluten-free food. http://www.bbc.co.uk/food/recipes/mostof_coeliac3.shtml (Last visited July 2006)
Campbell, I. (2003) Grain whisky distillation. In Whisky: Technology, Production and Marketing. Edited by Russell, I., Academic Press, London, UK, pp. 181-208
Campbell, J.A. (1988) The Question of Gluten in Distilled Alcohol. Journal of the Canadian Dietetic Association, 49:204
Coeliac Society UK Website (2006). Coeliac UK – The charity for people with coeliac disease and dermatitis herpetiformis. http://www.coeliac.co.uk (Last visited July 2006)
Commission Directive 2005/26/EC of 21 March 2005 establishing a list of food ingredients or substances provisionally excluded from Annex IIIa of Directive 2000/13/EC of the European Parliament and of the Council. Official Journal of the European Communities. No. L75:33 -34
Considine, D.M. and Considine, G.D. (1984) Van Nostrand Reinhold Encylcopedia of Chemistry, (Fourth Edition). Van Nostrand Reinhold Company Ltd., Wokingham, UK, pp. 326-329
Council Regulation (EEC) No. 1576/89 of 29 May 1989, laying down general rules on the definition, description and presentation of spirit drinks. Official Journal of the European Communities. No. L160:1 -17 (as amended) de Laat, W.T.A.M (2004) Analysis of the presence of gluten residues in neutral spirits production from cereals at Royal Nedalco B.V. Personal Communication
Dickey, L.C., Parris, N., Craig, J.C. and Kurantz, M.J. (2002) Serial batch extraction of zein from milled maize. Industrial Crops and Products, 15:33-42
Page 49 of 173
Directive 2000/13/EC of the European Parliament and of the Council of 20 March 2000 on the approximation of the laws of the Member States relating to the labelling, presentation and advertising of foodstuffs. Official Journal of the European Union. No L109:29-42.
Directive 2003/89/EC of the European Parliament and of the Council of 10 November 2003 amending Directive 2000/13/EC as regards indication of the ingredients present in foodstuffs. Official Journal of the European Union. No L308:15-18
Dolan, T.C.S. (1976) Some aspects of the impact of brewing science on Scotch malt whisky production. Journal of the Institute of Brewing, 82:177-181
EFSA (2004a) Opinion of the Scientific Panel on Dietetic Products, Nutrition and Allergies on a request from the Commission relating to the evaluation of allergenic foods for labelling purposes. The EFSA Journal, 32:1-197
EFSA (2004b) Opinion of the Scientific Panel on Dietetic Products, Nutrition and Allergies on a request from the Commission related to a notification from CEPS on distillates made from cereals pursuant to Article 6 paragraph 11 of Directive 2000/13/EC. The EFSA Journal, 130:1-6
Farrell, R.J. and Kelly, C.P. (2002) Celiac sprue. The New England Journal of Medicine, 346:180-188
Figueredo, E., Quirce, S. del Amo, A., Cuesta, J., Arrieta, I., Lahoz, C. and Sastre, J. (1999) Beer-induced anaphylaxis: identification of allergens. Allergy 54:630-634
Food Allergen Labeling and Consumer Protection Act of 2004 (Title II of Public Law 108-282), Aug. 2, 2004, 118 Stat.: 905-911
Gowland, H. (2004) Spirit Beverages and Anaphylaxis, Email Communication
Health Canada Website (2006). Proposed regulatory amendments to enhance the labelling of priority allergens in foods. http://www.hc-sc.gc.ca/fn-an/label-etiquet/allergen/allergy_label_letter- lettre_etiquetage_alergene_e.html (Last visited June 2006)
Jacques, K.A., Lyons, T.P and Kelsall, D.R. (Editors) (2003) The Alcohol Textbook: A reference for the beverage, fuel, and industrial alcohol industries (Fourth Edition). The Nottingham University Press, Nottingham, UK
Kasarda, D.D. (2003) Celiac Disease and Safe Grains http://wheat.pw.usda.gov/ggpages/topics/Celiac.vs.grains.html (last checked June 2006)
Lehotay, S.J. and Hajšlová, J. (2002) Application of gas chromatography in food analysis. Trends in Analytical Chemistry, 21(9+10):686-697
Page 50 of 173
Lioutas, T (2004), The renaissance of American Bourbons: developments and technical challenges of the production of premium Bourbon distillates by batch distillation. In Distilled spirits: Tradition and innovation. Edited by Bryce, J.H and Stewart, G.G., Nottingham University Press, Nottingham, UK, pp. 233-243
Lyons, T. P. (2003) Production of Scotch and Irish whiskies: their production and evolution. In: The Alcohol Textbook: A reference for the beverage, fuel, and industrial alcohol industries (Fourth Edition). Edited by Jacques, K.A., Lyons, T.P and Kelsall, D.R., The Nottingham University Press, Nottingham, UK: pp. 193-222
Lyons, T.P. and Rose, A.H. (1977) Whisky. In: Economic Microbiology, Volume 1. Edited by Rose, A.H., Academic Press Inc. (London) Ltd., London, pp. 635-692
Madson, P.W (2003) Ethanol distillation: the fundamentals. In: The Alcohol Textbook: A reference for the beverage, fuel, and industrial alcohol industries (Fourth Edition). Edited by Jacques, K.A., Lyons, T.P and Kelsall, D.R., The Nottingham University Press, Nottingham, UK, pp. 319-336
McGough, N. (2005) Diet Round-up. In: Coeliac UK Crossed Grain Magazine, Summer: 21
Ministry of Health, Labour and Welfare Ordinance No. 23 of 2001 amending (in part) the Food Sanitation Law Enforcement Regulations and the Ministerial Ordinance concerning Compositional Standards, etc. for Milk and Milk Products, and the Establishment of labelling standards for processed foods in which milk is used as raw material (Ministry of Health, Labour and Welfare Notification No. 71 of 2001). http://www.mhlw.go.jp/english/topics/qa/allergies/al1.html (Last visited June 2006)
Murtagh, J.E. (1995) Neutral Spirit Production (with notes on gin and vodka preparation). In: The Alcohol Textbook: Ethanol production by fermentation and distillation. Edited by Lyons, T.P., Kelsall, D.R. and Murtagh, J.E, The Nottingham University Press, Nottingham, UK, pp. 193-212
Mustafa, A.F., McKinnon J.J. and Christensen D.A. (2000) ABSTRACT: Chemical characterization and in situ nutrient degradability of wet distillers’ grains derived from barley-based ethanol production. Animal Feed Science and Technology, 83(3-4): 301-311
Mustafa, A.F., McKinnon J.J., Ingledew, M.W. and Christensen, D.A. (2000) ABSTRACT: The nutritive value for ruminants of thin stillage and distillers’ grains derived from wheat, rye, triticale and barley. Journal of the Science of Food and Agriculture 80(5):607-613
Nicol, D (1989) Batch Distillation. In: The Science and Technology of Whiskies. Edited by Piggott, J.R, Sharp, R. and Duncan, R.E.B, Longman Scientific and Technical, Essex, UK, pp. 118-147
Page 51 of 173
Ohtake, K., Yamasaki, H. and Kojima, K (1995) Evaluation of mass transfer by mist as a new parameter in the control of wash distillation. In: Proceedings of the Fourth Aviemore Conference on Malting, Brewing and Distilling. Edited by Campbell, I and Priest, F.G, Institute of Brewing, London, UK, pp. 202-208
Oldani, A., Iametti, S., Bonomi, F. and Scafuri, L. (2001) In alcoholic cereal derived beverages are there still gluten traces? Tecnica Molitoria 7:641-646
Palosuo, K., Alenius H., Varjonen, E., Kalkkinen N. and Reunala T. (2001) Rye gamma-70 and gamma-35 secalins and barley gamma-3 hordein cross-react with omega-5 gliadin, a major allergen in wheat-dependent, exercise induced anaphylaxis. Clinical & Experimental Allergy 31:466-473
Panek, R.J. and Boucher, A.R (1989) Continuous distillation. In: Science and Technology of Whiskies. Edited by Piggott, J.R, Sharp, R. and Duncan, R.E.B., Longman Scientific and Technical, Essex, UK, pp 150-181
Penguin Dictionary of Science (1986) 6th Edition edited by Uvarov, E.B. and Isaacs, A., Penguin, UK, p. 115
Piggott, J.R, Sharp, R. and Duncan, R.E.B. (Editors) (1989) The Science and Technology of Whiskies, Longman Scientific and Technical, Essex, UK
Piggott, J.R. and Conner, J.M. (2003). In: Fermented Beverage Production (2nd Edition) Edited by A Lea and JR Piggott, Kluwer Academic/Plenum Publishers, New York, USA, pp 239-262
Piggott, R. (2003) From pot stills to continuous stills: Flavour modification by distillation. In: The Alcohol Textbook: Ethanol production by fermentation and distillation. Edited by Lyons, T.P., Kelsall, D.R. and Murtagh, J.E, The Nottingham University Press, Nottingham, UK, pp. 255-266
Pyke, M. (1965) The Manufacture of Scotch Grain Whisky. Journal of the Institute of Brewing, 71:209-218
Ralph, R. (2003) Production of American whiskies: bourbon, corn, rye and Tennessee. In: The Alcohol Textbook: Ethanol production by fermentation and distillation. Edited by Lyons, T.P., Kelsall, D.R. and Murtagh, J.E, The Nottingham University Press, Nottingham, UK, pp. 275-285
Russell, I. (Editor) (2003) Whisky: Technology, Production and Marketing, Academic Press, London, UK
Rutherford, A.G. (1981) Mashing for malt and whisky production. The Brewer, March:82-86
Shan, L., Molberg, O., Parrot, I., Hausch, F., Filliz, F., Gray, G.M., Sollid, L.M. and Khosla, C. (2002) Structural basis for gluten intolerance in celiac sprue. Science 297:2275-2279
Page 52 of 173
Simonato, B., Pasini, G., Giannattasio, M., Peruffo, A.D., De Lazzari, F. and Curioni, A. (2001) Food allergy to wheat products: the effect of bread baking and in vitro digestion on wheat allergenic proteins. A study with bread dough, crumb and crust. Journal of Agriculture and Food Chemistry 49:5668-5673
Simpson, A. C. (1968) Manufacture of Scotch Whisky, Process Biochemistry, January:9-12
Standard 1.2.3 Mandatory warning and advisory statements (2006). Australian Food Standards Code. Anstat Pty Ltd, Australia
Stern M., Ciclitira P.J., Van Eckert R., Feighery C., Janssen, F.W., Mendez, E., Mothes, T., Troncone, R., Wieser, H. (2001) Analysis and clinical effects of gluten in coeliac disease. European Journal of Gastroenterology & Hepatology 13:741-747
Vally, H. and Thompson, P.J. (2003) Invited Review: Allergic and asthmatic reactions to alcoholic drinks. Addiction Biology 8:3-11
Vandenheede J.R. (1993) Immunoassay for the determination of “active gluten” in beers and other alcoholic drinks. Alimentaria 239:41-42
Watson, D.C. (1983) Factors influencing the congener composition of malt whisky new spirit. In Flavour of Distilled Beverages: Origin and development. Edited by Piggott J.R., Ellis Horwood Ltd, Chichester, UK, pp. 79-92
Wieser, H. (1996). Relation between gliadin structure and celiac disease. Acta Paediatr 85(Suppl 412):3-9
Wilkin, G.D., Webber, M.A and Lafferty E.A. (1983) Appraisal of Industrial Continuous still products. In Flavour of Distilled Beverages: Origin and development. Edited by Piggott J.R., Ellis Horwood Ltd, Chichester, UK, pp. 154- 164
Reference Reports Reproduced in Appendices
Brotherhood, A. (2006) Total Protein and Free Amino Acid Content, ANB 0325/12, Intertek TSG.
Flanagan, S. (2006) Certificate of Analysis – Gluten Results P6-01108, Reading Scientific Services Limited.
Flanagan, S. (2005) Development and validation of methods for the analyses allergens in spirits SWRI spike recovery project, Reading Scientific Services Limited.
Libmann, F. (2006) Report on a literature search for cases of allergy due to cereals, milk by-products, or almonds, in alcoholic beverages other than wine or beer. F. L. A. Consultants.
Sørensen, E. (2006) Fundamentals of Distillation, University College London.
Page 53 of 173
Wass, F. (2004) The determination of the presence of gluten and almonds using ELISA techniques in samples of various types of spirits, P4-06094, Reading Scientific Services Limited
Williams, D.S. (2006) Total protein content of spirits: feasibility study, ANB 0259/70, Intertek TSG
Page 54 of 173
Appendix 1 *Commissioned Physico-Chemical Expert – Dr. Eva Sørensen*
Dr Eva Sørensen Department of Chemical Engineering University College London Torrington Place London, WC1E 7JE UK
CV
Higher Education 1990- Ph.D. (Dr.ing.) from the Department of Chemical Engineering, 1994: University of Trondheim-NTH, Trondheim, Norway. Supervisor: Professor Sigurd Skogestad. Thesis: Studies on optimal operation and control of batch distillation columns. 1985- M.Sc. (Siv.ing.) from the Department of Chemical Engineering, 1989: University of Trondheim-NTH, Trondheim, Norway. Grade: 1.9 (Very Satisfactory).
Academic Career 2000-> Senior Lecturer at the Department of Chemical Engineering, University College London, London, UK. 1996- Lecturer at the Department of Chemical Engineering, University 2000: College London, London, UK. 1995- Postdoctoral researcher at the Centre for Process Systems Engineering, 1996: Imperial College of Science, Technology and Medicine, London, UK. 1990- Research assistant at the Department of Chemical Engineering, 1994: University of Trondheim-NTH, Trondheim, Norway. 1987- Part time teaching assistant at the Department of Inorganic Chemistry, 1989: University of Trondheim-NTH, Trondheim, Norway.
Membership of Professional Bodies Associate member of the Institution of Chemical Engineers (IchemE) Chair of the IchemE’s Fluid Separation Processes Subject Group Secretary of the European Federation of Chemical Engineers’ Working party on Distillation, Absorption and Extraction Member of European Membrane Society Member of Chromatographic Society
Page 55 of 173
Awards 2001 Faculty Teaching Award for the Faculty of Engineering for outstanding achievement in teaching, University College London (2001). Postdoctoral Research Scholarship awarded from the Norwegian Research Foundation (NFR) for postdoctoral research at Imperial College, London (1995- 1996). Research Scholarship awarded from the Norwegian Research Foundation (NFR) for graduate studies at the University of Trondheim, Norway (1993-1994). University Scholarship (Utdannelsesstipend) awarded from the University of Trondheim for graduate studies at the University of Trondheim, Norway (1990- 1993).
Page 56 of 173
Appendix 2 Dr. Eva Sørensen’s Report on the Fundamentals of Distillation Theory and its Application to the Likelihood of Allergen Transfer during Distillation
Page 57 of 173
Page 58 of 173
Page 59 of 173
Page 60 of 173
Page 61 of 173
Page 62 of 173
Page 63 of 173
Page 64 of 173
Page 65 of 173
Appendix 3 Detail of Experiments to Highlight the Non-Volatility of Proteins
Appendix 3.1 Introduction
The distillation concepts presented in Section 5.3 show how compounds such as proteins and peptides, which have no vapour pressure, will not transfer into the distillate during standard alcohol beverage distillations. A practical demonstration of this principle was performed in the laboratory by distilling two different proteins in a model ethanol/water matrix.
The first set of distillations used BSA (Bovine Serum Albumin, approximately 66 kDa, fraction V, approximately 99% pure, Sigma-Aldrich Company Limited, UK). A method for analysing the presence of BSA in the alcoholic distillate with a limit of detection of 0.5 mg/l had already been established (refer to Section 6.4). The second set of distillations used gluten (gluten from wheat, approx. 80% protein, Sigma- Aldrich Company Limited, UK). A method for analysing the presence of gluten in the alcoholic distillate with a reporting limit of 0.4 mg/l had already been established (refer to Section 7.4).
Whilst the protein concentration and the initial alcoholic strength in the gluten distillation were designed to simulate conditions in an industrial situation, the experiments were not planned to replicate the industrial distillation process on a smaller scale. Instead the distillations were designed to demonstrate the principle that proteins do not distil. The distillation feedstock was a simple model distillation matrix consisting of the protein in an alcoholic solution; the distillation was a simple one step batch distillation process. Hypothetically, if proteins exhibited volatility, the one step batch distillation was the most likely to see protein transfer into the distillate.
A very simple distillation protocol was established. The apparatus employed can be seen in Figure 1 of this Appendix.
250 ml of the model solution was placed in a 500 ml round bottom flask to which a neck was added through which distilled vapour would pass to the condenser. Heating was achieved by means of an electric isomantle. Anti-bumping granules were placed in the round bottom flask to help ensure even boiling.
Page 66 of 173
Condenser
Neck
Pot containing model solution Isomantle
Appendix 3 Figure 1 Experimental set-up for model protein distillations
Appendix 3.2 Distillation of BSA Model Solutions
The model solution was prepared by adding 250 mg of BSA (99%) to 250 ml of 10% by volume ethanol. The ethanol solution was prepared using ethanol (>99.7%, VWR International, UK) and ultra-high quality (UHQ) water. The distillation of pure protein solutions in the laboratory, without the presence of additional components typically found in industrial distillation matrices, can result in excessive foaming and make distillation hard to control. However, up to levels of 1000 mg/l of BSA in 10% by volume ethanol, this was not observed to be a problem and prevented the necessity of adding an antifoaming reagent, which could interfere with the total protein analysis of the resultant distillate. The choice of 10% alcohol was arbitrary and simply used to simulate an alcohol containing starting matrix. It also meant that over 99% of the ethanol present in the model solution would be collected in the distillate, which is typical of the industrial situation.
Duplicate distillations of the 1000 mg/l BSA solution were performed. The resulting distillates were analysed according to the protocol presented in Section 6.4. Results are recorded in Section 6.5.
No BSA protein was detected in the distillates collected. The limit of detection for this method was 0.5 mg/l.
Page 67 of 173
Appendix 3.3 Distillation of Gluten Model Solutions
Gluten content in wheat varies but according to research in the scientific literature typical concentrations of protein in distilling wheat would be expected to be approximately 10%, of which about 80% is represented by the gluten protein fractions. Therefore about 8% of wheat is gluten protein (see Appendix 4). As an example of industry practice one Scotch whisky distillery operates such that 84,400 kg of cereal (wheat and malt) contributes to 375 000 litres of the fermented product (wash). This is 225 g of cereal per litre of wash. Assuming all the cereal is wheat for convenience (barley has a lower prolamin concentration anyway), this is 225 g of wheat per litre of wash or 18 g of gluten protein per litre. Thus, to mimic this concentration in the laboratory, assuming all the gluten from the wheat is present in the wash, 4.5 g of gluten is required for a 250 ml distillation. Since the sample of wheat gluten supplied by Sigma-Aldrich is approximately 80% protein, this implies that 5.63 g of gluten from wheat as supplied is required for a 250 ml distillation. This figure is an over-estimation of gluten content, since pre-treatment during malting and fermentation can be assumed to reduce gluten content via such mechanisms as the action of proteases. An analysis of a Scotch grain whisky distillery wash, sampled 59 hours into fermentation (sample number S06-0322) confirms this, having a reported gluten content of 68 mg/l (see Appendix 14).
The model solution was prepared by adding 5.63 g of gluten (80%) to 250 ml of 10% by volume ethanol. The ethanol solution was prepared using ethanol (>99.7%, VWR International, UK) and UHQ water. The distillation of large quantities of protein solutions in the laboratory, without the presence of additional components typically found in industrial distillation matrices, can result in excessive foaming and make distillation hard to control. In this instance, since the use of antifoam did not interfere with subsequent analysis of the distillate, 3-4 drops of antifoam (Antifoam A Emulsion, 30% aqueous emulsion of silicon polymer, Sigma-Aldrich Company Limited, UK) were used to prevent excessive foaming. The choice of 10% alcohol was based on typical initial alcohol concentration for cereal based distillations. The starting and distillate volumes chosen meant that over 99% of the ethanol present in the model solution would be collected in the distillate, which is typical of the industrial situation.
Duplicate distillations of the 18 g/l gluten solution were performed. The resulting distillates were analysed according to the protocol presented in Section 7.2. Results are also recorded in Section 7.4.
No gluten protein was detected in the distillates collected. The limit of detection for this method was 0.4 mg/l.
Appendix 3.4 Distillation of Model Solutions – Conclusion
Both sets of model distillations, using BSA and gluten, have provided a practical demonstration that proteins do not distil.
This confirms the distillation concepts as presented by Dr. Eva Sørensen and discussed in Section 5.3. The level of gluten employed in the model distillation is an accurate estimate of the maximum concentration found in a cereal distillery wash.
Page 68 of 173
Appendix 4 A Brief Review of the Scientific Literature Relating to “Gluten” Content in Cereals
Appendix 4.1 Introduction
Protein in cereals has traditionally been classified according to a fairly crude fractionation with various solvents into four ‘Osborne’ fractions: albumin, globulin, prolamin and glutelin. The water/salt soluble fractions are characterised as albumin and globulin. These fractions are generally found in the aleurone, germ and bran of the grain, and contain physiologically active proteins such as enzymes.
Gluten is generally considered as the protein fractions remaining after the water soluble fractions (albumin and globulin) are removed and represents the storage proteins. The gluten fractions are prolamin, which is extracted in ethanol (normally 70%), and glutelin, which can be extracted in dilute acids or bases, or latterly using a detergent or chaotropic agent. The gluten fractions are confined to the starchy endosperm (Hoseney 1998a).
The accepted definitions of these fractions are very crude because in reality each fraction has a very complex composition. It is important to note that different cereals have a range of distributions (and terminologies) for prolamins and glutelins. The prolamin fraction is referred to as gliadin in wheat, hordein in barley and zein in maize.
Appendix 4.2 Wheat
Most of the information collated refers to breadmaking wheat, which has very different characteristics from distilling wheat, and which has much higher levels of gluten. In wheat, gluten comprises of two major fractions, gliadin (prolamin) and glutenin (glutelin).
Halveston and Zeleny (1988) supply the following protein levels for a range of different wheat types.
Appendix 4 Table 1
Product Wheat Protein Type (%) Pasta/Macaroni >13 (>2.28*) Durum Hearth Bread and Hard Rolls >13.5 (>2.37) Spring Pan Bread 11.5 – 13 (2.02 – 2.28) Winter Crackers 10 – 11 (1.75 – 1.93) Soft/Hard Biscuits 9 – 11 (1.58 – 1.93) Soft/Hard Cakes, Pies, Cookies 8 – 10 (1.40 – 1.75) Soft
(*N = % Protein/ 5.7)
Page 69 of 173
Halveston and Zeleny (1998) state that cake and cookie flour (similar to distilling quality) are low in gluten, and that gluten quality is a varietal characteristic.
According to Wrigley and Bietz (1998), the protein fractions in wheat gluten make up about 80% of the protein fraction of the flour. The flour fraction approximates to 85% of the original grain weight. The gluten fraction comprises about 80 % protein, with about 5 -10% lipid material. Gluten quantity is dependent on the nutrient levels resulting from growing conditions, while gluten quality is primarily influenced by genotype (variety).
Coultate (1996) reports wheat protein content at 7–15%, of which 85% is storage protein (i.e. gluten), and states that about 2/3 of this is gliadin, which is the most soluble component of this fraction, and 1/3 glutenin, which is largely insoluble.
Hoseney (1998a), in a wider review covering a range of cereals, gives a protein content of 8 – 16% for commercial wheat. High quality, speciality bread wheat contains around 20% gluten. This comprises 80% protein, and 8 percent lipids, with the remainder carbohydrate and ash (Hoseney 1998b). Gluten proteins contain about 35% glutamic acid.
Sarkki (1980) gives a table showing the breakdown of wheat flour proteins, stating that wheat contains 10 – 14% protein of which 80% is gluten. This is composed of around 75% protein, and the remainder is carbohydrate (15%), fat (6%) and ash (0.8%). The proteins of the flour represent about 85% gluten proteins with the remaining 15% non-gluten (60% albumin, 40% globulin), peptides and amino acids. No values are given for the relative amounts of gliadin and glutenin, but these are grouped in terms of relative molecular weight into high (≈100 kDa glutenin) and low (25–100 kDa gliadin) molecular weight fractions.
Feillet (1980) gives relative proportions of 15-20% for albumins and globulins and 40% for gliadins and 40% for glutenins.
Appendix 4.3 Barley
In barley, proteins account for about 8 – 15% of the total barley grain (Shewry, 1993). The water soluble albumins and globulins account for 8 – 21% of the total. The ‘gluten’ storage proteins consists of two fractions, hordein (prolamin), which comprises 35-55% of the total grain proteins and glutelin, which accounts for about 10% of the total grain nitrogen.
Duffus and Cochrane (2000) give a breakdown for barley of 10% percent total protein comprised of 30% hordein, 30% glutelin, 10% globulin and 30% albumin and free amino acids, based on the variety Bomi, which is a feed variety.
Hoseney (1998a) puts the hordein fraction at about 40% of barley proteins.
Page 70 of 173
Appendix 4.4 Corn (Maize)
Wilson (1999) gives a total grain protein content of 8.3% for corn at 15.5% moisture, giving a total of 9.8% on a dry weight basis. In corn, the prolamin fraction is composed of zein I and zein II, which together account for about 52 percent of the total kernel protein. This also includes a small fraction of glutelin (known as glutelin I). The proper glutelin fractions (glutelin II and III) make up a further 25 percent of the grain nitrogen.
Mossé and Landry (1980) state that the prolamin and glutelin roughly correspond to about 80% of the total seed nitrogen.
References
Coultate T.P. (1996) Food The Chemistry of its Components, 3rd Edition, Royal Society of Chemistry, Cambridge, pp. 129 – 137
Duffus C.M. and Cochrane M.P. (1993) The Formation of the Barley Grain- Morphology, Physiology and Biochemistry – Overall Changes in Chemical Composition of Grains During Development and Maturation. In: Barley: Chemistry and Technology. Edited by MacGregor A.W., American Society of Cereal Chemists, St Paul Minnesota, USA, pp. 61-65
Feillet P. (1980) Evaluation and Measurement of Wheat Quality - Introduction. In: Cereals for food and Beverages, Recent Progress in Cereal Chemistry, Proceedings of an International Conference on Cereals for Food and Beverages. Edited by G.E. Inglett and L Munck, Copenhagen, Denmark, Academic Press, New York, USA, pp. 183-184
Halveston J. and Zeleny, L. (1998) Criteria of Wheat Quality - Chemical Characteristics that Determine Wheat Quality. In: Wheat: Chemistry and Technology. Edited by Y Pomeranz, Y. Volume 1, American Association of Cereal Chemists, St Paul, Minnesota, USA, pp. 24 – 29
Hoseney R.C. (1998a) Principles of Cereal Science and Technology, American Association, St Paul, Minnesota, USA, pp. 65-79
Hoseney R.C. (1998b) Principles of Cereal Science and Technology, American Association, St Paul, Minnesota, USA, pp. 197-201
Mossé J. and Landry J. (1980) Recent Research on Major Maize Proteins: Zeins and glutelins. In: Cereals for food and Beverages, Recent Progress in Cereal Chemistry, Proceedings of an International Conference on Cereals for Food and Beverages. Edited by G.E. Inglett and L Munck, Copenhagen, Denmark, Academic Press, New York, USA, pp. 255-273
Sarkki M.L. (1980) Wheat gluten - Properties and Analysis of Gluten. In: Cereals for food and Beverages, Recent Progress in Cereal Chemistry, Proceedings of an
Page 71 of 173 International Conference on Cereals for Food and Beverages. Edited by G.E. Inglett and L Munck, Copenhagen, Denmark, Academic Press, New York, USA, pp. 156-158
Shewry P.R. (1993) Barley seed proteins. In: Barley: Chemistry and Technology. Edited by MacGregor, A.W., American Society of Cereal Chemists, St Paul Minnesota, USA, pp. 131-197
Wilson C.M. (1999) Proteins of the Kernel - Introduction and Protein Fractionation Overview. In: Corn: Chemistry and Technology, Edited by Watson, S.A, American Association of Cereal Chemists., St Paul, Minnesota, USA, pp. 273-277
Page 72 of 173 Appendix 5 *Details of Samples for which Analytical Data were Submitted as Part of the Request for Provisional Exemption – Bottled Products and Production Distillate Samples* (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Appendix 5 *Details of Samples for which Analytical Data were Submitted as Part of the Request for Provisional Exemption – Bottled Products and Production Distillate Samples* The EFSA Panel noted that sample details were not supplied for the cereal based distillate samples which were analysed for total protein in the application for provisional exemption. Some additional information for these samples is now provided in the following table. However, the level of detail is not as complete as the samples for which analytical data were collected using improved methodology as detailed in Appendices 6a+6b. In this case, specific requests were made for sample suppliers to provide as much background detail as possible.
Neutral spirit production will always use a multi-column still (3). Whilst traditionally, Scotch and Irish Grain Whisk(e)y production has used a 2 column still, as described in 4.3.3.2, some distilleries employ extra columns to provide more control. The products from such stills are distinguished from neutral spirit in that they retain the taste and flavour of grain whisky. The basis of design is typically the system as referred to in (2) and so this has been indicated.
Sample Analysed for Total Analysed for Number Distillate Type Distillation System Description Source Protein (39) Gluten (46) Scotch Malt Whisky – Bottled Products S04-0820 Scotch Malt Whisky Double Pot Stills (1) Glenfiddich Scotch Malt Whisky Bottled Product Scotch Malt Whisky S04-0821 Double Pot Stills (1) Glenfiddich Liqueur Bottled Product (Liqueur) Aultmore Scotch Malt Whisky 12 year S04-0868 Scotch Malt Whisky Pot Stills (1) Bottled Product old Dalwhinnie Scotch Malt Whisky 15 year S04-0902 Scotch Malt Whisky Double Pot Stills (1) Bottled Product old Lagavulin Scotch Malt Whisky 16 year S04-0905 Scotch Malt Whisky Double Pot Stills (1) Bottled Product old
Scotch Blended Whisky – Bottled Products Pot Stills and 2 Column Still Teachers Highland Cream Scotch S04-0707 Scotch Blended Whisky Bottled Product (1) and (2) Whisky Pot Stills and 2 Column Still Stewarts Cream of the Barley Blended S04-0708 Scotch Blended Whisky Bottled Product (1) and (2) Scotch Whisky Pot Stills and 2 Column Still S04-0709 Scotch Blended Whisky Ballantine’s Finest Scotch Whisky Bottled Product (1) and (2) Pot Stills and Coffey Still (1) Johnnie Walker Black Label Old Scotch S04-0714 Scotch Blended Whisky Bottled Product and (2) Whisky Pot Stills and Coffey Still (1) S04-0715 Scotch Blended Whisky Bells Scotch Whisky Bottled Product and (2) Pot Stills and 2 Column Still S04-0719 Scotch Blended Whisky Cutty Sark Scotch Whisky Bottled Product (1) and (2) Pot Stills and 2 Column Still S04-0720 Scotch Blended Whisky The Famous Grouse Bottled Product (1) and (2) Scotch Blended Whisky Pot Stills and 2 Column Still S04-0721 The Famous Grouse Liqueur Bottled Product (Liqueur) (1) and (2) Pot Stills and Multi-Column S04-0824 Scotch Blended Whisky Grant’s Family Reserve Bottled Product Still (1) and (2)
Page 73 of 173 Appendix 5 *Details of Samples for which Analytical Data were Submitted as Part of the Request for Provisional Exemption – Bottled Products and Production Distillate Samples* (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Sample Analysed for Total Analysed for Number Distillate Type Distillation System Description Source Protein (39) Gluten (46) Irish Whiskey – Bottled Products Pot Still (1) and Multi-column S04-0875 Irish Whiskey Jameson Irish Whisky Bottled Product still (2)
Vodka (cereal derived) – Bottled Products S04-0710 Vodka Multi-column Still (3) Altai Vodka Bottled Product S04-0711 Vodka Multi-column Still (3) Wodka Wyborowa Pure Rye Grain Bottled Product S04-0717 Vodka Multi-column Still (3) Smirnoff Triple Distilled Vodka Bottled Product Made Wine produced using S04-0825 Multi-column Still (3) Taboo Bottled Product Vodka Altia Corporation – Rajamaki Plant – S04-1065 Vodka Multi-column Still (3) Production Distillate Sample Dry Vodka
Gin (cereal derived) – Bottled Products Multi-column Still followed by S04-0716 Distilled Gin single Pot Still Distillation (3) Tanqueray Special Dry Gin Bottled Product and (1) Multi-column Still followed by S04-0727 Distilled Gin single Pot Still Distillation (3) Bombay Sapphire Dry Gin Bottled Product and (1) Multi-column Still followed by S04-0829 Distilled Gin single Pot Still Distillation (3) Grants Gin Bottled Product and (1) Multi-column Still followed by S04-0870 Distilled Gin single Pot Still Distillation (3) Greenalls Gin Finished Product Bottled Product and (1) Multi-column still followed by S04-1047 Distilled Gin single Pot Still Distillation (3) Hendricks Gin Bottled Product and (1)
Bourbon/Canadian Whisky/Other Whisk(e)y Products – Bottled Products Column Still and Doubler (2) S04-0726 Bourbon Jack Daniels Whiskey Bottled Product and (1) S04-0973 Korn Multi-column Still (3) Echter Nordhaufer Doppelkorn Bottled Product
Page 74 of 173 Appendix 5 *Details of Samples for which Analytical Data were Submitted as Part of the Request for Provisional Exemption – Bottled Products and Production Distillate Samples* (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Sample Analysed for Total Analysed for Number Distillate Type Distillation System Description Source Protein (39) Gluten (46) Neutral Spirit – Production Distillate Samples S04-0827 Neutral Spirit Multiple Column System (3) Girvan Distillery Grain Neutral Spirit Production Distillate Sample S04-0872 Neutral Spirit Multi-column Still (3) Greenalls Grain Spirit Production Distillate Sample S04-0900 Neutral Spirit Multi-column Still (3) Grain Neutral Spirit ex Velva (French) Production Distillate Sample S04-0901 Neutral Spirit Multi-column Still (3) Grain Neutral Spirit ex Velva (French) Production Distillate Sample Sedalcol Grain Neutral Alcohol S04-1034 Neutral Spirit Multi-column Still (3) Production Distillate Sample 23/03/04 Sedalcol Grain Neutral Alcohol S04-1035 Neutral Spirit Multi-column Still (3) Production Distillate Sample 07/06/04 Sedalcol Grain Neutral Alcohol S04-1036 Neutral Spirit Multi-column Still (3) Production Distillate Sample 11/05/04 Sedalcol Grain Neutral Alcohol S04-1038 Neutral Spirit Multi-column Still (3) Production Distillate Sample 14/04/04 Altia Corporation – Koskenkorva Plant – S04-1064 Neutral Spirit Multi-column Still (3) Production Distillate Sample Grain Ethanol Cameronbridge Scotch Grain Neutral S04-1066 Neutral Spirit Multi-column Still (3) Production Distillate Sample Spirit
Grain New Make Spirit (Scotch/Irish) – Production Distillate Samples S04-0826 Scotch New Make Grain Multi-column Still (2) Girvan Scotch Grain New Make Spirit Production Distillate Sample S04-0874 Irish New Make Grain Multi-column Still (2) Middleton Irish Grain New Make Spirit Production Distillate Sample Cameronbridge Scotch Grain New S04-1067 Scotch New Make Grain Coffey Still (2) Production Distillate Sample Make Spirit Strathclyde Scotch Grain New Make S04-1076 Scotch New Make Grain 2 Column Still (2) Production Distillate Sample Spirit
Malt New Make Spirit (Scotch/Irish) – Production Distillate Samples S04-0869 Scotch New Make Malt Pot Stills (1) Aultmore Scotch Malt New Make Spirit Production Distillate Sample S04-0873 Irish New Make Malt (Pot) Pot Stills (1) Middleton Irish Pot Still New Make Spirit Production Distillate Sample Dalwhinnie Scotch Malt New Make S04-0903 Scotch New Make Malt Double Pot Stills (1) Production Distillate Sample Spirit S04-0904 Scotch New Make Malt Double Pot Stills (1) Lagavulin Scotch Malt New Make Spirit Production Distillate Sample S04-0906 Scotch New Make Malt Pot Stills (1) Talisker Scotch Malt New Make Spirit Production Distillate Sample S04-0907 Scotch New Make Malt Double Pot Stills (1) Caol Ila Scotch Malt New Make Spirit Production Distillate Sample
Page 75 of 173 Appendix 5 *Details of Samples for which Analytical Data were Submitted as Part of the Request for Provisional Exemption – Bottled Products and Production Distillate Samples* (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Malt New Make Spirit (Scotch/Irish) – Production Distillate Samples S04-1071 Scotch New Make Malt Pot Stills (1) Tormore Scotch Malt New Make Spirit Production Distillate Sample S04-1072 Scotch New Make Malt Double Pot Stills (1) Laphroaig Scotch Malt New Make Spirit Production Distillate Sample Glentauchers Scotch Malt New Make S04-1073 Scotch New Make Malt Pot Stills (1) Production Distillate Sample Spirit S04-1074 Scotch New Make Malt Double Pot Stills (1) Miltonduff Scotch Malt New Make Spirit Production Distillate Sample Glenburgie Scotch Malt New Make S04-1075 Scotch New Make Malt Pot Stills (1) Production Distillate Sample Spirit
Gin (cereal derived) – Production Distillate Samples S04-0712 Gin Multi-column Still (3) Compounded Gin 193 Production Distillate Sample Multi-column Still followed by S04-0713 Distilled Gin single Pot Still Distillation (3) Gin Distillate (LGS) Production Distillate Sample and (1) Multi-column Still followed by S04-0871 Distilled Gin single Pot Still Distillation (3) Greenalls Gin Concentrate Production Distillate Sample and (1) Multi-column still followed by S04-0893 Distilled Gin single Pot Still Distillation (3) Distilled London Dry Gin Concentrate Production Distillate Sample and (1)
Page 76 of 173 Appendix 6a List of Samples Analysed Using Improved Analytical Methodology – Bottled Products (Sampling Time and Date were not requested for bottled samples, but in some instances they were provided; lot codes can be used to provide information on date of bottling.) (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Appendix 6a List of Samples Analysed Using Improved Analytical Methodology – Bottled Products Allergenic Sample Materials Prior Sampling Number Distillate Type Distillation System to Distillation Description Source Time+Date
Scotch Grain Whisky Scotch Grain Wheat, Barley Black Barrel Single Grain Whisky, 40% Produced and Matured at Girvan Distillery, S05-3232 Whisky Multi-column still (2) (Malted) vol Bottled at Strathclyde Business Park NA Scotch Grain Wheat, Barley Cameron Brig Single Grain Whisky, S05-3236 Whisky 2 column Coffey Still (2) (Malted) 40% vol, L5284CR000 Leven Bottling Plant, Fife NA Scotch Grain The North British Cask Strength Single S05-3439 Whisky 2 column Still (2) Barley (Malted) Grain 1980, 60.3% vol The North British Distillery NA
Scotch Malt Whisky Scotch Malt Double pot stills with “purifiers”, balanced Glen Grant 5 Year Old Single Malt Chivas Brothers Limited, Newbridge, S05-1977 Whisky system, shell+tube condensers (1) Barley (Malted) Scotch Whisky, 40% vol Edinburgh, Scotland NA Scotch Malt S05-2443 Whisky Double pot stills (1) Barley (Malted) Miltonduff, 40% vol Miltonduff Distillery, Elgin, Morayshire NA Scotch Malt Laphroaig Single Islay Malt Whisky, S05-3069 Whisky Double pot stills (1) Barley (Malted) 40% vol, Lot Code LF255R2K Allied Distiller’s Limited, Dumbarton NA Dalwhinnie, 15yr old, 43% Scotch Malt Double pot stills, worm tub condenser, L5248CM000 03312475, Rotation S05-3286 Whisky unbalanced system (1) Barley (Malted) Number 105/841 Leven Bottling Plant, Fife NA Scotch Malt Auchestoshan 10 yr old, 40%, S05-3485 Whisky Triple pot stills (1) Barley (Malted) Sampled at bottling hall Morrison Bowmore, Springburn, Glasgow 08/11/2005
Scotch Blended Whisky Scotch Blended Wheat, Barley Chivas Regal 12 Year Old Blended S05-1976 Whisky 2 column stills and pot stills (1) and (2) (Malted) Scotch Whisky, 40% vol Chivas Brothers Limited, Paisley, Scotland NA Scotch Blended Wheat, Barley Dewars’ White Label, 6.286.01, 40% John Dewar & Sons Ltd Bottling Facility, S05-2426 Whisky 2 column stills and pot stills (1) and (2) (Malted) vol 1700 London Road, Glasgow 21/09/2005 Ballantine’s Finest Scotch Whisky, Scotch Blended Wheat, Barley 40% vol, metal presentation, Lot Code S05-3061 Whisky 2 column stills and pot stills (1) and (2) (Malted) LU29740 Allied Distiller’s Limited, Dumbarton NA Scotch Blended Wheat, Barley Johnnie Walker Red Label, 40% vol, S05-3993 Whisky 2 column stills and pot stills (1) and (2) (Malted) L5179CP000 Leven Bottling Plant, Fife NA
Page 77 of 173 Appendix 6a List of Samples Analysed Using Improved Analytical Methodology – Bottled Products (Sampling Time and Date were not requested for bottled samples, but in some instances they were provided; lot codes can be used to provide information on date of bottling.) (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Allergenic Sample Materials Prior Sampling Number Distillate Type Distillation System to Distillation Description Source Time+Date
Irish Whiskey Barley (Unmalted), Powers Gold Label, 40% vol, Irish Distillers Ltd, Fox and Geese, August S05-2196 Irish Whiskey Multi column still and pot stills (3) and (1) Barley (Malted) L42501433 Robinhood Road, Clondalkin, Dublin 22 2004 Barley (Unmalted), Irish Distillers Ltd, Fox and Geese, S05-2200 Irish Whiskey Multi column still and pot stills (3) and (1) Barley (Malted) Jameson, 40% vol, L52101369 Robinhood Road, Clondalkin, Dublin 22 July 2005 Bushmills Single Malt 10 Year Old, Irish Distillers Ltd, Fox and Geese, February S05-2201 Irish Whiskey Pot Stills (1) Barley (Malted) 40% vol, L50555048 Robinhood Road, Clondalkin, Dublin 22 2005 Redbreast 12 Year Old, 40% vol, 3036 Irish Distillers Ltd, Fox and Geese, S05-2202 Irish Whiskey Pot Stills (1) Barley (Malted) 5129 Robinhood Road, Clondalkin, Dublin 22 April 2005
Vodka Vodka (cereal S05-1978 derived) 3-column stills (3) Rye Wyborowa Wodka, 40% vol Wyborowa S.A., Poznan, Poland NA Vodka (cereal S05-2457 derived) Multi-column stills (3) Wheat Grey Goose vodka, 40% vol Grey Goose Plant 26/05/2005 Vodka (cereal Stolichnaya Russian Vodka, 37.5% Lot S05-3065 derived) Multi-column stills (3) Wheat and Rye Code L30.12.04, production time 10:23 Allied Distillers Limited, Dumbarton NA Vodka (cereal S05-3125 derived) Multi-column stills (3) Wheat Absolut Vodka, 40% vol V&S Absolut Spirits Distillery NA Vodka (cereal Wheat, Barley Smirnoff, 37.5% vol, L5244CY000 S05-3287 derived) 5 column stills (3) (Malted) 16537011, Rotation Number 505/5396 Leven Bottling Plant, Fife NA
Distilled Gin Pot Still Flavouring of Alcohol Using Essential Oils Contained Within Natural Botanicals, Alcohol Sourced From Greenwich Distillers, London – Plymouth Gin, 41.2% vol, Identification Plymouth Gin, Black Friars Distillery, 60 S05-2015 Distilled Gin multicolumn system (1) and (3) Wheat Number – LS193-13:59, Southside St, Plymouth PL1 2LQ NA Multi-column still and Carterhead Still (3) Wheat, Barley Bombay Sapphire gin, 47% vol, L05 S05-2461 Distilled Gin and (1) (Malted) 255 W5 15:17, G+J Greenall NA Wheat, Barley Gordons Gin, 37.5% vol, Rotation S05-3285 Distilled Gin 5 column still + pot still (3) and (1) (Malted) Number 805/5711 Leven Bottling Plant, Fife NA Beefeeter London Dry Gin, 40% vol, Wheat, Barley 700 ml clear glass bottle, Lot Code S05-3063 Distilled Gin Multi-column + Pot Still (3) and (1) (Malted) LF237H4K, production time 14:27 Allied Distillers Limited, Dumbarton NA
Page 78 of 173 Appendix 6a List of Samples Analysed Using Improved Analytical Methodology – Bottled Products (Sampling Time and Date were not requested for bottled samples, but in some instances they were provided; lot codes can be used to provide information on date of bottling.) (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Allergenic Sample Materials Prior Sampling Number Distillate Type Distillation System to Distillation Description Source Time+Date
Bourbon/Canadian Whisky and Other Whisk(e)y Products Wheat, Barley Maker’s Mark – Finished Product, Maker’s Mark Distillery, Loretto, KY, USA, July 11, S05-2014 Bourbon Column Still and Doubler (2) and (1) (Malted) 90°proof, 45% vol Bottling Line “A” 2005 Rye, Barley Jack Daniel Distillery, From Bottling Tank 05/08/2005 S05-2078 Bourbon Column Still and Doubler (2) and (1) (Malted) Jack Daniels Batch 5-5702, 40% vol 11 09:15 Blending from Multicolumn still and pot still (Maturation more than 12 months) Edelkorn Vom alten Fass, 38.5% vol, Berentzen-Gruppe AG, Ritterstrasse 7, S05-2151 Korn (3) and (1) Wheat L5080 47940 Haselünge 21/03/2005 Rye, Barley Jim Beam Kentucky Straight Whiskey, S05-2991 Bourbon Column Still and Doubler (2) and (1) (Malted) 40% vol Jim Beam Brands Co, Clermont, Kentucky NA Blending from Pot stills and Column Still Bottling Plant Valverde del Majano. Line S05-3036 Spanish Whisky (1) and (2) Barley (Malted) Blend Whisky, 40% vol, LF 266C1 1., Destilerias DYC, Spain 23/09/2005 Canadian Club Barrel Blended Whisky, Canadian Blending from 1 column still, multicolumn Rye, Barley 40% vol, Lot Code LF194, production S05-3066 Whisky still and pot stills (2), (3) and (1) (Malted) time 18:26 Allied Distiller’s Limited, Dumbarton NA Blending from 1 column still and double S05-3121 French Whisky pot stills (2) and (1) Barley (Malted) Breton Whisky, 40% vol Distillerie Warenghem, Lannion NA LaSalle, Blending and Production Canadian Blending from batch stills, beer stills and Rye, Barley Crown Royal De Luxe Canadian Planning, Diageo Global Supply, Montreal, S05-4026 Whisky batch kettle stills (2) and (1) (Malted) Whisky, 40% vol Canada NA Barley (Malted), Wild Turkey Bourbon Whiskey, 50.5% S05-4343 Bourbon Beer Still and Doubler (2) and (1) Rye vol Wild Turkey Distillery Lawrenceburg 01/08/2005 Barley (Malted), S05-4345 Bourbon Beer Still and Doubler (2) and (1) Rye Wild Turkey Rye Whiskey, 50.5% vol Wild Turkey Distillery Lawrenceburg 01/08/2005 Wheat, Barley French Distillery (via LaSaffre), prior to S06-0061 Genever 2 column still and pot still (2) and (1) (malted), Rye Genever Distillate, 56.9% vol bottling 29/11/2005 Japanese Hakashu 12 year old Japanese whisky, Suntory Ltd. via Morrison Bowmore S06-0304 Whisky Double pot stills (1) Barley (Malted) 43.5% vol Distillers Ltd. From ex-duty paid stock 23/01/2006 Japanese Yamazaki 10 year old Japanese Suntory Ltd. via Morrison Bowmore S06-0305 Whisky Double pot stills (1) Barley (Malted) whisky, 40% vol Distillers Ltd. From ex-duty paid stock 23/01/2006 Mixture of column and pot stills (2) and Wheat, Barley Korenwijn 10 jaar vatgelagerd, 38% 19/05/2005 S06-0077 Genever (1) (malted), Rye vol, 1359ZU Korenwijn 10J Zuidam Distillers BV, The Netherlands 16:33 NA = Not Applicable (Dates and times of sampling were not specifically requested for bottled products)
Page 79 of 173 Appendix 6b List of Samples Analysed Using Improved Analytical Methodology–Production Distillate Samples (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Appendix 6b List of Samples Analysed Using Improved Analytical Methodology – Production Distillate Samples Allergenic Sample Materials Prior Sampling Number Distillate Type Distillation System to Distillation Description Source Time+Date
Neutral Spirit The North British Distillery Co. Ltd, 24/8/05 S05-2208 Neutral Spirit Combination of still systems (3) Wheat Grain Neutral Spirit, 96.6% vol Scotland, Vat NV1 0345 hours 2 column stills to produce feedstock, 2 column stills to produce finished product Wheat, Barley Invergordon Distillery, Scotland, Vat NGR 22/08/2005 S05-2243 Neutral Spirit (3) (Malted) Grain Neutral Spirit, 96.3% vol 6 14:00 Ethyl Alcohol of an Agricultural Origin Grain Distillery of NORDBRAND (Neutral Spirit), 96.3% vol, Material No. NORDHAUSEN GmbH, Bahnhofst. 25, 25/8/005 S05-2280 Neutral Spirit Multi-column still (3) Wheat 04100243 99734 Nordhausen, Germany 11:00 Ryssen, Ethyl Alcohol of French origin, Ethyl Alcohol used for production of produced as a by-product of starch S05-2455 Neutral Spirit Multi-column still (3) Wheat vodka, 96.2% vol processing 21/09/2005 Ethyl Alcohol of French Origin used for Sampled at reception (Bacardi-Martini production of Grey Goose Vodka, 96.2 Production) from Supplier (Tereos – Lille, S05-2456 Neutral Spirit Multi-column still (3) Wheat % vol France) 14/09/2005 G + J Greenall, Warrington, England. Sample taken from tank on site. The tank is topped up via a solero system from various suppliers. It is therefore S05-2458/ Wheat, Barley impossible to determine the source of the S05-2459 Neutral Spirit Multi-column still (3) (Malted) Grain Neutral Spirit, 96% vol. sample. Sep-05 Finsprit – Cistern, 96.0% vol, 1-2005- V&S Absolut Spirits Distillery, Denmark, S05-3122 Neutral Spirit Multi-column still (3) Wheat 09-29 Final Spirit, Still 1 29/09/2005 S05-3123/ V&S Absolut Spirits Distillery, Denmark, S05-3124 Neutral Spirit Multi-column still (3) Wheat Grain alcohol, 96.2% vol, 050919-1 Tank 1 (containing several productions), 20/09/2005 Girvan Distillery, Scotland, No.4 S05-3234 Neutral Spirit Multi-column still (3) Wheat Grain Neutral Spirit, 96.5% vol Apparatus, Vat Sample 2 05/10/2005 Wheat, Barley Grain Neutral Spirit, 96.5% vol, Strathclyde Distillery, Scotland, ex Column 18/10/2005 S05-3310 Neutral Spirit 4 column grain neutral spirit still (3) (Malted) Certificate 5-17 Sample Point 14:30 Wheat, Barley Grain Neutral Spirit, 96.5% vol, Strathclyde Distillery, Scotland, ex Column 20/08/2005 S05-3312 Neutral Spirit 4 column grain neutral spirit still (3) (Malted) Certificate 5-9 Sample Point 14:00 S05-3279/ Wheat, Barley Spirit Receiver 4, Neutral Spirit Still, S05-3281 Neutral Spirit 5 column still (3) (Malted) Grain Neutral Spirit, 96.3% vol Cameron Bridge, Scotland 06/09/2005 S05-3323 Neutral Spirit Multi-column still (3) Barley (Malted) Grain Neutral Spirit, 96.4% vol Altia Corporation, Grain Division, Finland 18/10/2005
Page 80 of 173 Appendix 6b List of Samples Analysed Using Improved Analytical Methodology–Production Distillate Samples (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Allergenic Sample Materials Prior Sampling Number Distillate Type Distillation System to Distillation Description Source Time+Date
Neutral Spirit S05-3248 Neutral Spirit Multi-column still (3) Wheat Sample no 5004362, 96.3% vol. Sedamyl S.p.A – Saluzzo (CN), Italy 38615 Ethyl Alcohol of Agricultural Origin – S05-4337 Neutral Spirit Multi-column still (3) Wheat SILO A Sample, 96.2% vol Sedamyl S.p.A – Saluzzo (CN), Italy 05/12/2005 S06-0070- S06-0073 Neutral Spirit Multi-column still (3) Wheat Grain Neutral Spirit, 96.1% vol Greenwich Distillers, London, England 05/12/2005 S05-4395- Ethyl Alcohol of Agricultural Origin – 96 Sedalcol SNC, Mesnil-Saint-Nicaise, S05-4397 Neutral Spirit Multi-column still (3) Wheat degrees minimum France 12/12/2005 Grain Neutral Spirit, Sas van Gent, V- Royal Nedalco, Sas Van Gent Distillery, S06-0057 Neutral Spirit Multi-column still (3) Wheat 05.255, 96.3 % vol The Netherlands 14/12/2005 Wheat, Barley Grain Neutral Spirit, Delfzyl, V-05.256, Royal Nedalco, Delfzyl Distillery, The mix 27/6/05 S06-0058 Neutral Spirit Multi-column still (3) (Malted), Rye ca 96% Netherlands and 24/1/05 Wheat, Barley Grain Neutral Spirit, Heibronn, V- Royal Nedalco, Heibronn Distillery, mix 7/4/05 S06-0059 Neutral Spirit Multi-column still (3) (Malted), Rye 05.257, ca 96% Germany and 3/4/05
Grain New Make Spirit (Scotch/Irish) Irish New Make Middleton New Make Grain S05-2203 Grain Multi-column still (2) Barley (Malted) Distillate/G11, 94.47% vol Middleton Distillery, Cork, Ireland June 2005 Irish New Make Middleton New Make Grain S05-2204 Grain Multi-column still (2) Barley (Malted) Distillate/G11, 94.45% vol Middleton Distillery, Cork, Ireland June 2005 Scotch New Wheat, Barley NB New Make Wheat Whisky Spirit, The North British Distillery Co. Ltd, S05-2206 Make Grain 2 column still (2) (Malted) 94.5% vol, 05E0011(A) Scotland, Ex bulk vat 13/06/2003 Scotch New NB New Make Maize Whisky Spirit, The North British Distillery Co. Ltd, 25/8/05 S05-2207 Make Grain 2 column still (2) Barley (Malted) 94.6% vol, 05E0011 (B) Scotland, Ex. No1 still 1200 hours Scotch New Wheat, Barley S05-2212 Make Grain 2 column still (2) (Malted) Loch Lomond Grain, 94.2% vol, ex ISR Loch Lomond Distillery Co Ltd, Scotland 24/08/2005 Scotch New Wheat, Barley Invergordon Scotch Grain Spirit, 94.4% 15/07/2005 S05-2244 Make Grain 2 column still (2) (Malted) vol Invergordon Distillery, Scotland, ISR 18 16:00 Scotch New Wheat, Barley Invergordon Scotch Grain Spirit, 94.3% 22/08/2005 S05-2245 Make Grain 2 column still (2) (Malted) vol Invergordon Distillery, Scotland, ISR 10 14:00 Scotch New Wheat, Barley Girvan New Make Grain Spirit, 94.5% Girvan Distillery, Scotland, No.1 05/10/2005 S05-3233 Make Grain Multi-column still (2) (Malted) vol Apparatus, Vat Sample 11 15:00 Scotch New Wheat, Barley Girvan New Make Grain Spirit, 94.5% Girvan Distillery, Scotland, No.5 27/09/2005 S05-3235 Make Grain Multi-column still (2) (Malted) vol Apparatus, Running Sample 16:00 Scotch New Wheat, Barley Strathclyde New Make Grain Spirit, Strathclyde Distillery, Scotland, ex Column 18/10/2005 S05-3313 Make Grain 2 column still (2) (Malted) PBS, 94.4% vol, Certificate 5-16 Sample Point 14:30 Scotch New Wheat, Barley Strathclyde New Make Grain Spirit, Strathclyde Distillery, Scotland, ex Column 20/08/2005 S05-3311 Make Grain 2 column still (2) (Malted) PBS, 94.4% vol, Certificate 5-8 Sample Point 14:00
Page 81 of 173 Appendix 6b List of Samples Analysed Using Improved Analytical Methodology–Production Distillate Samples (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Allergenic Sample Materials Prior Sampling Number Distillate Type Distillation System to Distillation Description Source Time+Date
Grain New Make Spirit (Scotch/Irish) S05-3282 Scotch New Wheat, Barley Cameron Bridge New Make Grain Cameron Bridge Distillery, Scotland, Spirit S05-3284 Make Grain 2 column Coffey still (2) (Malted) Spirit, 94.4% vol Receiver 6 (Still No. 3) 06/09/2005 Product 13 Port Dundas Grain Distillery, Scotland, S05-3994 Scotch New Wheat, Barley Port Dundas New Make Grain Spirit, Receiver 4, Period 47, Charge 6, Stills S05-3995 Make Grain 2 column Coffey still (2) (Malted) 94.3% vol 1+2 17/11/2005 S05-3996 Scotch New Wheat, Barley Port Dundas New Make Grain Spirit, Port Dundas Grain Distillery, Scotland, S05-3997 Make Grain 2 column Coffey still (2) (Malted) 94.3% vol Receiver 3, Period 47, Charge 5, Still 3 17/11/2005 Irish New Make Cooleys Grain Spirit, 94.6% vol, Cooley Distillery, Dundalk, Ireland, Ex Vat 07/12/2005 S05-4352 Grain 2 column still (2) Barley (Malted) CD05/117 1 10:00 Scotch New Wheat, Barley Cameron Bridge New Make Grain S06-0150 Make Grain 2 column Coffey still (2) (Malted) Spirit, 68.5% vol Cameron Bridge Distillery, Scotland 06/12/2005
Malt New Make Spirit (Scotch/Irish) Scotch New Double pot stills, shell and tube S05-0166 Make Malt condensers (1) Barley (Malted) MACALLAN 2005 SR/WV 1, 71% vol Macallan Distillery, Scotland 01/02/2005 Scotch New Double Pot Still, shell and tube Glenmorangie New Make Spirit, 63.5% S05-2061 Make Malt condensers (1) Barley (Malted) vol Glenmorangie Distillery, Tain, Scotland 19/07/2005 Scotch New Double pot stills, paired, with vertical Aberlour New Make Distillate, 68.8% Aberlour Malt Distillery, Aberlour, S05-2107 Make Malt shell and tube condensers (1) Barley (Malted) vol, Ref: AL05/62/4 Scotland, Spirit Receiver Sample 28/06/2005 Scotch New Double pot stills, paired, with purifiers + Glen Grant New Make Distillate, 68.7% Glen Grant Malt Distillery, Rothes, S05-2108 Make Malt vertical shell and tube condensers (1) Barley (Malted) vol, Ref: GG05/94/2 Scotland, Spirit Receiver Sample 07/07/2005 The Glenlivet Malt Distillery, Glenlivet, Scotch New Double pot stills, paired, with vertical The Glenlivet New Make Distillate, Ballindalloch, Scotland, Spirit Receiver S05-2109 Make Malt shell and tube condensers (1) Barley (Malted) 68.8% vol, Ref: GL05/86/1 Sample 15/07/2005 Scotch New Double pot stills, paired, with vertical Longmorn New Make Distillate, 68.3% Longmorn Malt Distillery, Elgin, Scotland, S05-2110 Make Malt shell and tube condensers (1) Barley (Malted) vol, Ref: LM05/55/3 Spirit Receiver Sample 05/07/2005 Barley Irish New Make (Unmalted), Middleton New make pot still S05-2197 Malt Pot stills (1) Barley (Malted) distillate/MMP, 85.13% vol Middleton Distillery, Cork, Ireland June 2005 Barley Irish New Make (Unmalted), Middleton New make pot still S05-2198 Malt Pot stills (1) Barley (Malted) distillate/LMP, 86.12% vol Middleton Distillery, Cork, Ireland June 2005 Irish New Make Bushmills New make single malt Bushmills Distillery, Bushmills, Co. Antrim, August S05-2199 Malt Pot stills (1) Barley (Malted) distillate/B3D, 84.0% vol Northern Ireland 2005 Scotch New Loch Lomond Malt, 71.3% v/v, Ex Spirit S05-2213 Make Malt Pot stills (1) Barley (Malted) Receiver 7 Loch Lomond Distillery Co Ltd, Scotland 25/08/2005 Scotch New Pot stills with column head and water Loch Lomond Malt, 86.0% v/v, Ex Spirit S05-2214 Make Malt cooled reflux (1) Barley (Malted) Receivers 1 and 2 Loch Lomond Distillery Co Ltd, Scotland 19/08/2005
Page 82 of 173 Appendix 6b List of Samples Analysed Using Improved Analytical Methodology–Production Distillate Samples (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Allergenic Sample Materials Prior Sampling Number Distillate Type Distillation System to Distillation Description Source Time+Date
Malt New Make Spirit (Scotch/Irish) Scotch New Laphroaig Distillery, Scotland, No 1 Spirit 07/09/2005 S05-2276 Make Malt Double pot stills (1) Barley (Malted) Laphroaig New Make Spirit, 68% vol Still 11:30 Scotch New Miltonduff Distillery, Scotland, spirit 22/09/2005 S05-2442 Make Malt Double pot stills (1) Barley (Malted) Miltonduff New Make Spirit, 65.8% vol receiver ISR Sample 13:00 S05-2897- Scotch New Craigellachie new-make spirit, Period 26/09/2005 S05-2899 Make Malt Pot stills (1) Barley (Malted) No. 32 Charge No. 2, 69.9% vol Craigellachie Distillery 16:30 S05-3073 Scotch New Double pot stills, worm tub condenser, S05-3074 Make Malt unbalanced(almost balanced) system (1) Barley (Malted) Talisker New Make Spirit, 70.2% vol Talisker Distillery 16/08/2005 S05-3075/ Scotch New Double pot stills, worm tub condenser, Dalwhinnie New Make Spirit, 67.6 % S05-3076 Make Malt unbalanced system (1) Barley (Malted) vol Dalwhinnie Distillery 17/08/2005 Triple/quadruple pot still distillation, S05-3077/ Scotch New Worm tub condenser, unbalanced S05-3078 Make Malt system (1) Barley (Malted) Mortlach New Make Spirit, 71.9% vol Mortlach Distillery 23/09/2005 Glenfiddich Distillery New Make Spirit, Scotch New 74.50% vol, No 1 Stillhouse, Spirit Still S05-3321 Make Malt Double pot stills (1) Barley (Malted) 7, Sampled from Safe, Glenfiddich Distillery 03/10/2005 Scotch New Auchentoshan New Make Spirit, 79.9% Auchentoshan Distillery, via Morrison S05-3486 Make Malt Triple pot stills (1) Barley (Malted) vol Bowmore, Springburn 08/11/2005 Irish New Make Cooleys New Make Malt Spirit, 64.8% Cooley Distillery, Dundalk, Ireland, New 07/12/2005 S05-4353 Malt Double pot stills (1) Barley (Malted) vol, CD05/118 Filling 15:15 Scotch New S06-0149 Make Malt Pot Stills (1) Barley (Malted) Cardhu New Make Spirit, 63.4% Cardhu Distillery, Knockando, Aberlour 06/12/2005
Bourbon/Canadian/Other Whisk(e)y Distillates August 8th, Bourbon Wheat, Barley Maker’s Mark New Make Distillate Maker’s Mark Distillery, Loretto, KY, USA, 2005 S05-2013 Distillate Column still and doubler (2) and (1) (Malted) (Maker’s Mark distillate), 65.00% vol Still System #2, High Wine Tank 9.00am Rye, Barley Canadian (Malted), Rye Star – S17, Sample #05-0089-01, Canadian Club Distillery (Hiram Walker S05-2062 Whisky Distillate Column Still (2) and (1) (Malted) 67.4% vol, (Canadian Club Distillate 1) and Sons Ltd), Tank S17 19/09/2003 Rye, Barley Star Special – SS14, Sample #05- Canadian (Malted), Rye 0089-02, 80.9% vol, (Canadian Club Canadian Club Distillery (Hiram Walker S05-2063 Whisky Distillate Column Still and Pot Still (2) and (1) (Malted) Distillate 2) and Sons Ltd), Tank SS14 05/03/2004 Brown Forman Cistern Room Distillate Bourbon Rye, Barley 6/8/05 production, 62.5% vol, (Bourbon Brown Forman Distillery, Cistern Room S05-2076 Distillate Column Still and Thumper (2) and (1) (Malted) Distillate) Sample (after distillation) 10/08/2005 Bourbon Rye, Barley Jack Daniels 2,5, 70% vol, (Jack Jack Daniel Distillery, Sample taken at 5/8/2005, S05-2077 Distillate Column Still and Doubler (2) and (1) (Malted) Daniels Distillate) Composite Still Micromotion Meter 8:45 am
Page 83 of 173 Appendix 6b List of Samples Analysed Using Improved Analytical Methodology–Production Distillate Samples (1) – Refer to Section 4.3.3.1 (2) Refer to Section 4.3.3.2 (3) See Section 4.3.3.3
Allergenic Sample Materials Prior Sampling Number Distillate Type Distillation System to Distillation Description Source Time+Date
Bourbon/Canadian/Other Whisk(e)y Distillates Blending from Multicolumn still and pot Berentzen Traditionskorn, 32% vol, Berentzen-Gruppe AG, Ritterstrasse 7, S05-2150 Korn Distillate still (3) and (1) Wheat L5159 (Korn Distillate) 47940 Haselünge 08/06/2005 Oldesloer Doppelkorn, 38% vol, August Ernst GmbH & Co. KG, Bad S05-2194 Korn Distillate 2 column still (2) Wheat L1185031 (Korn Distillate) Oldesloe Grain Distillery of NORDBRAND Barley (malted), Roggen-Kornfeindestillate, 96.1% vol, NORDHAUSEN GmbH, Bahnhofst. 25, 25/8/005 S05-2279 Korn Distillate Multi-column still (3) Rye Material No. 04100244 (Korn Distillate) 99734 Nordhausen 10:00 S05-2993/ Bourbon Jim Beam Hi Wine (Jim Beam S05-2994 Distillate Distillate), 64.02% vol Jim Beam Brands Co, Clermont, Kentucky Spanish New Grain Spirit, 94.7% vol, (Spanish Grain 23/09/2005 S05-3032 Make Grain 1 column still (2) Barley (Malted) Whisky Distillate) Destilerias DYC, Spain 12:30 Spanish New DYC, New make spirit, 68.5% vol, S05-3034 Make Malt Pot stills (1) Barley (Malted) (Spanish Malt Whisky Distillate) Pot still tank, Destilerias DYC, Spain 23/09/2005 Crown Royal – Four Roses Bourbon LaSalle, Blending and Production S05-3269/ Canadian Barley (malted), Composite Flavouring Whisky 69.9% Planning, Diageo Global Supply, Montreal, S05-3070 Distillate regular, doubled Rye vol, (Crown Royal Distillate 1) Canada 02/09/2005 Crown Royal – New Make Bourbon LaSalle, Blending and Production Canadian Barley (malted), Whisky (Gimli) 69.0% vol, (Crown Planning, Diageo Global Supply, Montreal, S05-3254 Distillate Beer still (2) Rye Royal Distillate 3) Batch 1, 405078785 Canada 25/08/2005 Crown Royal – New Make Rye Whisky LaSalle, Blending and Production Canadian Barley (malted), (Gimli), 69.9% vol, (Crown Royal Planning, Diageo Global Supply, Montreal, S05-3276 Distillate Coffey still (2) Rye Distillate 3) Batch 4, 405038133 Canada 25/08/2005 Bourbon Barley (malted), Wild Turkey Bourbon Whiskey S05-4342 Distillate Beer Still and Doubler (2) and (1) Rye Distillate, 55% vol Wild Turkey Distillery Lawrenceburg 01/08/2005 Bourbon Barley (malted), Wild Turkey Rye Whiskey Distillate, S05-4344 Distillate Beer Still and Doubler (2) and (1) Rye 55% vol Wild Turkey Distillery Lawrenceburg 01/08/2005 Genever Wheat, Barley French Distillery (via LaSaffre), Prior to S06-0060 Distillate 2 column still (2) (Malted), Rye Genever Distillate, 51.39% vol, addition of botanicals, ex still (zud column) 29/11/2005 S06-0151/ Swedish New Mackmyra Swedish Whisky, P1901, Mackmyra Distillery, Bruksgatan 4, SE- S06-0152 Make Malt Pot stills (1) Barley (Malted) 68.9% vol. 818 32 Valbo 20/12/2005 Grain Distillery of NORDBRAND Weizen-Kornfeindestillate, 96.1% vol, NORDHAUSEN GmbH, Bahnhofst. 25, 25/8/005 S05-2278 Korn Distillate Multi-column still (3) Wheat Material No. 04100257 99734 Nordhausen 09:00
Page 84 of 173
Appendix 7 Extended Analytical Survey - Sample Collection and Traceability
Sample Collection
A proforma was sent out with each sample request to provide details on the samples collected for the extended analytical survey (Appendices 6a and 6b). The following information was requested:
Identification of the potentially allergenic material used prior to distillation. The source of the sample. For a production distillate sample, the location of sampling was requested. The date and time of sampling, for production distillate samples. The type of distillation system employed to produce the distillate.
It was stressed that all distillates/bottled products collected must be produced using one or more of the cereals identified in Annex IIIa of Directive 2003/89/EC (that is, wheat, rye, barley, oats, spelt, kamut or their hybridised strains) and that details regarding the distillation systems employed for each distillate and bottled product must be provided to ensure that the range of distillation systems used within the industry was fully covered. Most samples were accompanied by a completed proforma; although in some instances certain details had to be obtained from the distillery contact after receipt of the sample. A request that clean, unused bottles be used for the purpose of sampling was also placed on the proforma.
The sampling proforma that was issued is reproduced in Figure 1 of this Appendix.
Traceability
All samples were received at The Scotch Whisky Research Institute (SWRI) and automatically assigned an individual numerical reference code using the Laboratory Information Management System (LIMS). All details were recorded in the database and an adhesive label, printed with the sample reference code, was attached to the sample. The individual reference code was used to trace the sample through storage and all subsequent in-house analyses.
To ensure data of the highest standard the Institute has UKAS (United Kingdom Accreditation Service) certification for its principal analysis methods, giving independent confirmation that the quality of this analytical work, based on an established quality system, is compliant with BS EN ISO/IEC 17025. For more information about the research and analysis work undertaken by the SWRI, consult the SWRI public website. The SWRI is registered as Accredited Testing Laboratory No. 1690.
Scotch Whisky Research Institute Research Avenue North Riccarton Edinburgh EH14 4AP www.swri.co.uk
Page 85 of 173
Two external contract laboratories were also used for the analysis of collected samples. All samples analysed externally were sub-sampled into clean, unused sample containers to which were affixed labels with the assigned Scotch Whisky Research Institute sample numbers. All externally produced results were referred back to this sample number to ensure complete traceability. The two contract laboratories employed were Reading Scientific Services Limited (RSSL) and Intertek ASG.
RSSL's laboratories perform testing that is accredited by UKAS as being compliant with BS EN ISO/IEC 17025. It is RSSL's policy to accredit all testing that is carried out routinely by the laboratories. However, since the Quality System operates in all laboratories and many of the procedures are common for work performed, even work that is not accredited benefits significantly from the UKAS accreditation. RSSL have achieved UKAS accreditation for residual allergens in foods using Enzyme Linked Immunosorbent Assay (ELISA) kit methods. RSSL is registered as Accredited Testing Laboratory No. 1216. RSSL also participates in the FAPAS food analytical proficiency testing scheme for gluten analysis.
Reading Scientific Services Limited The Lord Zuckerman Research Centre Whiteknights Campus Reading RG6 6LA England www.rssl.com
The work conducted by Intertek ASG was carried out in a laboratory which is a member of the UK Good Laboratory Practice (GLP) compliance programme. Intertek ASG has been a member of the UK GLP compliance programme since 1987 for the provision of chemical characterisation and physical, chemical data to support the worldwide registration and notification of new chemical products and intermediates, including submissions to the US EPA and Japanese MITI regulatory agencies. A statement of GLP compliance, issued by the monitoring authority via their biennial inspection programme has been maintained since 1987.
Intertek ASG PO Box 42 Hexagon Tower Blackley Manchester M9 8ZS United Kingdom http://www.intertek-cb.com/newsitetest/services/asg/index.shtml
Page 86 of 173
Appendix 7 Figure 1 Reproduction of Sample Proforma Issued with Each Sample Request
Page 87 of 173
Appendix 8 *Total Protein Based Analytical Evidence as Presented in the Application for Provisional Exemption*
Appendix 8.1 *Introduction*
In the application for provisional exemption, 39 samples of new make whisky and white spirits were analysed for total protein levels using a Total Protein Assay by Randox. This method is based on the principle that pyrogallol red complexes with proteins in an acid environment containing molybdate ions, to form a blue-coloured complex. This coloured complex, which absorbs maximally at 600 nm, is measured by spectrophotometer. The colour intensity is directly proportional to the protein concentration in the samples.
Appendix 8.2 *Sampling Protocol*
Samples were received at The Scotch Whisky Research Institute from a variety of different sources. These samples were selected to cover a range of the types of distilled beverages produced using cereal as a raw material prior to distillation. Each sample was assigned an individual numerical reference code using the Laboratory Information Management System (LIMS), recording all the details provided in the database and an adhesive label was attached. Information regarding the samples for which analytical data were submitted as part of the provisional exemption can be found in Appendix 5.
The EFSA NDA Panel in their opinion on the provisional application (Section 3.3.2) noted that sample details were not supplied for the cereal based distillate samples which were analysed for total protein in the application for provisional exemption. Some additional information for these samples is now provided in Appendix 5. However, the level of detail is not as complete as the samples for which analytical data were collected using improved methodology as detailed in Appendices 6a+6b. In this case, specific requests were made for sample suppliers to provide as much background detail as possible.
Whilst all bottled samples analysed in this section were analysed at the alcohol strength at which the product is sold to the consumer, production samples were often analysed at elevated alcohol strengths. Consequently, many production distillate samples would have been reduced with water prior to being sold, and thus any limit of detection applied to a production sample, when adjusted to reflect the final product, will be improved by this level of dilution. (A typical bottling strength for a spirit is 40% alcohol by volume.)
Appendix 8.3 *Materials and Method – Total Protein, Provisional Exemption*
A diagnostic kit for the determination of total protein was supplied by Randox. The kit consisted of the following reagents;
Page 88 of 173
Solution 1 Colour Reagent Pyrogallol red 60 µmol/l Disodium molybdate 40 µmol/l Succinic acid 150 mmol/l Sodium oxalate 1 mmol/l Sodium benzoate 3 mmol/l
Solution 2 Protein standard Human Serum Albumin 1000 mg/l
Sample Preparation
All spirit samples were concentrated by a factor of 10 by evaporation of 10 ml of sample in a Genevac SF60. The resultant residue was reconstituted in 1 ml ultra-high quality (UHQ) water.
Samples, standards and blanks were prepared as follows;
Reagent blank: 50 µl UHQ H2O + 3.0 ml Colour Reagent Samples: 50 µl Spirit sample + 3.0 ml Colour Reagent Standards: 50 µl Protein standard + 3.0 ml Colour Reagent
All of the solutions were vortex mixed and incubated at 25ºC for 10 minutes. Resultant absorbances were measured against the reagent blank at 600 nm using a Cecil 3000 spectrophotometer.
Protein Calibration Curve
A series of dilutions was prepared to assess the linearity and sensitivity of the method. The 1000 mg/l protein standard was diluted with high purity water to 25, 50, 75 and 100 mg/l. The calibration curve showed good linearity over the range 0 – 100 mg/l.
Recovery Experiment to Determine a Limit of Detection for Protein in Spirit Samples
Spirit samples underwent a tenfold concentration step during sample preparation. Recovery experiments were carried out to determine the detection limit for protein in the original sample matrices, given the preparation procedure. Duplicate spiked samples were prepared from a 1000 mg/l protein standard as shown below in Table 1 of this Appendix.
Appendix 8 Table 1: Preparation of Spiked Samples
Standard High Purity H20 Alcohol * Final Concentration (mg/l) 50µl 950µl 9ml 5 100µl 900µl 9ml 10 200µl 800µl 9ml 20
* Each spike was made up in 100% ethanol, gin, vodka and whisky.
Page 89 of 173
The results of the recovery of the spiked samples are shown in Table 2 of this Appendix.
As can be seen, recovery of protein is greater than 80% for most sample matrices down to the 5 mg/l level. Protein was not detected in samples prepared from the gin samples at 5 mg/l, suggesting some interaction between the protein and the sample matrix. Protein was successfully recovered in the gin samples at 20 mg/l, and at a detectable level at 10 mg/l.
Appendix 8 Table 2: Recovery of Spiked Protein from Ethanol, Gin, Vodka and Whisky Matrices
Spike Concentration (mg/l) % Recovery 5 100.9 5 100.9 10 91.3 Ethanol 10 91.3 20 94.1 20 96.7
5 ND 5 ND 10 65.8 Gin 10 65.8 20 86.5 20 91.6
5 80.5 5 90.7 10 96.4 Vodka 10 91.3 20 96.7 20 101.8
5 80.5 5 90.7 10 86.2 Whisky 10 76.0 20 94.1 20 96.7
The results of the recovery experiment established a limit of detection of 5 mg/l for protein in ethanol, vodka and whisky samples. The limit of detection established for protein in gin was 10 mg/l.
Appendix 8.4 *Results and Discussion – Total Protein, Provision Exemption*
Analysis of all spirit samples gave absorbance readings which were within the 99% confidence interval (standard deviation x 2.58) of the reagent blank.
Levels of measured protein are shown in Table 3 of this Appendix.
Page 90 of 173
Appendix 8 Table 3: Protein in Samples of Ethyl Alcohol of Agricultural Origin, New Make Whisky, Gin and Vodka.
Sample Type and Number. Description Protein Concentration (mg/l) Vodka (cereal derived) Bottled Products S04-0710 Altai Vodka ND S04-0711 Wodka Wyborowa Pure Rye Grain ND S04-0717 Smirnoff Triple Distilled Vodka ND Altia Corporation – Rajamaki Plant – Dry S04-1065 ND Vodka
Gin (cereal derived) Bottled Products S04-0716 Tanqueray Special Dry Gin ND S04-0727 Bombay Sapphire Dry Gin ND S04-0829 Grants Gin ND S04-0870 Greenalls Gin Finished Product ND S04-1047 Hendricks Gin ND
Bourbon/ Canadian Whisky/Other Bottled Products Whisk(e)y Products S04-0973 Echter Nordhaufer Doppelkorn ND
Neutral Spirit Production Distillate Samples S04-0827 Girvan Distillery Grain Neutral Spirit ND S04-0872 Greenalls Grain Spirit ND S04-0900 Grain Neutral Spirit ex Velva (French) ND S04-0901 Grain Neutral Spirit ex Velva (French) ND S04-1034 Sedalcol Grain Neutral Alcohol 23/03/04 ND S04-1035 Sedalcol Grain Neutral Alcohol 07/06/04 ND S04-1036 Sedalcol Grain Neutral Alcohol 11/05/04 ND S04-1038 Sedalcol Grain Neutral Alcohol 14/04/04 ND Altia Corporation – Koskenkorva Plant – S04-1064 ND Grain Ethanol Cameronbridge Scotch Grain Neutral S04-1066 ND Spirit
Grain New Make Spirit Production Distillate Samples (Scotch/Irish) S04-0826 Girvan Scotch Grain New Make Spirit ND S04-0874 Middleton Irish Grain New Make Spirit ND Cameronbridge Scotch Grain New Make S04-1067 ND Spirit S04-1076 Strathclyde Scotch Grain New Make Spirit ND
Malt New Make Spirit Production Distillate Samples (Scotch/Irish) S04-0869 Aultmore Scotch Malt New Make Spirit ND S04-0873 Middleton Irish Pot Still New Make Spirit ND S04-0903 Dalwhinnie Scotch Malt New Make Spirit ND S04-0904 Lagavulin Scotch Malt New Make Spirit ND S04-0906 Talisker Scotch Malt New Make Spirit ND S04-0907 Caol Ila Scotch Malt New Make Spirit ND S04-1071 Tormore Scotch Malt New Make Spirit ND S04-1072 Laphroaig Scotch Malt New Make Spirit ND Glentauchers Scotch Malt New Make S04-1073 ND Spirit S04-1074 Miltonduff Scotch Malt New Make Spirit ND
Page 91 of 173
Sample Type and Number. Description Protein Concentration (mg/l) S04-1075 Glenburgie Scotch Malt New Make Spirit ND
Gin (cereal derived) Production Distillate Samples S04-0712 Compounded Gin 193 ND S04-0713 Gin Distillate (LGS) ND S04-0871 Greenalls Gin Concentrate ND S04-0893 Distilled London Dry Gin Concentrate ND
ND = not detected. Limit of Detection (LOD) = 5 mg/l for all samples except the gin samples, for which the limit of detection is 10 mg/l.
Appendix 8.5 *Analytical Conclusions – Total Protein, Provisional Exemption*
The results show that no protein was detected in all the samples tested.
Page 92 of 173
Appendix 9 Development and Validation Work to Improve the Sensitivity of the Total Protein Method
Appendix 9.1 Introduction
In the application for provisional exemption, Dr. Philip Meaden and Dr. Derek Jamieson of Heriot Watt University put forward and discussed a number of methods for measuring total proteins. These varied in terms of suggested detection limits and protein specificity. The Bradford (Bradford, 1976), Lowry (Lowry et al., 1951) and BCA (Smith et al., 1985) methods, which are all colorimetric methods, claim detection limits close to 1 mg/l, which were better than the 5-10 mg/l level of sensitivity provided by the Randox method described in Section 6.1. It was suggested, however, that the NanoOrange® fluorescent technique (Jones et al., 2003), be trialled initially since this method claimed lower detection limits. It was noted in the discussion of the relative merits of these techniques that no published literature was known that describes an attempt to measure the content of protein in any spirit. The following five sections detail experimentation to develop an improved analysis technique for the determination of total protein content in a range of spirit samples. This discussion highlights a number of the problems experienced with the methods available and the difficulties of applying them to the matrices under consideration.
The total protein analysis method was developed so that it could be applied to distillates made using cereals, nuts or whey as raw materials prior to distillation. Specific application to nut or whey distillates can be found in other applications for permanent labelling exemptions, although references to non-cereal derived distillates will be found in this Appendix.
The goal of the method development was to identify the best method available for measuring total protein in spirit matrices and validate the protocol which would be used in the extended survey of production distillate samples and bottled products.
Appendix 9.2 Method 1 – Total Protein Analysis Using NanoOrange®
Principle
NanoOrange® reagent is non-fluorescent in aqueous solution. However upon reaction with proteins in the diluent provided, it becomes strongly fluorescent with a broad excitation around 470 nm and emission centred about 570 nm. Technical information provided with the kit claimed accurate detection of BSA between 10 μg/l–10 mg/l, and low protein to protein variability. For more information consult NanoOrange® Protein Quantitation Kit (N-6666) – Product Information Sheet (Molecular Probes, 2001).
Method Development
Initial investigations involved evaluating the sensitivity and linearity achievable with the kit. Linearity was found to be good from 1-6 mg/l though fluctuations in the
Page 93 of 173
lower region of the curve were observed (0.01–1 mg/l). A linear working range of 0.1-3 mg/l was found to be readily repeatable (Appendix 9 Figure 1).
3.50
3.00
2.50 2.00
1.50
1.00 Fluorescnece 0.50
0.00 0 0.5 1 1.5 2 2.5 3 -0.50 Concentration (ug/ml)
Appendix 9 Figure 1 Typical BSA calibration curve
The linearities of gluten, almond and whey protein calibrations were determined and compared to the BSA calibration (Appendix 9 Figure 2). As can be seen, there is some protein to protein variation, though calibration is generally linear. Gluten response is low, the likely cause of this being its poor solubility in the aqueous NanoOrange® regent. For the remainder of the validation exercise, BSA was employed as the protein calibration standard.
1.800 1.600 1.400 1.200 BSA 1.000 Almond 0.800 Gluten
fluoresence 0.600 Whey 0.400 0.200 0.000 0 0.2 0.4 0.6 0.8 1 conc (mg/l)
Appendix 9 Figure 2 Almond, Gluten and Whey Calibration
The NanoOrange® kit recommended that the sample should constitute no more than 4% of the total reaction volume, thus effectively diluting samples 25 times. This would have dramatically reduced sensitivity, and significantly raised the Limit of Detection. In order to avoid this, it was decided to evaporate samples to dryness using a Genevac SF60. In order to validate the efficiency of this stage, BSA spikes were prepared in ethanol/water solutions, evaporated in the Genevac to dryness and subsequently reconstituted in the NanoOrange® reagent. The results revealed background fluorescence, caused either from external contamination or from the ethanol itself. Further investigation of ethanol and water blanks prepared using the Genevac evaporation stage (using a variety of purity grades) indicated interferences resulting from the evaporation of ethanol and water.
A selection of samples was also screened at this stage, with some giving fluorescence values above those observed for ethanol blanks. Given the positive responses from
Page 94 of 173
the ethanol and water ‘blanks’, assigning such values to the measurement of protein was unjustified. It was also evident that matured samples could not be analysed by fluorimetry as the colour present affected readings.
A number of measures were taken to try and improve the method by reducing the effect of fluorescence from the preparation of ethanol and ultra-high quality (UHQ) water blanks and remove ‘false positive’ results. Such measures included:
Replacement of bandpass filters (excitation 380-500nm; emission 480-580nm) with interference filters (470nm and 568nm) to improve specificity and eliminate the possibility of fluorescence from non protein material. Some improvement was shown but this was not significant.
Investigation of ‘auto fluorescence’ in samples by analysis without the addition of the NanoOrange® reagent. Some evidence was gathered as to its occurrence in heavily peated samples.
Preparation of BSA calibration standard solutions in ethanol, followed by a 1:25 dilution in the NanoOrange® reagent and preparation of samples in the same manner.
Whilst the latter measure effectively reduced sensitivity by a factor of 25, it meant that standards and samples were prepared in the same manner. However, analysis of a range of blanks (ethanol and water) still showed unacceptable variations in response, which when magnified by a factor of 25 meant that sensitivity was severely compromised. Once again, it was not possible to determine whether analysed samples were producing actual or false positive results (see Appendix 9 Figure 3 for positive results from ‘blank’ ethanol samples as well as a range of values for actual samples).
1
0.8
0.6
0.4
fluorescence 0.2
0 ater ore hiskey hiskey Caol Ila Caol
-0.2 Talisker Torm Lagavulin Laphroaig UHQ W l QC check Dalwhinnie Grain Spirit Grain ot W Strathclyde eronbridge Port Dundas Cam 1ug/m New P Analar Ethanol 60% New Grai n W Analar Ethanol 100% Appendix 9 Sample Analysis (1:25 Dilution) against BSA in Ethanol Calibration (duplicates). Figure 3 Fluorescence is equivalent to concentration in mg/l.
Page 95 of 173
Appendix 9 Figure 3 includes a number of malt new make samples (Dalwhinnie, Lagavulin, Talisker, Caol Ila, Tormore, Laphroaig). These samples showed high fluorescence, ignoring the 25 times dilution factor. It was hypothesised at this stage that other constituents such as fats, phenols and copper may have been causing interferences with the assay. Samples such as the grain new make spirits, which contain significantly reduced levels of these constituents, generally did not give positive results, although this did not explain all trial sample results. It was also noted that alcohol strength affected the fluorescence reading, with 100% Analar ethanol blanks giving higher readings that 60% ethanol blanks.
At this stage it was apparent that neither the Genevac evaporation stage, nor analysis after a 1:25 sample dilution, were going to be suitable for use with this method. Variability and positive results from sample blanks cast doubt upon the validity of results from actual samples.
In order to try and reduce or remove interferences (in the form of ethanol, copper, fats), it was decided to look at the application of protein concentrators (2kDa MWCO Hydrosart®, 2ml, VivaScience, Sartorius, UK). These devices are disposable ultrafiltration devices for the concentration of biological samples. The Hydrosart membrane was chosen since it is extremely hydrophilic, making it non-protein binding. The very low molecular weight cut off (MWCO) was chosen to ensure retention of all protein material. The instructions for use indicate that for maximum recovery an MWCO at least 50% smaller than the molecular size of interest should be chosen. For more details refer to the Vivaspin 500 μl and 2 ml “Technical data and operating instructions” issued by VivaScience.
A selection of samples which had proved problematic in previous assays were analysed along with BSA recovery spikes. Results from the analysis of the concentrates can be seen in Appendix 9 Figure 4.
1.60 1.40 1.20 1.00 0.80 0.60
fluorescence 0.40 0.20 0.00 r A A H e lin a in S S tO k u k G B B is v od /l /l E al a ts g g % T g V n m m 0 a ey ra 1 6 L l G .6 ar 0 B
Appendix 9 Results from analysis of Vivaspin concentrated samples (duplicate analyses). Figure 4 Fluorescence is equivalent to concentration in mg/l
Whilst recovery for standards proved promising, results for blanks (and samples) still gave elevated results. On analysis of the filtrate which had passed through the concentrators, it was found that these too gave ‘positive’ results using the assay. Since this fluorescence was found in both the concentrate and filtrate of 60% ethanol,
Page 96 of 173
this was assumed to be due to compounds leached from the plastic concentrator by the ethanol. Investigation demonstrated an increase in response of ethanol blanks evaporated by the Genevac in plastic tubes compared to glass ones. Furthermore, on transfer of the blanks (glass) to a plastic tube and heating in the same manner as required in one of the steps in the assay, a marked increase in fluorescence was seen. Thus not only was the fluorescence reading affected by the use of plastic materials during preparation, but the response is accelerated by the application of heat.
The final stage of the development involved carrying out the assay completely in glass. However, once again, results obtained remained high for ethanol blanks and a number of samples.
Conclusions
A number of factors contributed to the problems encountered in measuring total protein in spirit samples by the NanoOrange® fluorimetric technique;
Interference from colour in mature spirit samples. Background fluorescence in ethanol/water blanks. Probable leaching of materials (plasticisers?) causing an increase in fluorescence, an effect that was enhanced on heating and which meant that plastic protein concentrators were unsuitable. Possible interferences of fats, copper and phenols, thus being unsuitable for a number of malt new make spirits.
The very nature of the alcoholic samples to be analysed rendered them unsuitable for this fluorimetric technique.
Appendix 9.3 Method 2 – Total Protein Analysis Using CBQCATM
An alternative method to the NanoOrange® was tested to determine whether any improvements could be made. This was the analysis of proteins using the CBQCATM Protein Quantification Kit from Molecular probes. Like the NanoOrange® analysis method it is based on the measurement of fluorescence of the reagents in the presence of proteins.
Principle
Molecular Probes CBQCA Protein Quantitation Kit utilises the ATTO-TAG CBQCA reagent (3-(4-carboxybenzoyl)quinoline-2-carboxaldehyde) originally developed as a chromatographic derivatisation reagent for amines. The ATTO-TAG CBQCA reagent is virtually non-fluorescent in aqueous solution; however, in the presence of cyanide, it reacts with primary amines such as those found in proteins to form highly fluorescent derivatives. For more information, please consult the CBQCA Protein Quantitation Kit (C-6667) – Product Information Sheet (Molecular Probes, 2001).
Method Development
Compared to the NanoOrange® analysis the CBQCA kit underwent limited method development. The use of a Genevac evaporation step and other developments may
Page 97 of 173
have seen a more successful analysis method emerge. However, a number of factors weighed against development. Initial results established that the method did not provide enhanced sensitivity over the widely reported colorimetric methods, and the variability in fluorescent measurements reduced confidence in the results.
Initial tests demonstrated that the limit of detection for BSA protein in water was approximately 0.5 mg/l, although full validation of this figure was not carried out. This was not dissimilar to the best sensitivities indicated in the literature by the commonly used Lowry and Bradford methods. To achieve lower levels of sensitivity, the method indicated that silanised glassware was required, which was not felt practicable for a desired routine analytical method. The method indicated that it was also possible to increase sensitivity by increasing the concentration of the CBQCA reagent used in each analysis. However, using a macro assay in a standard fluorimeter meant that to achieve the level of detection achieved a high concentration of the CBQCA reagent was required thus making analysis expensive.
As with the NanoOrange® method, fluorescence readings were variable with the CBQCATM method. It had also been hoped that the method would be compatible with the alcoholic sample matrix. However, variations in alcohol strength affected absorbance readings and precipitation of reagents could lead to unexpected results.
Conclusions
Whilst the CBQCATM kit could probably be developed to suit the measurement of protein in alcoholic distillates, by the removal of ethanol and the use of a microplate reader, the indicated benefits did not justify the development time.
Appendix 9.4 Method 3 – Total Protein Analysis Using Bradford MicroAssay Protocol Principle
As indicated in the introduction, most of the colorimetric methods for protein analysis were quoted as having similar limits of detection. The initial choice of the Bradford assay (Bradford, 1976) was based on its increased robustness with respect to chemical interferences, which seemed to cause problems for the fluorimetric analysis methods. The Protein Assay kit was purchased from Bio-Rad Laboratories, UK (Protein Kit II). The Bio-Rad Protein Assay is based on the Bradford method, in which a differential colour change of a dye occurs in response to various concentrations of protein. The absorbance maximum for an acidic solution of Coomassie® Brilliant Blue G-250 dye shifts from 465 nm to 595 nm when binding to protein occurs. The Coomassie blue dye binds to primarily basic and aromatic amino acid residues, especially arginine. For more information, refer to the Bio-Rad Protein Assay Product Information Booklet.
Method Development
Initial tests demonstrated the linearity of protein measurement at various concentrations (Appendix 9 Figure 5 and Appendix 9 Figures 6). Good linearity was shown down to a concentration on 0.6 mg/l using the BSA calibration standard
Page 98 of 173
included in the test kit. Deviations from linearity were noted below this concentration.
BSA Standard Curve for the Bio-Rad Protein Micro Assay
0.70
0.60 y = 0.0465x + 0.0061
0.50
0.40
O.D.595nm 0.30
0.20
0.10
0.00 02468101214 Protein (ug/ml)
Appendix 9 BSA at high and low concentrations (repeat readings, all taken within timescale Figure 5 indicated acceptable by the method)
BSA Standard Curve for the Bio-Rad Protein Micro Assay
0.07
0.06
0.05
0.04
0.03 O.D.595nm
0.02
0.01
0.00 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
-0.01 Protein (ug/ml)
Appendix 9 BSA at low concentration (repeat readings, all taken within timescale indicated Figure 6 acceptable by the method)
An initial indication of 0.6 mg/l for a limit of detection looked promising and certainly comparable with the CBQCA methodology
Previously reported work has indicated the relative concentrations of amino acid content in various wheat proteins compared to BSA (Eynard et al, 1994) when measured using the Bradford method. On this basis estimates can be made between
Page 99 of 173
responses for BSA and those for typical wheat proteins. According to Eynard et al (1994) the colour developed in the method best complied with the sum of all the Coomassie Blue-reactive amino acids or with the sum of the lysine and arginine residues. Taking the Coomassie Blue-reactive amino acids the ratio of responses expected for BSA:albumin, BSA:gliadin and BSA:glutenin are 1.8, 2.7 and 2.2 respectively. To account for all wheat proteins a conversion factor of 2.2 was assumed so that BSA results could be equated to wheat proteins.
As with all tests conducted, the presence of alcohol interfered with the analysis. This was noted for any sample containing over 2% alcohol by volume. Thus ethanol had to be removed by means of evaporating samples to dryness. This was carried out using a Genevac SF60. To improve the analysis, a concentration step was introduced. 9.6 ml of sample was evaporated to dryness and then made up in 2.4 ml of water.
A number of recovery experiments were conducted. BSA was spiked into various cereal derived alcoholic beverage matrices which were then evaporated to dryness. Analysis of the sample having been reconstituted in UHQ water indicated the level of recovery. It was soon determined that the level of recovery was dependent on the spirit type and its alcohol strength. The results for a range of spirit matrices can be seen in Table 1 of this Appendix.
Vodka/Neutral Spirit Recovery Data
BSA was spiked into a vodka at a range of concentrations from 0.2 mg/l to 10 mg/l. Recovery values can be seen in Table 1 of this Appendix. Recoveries varying between 65 and 105% could be determined for all concentrations, so long as the strength was at approximately 40% alcohol volume or lower. At the lower spike concentrations, increased alcohol strengths resulted in poor recoveries.
Gin Recovery Data
BSA was spiked into a gin at a range of concentrations from 0.2 mg/l to 10 mg/l Initial analysis of gin recovery data at approximately 40% alcohol by volume yielded poor results, particularly at low spike concentrations. Reducing the strength of the matrix to 20% alcohol by volume improved the recoveries observed. Recoveries ranging from 56% to 94% were observed at levels of 0.5 mg/l and above. Experimentation with a range of gin brands at spike levels of 2.0 and 5.0 mg/l showed very reproducible recovery factors.
Whisky Recovery Data
BSA was spiked into a whisky grain spirit at a range of concentrations from 0.2 mg/l to 1 mg/l. Initial analysis of whisky recovery data at a sample strength of 68.5% gave no measurable recovery at spiking concentration less than 5 mg/l. By reducing the alcohol strength better recoveries could be obtained. Good recoveries are observed for the whisky matrix at a spiking concentration of 0.5 mg/l or greater, so long as the alcohol concentration is 40% or lower.
Page 100 of 173
Appendix 9 Table 1: Recovery of Spiked Protein from Ethanol, Gin, Vodka and Whisky Matrices
Spike Alcohol Concentration Strength (mg/l) % Recovery 37.5 0.2 65 36.9 0.2 105 37.5 0.5 76 37.5 0.5 92 37.5 0.5 105 37.5 0.5 72 36.1 0.5 88 Vodka/Neutral Spirit 37.5 1.0 81 34.4 1.0 94 37.5 2.0 75 32.0 2.0 78 37.5 5.0 90 23.4 5.0 81 37.5 10.0 75 9.4 10.0 71 20 0.2 19 20 0.5 56 20 0.5 73 20 0.5 94 20 0.5 90 Gin 20 0.5 60 20 0.5 67 20 0.5 66 20 1.0 70 20 2.0 93 20 5.0 87 20 0.2 37 34.3 0.2 82 20 0.5 66 34.3 0.5 63 Whisky 40 0.5 64 20 1.0 75 34.3 1.0 70 40 1.0 68 Italicised figures were calculated using absorbances at 595 nm which fell within 3 standard deviations of the blank
From an analysis of the standard deviation of a number of blank samples (UHQ water and Analar ethanol), it was evident that absorbance values for protein spikes at a concentration of 0.2 mg/l typically fell within the variability of the blanks (average ± 3 the standard deviation). All absorbance values (at 595 nm) for 0.5 mg/l protein spikes, however, corresponded to protein concentrations greater than this limit and could be used to safely calculate recovery values.
Conclusions
For all cereal based products, suitable recoveries (between 56 and 104%) could be obtained for vodka/neutral alcohol/gin and whisky cereal based spirits, at 0.5 mg/l.
Page 101 of 173
Absorbance values obtained were all greater than that corresponding to the upper three standard deviation limit defined for the measurement of the blanks.
For vodka and whisky matrices, measurement could take place at the typical bottling strengths at which these products are sold (40% alcohol by volume). Hence, BSA protein present at 0.5 mg/l can reproducibly be determined in these spirit matrices using this method. Since gin is typically sold at around twice the alcoholic strength (approximately 40%) at which measurement takes place in this method, the limit of detection above which BSA protein can reproducibly be determined at typical bottling strength is 1.0 mg/l.
These limits of detection, 0.5 mg/l for whisky/vodka/neutral spirit and 1.0 mg/l for gin, correspond to a 10-fold increase in sensitivity compared to the previously reported Randox method. Only non-matured samples can be analysed using this technique.
One drawback to the methodology was noted, but this was considered to be a problem for all fluorimetric/colorimetric methods which are affected by the presence of alcohol and require its removal prior to analysis. Some cereal based spirits contain levels of long chain fatty acid esters (ethyl hexadecanoate and ethyl hexadecenoate for example) which are only sparingly soluble in aqueous solution. In the case of Scotch malt new make spirits it was often (although not always) observed that residual precipitates after evaporation could produce a haze upon reconstitution in UHQ water. Sometimes these could be coloured by the presence of copper. In such circumstances the assay method returned values indicative of the presence of protein. Further development could probably address the issue of this residue, but it was felt that the Bradford method, as established, provided a suitable method for screening spirits for the presence of protein. For certain matrices, the presence of a positive result would have to be investigated further.
Appendix 9.5 Method 4 - Total Protein Analysis Using Mass Spectrometry
Two alternative methods to the fluorimetric and colorimetric examples given above were considered for the analysis of protein in spirit samples. The first of these was the use of Mass Spectrometry. It was felt that the development of a method to detect a range of proteins in a variety of spirit matrices, whilst possible, would be a considerable analytical challenge. An exploratory experiment was conducted to determine whether the application of MALDI-TOF would prove a simple indicator of the absence of protein. A number of model alcoholic solutions (60% alcohol by volume) spiked individually with almond (1.4 mg/l), whey (1 mg/l) and gluten (1 mg/l) proteins were submitted for protein analysis by MALDI-TOF. MALDI-TOF is an acronym for Matrix-Assisted Laser Desorption/Ionisation – Time of Flight. Proteins are desorbed by a laser, and then separated according to their masses. The initial work failed to observe the presence of any protein (20-100 kDa) in the gluten, whey or almond samples. Whilst this approach would have benefited from further development, an established method which it would be easy to adapt to alcoholic matrices was required, given time constraints. The tested technology did not provide the required sensitivity for relevant proteins in the matrices tested.
Page 102 of 173
Appendix 9.6 Method 5 – Total Protein Analysis Using AAA-DirectTM
Principle
AAA-Direct™ is a liquid chromatographic method with integrated pulsed amperometry which analyses the amino acid content of a sample, following hydrolysis, to evaluate total protein content. One of the advantages of the method over the popular UV absorbance or dye-binding assays as discussed previously, is that is much less susceptible to the interferences already indicated (ethanol, fats, phenols) which can adversely affect these assay techniques.
The principle of AAA-Direct™ is to hydrolyse any proteins/peptides of interest to liberate their free amino acids. These are then separated using high-performance anion exchange chromatography and analysed and integrated with pulsed amperometric detection (HPAE-IPAD). The protein or peptide concentration is calculated from the sum of the individual amino acids (Dionex Corporation, 2004 and Hanko and Rohrer, 2005).
Method and Results
A contract laboratory, Intertek ASG, experienced in the analysis of proteins and peptides using the AAA-Direct™ method was asked to carry out a feasibility study on the application of this technique to spirit sample matrices. Four sample matrices were investigated: 2 new make Scotch malt whisky samples (which contained interferences detrimental to the Bradford based method), a gin sample and a matured Scotch whisky. Since colour is not an issue for this type of analysis the possibility of analysing protein in mature samples merited investigation. The feasibility study conducted by Intertek is reproduced at the end of this Appendix.
The feasibility study indicated the following:
Apart from the matured whisky sample, low levels of ‘total protein’ were detected in the spirits. Results obtained were similar to the water blanks indicating that the amino acids detected are from contamination rather than protein present in the spirits. No levels of proline or glutamate can be detected. These are the principal wheat protein amino acids, making up approximately 20%-55% of the amino acid content (Wrigley and Bietz, 1998). Similarly, no glycine and alanine can be detected: glutamate, glycine and alanine are the principal protein amino acids (approximately 30 mol %) in barley (Shewry, 1993). These absences confirm that no proteins derived from the raw materials prior to distillation are contained in the analysed distillates. Since levels of histidine and tyrosine have been detected at levels of 0.05 to 0.1 mg/l, this supports the absence of wheat proteins in the distillates, certainly down to a limit of 0.5 mg/l. A large component detected in the matured whisky is due to either free arginine present in the sample or another component that co-elutes with arginine and not from any protein present. Analysis of another matured spirit demonstrated the same intense peak. The fact that the peak is reduced in size following hydrolysis indicates the presence of an interference rather that the presence of arginine.
Page 103 of 173
Conclusions
On the basis of the results obtained from the feasibility study, the AAA-DirectTM method is applicable to alcoholic distillates. An estimated limit of detection for the analysis is 0.5 mg/l protein, but this would be revised in any full scale analysis. Trace levels of amino acids are detectable in the hydrolysed samples. However, once these are adjusted for with a correction for apparent free amino acids in samples measured without hydrolysis, residual values cannot be associated with the presence of any cereal protein.
The presence of a particularly large peak assigned to arginine in one sample is probably the result of an interference resulting from maturation. Whilst the AAA- DirectTM method can cope with the interferences in the Scotch new make spirit samples, the further complexity of mature spirits presents difficulties. This was not a major issue for the analysis, since the exemption is being sought for the distillates, and if it can be demonstrated that these ingredients are free from cereal proteins, they will not be introduced as a result of maturation in wood.
The AAA-DirectTM method complimented the simpler Bradford method described in Section 9.3 within this Appendix. The Bradford method could act as a screening technique. Those samples for which a positive result was determined from the Bradford method could be analysed by the AAA-DirectTM method if it was suspected that the positive results was due to the presence of interferences.
The pages following the reference section in this Appendix reproduce the reports from Intertek TSG which provide information on the feasibility study to determine whether alcohol distillates could be analysed for total protein content using the AAA-DirectTM method.
References
Bio-Rad Laboratories Life Science Group. Bio-Rad Protein Assay.
Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72:248-254
Dionex Corporation (2004) Determination of Protein Concentrations Using AAA- DirectTM, Application Note 163
Eynard, L. (1994) Determination of Wheat Proteins in Solutions by Dye Binding, Cereal Chemistry 71(5):434-438.
Hanko, V.P. Rohrer, J S (2005) Determination of protein and peptide concentrations by amino acid analysis using anion-exchange chromatography with electrochemical detection. LC GC North America, 23(6) SUPP:20
Jones, L.J., Haugland , R.P., & Singer V.L. (2003) Development and characterization of the NanoOrange® protein quantitation assay: a fluorescence-based assay of proteins in solution. BioTechniques 34:850-861
Page 104 of 173
Lowry, O.H., Rosebrough, N.J., Farr, A.L., & Randall, R.J (1951) Protein measurement with the Folin-phenol reagent. Journal of Biological Chemistry 193:265-275
Molecular Probes (2001) NanoOrange® Protein Quantitation Kit (N-6666) Product information. 22 October 2001:1-4
Molecular Probes (2001) CBQCA Protein Quantitation Kit (C-6667). 11 October 2001 :1-4
Shewry P.R. (1993) Barley seed proteins. In: Barley: Chemistry and Technology. Edited by MacGregor, A.W., American Society of Cereal Chemists, St Paul Minnesota, USA, pp. 131-197
Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J., & Klenk, D.C. (1985) Measurement of protein using bicinchoninic acid. Analytical Biochemistry 150, 76-85
Vivaspin 500 μl and 2 ml “Technical data and operating instructions” issued by VivaScience.
Wrigley C.W. and Bietz J.A. (1998) Proteins and Amino Acids. In: Wheat: Chemistry and Technology. Edited by Pomeranz, Y., American Association of Cereal Chemists, St Paul, Minnesota, USA, Volume 1, pp. 159-275
Page 105 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 1 of 14
STUDY TITLE
TOTAL PROTEIN CONTENT OF SPIRITS: FEASIBILITY STUDY
AUTHOR
D G Williams
COMPLETED
January 2006
TEST FACILITY
Intertek ASG P.O. Box 42 Hexagon House Blackley Manchester M9 8ZS UK
Page 106 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 2 of 14
CUSTOMER
Craig Owen The Scotch Whisky Research Institute The Robertson Trust Building Research Park North Riccarton Edinburgh EH14 4AP. UK
INTERTEK ASG PERSONNEL
The following personnel were involved in the study.
D G Williams, Senior Technologist, Technology Group. A Brotherhood, Senior Experimental Chemist, Technology Group. I Bartkowiak, Senior Experimental Chemist, Technology Group.
Page 107 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 3 of 14
CONTENTS
PAGE NUMBER SECTION DESCRIPTION
1 Study Title 2 Customer. 2 Intertek ASG Personnel. 3 Contents. 4 Samples Involved. 4 Introduction. 5 Methodology. 5 Sample Preparation. 5 Acid Hydrolysis for Total Amino Acids. 5 Alkaline hydrolysis for Tryptophan. 5 No Hydrolysis for Free Amino Acid Content. 6-7 Chromatographic Conditions for Total and Free Amino Acids. 7-8 Chromatographic Conditions for Tryptophan Content. 9 Calibration Standards. 9 Results. 9 Recovery Experiments. 10 Total Protein Content after Acid Hydrolysis. 11 Figure 1. 5ppm Mixed Amino Acid Standard. 12 Figure 2. Laphroaig Hydrolysed Sample. 13 Discussion of Results. 13 Free Amino Acid Content. 14 Determination of Tryptophan Content. 14 Investigation of Source of ‘Arginine’ Contamination.
Page 108 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 4 of 14
SAMPLES INVOLVED
SAMPLE REFERENCE ASG NUMBER
JOHNNIE WALKER BLACK S04-0714 10376639 LABEL
LAPHROAIG NEW MAKE S04-1072 10376641
TANQUERAY GIN S04-0716 10376642
TORMORE NEW MAKE S04-1071 10376643
MACALLAN NEW MAKE SPIRIT S05-3237 10377270
MACALLAN 1992 99 2552 10377271
CARAMEL COLOR 01-1201 10377272
INTRODUCTION
A feasibility study was carried out to check if the total protein content of spirits could be determined by means of the acid hydrolysis of any protein present in the spirits to form the constituent amino acids. A comparison was made of samples both before and after hydrolysis to obtain both the total and free amino acid content of the spirits. Alkaline hydrolysis tests were also undertaken to check for the presence of any tryptophan which is decomposed under the acid hydrolysis conditions. The amino acid content is determined by AAA-Direct (amino acid analysis direct) using ion chromatography. This is a highly sensitive method of detection of amino acids which uses integrated pulsed amperometry (IPAD), a type of electrochemical detection, to directly detect amino acids. AAA-Direct also provides separation of all common amino acids using anion exchange chromatography.
Page 109 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 5 of 14
METHODOLOGY
Sample Preparation
Acid Hydrolysis for Total Amino Acids
1.0ml of sample was pipetted into an Eppendorf vial (1.5ml) and evaporated to dryness using a SpeedVac Evaporator system at 45°C. The residues were dissolved in 150µl of 6N hydrochloric acid (HCl) and transferred to a vacuum hydrolysis tube. Hydrolysis tubes were evacuated and the headspace replaced with argon. The tubes were sealed and placed in a Reacti-Therm heating module for 16 hours at 105°C. After allowing to cool to ambient temperature the HCl was transferred to an Eppendorf vial and evaporated to dryness using the SpeedVac Evaporator system at 45°C. The residues were then dissolved in 1.0ml of deionised water. Blank tests were prepared in the same way using 1.0ml of deionised water. Sample and blank tests were then examined by ion chromatography using the conditions shown.
Alkaline Hydrolysis for Tryptophan
1.0ml of sample was pipetted into an Eppendorf vial (1.5ml) and evaporated to dryness using a SpeedVac Evaporator system at 45°C. The residues were dissolved in 4.4M sodium hydroxide (NaOH) and transferred to a vacuum hydrolysis tube. Hydrolysis tubes were evacuated and the headspace replaced with argon. The tubes were sealed and placed in a Reacti-Therm heating module for 2 hours at 105°C. After allowing to cool to ambient temperature the NaOH was transferred to an autosampler vial. Blank tests were prepared in the same way using 1.0ml of deionised water. Sample and blank tests were then examined by ion chromatography using the conditions shown.
No Hydrolysis for Free Amino Acid Content
1.0ml of sample was pipetted into an Eppendorf vial (1.5ml) and evaporated to dryness using a SpeedVac Evaporator system at 45°C. The residues were then dissolved in 1.0ml of deionised water. Blank tests were prepared in the same way using 1.0ml of deionised water. Sample and blank tests were then examined by ion chromatography using the conditions shown.
Page 110 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 6 of 14
Chromatographic Conditions for Total and Free Amino Acids
Instrument: Dionex ICS 3000 Ion Chromatograph with Chromeleon Data System
Columns: Dionex AminoPac PA10 Analytical (2mm x 250mm) Dionex AminoPac PA10 Guard column (2mm x 50mm)
Detection: IPAD
Waveform for Electrochemical Detector
Time (s) Potential (volts vs pH) Integration
0.00 +0.13 0.04 +0.13 0.05 +0.28 0.11 +0.28 Begin 0.12 +0.55 0.41 +0.55 0.42 +0.28 0.56 +0.28 End 0.57 -1.67 0.58 -1.67 0.59 +0.93 0.60 +0.13
Flow: 0.25 ml/minute
Injection Volume: 10µl
Oven Temperature: 30°C
Eluents: A. Water B. 250 mM sodium hydroxide C. 1.0 M sodium acetate
Page 111 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 7 of 14
Programmed Method
Time %A %B %C Curve Comments (min)
Init. 76 24 0 - Autosampler fills sample loop 0.0 76 24 0 - Valve from load to inject 2.0 76 24 0 1 Begin hydroxide gradient 8.0 64 36 0 8 11.0 64 36 0 8 Begin acetate gradient 18.0 40 20 40 8 21.0 44 16 40 5 23.0 14 16 70 8 42.0 14 16 70 8 42.1 20 80 0 5 Column wash with hydroxide 44.1 20 80 0 5 44.2 76 24 0 5 Equilibrate to start condition 75.0 76 24 0 5 End of run
Typical System Operating Backpressure: 2300 – 3000 psi
On-Line Degas: 30s every 4 minutes
Chromatographic Conditions for Tryptophan Content
Instrument: Dionex ICS 3000 Ion Chromatograph with Chromeleon Data System
Columns: Dionex AminoPac PA10 Analytical (2mm x 250mm) Dionex AminoPac PA10 Guard column (2mm x 50mm)
Detection: IPAD
Page 112 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 8 of 14
Waveform for Electrochemical Detector
Time (s) Potential (volts vs pH) Integration
0.00 +0.13 0.04 +0.13 0.05 +0.33 0.21 +0.33 Begin 0.22 +0.60 0.46 +0.60 0.47 +0.33 0.56 +0.33 End 0.57 -1.67 0.58 -1.67 0.59 +0.93 0.60 +0.13
Flow: 0.25 ml/minute
Injection Volume: 10µl
Oven Temperature: 30°C
Eluents: A. Water B. 250 mM sodium hydroxide C. 1.0 M sodium acetate
Programmed Method
Time %A %B %C Comments (min)
Init. 0 20 80 Autosampler fills sample loop 0.0 0 20 80 Valve from load to inject 2.0 0 20 80 End of run
Typical System Operating Backpressure: 2300 – 2540 psi
On-Line Degas: 30s every 4 minutes
Page 113 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 9 of 14
Calibration Standards
The system is calibrated for 19 amino acids over the range 0 to 5µg/ml.
Amino Acids
Arginine # Lysine # Alanine # Threonine # Glycine # Valine # Hydroxyproline # Serine Proline Isoleucine Leucine # Methionine Histidine # Phenylalanine # Glutamate Aspartate Cysteine Tyrosine # Tryptophan #
RESULTS
Recovery Experiments
Recovery experiments were carried out by spiking spirit samples with 4ppm of selected amino acids (listed above #). Samples prepared in triplicate, two tests and a third to which the amino acid spike was made.
Page 114 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 10 of 14
Sample % Average Recovery
Johnnie Walker Black Label 76
Laphroaig New Make 85
Tanqueray Gin 78
Tormore New Make 58
Total Protein Content after Acid Hydrolysis
The total protein content is calculated by summing the concentration of all the amino acids detected after acid hydrolysis of the spirits. Water blanks were also processed by acid hydrolysis to determine the amino acid background level due to contamination from the equipment and reagents used. Examples of the chromatograms obtained are given in Figures 1 and 2.
Sample Total Protein (g/ml)
Water Blank 0.7
Johnnie Walker Black Label 53.4*
Laphroaig New Make 1.1
Tanqueray Gin 2.4
Tormore New Make 1.2
*Mainly due to a large component eluting where arginine is detected, present at consistent levels in all three solution preparations.
Page 115 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 11 of 14
Figure 1 5ppm Mixed Amino Acid Standard
350 Proteins 091205 #1 [modified by cmadmin, 9 peaks manually assigned] ED_1 nC
300
250 4 - Threonine 6.484
200 7 - Hydroxyproline 9.017 150 11 - Tyrosine 27.650 1 - Arginine 1.700 5 - Glycine 7.134 9 - Histidine 21.367
100 3 - Alanine 6.167 2 - Lysine 3.450 10 - Phenylalanine 22.634 6 - Valine 8.367 50 8 - Leucine 13.784 12 - Tryptophan 41.434
0
-50 min 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0
The set of unresolved peaks at around retention time 20 to 23 minutes is due to the gradient employed to elute all amino acids, hence the baseline drift. The peak at ~20 minutes is a system peak (instrument manufacturer literature confirms this peak under the same conditions). Standards were stable over the course of the analysis.
Page 116 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 12 of 14
Figure 2 Laphroaig Acid Hydrolysed Sample
70.0 Proteins 091205 #12 [modified by cmadmin, 3 peaks manually assign ED_1 nC
60.0
50.0
40.0
30.0 5 - 6.534
20.0 2 - Arginine 1.684 1 - 1.550 9 - Histidine 21.467
10.0 6 - Glycine 8.600 3 - 2.150 4 - 4.300 7 - 10.984
0.0 8 - 13.717 10 - Phenylalanine 22.600 11 - Tyrosine 27.600
-10.0
-20.0 min 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0
Page 117 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 13 of 14
Discussion of Results
Apart from the Johnnie Walker Black Label sample low levels of total protein were detected in the spirits. Results obtained are similar to the water blanks indicating that the amino acids detected are from contamination rather than protein present in the spirits. The large component detected in the Johnnie Walker Black Label is most probably due to free arginine present in the sample or another component that co- elutes with arginine and not from any protein present. To check if free amino acids are present the samples have been examined directly without hydrolysis.
Free Amino Acid Content
Sample Total Free Amino Acids (g/ml)
Water Blank 0.5, 0.5, 0.5
Johnnie Walker Black Label 325.7, 305.4
Laphroaig New Make 1.7, 1.2
Tanqueray Gin 1.1, 0.8
Tormore New Make 0.7, 0.8
These results show that by direct analysis without any hydrolysis that a large component is detected in Johnnie Walker Black label. This component is detected at the same retention time as arginine. Again only low levels of amino acids detected in the other spirits and blanks.
Page 118 of 173
Report Ref: ANB 0259/70 Submission: 1299162. Page 14 of 14
Determination of Tryptophan Content
Tryptophan is decomposed under the acid hydrolysis conditions so an alkaline hydrolysis of the spirit samples was undertaken. No tryptophan was detected in Johnnie Walker and Laphroaig whisky or Tanqueray gin. From this it was decided that the best analysis strategy for the spirit samples would be to do two tests. One acid hydrolysis for amino acids and a direct test for free amino acid content with corresponding blank tests to check for background contamination from the reagents used.
Investigation of Source of ‘Arginine’ Contamination
It was thought that the large component that was detected in the Johnnie Walker whisky was possibly from the caramel which is added as a colorant. Samples of Macallan whisky, new make spirit and caramel colorant were analysed directly and after acid hydrolysis. The Macallan new make spirit was also spiked with caramel to check if any components were detected at the same retention time as arginine. The following results were obtained.
Sample Total Protein (g/ml) Total Free Amino Acids (g/ml)
Water Blank 0.6, 0.6 0.1, 0.1
Macallan Spirit New Make 1.3 0.1
Macallan Whisky 1992 224.9 272.7
Caramel 1.1 3.4
Macallan Spirit New Make 1.3 2.7 Spiked with Caramel
A large component was detected in the Macallan whisky which eluted in the same place as arginine. Very little arginine was detected in the caramel so the source of the large component is still unknown.
Page 119 of 173
Appendix 10 Analytical Protocol for Total Protein Analysis of Alcohol Distillate Samples Using the Bradford Analysis Microassay
A diagnostic kit for the determination of total protein was supplied by Bio-Rad. The kit consisted of a dye reagent concentrate which was used without treatment for the analysis and a protein BSA standard (1.48 mg/ml) which was diluted appropriately in ultra-high quality (UHQ) water to provide calibration standards.
Sample Preparation
All samples analysed with an alcohol strength in excess of 40% alcohol by volume were adjusted to 40% by volume (typical bottling strength) with UHQ water. In some instances this may be less than the actual bottling strength (e.g. 50.5% is a common strength for bottling some bourbon varieties) but for ease of comparison these minor variations were ignored. All samples at, or slightly below 40% alcohol by volume, were not diluted. The exceptions were the gin samples which were diluted to an alcohol strength of 20% by volume. These sample strengths were shown to be the best conditions for achieving good recoveries.
All samples were concentrated by a factor of 4 by evaporation of 9.6 ml of sample in new borosilicate Pyrex (10016 mm, Fischer Scientific, UK) tubes using a Genevac SF60. The resultant residue was reconstituted in 2.4 ml UHQ water. A vigorous reconstitution process was employed as follows:
1) Add 2.4 ml of UHQ water 2) Vortex mix for approximately 10 seconds 3) Place in an ultrasonic bath for 60 minutes 4) Vortex mix for approximately 10 seconds 5) Place in a rotary bench top mixer set at 600 rpm (Status S20, Novara, Germany) for 60 minutes 6) Vortex mix for approximately 10 seconds
Following this reconstitution, 0.6 ml of dye concentrate was added to the Pyrex tubes and the solution was vortex mixed and incubated at for at least 5 minutes (no longer than 60 minutes). Resultant absorbances were measured at 595 nm using a Cecil 3000 spectrophotometer and disposable cuvettes (Plastibrand, UV-cuvette macro, 2.5 ml, Germany). Care was taken to ensure cuvettes were positioned within the sample cell in the same manner each time, assuming there were distinguishing marks to identify orientation. This removed a small variation in the absolute value of the measurements, which given the need to detect small absorbances reliably was judged sufficiently important.
Calibration Preparation
Calibration standards were prepared by diluting the BSA protein standard provided with the kit (1.48 mg/ml) to the following concentrations with UHQ Water:
10 mg/l, 5 mg/l, 1 mg/l, 0.5 mg/l, 0.2 mg/l
Page 120 of 173
Based on these values and that of the reagent blank, a quadratic calibration was constructed. All calibration standards were prepared in the same borosilicate tubes used for the purpose of evaporation of samples to dryness. Preparation in alternative plastic vials could partially bind some of the reagent dye and hence reduce the absorbance values of the calibration standards in relation to the samples. Each time a new batch of disposable cuvettes or dye was employed, a new calibration curve was constructed.
A limit of detection was constructed by measuring the absorbance variation of a number of reagent blanks and evaluating the concentration equivalent to the average absorbance value of the reagent blanks + 3 times the standard deviation. For the purpose of this experiment, reagent blanks were taken through the whole preparation procedure, as if samples. Two types of reagent blanks were employed, UHQ water, and 60% ethanol (Analar, >99.7% , BDH, UK). Over one batch of reagent dye and UV-cuvettes it was shown that the average reading for the UHQ blanks was 0.512, with a standard deviation of 0.009. For this calibration, 595 nm absorbance readings of 0.540 (average plus three standard deviations) or greater were regarded as positive results. Those results falling between 0.531 and 0.540 (two and three standard deviations) were reanalysed. The 60% ethanol blanks had a similar distribution (average = 0.511, average + 2 standard deviations = 0.528, average +3 standard deviations = 0.536).
The calibration graph associated with the set of analysis which produced the range of blanks detailed above can be seen in Figure 1 of this Appendix. From the equation for the graph it can be calculated that the cut off point below which all absorbances are treated as being due to a blank (0.540) equates to 0.22 mg/l of BSA, assuming 100% recovery. However, as shown in Appendix 9 Table 1, recoveries estimated close to this limit will be based on absorbance values below the limit of detection. The method was validated at 0.5 mg/l, where the average recovery was approximately 75%.
Using a conversion factor of 2.2 to account for an estimated difference in response between BSA and wheat proteins (see Appendix 9.4), a limit of detection of 1 mg/l or less is still achieved.
1.000 0.950 y = 0.0005x2 + 0.0388x + 0.5222 0.900 0.850 0.800 0.750 0.700 0.650 0.600 0.550 0.500 Absorbance reading at 595nm 024681012 Protein Concentration, BSA, mg/l
Appendix 10 Calibration curve for absorbance measurements (595nm) against protein concentration
Page 121 of 173
Figure 1 However, when an absorbance greater than the upper limit of the blank range is detected, the quadratic equation provides some means of estimating the perceived protein content.
Page 122 of 173
Appendix 11 Analytical Protocol and Results for the Total Protein Analysis of Alcohol Distillate Samples Using the AAA-DirectTM Method
The following report details the application of the AAA-DirectTM method to the analysis of 20 samples which gave a positive response to the Bradford Microassay protein determination.
Full details on the methodology employed for the analysis of amino acids with and without an acid hydrolysis step can be found in the feasibility study reported in Section 6 of Appendix 9. On the recommendation of the feasibility study, no base hydrolysis analysis was carried out to ensure the correct determination of tryptophan (it is degraded as a result of acid hydrolysis). Tryptophan is a minor amino acid component of barley and wheat proteins (Shewry, 1993 and Wrigley and Bietz, 1998).
The estimation of total protein is made by subtracting the free amino acid content (measured without an acid hydrolysis step) from the amino acid content following acid hydrolysis. The results demonstrate the following points:
1) Tyrosine is at a consistently low level in all the samples (at or around the reporting level of 0.1 mg/l). It was also detected in the water blanks. The presence of tyrosine can thus be assigned to an analysis artefact.
2) Aside from the very low serine peak in 10380214, which is also detected occasionally in the non-hydrolysed samples and is also probably an analytical artefact, the only amino acid detected in 19 out of the 20 samples was arginine. In the feasibility study, a peak which eluted at the same time as arginine was seen to increase dramatically in a matured sample. The fact that the peak is reduced in size following hydrolysis indicated the presence of an interference rather that the presence of arginine.
3) The presence of only one amino acid peak, which may be the result of an interference, in 19 out of the 20 samples is atypical of any known protein or peptide sequence. The absence of the principal amino acids found in wheat and barely proteins, glutamate and proline for wheat and aspartate, glutamate, glycine and alanine for barley (Shewry, 1993 and Wrigley and Bietz, 1998) also confirms that no cereal proteins or peptides could be detected in the these 19 samples. As expected, the positive results provided by the Bradford Microassay for these samples, were due to interferences.
4) Sample 10380232 has an amino acid profile which is typical of a protein or peptide. The estimated total protein content is 1.3 mg/l.
References
Shewry P.R. (1993) Barley seed proteins. In: Barley: Chemistry and Technology. Edited by MacGregor, A.W., American Society of Cereal Chemists, St Paul Minnesota, USA, pp. 131-197
Page 123 of 173
Wrigley C.W. and Bietz J.A. (1998) Proteins and Amino Acids. In: Wheat: Chemistry and Technology. Edited by Pomeranz, Y., American Association of Cereal Chemists, St Paul, Minnesota, USA, Volume 1, pp. 159-275
Page 124 of 173
Page 125 of 173
Page 126 of 173
Page 127 of 173 Page 128 of 173
Page 129 of 173
Appendix 12 *ELISA Based Gluten Analytical Evidence as Presented in the Application for Provisional Exemption*
Appendix 12.1 *Introduction*
In the application for provisional exemption 46 cereal derived production distillate samples, matured spirit samples and liqueurs were analysed for gluten levels using an ELISA based analytical method.
Appendix 12.2 *Sampling Protocol*
Samples were received at The Scotch Whisky Research Institute from a variety of different sources. These samples were selected to cover a broad range of the types of distilled beverages produced using cereal as a raw material prior to distillation. Each sample was assigned an individual numerical reference code using the Laboratory Information Management System (LIMS), recording all the details provided in the database and an adhesive label was attached.
The EFSA NDA Panel in their opinion on the provisional application (Section 3.3.2) noted that sample details were not supplied for the cereal based distillate samples which were analysed for gluten in the application for provisional exemption. Sample information for these samples is now provided in Appendix 5, although this is not as complete as that obtained for the extended survey as reported in Appendices 6 and 6b.
Prior to analysis each sample was sub-sampled into a 100 ml bottle, labelled with the individual sample code by a member of the Institute staff. These samples were then dispatched by courier to Reading Scientific Services Limited (RSSL, The Lord Zuckerman Research Centre, Whiteknights Campus, Reading, RG6 6LA, UK) where they were analysed using the methodology described in Section 12.3 of this Appendix.
It should be remembered that whilst all bottled samples were analysed at the alcohol strength at which the product is sold, production samples were often analysed at elevated alcohol strengths. Thus, actual limits of detection for certain samples will be improved because the samples will be diluted prior to bottling.
Appendix 12.3 *Materials and Method*
Introduction
The test is an antigen-antibody reaction and it is based on the ELISA method using the single monoclonal antibody (R5). The wells of the microtiter strips are coated with specific antibodies to gliadins. By adding the standard or sample solution to the wells, gliadin present will bind to the specific capture antibodies. The result is an antibody- antigen-complex. Sample components not bound by the antibodies are then removed in a washing step. The bound gliadin is detected by an antibody conjugated to peroxidase (enzyme conjugate).
Page 130 of 173
Any unbound enzyme conjugate is then removed in a washing step. Enzyme substrate (urea peroxide) and chromogen (tetramethylbenzidine) are added to the wells and incubated. Bound enzyme conjugate converts the colourless chromogen into a blue product. The addition of the stop reagent leads to a colour change from blue to yellow. The measurement is made photometrically at 450 nm.
Equipment
Gluten: Commercial Immunoassay supplied by R-Biopharm Kit. The Limit of Detection (LOD) and reporting limit were determined to be 10 mg/kg. The ELISA kit detects monomers of gliadin, secalins and hordeins.
Preparation of Samples
Extraction with the extraction buffer (Mendez Cocktail). This extraction method is especially recommended for heat-treated foods & beverages.
2 g of each sample was approximately but accurately weighed and dried at about 50°C under a stream of nitrogen. The dried substance was reconstituted with Milli-Q water to 0.5 ml. 0.25 g of sample was weighed into a vial and 2.5 ml of the extraction buffer was added. The vial was closed and mixed well prior to incubation for 45 min at 50°C. After being allowed to cool the samples were then mixed with 7.5 ml ethanol (80 %) and mixed again. The samples were shaken for 1 h at room temperature (20 – 25 °C) and centrifuged for 10 minutes ca 3000 rpm. The supernatant was transferred to a screw top vial and diluted 1:12.5. 100 µl of each sample was used per well in the assay.
Appendix 12.4 *Results*
The full reports produced by RSSL on the analysis for gluten in the cereal distillates and spirit drinks produced from cereal distillates, as presented in the provisional application for exemption, can be found at the end of this Appendix. A summary table of the relevant results can be found in Appendix 12 Table 1.
Appendix 12 Table 1: Gluten in Cereal Distillates and Spirit Drinks produced from Cereal Distillates (Bottled Spirits and Production Samples)
Sample Type and Number. Description Protein Concentration (mg/l) Scotch Malt Whisky Bottled Products S04-0820 Glenfiddich Scotch Malt Whisky ND S04-0821 Glenfiddich Liqueur ND S04-0868 Aultmore Scotch Malt Whisky 12 year old ND Dalwhinnie Scotch Malt Whisky 15 year S04-0902 ND old S04-0905 Lagavulin Scotch Malt Whisky 16 year old ND
Scotch Blended Whisky Bottled Products S04-0707 Teachers Highland Cream Scotch Whisky ND Stewarts Cream of the Barley Blended S04-0708 ND Scotch Whisky S04-0709 Ballantine’s Finest Scotch Whisky ND Johnnie Walker Black Label Old Scotch S04-0714 ND Whisky
Page 131 of 173
Sample Type and Number. Description Protein Concentration (mg/l) S04-0715 Bells Scotch Whisky ND S04-0719 Cutty Sark Scotch Whisky ND S04-0720 The Famous Grouse ND S04-0721 The Famous Grouse Liqueur ND S04-0824 Grant’s Family Reserve ND
Irish Whiskey Bottled Products S04-0875 Jameson Irish Whisky ND
Vodka (cereal derived) Bottled Products S04-0710 Altai Vodka ND S04-0711 Wodka Wyborowa Pure Rye Grain ND S04-0717 Smirnoff Triple Distilled Vodka ND S04-0825 Taboo ND
Gin (cereal derived) Bottled Products S04-0716 Tanqueray Special Dry Gin ND S04-0727 Bombay Sapphire Dry Gin ND S04-0829 Grants Gin ND S04-0870 Greenalls Gin Finished Product ND S04-1047 Hendricks Gin ND
Bourbon/ Canadian Whisky/Other Bottled Products Whisk(e)y Products S04-0973 Echter Nordhaufer Doppelkorn ND
Neutral Spirit Production Distillate Samples S04-0827 Girvan Distillery Grain Neutral Spirit ND S04-0872 Greenalls Grain Spirit ND S04-0900 Grain Neutral Spirit ex Velva (French) ND S04-0901 Grain Neutral Spirit ex Velva (French) ND S04-1034 Sedalcol Grain Neutral Alcohol 23/03/04 ND S04-1035 Sedalcol Grain Neutral Alcohol 07/06/04 ND S04-1036 Sedalcol Grain Neutral Alcohol 11/05/04 ND S04-1038 Sedalcol Grain Neutral Alcohol 14/04/04 ND
Grain New Make Spirit Production Distillate Samples (Scotch/Irish) S04-0826 Girvan Scotch Grain New Make Spirit ND S04-0874 Middleton Irish Grain New Make Spirit ND
Malt New Make Spirit Production Distillate Samples (Scotch/Irish) S04-0869 Aultmore Scotch Malt New Make Spirit ND S04-0873 Middleton Irish Pot Still New Make Spirit ND S04-0903 Dalwhinnie Scotch Malt New Make Spirit ND S04-0904 Lagavulin Scotch Malt New Make Spirit ND S04-0906 Talisker Scotch Malt New Make Spirit ND S04-0907 Caol Ila Scotch Malt New Make Spirit ND
Gin (cereal derived) Production Distillate Samples S04-0712 Compounded Gin 193 ND
Page 132 of 173
Sample Type and Number. Description Protein Concentration (mg/l) S04-0713 Gin Distillate (LGS) ND S04-0871 Greenalls Gin Concentrate ND S04-0893 Distilled London Dry Gin Concentrate ND
Bourbon/ Canadian Whisky/Other Production Distillate Samples Whisk(e)y Products S04-0726 Jack Daniels Whiskey ND ND= not detected Reporting Limit/Limit of Detection = 10 mg/kg (approximately equivalent to 10 mg/l)
Appendix 12.5 *Analytical Conclusions*
The results show that no gluten was detected in all the samples tested.
Appendix 12.6 *RSSL Reports on the Determination of Gluten in Cereal Distillates and Spirits Produced from Cereal Distillates Presented for the Application for Provisional Exemption*
The samples which were analysed for gluten content are detailed in Table 1 of this Appendix and the certificates of analysis are reproduced on the following pages. These pages also include information on a method of analysis for almond allergen, the certificates of analysis for samples analysed using this method and the certificate of analysis for an alcohol sample analysed for whey protein. The results from the almond allergen and whey protein analyses have been submitted separately as part of two additional dossiers requesting labelling exemptions for distillates which have been flavoured with nuts added prior to distillation and for distillates which have been made from whey used before distillation.
Whilst the reports from RSSL have been introduced in this Appendix, hard copies are included with the supporting references to this document.
Page 133 of 173
THE DETERMINATION OF THE PRESENCE OF GLUTEN AND ALMONDS USING ELISA TECHNIQUES IN SAMPLES OF VARIOUS TYPES OF SPIRITS
Page 134 of 173
P4-05811
Aims and Objective To determine the presence of gluten and almond in various types of spirits using ELISA techniques.
Summary & Conclusions In order to reliably determine low levels of gluten for these types of sample matrices, the samples need to be evaporated to dryness prior to extraction procedure. It was found that the high alcohol contents of the samples interfere with the antigen-antibody reaction in the assay thus resulting in low recovery of gluten from the samples. The recovery obtained from the samples that were evaporated to dryness spiked at 10mg/kg gluten extracts was between 80- 109%, whereas those that had not been evaporated to dryness gave recoveries between 47-70%. Due to much better recoveries obtained at 10mg/kg, the detection and reporting limits were set at this level for these types of sample matrices.
In contrast to the gluten test, the almond test can reliably detect low levels of almond in the samples using untreated samples (i.e. samples not evaporated to dryness). The samples that were spiked gave very good recoveries at all spiked levels: 1.25mg/kg spiked = 138%, 2.5 mg/kg spiked = 98%, 3.75 mg/kg spiked = 82%. This means that the alcohol content of the sample does not interfere in the assay.
Samples that were evaporated to dryness also gave very low recoveries at all spike levels. This could be due to other components in the samples having been concentrated down to levels that could lead to interference in the assay.
Page 135 of 173
P4-05811
GLUTEN Gluten is the specific name for proteins found in wheat, barley, rye, oats, triticale, splet and kamut. Glutens are functional proteins that have visco-elastic properties. Gluten is composed of two groups of proteins known as prolamines (gliadins) and glutenins. Gliadins and glutenins each form 50% of the protein. Prolamines (gliadins) are water insoluble but soluble in alcohol. They have different names in different cereals, e.g. gliadin in wheat; avenins in oats; secalins in rye and hordein in barley.
The method is a competitive sandwich type ELISA. It is based on the antigen-antibody reaction. The R-5 antibody used in the test kit specifically detects prolamines from wheat, rye and barley. The wells of the microtiter strips are coated with specific antibodies to gliadins. By adding the standard or sample solution to the wells, only gliadin present will bind to the specific capture antibodies. The result is an antibody-antigen-complex. Sample components not bound by the antibodies are then removed by washing. A second antibody conjugated to peroxidase detects the bound gliadin complex. Any unbound enzyme conjugate is then removed in a washing step. Enzyme substrate and chromogen are added to the wells and incubated. Bound enzyme conjugate converts the colourless chromogen into a blue product.
The addition of the stop reagent leads to a colour change from blue to yellow. The measurement is made photometrically at 450 nm.
Method A known amount of each sample was evaporated to dryness at 50ºC under a flow of nitrogen. The dried materials obtained were re-suspended in MQ water and vortexed to dissolve dried residues.
Sample extraction procedure: 0.25g of the suspended sample was weighed and 2.5ml extraction buffer was added. The samples were incubated in a water bath at 50ºC for 45 minutes. After
Page 136 of 173
P4-05811
cooling, 7.5ml of 80% ethanol was added. Samples were then shaken for one hour at room temperature. The samples were then diluted with sample diluent 1:12.5 prior to assay.
Gluten Recovery Test 1000 mg/kg gluten solution was made up with 80% ethanol. By serial dilution, using the sample diluent, a 10mg/kg working gluten spike solution was made up. Three of the samples (S04-0710, S04-0720 and S04-0829) were spiked at 3.0, 5.0 and 10.0 gluten levels. The samples were then prepared following the sample extraction procedure detailed above.
A recovery test was also conducted in samples that were not dried.
The sample was assayed following the protocol given in the in-house method TM-178.
Calculation of Results The optical densities of the sample and standards were read using the laboratory microplate reader, which is directly controlled from the PC using KC 4 software, which also collects the data. The plate reader used to read the absorbance values of samples and standards also automatically calculates the amount of gliadin present in each sample in the microwells. In order to calculate the amount of gluten found in the sample, the concentrations obtained by KC4 in the extracts were then entered on the Excel spreadsheet and multiplied by the dilution factor of the respective samples to get the gliadin levels in the samples.
According to the test kit’s manufacturer, 80% of wheat proteins are made up of glutens, of which 50% are prolamines (gliadins) and 50% are glutelins. Therefore, in order to obtain the amount of gluten in the sample, the calculated results obtained were further multiplied by 2.
Results and Discussion
Page 137 of 173
P4-05811
The recoveries obtained for samples that were not dried, were found to be very low. Gliadins are soluble in 40-70% ethanol. As the samples analysed contained about this much ethanol, any gluten found in the sample could have been further diluted during extraction. Consequently very low recoveries were obtained for the samples analysed.
In order to eliminate the effect of alcohol in the sample during assay, it was decided to evaporate the samples to dryness. The recoveries obtained in the dried samples were much improved as can be seen in the table below.
Sample Code/Sample Reference % Recovery
Undried Sample Dried Samples P4-05811-04/SO4-0710 3 mg/kg gluten spike 14 73 5 mg/kg gluten spike 24 65 10 mg/kg gluten spike 62 92 P4-05811-14/S04-0720 3 mg/kg gluten spike 14 53 5 mg/kg gluten spike 25 60 10 mg/kg gluten spike 47 109 P4-05811-27/S04-0829 3 mg/kg gluten spike 59 60 5 mg/kg gluten spike 55 60 10 mg/kg gluten spike 70 80
As can be seen from the table above, of the three spike levels, an improved recovery was obtained at 10mg/kg. As recoveries obtained below this level were not as satisfactory, it was decided that detection and reporting limit for these types of sample matrices would be 10mg/kg gluten.
Page 138 of 173
P4-05811
ALMOND The almond assay has been developed to detect almond traces in foodstuffs, either as a component ingredient or as a contaminant. This competitive immunoassay is proposed as suitable for the detection of almond protein in either cooked or raw foodstuffs. Almond proteins are extracted from the homogenized food samples using an aqueous buffer at 60°C. The diluted sample extract is applied to anti-almond coated microtiter wells. Conjugated antibody then binds captured almond protein from the samples. The binding of antigen (almond) is visualized by incubation with substrate / chromogen which turns to deep blue in the presence of almond. The addition of the stop reagent leads to a colour change from blue to yellow. The measurement is read photometrically against air at 450 nm with reference filter set between 620-650mn.
Method Approximately 5.0g of each sample was weighed and 10x volumes of the pre-heated (60ºC) extraction buffer were added. The samples were incubated for 15 minutes in a 60ºC water bath with shaking for one minute every five minutes. The samples were then cooled to room temperature and centrifuged (10minutes 3000rpm). The filtrates were used in the assay.
Recovery Test Procedure A 1000ppm almond stock spike solution was made up from finely ground almond using the extraction buffer. The solution was shaken for 20 minutes and centrifuged. The supernatant was serially further diluted with the extraction buffer to give a 10mg/kg working spike solution.
A known amount of the working spike solution was added to each 2.5g samples to give a 1.25, 2.5 and 3.75 mg/kg spike levels. 25ml pre-warmed buffer was added to each spike sample and the samples were incubated for 15 minutes with shaking for one minute in every five minutes. The samples were cooled and centrifuged. The supernatants were used in the assay.
Page 139 of 173
P4-05811
In order to ascertain if the alcohol content of the samples would interfere with the antigen- antibody binding in the assay, samples that were evaporated to dryness in the Gluten test were also used in the Almond recovery test.
The assay procedure given in the in-house test method TM-186 was followed.
Calculation of Results The optical densities of the sample and standards were read using the laboratory microplate reader, which is directly controlled from the PC using KC 4 software, which also collects the data. The KC4 automatically generated the standard curve based on the optical density of each standard obtained and also automatically calculated the amount of almond present in the sample.
Results and Discussion Shown in the table below are the recoveries for samples that were spiked with almond extracts at various levels. Good recoveries were obtained for the samples that were not evaporated to dryness at all spike levels. The experiment showed that the method could reliably detect almonds at low levels. The average recovery obtained at 1.25mg/kg spike level were found to be 138%. This means that, for the Almond Test, the alcohol content of the samples does not interfere in the antigen-antibody binding in the assay.
The low recoveries levels obtained for the samples that were evaporated to dryness could be due to the interference of other components found in the samples that had been concentrated down during the drying process.
Page 140 of 173
P4-05811
% Recovery Sample Code/Sample Reference Undried Sample Dried Samples P4-05811-04/SO4-0710
1.25 mg/kg Almond Spike 136 77 2.5 mg/kg Almond Spike 100 67 3.75 mg/kg Almond spike 83 68 P4-05811-27/S04-0829 1.25 mg/kg Almond Spike 139 23 2.5 mg/kg Almond Spike 95 52 3.75 mg/kg Almond spike 80 51
Conclusions For these type of sample matrices, the samples had to be evaporated to dryness in order to reliably detect low levels of gluten as the alcohol content of the samples were found to interfere in the assay. The detection and reporting limit was found to be at 10mg/kg gluten. The recovery found in the three different sample matrices spiked at 10mg/kg gluten level was found to be between 80-109%.
Analysis for the presence of almond can reliably be conducted on normal samples (i. e. samples that have not been evaporated to dryness). The low recoveries obtained for the treated samples could be due to the other components that had been concentrated down during evaporation to dryness. These components could have caused interference in the assay, thus giving much lower recoveries as compared to samples that had not undergone such treatment.
Page 141 of 173
P4-05811
Issued By: Date:
...... 30.06.04.
DR FEBE WASS Technical Manager Immuno Chemistry
Whilst every care has been taken in the preparation of this report, Reading Scientific Services Limited cannot be held responsible or liable in respect of the use to which the information, contained in this report, is put. This report may only be reproduced in its entirety unless permission to do otherwise has been obtained from Reading Scientific Services Ltd.
Where samples have been analysed, the results obtained relate only to the samples tested and Reading Scientific Services Ltd does not guarantee the bulk of the material to be of equal quality.
Page 142 of 173
C e r t i f i c a t e o f A n a l y s i s
MR CRAIG OWEN Reference No: P4-05811 SWRI Date Received: 9th June 2004 THE ROBERTSON TRUST BUILDING Date Started: 29th June 2004 RESEARCH PARK NORTH Page 1 of 4 RICCERTON EDINBURGH, EH14 4AP
Analysis: ALMOND USING METHOD TM 186
RESULTS:
Sample ID Sample Description Results (mg/kg) P4-05811-04 S04-0710 <1 P4-05811-06 S04-0712 <1 P4-05811-07 S04-0713 <1 P4-05811-10 S04-0716 <1 P4-05811-12 S04-0718 <1 P4-05811-14 S04-0720 <1 P4-05811-17 S04-0727 <1 P4-05811-26 S04-0829 <1 P4-05811-27 S04-0836 <1 P4-05811-28 S04-0837 <1 P4-05811-29 S04-0838 <1 P4-05811-32 S04-0870 <1 P4-05811-33 S04-0871 <1
Signatories: Date: 30 June 2004
DR F G WASS (Technical Manager, Immuno Chemistry)
These results relate only to the sample(s) tested and do not guarantee the bulk of the material to be of equal quality. RSSL staff were not responsible for sampling and cannot be held liable in respect of the use to which this information is put. ______READING SCIENTIFIC SERVICES LTD The Lord Zuckerman Research Centre Whiteknights Campus, Pepper Lane, Reading RG6 6LA Tel: 0118 9868541 (Int: +44 118 9868541) Fax: 0118 9868932 (Int: +44 118 9868932)
Page 143 of 173
C e r t i f i c a t e o f A n a l y s i s
MR CRAIG OWEN Reference No: P4-06094 SWRI Date Received: 16th June 2004 THE ROBERTSON TRUST BUILDING Date Started: 29th June 2004 RESEARCH PARK NORTH Page 2 of 4 RICCERTON EDINBURGH, EH14 4AP
Analysis: GLUTEN USING METHOD TM 178
RESULTS:
Sample ID Sample Description Results (mg/kg) P4-05811-1 S04-0707 <10 P4-05811-2 S04-0708 <10 P4-05811-3 S04-0709 <10 P4-05811-4 S04-0710 <10 P4-05811-5 S04-0711 <10 P4-05811-6 S04-0712 <10 P4-05811-7 S04-0713 <10 P4-05811-8 S04-0714 <10 P4-05811-9 S04-0715 <10 P4-05811-10 S04-0716 <10 P4-05811-11 S04-0717 <10 P4-05811-12 S04-0718 <10 P4-05811-13 S04-0719 <10 P4-05811-14 S04-0720 <10 P4-05811-15 S04-0721 <10
Signatories: Date: 30 June 2004
DR F G WASS (Technical Manager, Immuno Chemistry)
These results relate only to the sample(s) tested and do not guarantee the bulk of the material to be of equal quality. RSSL staff were not responsible for sampling and cannot be held liable in respect of the use to which this information is put. ______READING SCIENTIFIC SERVICES LTD The Lord Zuckerman Research Centre Whiteknights Campus, Pepper Lane, Reading RG6 6LA Tel: 0118 9868541 (Int: +44 118 9868541) Fax: 0118 9868932 (Int: +44 118 9868932)
Page 144 of 173
C e r t i f i c a t e o f A n a l y s i s
MR CRAIG OWEN Reference No: P4-06094 SWRI Date Received: 16th June 2004 THE ROBERTSON TRUST BUILDING Date Started: 29th June 2004 RESEARCH PARK NORTH Page 3 of 4 RICCERTON EDINBURGH, EH14 4AP
Analysis: GLUTEN USING METHOD TM 178
RESULTS:
Sample ID Sample Description Results (mg/kg) P4-05811-16 S04-0726 <10 P4-05811-17 S04-0727 <10 P4-05811-18 S04-0820 <10 P4-05811-19 S04-0821 <10 P4-05811-20 S04-0822 <10 P4-05811-21 S04-0823 <10 P4-05811-22 S04-0824 <10 P4-05811-23 S04-0825 <10 P4-05811-24 S04-0826 <10 P4-05811-25 S04-0827 <10 P4-05811-26 S04-0829 <10 P4-05811-30 S04-0869 <10 P4-05811-31 S04-0868 <10
Signatories: Date: 30 June 2004
DR F G WASS (Technical Manager, Immuno Chemistry)
These results relate only to the sample(s) tested and do not guarantee the bulk of the material to be of equal quality. RSSL staff were not responsible for sampling and cannot be held liable in respect of the use to which this information is put. ______READING SCIENTIFIC SERVICES LTD The Lord Zuckerman Research Centre Whiteknights Campus, Pepper Lane, Reading RG6 6LA Tel: 0118 9868541 (Int: +44 118 9868541) Fax: 0118 9868932 (Int: +44 118 9868932)
Page 145 of 173
C e r t i f i c a t e o f A n a l y s i s
MR CRAIG OWEN Reference No: P4-06094 SWRI Date Received: 16th June 2004 THE ROBERTSON TRUST BUILDING Date Started: 29th June 2004 RESEARCH PARK NORTH Page 4 of 4 RICCERTON EDINBURGH, EH14 4AP
Analysis: GLUTEN USING METHOD TM 178
RESULTS:
Sample ID Sample Description Results (mg/kg) P4-05811-32 S04-0870 <10 P4-05811-33 S04-0871 <10 P4-05811-34 S04-0872 <10 P4-05811-35 S04-0873 <10 P4-05811-36 S04-0874 <10 P4-05811-37 S04-0875 <10 P4-05811-38 S04-0900 <10 P4-05811-39 S04-0901 <10 P4-05811-40 S04-0902 <10 P4-05811-41 S04-0903 <10 P4-05811-42 S04-0904 <10 P4-05811-43 S04-0905 <10 P4-05811-44 S04-0906 <10 P4-05811-45 S04-0907 <10 P4-05811-46 S04-0973 <10
Signatories: Date: 30 June 2004
DR F G WASS (Technical Manager, Immuno Chemistry)
These results relate only to the sample(s) tested and do not guarantee the bulk of the material to be of equal quality. RSSL staff were not responsible for sampling and cannot be held liable in respect of the use to which this information is put. ______READING SCIENTIFIC SERVICES LTD The Lord Zuckerman Research Centre Whiteknights Campus, Pepper Lane, Reading RG6 6LA Tel: 0118 9868541 (Int: +44 118 9868541) Fax: 0118 9868932 (Int: +44 118 9868932)
Page 146 of 173
C e r t i f i c a t e o f A n a l y s i s
MR CRAIG OWEN Reference No: P4-06094 SWRI Date Received: 16th June 2004 THE ROBERTSON TRUST BUILDING Date Started: 29th June 2004 RESEARCH PARK NORTH Page 1 of 3 RICCERTON EDINBURGH, EH14 4AP
Analysis: ß-LACTOGLOBULIN USING METHOD TM 180.
RESULTS:
Sample ID Sample Description Results (mg/kg)
P4-06094-5 SO4-1037 <5
Signatories: Date: 30 June 2004
DR F G WASS (Technical Manager, Immuno Chemistry)
These results relate only to the sample(s) tested and do not guarantee the bulk of the material to be of equal quality. RSSL staff were not responsible for sampling and cannot be held liable in respect of the use to which this information is put. ______READING SCIENTIFIC SERVICES LTD The Lord Zuckerman Research Centre Whiteknights Campus, Pepper Lane, Reading RG6 6LA Tel: 0118 9868541 (Int: +44 118 9868541) Fax: 0118 9868932 (Int: +44 118 9868932)
Page 147 of 173
C e r t i f i c a t e o f A n a l y s i s
MR CRAIG OWEN Reference No: P4-06094 SWRI Date Received: 16th June 2004 THE ROBERTSON TRUST BUILDING Date Started: 29th June 2004 RESEARCH PARK NORTH Page 2 of 3 RICCERTON EDINBURGH, EH14 4AP
Analysis: GLUTEN USING METHOD TM 178
RESULTS:
Sample ID Sample Description Results (mg/kg) P4-06094-1 SO4-0893 <10 P4-06094-2 SO4-1034 <10 P4-06094-3 SO4-1035 <10 P4-06094-4 SO4-1036 <10 P4-06094-6 SO4-1038 <10 P4-06094-7 SO4-1047 <10
Signatories: Date: 30 June 2004
DR F G WASS (Technical Manager, Immuno Chemistry)
These results relate only to the sample(s) tested and do not guarantee the bulk of the material to be of equal quality. RSSL staff were not responsible for sampling and cannot be held liable in respect of the use to which this information is put. ______READING SCIENTIFIC SERVICES LTD The Lord Zuckerman Research Centre Whiteknights Campus, Pepper Lane, Reading RG6 6LA Tel: 0118 9868541 (Int: +44 118 9868541)
Page 148 of 173
Fax: 0118 9868932 (Int: +44 118 9868932)
C e r t i f i c a t e o f A n a l y s i s
MR CRAIG OWEN Reference No: P4-06094 SWRI Date Received: 16th June 2004 THE ROBERTSON TRUST BUILDING Date Started: 29th June 2004 RESEARCH PARK NORTH Page 3 of 3 RICCERTON EDINBURGH, EH14 4AP
Analysis: Almond using method TM 186
Results:
Sample ID Sample Description Results (mg/kg)
P4-06094-1 SO4-0893 <1
P4-06094-7 SO4-1047 <1
Signatories: Date: 30 June 2004
DR F G WASS (Technical Manager, Immuno Chemistry)
These results relate only to the sample(s) tested and do not guarantee the bulk of the material to be of equal quality. RSSL staff were not responsible for sampling and cannot be held liable in respect of the use to which this information is put. ______READING SCIENTIFIC SERVICES LTD The Lord Zuckerman Research Centre Whiteknights Campus, Pepper Lane, Reading RG6 6LA Tel: 0118 9868541 (Int: +44 118 9868541) Fax: 0118 9868932 (Int: +44 118 9868932)
Page 149 of 173
Appendix 13 RSSL Report Detailing Development Work to Increase Sensitivity of ELISA Based Methods
The following report deals with the exercise to validate improvements in the sensitivity of an ELISA based method used to quantify the level of gluten in alcoholic distillates and products thereof. This report also includes information regarding the development of ELISA based methods for detecting almond protein and β- lactoglobulin. The results from the almond protein and β-lactoglobulin analyses have been submitted separately as part of two additional dossiers requesting labelling exemptions for distillates which have been flavoured with nuts added prior to distillation and for distillates which have been made from whey used before distillation.
(The report details the second phase of the ELISA development and validation work. The first phase looked at validating methods at similar protein concentrations but using evaporation techniques to concentrate the proteins prior to assay. Such concentration techniques, as briefly described, proved unsuccessful, with variable and/or poor recoveries.)
Page 150 of 173
Page 151 of 173
Page 152 of 173
Page 153 of 173
Page 154 of 173
Page 155 of 173 Page 156 of 173
Page 157 of 173
Page 158 of 173
Page 159 of 173
Page 160 of 173
Appendix 14 RSSL Reports on the Determination of Gluten in Cereal Alcohol Distillates using Improved Methodology
Appendix 14.1 Introduction
The certificates of analysis provided by Reading Scientific Services Limited (RSSL) in Section 14.3 of this Appendix detail the results from the application of the improved gluten analysis method, to 115 distillate samples, following the introduction of a concentration step (see Section 7.2 for further details).
Appendix 14.2 Materials and Method
All samples, apart from the wash sample were first pre-concentrated using Vivaspin® 15R Hydrosart 2kDa Molecular Weight Cut-off ultrafiltration devices (VivaScience, Sartorius Ltd, UK). Approximately 15 ml of a 25 ml subsample of each distillate was transferred into an individual Vivaspin® concentrator. The concentrator was then sealed and placed into a swing bucket rotor centrifuge for a minimum of 50 minutes at 3000g. Following centrifugation, the individual filtrate container was emptied and the remainder of the sample added to their respective concentrators. Centrifugation was then repeated for another minimum 50 minute cycle at 3000g. Following secondary centrifugation, the retained sample was removed from the concentration pocket with a Pasteur pipette and placed into a separate glass vessel. The concentration pocket was washed with the sample diluent used in the ELISA gluten analysis method, RIDASCREEN extraction buffer, and this washing solution also transferred to the separate glass vessel. The sample volume was made up to 1 ml using the RIDASCREEN extraction buffer.
The 1 ml volume of 25 times concentrated sample was analysed using the Gluten – Ridascreen® Gliadin analysis kit supplied by R-Biopharm Rhone, according to the test kit manufacturer’s instructions. Samples were diluted by a factor of 500 using ELISA sample diluent prior to assay.
The limit of quantification for the method is 10 mg/kg; this equates to 0.4 mg/kg when the concentration factor is taken into account. (N.B. Concentrations quoted in mg/kg are equivalent to concentrations in mg/l).
An RSSL in-house limit of detection determined for the ELISA assay (conducted using standard extraction methodology on solid sample matrices) was quoted as 2.5 mg/kg. When applying the concentration factor of 25 this would give an estimated limit of detection of 0.1 mg/kg. Whilst no samples (aside from the pre-distillation distillery wash) demonstrated a detectable level of gluten, due to the uncertainty of measurement at this level, standard practice was to quote all such results as less than the limit of quantification (0.4 mg/kg), not at less than the limit of detection (0.1 mg/kg).
Appendix 14.3 Results
The following pages replicate the certificates of analysis for the gluten determination of the 115 distillate samples in the extended product survey (Samples S05-3234 and S05- 3322 were both analysed in duplicate.) Also included are the results from the laboratory
Page 161 of 173 distillations as described in Section 5.4, and the analysis of a sample of distillery wash. The laboratory distillations were designed to illustrate the principle that gluten is non- volatile. The sample of wash was a representative sample of the feed material which would be distilled in a whisky distillery. The wash was analysed according to the standard protocol for the Ridascreen® Gliadin analysis kit (0.25g of the sample was weighed and 2.5 ml extraction buffer was added. The sample was incubated in a water bath at 50ºC for 45 minutes. After cooling, 7.5ml of 80% ethanol was added. The samples was then shaken for one hour at room temperature, then diluted with sample diluent 1:12.5 prior to assay.)
Page 162 of 173
Page 163 of 173
Page 164 of 173
Page 165 of 173
Page 166 of 173
Page 167 of 173
Page 168 of 173
Page 169 of 173 Page 170 of 173 Page 171 of 173 Appendix 15 Independent Literature Review to Identify Occurrence of Allergic Reactions to Distillates as a Result of Specific Raw Materials Used Pre-Distillation
Page 172 of 173 Page 173 of 173 Attachment B
The EFSA Journal (2007) 484, 1-7
Opinion of the Scientific Panel on Dietetic Products, Nutrition and Allergies on a request from the Commission related to a notification from CEPS on cereals used in distillates for spirits, pursuant to Article 6 paragraph 11 of Directive 2000/13/EC
(Request Nº EFSA-Q-2006-143)
(adopted on 3 May 2007)
SUMMARY
In addition to the information submitted to obtain temporary exemption, the applicant provided further information regarding distillates made from cereals which include whisky, Kornbrand, gin, vodka and “made wine” produced using vodka, liqueur and similar beverages.
The beverages in question are widely consumed in the European Union. A literature review that includes information up to April 2006 failed to reveal allergic reactions after consumption of distillates made from cereals, although underreporting cannot be excluded. Further evidence of the unlikelihood of distillates made from cereals to elicit allergic reactions stems from additional analytical data on potential residual proteins and their allergenicity in the distillates and the distillation process. The analytical methodology did not address the allergenic activity of residual protein levels in the final product using appropriate human sera. Neither epidemiological studies nor double-blind placebo-controlled food challenge studies in clinical settings have been carried out to address possible adverse allergic reactions to distilled spirit drinks due to cereal allergens.
Based on the data submitted by the applicant, the Panel notes that proteins and peptides are not carried over into the distillate during a properly controlled distillation process, at least not in amounts higher than 1 mg/L for total proteins and 0.4 mg/kg for gluten. The Panel considers that distillates made from cereals are unlikely to trigger a severe allergic reaction in susceptible individuals.
KEY WORDS
Allergenicity, distillates, cereals, protein content, gluten, celiac disease, wheat allergy.
© European Food Safety Authority, 2007 Page 1 of 7
BACKGROUND
In November 2003, the European Parliament and the Council adopted Directive 2003/89/EC1 amending Directive 2000/13/EC, as regards indication of the ingredients present in foodstuffs.
Annex IIIa of the Directive specifies a list of food ingredients or substances that are known to trigger allergic reactions or intolerances in sensitive individuals for which no labelling exemptions are allowed. Whenever the listed ingredients/substances or their derivatives are used in the production of foodstuffs, they must be labelled.
Article 1, paragraph 11, subparagraph 2 of the Directive establishes a procedure allowing for temporary labelling exemption of derivatives from ingredients listed in Annex IIIa for which it has been scientifically established that it is not possible for them to cause adverse reactions. In accordance with this provision, submissions of requests for temporary labelling exemption were notified to the Commission before 25 August 2004. The Commission, after consultation with the European Food Safety Authority, adopted a list (Directive 2005/26/EC2) of those ingredients which are temporarily excluded from Annex IIIa until 25 November 2007, pending the final results of the notified studies.
Consequently, applicants who submitted a dossier in 2004 on the basis of subparagraph 2, resulting in the inclusion of a product in the list of Directive 2005/26/EC, and who are seeking exclusion of that product from Annex IIIa beyond 25 November 2007 will have to submit a request enclosing the final results of the notified scientific studies. Therefore in the context of the permanent labelling exemption procedure, the European Food Safety Authority is asked to provide scientific opinions on the submissions in accordance with the present terms of reference.
TERMS OF REFERENCE
In accordance with Article 29 (1) (a) of Regulation (EC) N° 178/2002, the European Commission requests the European Food Safety Authority to evaluate the scientific data submitted by European Spirits Organisation (CEPS) in the framework of the procedure laid down in Article 6, paragraph 11 of Directive 2000/13/EC. On the basis of that evaluation, EFSA is requested to issue an opinion on the information provided, and particularly to consider the likelihood of adverse reactions triggered in susceptible individuals by the consumption of the following ingredients/substances used under the conditions specified by the applicant: cereals used in distillates for spirits
ASSESSMENT
Taking account of the potential allergen content and well documented clinical allergic reaction in individuals sensitive to the stated material (cereals) (NDA, 2004a), it is appropriate for the
1 Directive 2003/89/EC of the European Parliament and of the Council amending Directive 2000/13/EC as regards indication of the ingredients present in foodstuffs. OJ L 308. 25.11.2003, p. 15. 2 Commission Directive 2005/26/EC of 21 March 2005 establishing a list of food ingredients or substances provisionally excluded from Annex IIIa of Directive 2000/13/EC of the European Parliament and of the Council. OJ L 75, 22.03.2005, p. 33-34.
© European Food Safety Authority, 2007 Page 2 of 7
Panel to assess the likelihood that the finished product may cause a reaction in a cereal allergic consumer.
In 2004, the Panel issued an Opinion on a notification submitted by European Spirits Organisation (CEPS) to the European Commission pursuant to Article 6, paragraph 11 of Directive 2000/13/EC as amended by Directive 2003/89/EC, for temporary exemption from labelling (NDA, 2004b).
Under the framework of permanent exemption from labelling, the present Opinion is based on the assessment of an updated dossier from CEPS, which contains additional information and data mainly with regard to literature review and laboratory-based, in vitro tests.
1. Manufacturing process
Distillates made from gluten-containing cereals (e.g. wheat, rye, barley, oats) include whisky, Kornbrand, gin, vodka, “made wine” produced from vodka, liqueur, and similar beverages.
The processes involved in production of cereal-based distillates comprise cereal processing (i.e. extraction of starch from cereals and conversion to fermentable sugars), fermentation (conversion of sugars to alcohol by action of yeast), distillation (concentration of alcohol, removal of non-volatiles, control of flavour) and post-distillation processes (flavouring, maturation, etc.). Distillation can be achieved in two ways, by simple (or batch) distillation or by continuous (fractional) distillation. In principle, both systems should prevent transfer of non- volatile material from the feedstock into the final product. In batch distillation for malt whisky production, considerable amounts of non volatile and high boiling point compounds are transferred from the wash to the low wines. This mass transfer phenomenon is not explained by the vapour-liquid equilibrium distillation theory, but it is rather related to the generation of the “mist” formed by a burst of the foam surface in the wash still (Ohtake et al., 1995). However, the modern distillation apparatus are multi-stage systems involving one or more rectifying columns or continuous counter current vapour liquid contacting systems which should make it unlikely that significant levels of high molecular weight, high boiling point compounds such as allergenic proteins and peptides will be carried over into the distillate.
The applicant has provided a model study performed in the laboratory with a standard distillation apparatus showing that bovine serum albumin (BSA) does not distil. Although the results are those expected, they cannot be taken as the definitive proof, since the conditions of the experiment are different from those applied in the industrial production process (number of theoretical plates, dimensions of the columns, steam flow). In addition, the distillation of pure protein solutions in the laboratory, without the presence of additional components typically found in industrial distillation matrices, can result in different foaming, boiling homogeneity, steam carry over, and do not represent the same industrial conditions. However, no studies presented in the open literature provided evidence that proteins do have a vapour pressure and there is general agreement in the scientific community that they do not distil (Ohtake et al., 1995, Oldani et al., 2001).
2. Evidence of non-allergenicity
2.1 History of non-allergenicity of the product
© European Food Safety Authority, 2007 Page 3 of 7
The applicant has performed a literature review using the PubMed database of the US National Library of Medicine, which includes information up to April 2006, and failed to identify any reports linking celiac symptoms or wheat allergy with the consumption of distilled beverages. The PubMed database included over 14 million citations for biomedical articles from MEDLINE and additional life science journals dating back to the 1950s. Titles and abstracts were searched for COELIAC or CELIAC and WHEAT ALLERGY, which identified a large number of literature reports (11,301 and 236 respectively). Linking of these search terms with specific alcoholic beverages (WHISKY or WHISKEY, GIN, VODKA) and with DISTILLED SPIRITS yielded no published reports. Use of the more general term ALCOHOL with COELIAC or CELIAC identified a number of literature reports, but of those referring to celiac disease, none reported distilled beverage alcohol causing celiac symptoms but referred to the alcohol solubility of gliadins. Combination of ALCOHOL with WHEAT ALLERGY gave no reported literature.
Combination of ALLERG* and DISTILLED SPIRITS identified one reference which was a review of allergic and asthmatic reactions to alcoholic drinks (Vally and Thompson, 2003). From this review most sensitivities to alcoholic drinks do not appear to be immune mediated, but are more frequently pharmacological intolerances to specific chemicals in these drinks. Where allergic and asthmatic reactions to specific non-alcohol components have been reported, these are almost wholly concerned with non-distilled drinks. Only one reported investigation of spirit consumption triggering asthmatic attacks could be found, however the substance causing the reaction was not identified (Breslin et al., 1973). Patients reacted to specific drinks but not to the equivalent amount of ethanol. Skin prick tests for routine common allergens gave no reaction to wheat.
The applicant’s literature search did not reveal allergic reactions after consumption of distillates made from cereals. It remains possible that adverse effects due to drinking of distillates made from cereals may not be perceived as due to wheat allergens or gluten but rather be attributed to alcohol, and under-reporting may thus have occurred.
2.2 Laboratory-based tests
2.2.1 In vitro studies
The applicant asserts that the physico-chemical properties of the allergenic materials of cereals will result in their complete lack of volatility, and hence their absence from products obtained by distillation. The applicant supports this assertion with information from several specific reports of small scale or ad hoc analyses indicating there is a very low level of cereal-derived protein in spirits derived from cereals:
• Campbell (1988) described a crude gravimetric approach to the detection of potential allergens in two alcohol distillates which concluded that the levels of protein would be minimal (less than 0.05 mg/L).
• Oldani et al. (2001) reported an enzyme-linked immunosorbent assay (ELISA) based study of several common cereal-derived spirits which concluded that the level of gluten was less than 0.006 mg/L.
© European Food Safety Authority, 2007 Page 4 of 7
• An ELISA-based study of samples from several points in the distillation production process for grain neutral spirits, and 3 crude alcohol samples, which concluded that the levels of cereal proteins were less than 2.4-3.1 mg/kg.
The analytical studies described report the examination of 39 bottled products and 76 samples of distillates produced using cereal as a raw material prior to distillation. The samples comprised ethyl alcohol of agricultural origin, whisk(e)y, Kornbrand, gin/gin concentrate, vodka, “made wine” produced using vodka and liqueur samples. Of these, 86 samples were analysed for total protein content and 45 samples were analysed for gluten.
Two analytical methodologies were employed to determine the presence of proteins in the spirit samples: the Bradford Analysis Microassay, with a limit of detection (LOD) of 0.5 mg/L for BSA, which accounted for a LOD of 1mg/L for wheat proteins, and the AAA-Direct Analysis with a LOD of 0.05 mg/L for amino acids and 0.5 mg/L for proteins. Samples for which a positive response was detected using the Bradford Microassay were tested with the AAA- Direct Analysis. 15 positive responses with the Bradford Microassay method for Malt New Make Spirits/Bourbon distillates were confirmed by the AAA-Direct Analysis at a protein level of 0.5-1 mg/L. Only in one sample the protein content was higher than 1 mg/L (1.3 mg/L), but it was not confirmed in a repeated experiment. These two methods are appropriate for their intended purpose.
No gluten was detected in any of the 115 samples analysed by ELISA with a LOD of 10 mg/kg. With a pre-concentration step, the LOD can reach 1 mg/kg, and 0.4 mg/kg when the concentration factor is taken into account. The calibration and recovery factor experiments were satisfactory.
The analytical data presented indicate that cereal proteins are not transported into the products of distillation during spirit manufacture at a level above 1 mg/kg. The analytical methodology did not address the allergenic activity of residual protein levels in the final product using appropriate human sera.
2.2.2 Animal studies
No animal studies were provided or referred to in the data submitted.
2.3 Clinical studies
2.3.1 Skin tests
No systematic skin prick testing studies in cereal allergic individuals exposed to distillates have been reported. Only one case report of spirit consumption triggering asthmatic attacks could be found, however the substance causing the reaction was not identified (Breslin et al., 1973). Skin prick tests for routine common allergens gave no reaction to wheat. Hence, this case report does not indicate gluten or cereal allergen involvement in the responses noted.
2.3.2 Double-blind placebo-controlled food challenge (DBPCFC)
No DBPCFC studies in sensitive individuals have been performed.
© European Food Safety Authority, 2007 Page 5 of 7 Figueredo et al. (1999) report a case of a Type 1 allergy to barley malt and corn triggered by beer consumption, but the patient showed no reaction to oral provocation tests with a distilled spirit (10 mL-50 mL). This indicates that responses due to drinking beverages that contain cereals are possible, but in this case the distillate did not appear to cause such effects. Firm conclusions cannot be drawn, as the data pertain to one case report only.
2.3.3 Epidemiological studies
The applicant has not carried out epidemiological studies in order to investigate possible adverse reactions to distilled spirit drinks due to cereal allergens or gluten.
CONCLUSIONS
Based on the data submitted by the applicant, the Panel notes that proteins and peptides are not carried over into the distillate during a properly controlled distillation process, at least not in amounts higher than 1 mg/L for total proteins and 0.4 mg/kg for gluten. The Panel considers that distillates made from cereals are unlikely to trigger a severe allergic reaction in susceptible individuals.
DOCUMENTATION PROVIDED TO EFSA
Dossier submitted by the European Spirits Organisation (CEPS) to the European Commission pursuant to Article 6, paragraph 11 of Directive 2000/13/EC on 25 August 2006.
REFERENCES
Breslin ABX, Hendrick DJ, Pepys J (1973). Effect of disodium cromoglycate on asthmatic reactions to alcoholic beverages. Clinical Allergy 3: 71-82.
Campbell JA (1988). The Question of Gluten in Distilled Alcohol. Journal of the Canadian Dietetic Association, 49:204.
Figueredo E, Quirce S, Del Amo A, Cuesta J, Arrieta I, Lahoz C, Sastre J (1999). Beer-induced anaphylaxis: identification of allergens. Allergy 54: 630-634.
NDA (Scientific Panel on Dietetic Products, Nutrition and Allergies) (2004a). Opinion of the Scientific Panel on Dietetic Products, Nutrition and Allergies on a request from the Commission relating to the evaluation of allergenic foods for labelling purposes. The EFSA Journal 32, 1-197. http://www.efsa.europa.eu/en/science/nda/nda_opinions/food_allergy/341.html
NDA (Scientific Panel on Dietetic Products, Nutrition and Allergies) (2004b). Opinion of the Scientific Panel on Dietetic Products, Nutrition and Allergies related to a notification from CEPS on distillates made from cereals pursuant to Article 6 paragraph 11 of Directive 2000/13/EC. The EFSA Journal 130, 1-6. http://www.efsa.europa.eu/en/science/nda/nda_opinions/food_allergy/684.html
© European Food Safety Authority, 2007 Page 6 of 7 Ohtake K, Yamasaki H, Kojima K (1995). Evaluation of mass-transfer by mist as a new parameter in the control of wash distillation. Proceedings of the Fourth Aviemore Conference on Malting, Brewing and Distilling. Ed. By Campbell I and Priest FG. Institute of Brewing, London, UK, pp. 202-208.
Oldani A, Iametti S, Bonomi F, Scafuri L (2001). In alcoholic cereal derived beverages are there still gluten traces? Tecnica Molitoria 7:641-646
Vally H and Thompson PJ (2003). Invited Review: Allergic and asthmatic reactions to alcoholic drinks. Addiction Biology 8: 3-11.
PANEL MEMBERS
Jean-Louis Bresson, Albert Flynn, Marina Heinonen, Karin Hulshof, Hannu Korhonen, Pagona Lagiou, Martinus Løvik, Rosangela Marchelli, Ambroise Martin, Bevan Moseley, Andreu Palou, Hildegard Przyrembel, Seppo Salminen, John (Sean) J Strain, Stephan Strobel, Inge Tetens, Henk van den Berg, Hendrik van Loveren and Hans Verhagen.
ACKNOWLEDGEMENT
The Scientific Panel on Dietetic Products, Nutrition and Allergies wishes to thank Taraneh Dean, Martin Stern and Jean-Michel Wal for their contributions to the draft opinion.
© European Food Safety Authority, 2007 Page 7 of 7 Attachment C