ACNFP response to the food safety issues raised in the evidence submitted to the Chardon LL hearing

2002: The ACNFP has considered the representations made to the Chardon LL public hearing and has concluded that no new evidence has been submitted that would question the safety of food products derived from this genetically modified maize line.

Summary
1. The Advisory Committee on Novel Foods and Processes (ACNFP) has agreed to provide advice on the food safety related issues raised from the recent Chardon LL Public Hearing regarding processed products derived from this genetically modified maize line.

2. The Committee is content that no new evidence was submitted to the Hearing that would question the safety of foods derived from Chardon LL maize.

Introduction

3. The ACNFP is an independent body of experts, whose remit is to advise the central authorities responsible, in England, Scotland, Wales and Northern Ireland respectively on any matters relating to novel foods, having regard, where appropriate, to the views of relevant expert bodies.

4. The Committee has considered the documents submitted to the T25 Public Hearing. The Committee has also addressed criticisms levied by some of those giving evidence to the Hearing, regarding the manner in which the safety assessment was conducted. These fall into the broad areas of the regulatory procedure, substantial equivalence and composition, the scientific assessment, the use of GM technology and the possible presence of GM pollen in honey.

5. The Advisory Committee on Animal Feedingstuffs (ACAF) has recently considered the safety of T25 maize as an animal feed. Their conclusions were addressed in a letter to the ACRE Secretariat, dated 5th September 2001. The implications of this GM variety entering the human food chain via the human consumption of products derived from animals fed on T25 maize was also addressed by ACAF. Points arising from the Hearing relating to animal feed issues have not been addressed by the ACNFP in this advice.

6. The Advisory Committee on Releases to the Environment (who consulted with the Advisory Committee on Animal Feedingstuffs) is conducting a similar exercise on the evidence presented at the Hearing on environmental issues, and other matters relating to the assessment carried out as part of the Part C clearance under 90/220/EEC. DEFRA's Plant Varieties and Seeds Division are also commenting on issues relating to the listing of T25 on the national seeds list.

Background

7. In 1995, the ACNFP considered the food safety of products derived from the genetically modified maize line Chardon LL (T25), which had been developed by AgrEvo USA to be tolerant to glufosinate–ammonium based herbicides. The request related to products derived from the seeds of this line, together with products derived from inbred and hybrid lines developed using conventional cross breeding. The ACNFP approved the application in 1996 and provided AgrEvo with a positive scientific opinion. The Company subsequently made a notification on the 8th January 1998, under Articles 4(3) and 5 of the Novel Foods Regulation, citing the ACNFP's scientific opinion. In accordance with Article 5, this notification was copied to all other Member States within 60 days of the Company notifying the Commission.

8. Following comments raised by one Member State regarding the notification route used for oils and other processed products derived from this and other GM crops, the Scientific Committee on Food reviewed the ACNFP opinion on T25 maize. The SCF concluded in September 2000, that the use of products derived from T25 maize did not endanger human health.

9. In 1996, AgrEvo applied to the French Competent Authority for marketing consent for T25 maize seed under Directive 90/220/EEC. Scientific advisory committees in other Member States, including ACRE in the UK, conducted their own independent safety assessments. Following this assessment, it was agreed that marketing consent should be granted, and in 1998, this was issued by France.

10. In 2000, the Company submitted an application to have Chardon LL placed on the UK National List. Under the National List regulations, parties affected by the proposed decision to 'list' a variety may make written representations to Ministers and request a public Hearing. Friends of the Earth challenged this application, and it was suspended. A public Hearing began in October 2000 to enable interested parties to present evidence for and against T25 being placed on the National List. The Hearing was chaired by Alun Aylesbury, a senior barrister, who was appointed by Ministers. The Hearing finished in June 2002.

Response

Regulatory procedure
11. AgrEvo applied to the ACNFP for clearance of food and feed products derived from this line in 1995, two years before the implementation of the Novel Foods Regulation (EC 258/97). Clearance was sought for starch, oil, and all heat processed or fermented food products obtained from maize meal and flour derived from the T25 maize line.

12. Until Regulation EC 258/97 came into effect, the UK operated a voluntary scheme under which companies could submit applications for food safety assessment. This system had been in place since the ACNFP was set up in 1988, to provide a forum in which the safety of novel foods could be assessed. In 1994 the ACNFP developed a structured decision tree for use in the safety assessment of novel foods. This was adopted following extensive public consultation.

13. The ACNFP carried out a full safety assessment and provided a scientific opinion that concluded that the processed products obtained from the GM maize line T25 were safe to use in food and any differences in composition from those of conventionally-bred maize were not considered to be of biological significance.

14. On the 15th May 1997, the EC Novel Foods and Novel Food Ingredients Regulation (EC 258/97) came into effect, introducing a statutory pre-market approval system for novel foods across the whole of the European Union. To ensure that all Member States follow a consistent approach to the safety assessment for novel foods, the Commission published a series of guidelines to accompany the regulation. These guidelines were largely based on the earlier approach taken by the ACNFP, in its voluntary assessments.

15. Articles 3(4) and 5 of the Novel Foods Regulation detail a simplified procedure whereby it is possible for companies to notify the Commission that they are placing a novel food on the market. To make use of this simplified procedure, a company requires a scientific opinion delivered by a Competent Authority in one of the Member States concluding that their novel food is substantially equivalent to existing foods or food ingredients, as regards its composition, nutritional value, metabolism, intended use and level of undesirable substances contained therein. The responsibility lies with the company to notify the Commission when the food is placed on the market.

16. Article 5 states that when a company submits such a notification, the Commission shall forward a copy to other Member States within 60 days. Member States have the option to request further details, such as the supporting scientific opinion and any data supplied by the company.

17. On the 15th December 1997, the ACNFP had written to the European Commission suggesting that in the future, opinions in accordance with Article 3(4) of the Novel Food Regulation, should only be provided for products where it could be demonstrated that novel DNA or protein was not present. The letter went on to say that for certain products, such as flour, a full safety assessment would be required since the processing that they undergo may not completely remove or destroy DNA and protein. In the case of the food products derived from T25 maize, these had already undergone a full safety assessment by the ACNFP.

18. On the 8th of January 1998, AgrEvo notified the Commission of their intention to market products derived from T25 maize, and used the scientific opinion of the ACNFP to support their notification. Although regarded by some as a 'fast track route' to approval, the Article 5 notification procedure involves a full safety assessment by the Competent Authority responsible for the initial opinion. In the case of T25, a full safety assessment was carried out by the ACNFP.

19. AgrEvo developed T25 maize as a processing line, and, as such, it would be unsuitable for human consumption in an unprocessed form. The ACNFP considered and gave food safety clearance only to products derived from this line and so, under regulation 5(1) of the Novel Foods and Novel Food Ingredients Regulations 1997 (UK Statutory Instrument), it would be an offence to sell unprocessed T25 maize for human consumption in the UK.

20. It is within ACAF's remit to assess any risk to humans of eating meat and / or dairy products from animals fed GM feed. Therefore, this response does not address animal feed issues. ACAF recently addressed the safety of T25 maize as an animal feed in a letter to the ACRE Secretariat dated 5th September 2001. ACAF was content that there was no evidence to indicate that T25 maize grain or its products pose any more risk to animals or humans if used in animal feed than non-GM maize varieties.

21. It has been suggested by some of those giving evidence to the Hearing that the safety assessment of T25 maize should have been carried out using independent data, and should not have been based on data submitted by the Company. However, the role of the scientific experts on the ACNFP is to scrutinise data supplied in support of an application, and this was the case with data submitted by AgrEvo to determine the safety of products derived from T25 maize. It is common practice in many areas of safety evaluation (such as for medicines, pesticides, food additives and industrial chemicals) to use toxicological data generated by the company seeking approval. Laboratories conducting such studies are subject to regular inspections to ensure that operating procedures are satisfactory and individual studies may be audited to ensure compliance with such procedures. Thus data produced to these standards are considered to be robust. The T25 application dossier has been deposited in the British Library, and so is available for scrutiny by interested parties.

22. The ACNFP's original assessment requested that the seed composition, including the amino acid profile and the fatty acid profile of the oil from the GM line and hybrid lines, should be monitored over time, to demonstrate the stable inheritance of the introduced trait and to determine any possible effect of genetic drift over successive generations on the plant's metabolism.

23. Seed composition data, collected from field trials carried out over two years, were submitted to the Committee in September 2002. These data demonstrated that some statistically significant differences were seen for seeds from non-transgenic and herbicide treated and untreated transgenic lines. However, in all cases, the nutritional impact of these differences is considered to be negligible, since the levels of the analysed components generally fall within previously reported reference ranges for conventionally-bred maize. Conventionally-bred plant varieties have a spectrum of chemical and nutritional compositions, and no regulatory process is in place to analyse or monitor these. The Committee was content that these data demonstrated the stability of the introduced trait and assessed the impact of the genetic modification event on the composition of the seed. The Committee is satisfied that this information does not alter the original safety assessment of T25 maize.

24. In May 2001, the Company supplied additional characterisation data for the Committee to consider, These data consisted of detailed sequence information of the flanking regions of the inserted DNA, and Northern blot analysis. A letter from Aventis, describing the findings of this analysis can be found on the ACNFP pages of the FSA website. The Committee was content that this information did not alter the original safety assessment of T25 maize derived products, and therefore a re-assessment of the dossier was not required at that time.

Substantial equivalence and composition

25. The concept of substantial equivalence is used to structure the safety assessment; it is not a safety assessment in itself. The safety assessment of GM foods requires an integrated and stepwise, case by case approach. An approach in which a GM food is compared with a conventional counterpart having a history of safe use is considered the most appropriate strategy for the safety assessment of GM foods.

26. The FAO/WHO expert consultation (May 2000) considered the concept of substantial equivalence as part of its review on the current safety assessment procedure for GM foods and recognised 'that there were presently no alternative strategies that would provide a better assurance of safety for genetically modified foods than the appropriate use of the concept of substantial equivalence' and that 'substantial equivalence should be seen as a key step in the safety assessment process'. The ACNFP fully supports that view.

27. As part of its initial application, AgrEvo supplied compositional and nutritional analysis data from material harvested from 2 US field sites (4 pairs of trials, with at least 3 samples taken from each pair) in 1994, which demonstrated that the silage and grain were not materially different from current commercial varieties in essential nutrients, including moisture, crude fat, crude protein, carbohydrate, acid detergent fibre, neutral fibre and ash. The approach adopted by the Company in its application was to demonstrate that the GM grain was equivalent in compositional terms to conventional grain and then to conclude that any products derived from such grain would also be comparable.

28. The Committee recognised that there were some statistically significant differences between the fat, carbohydrate, amino acid and fatty acid contents of the GM and non-GM lines. However, Members considered that although the differences were occasionally statistically significant, the levels of the components still fell within previously published reference ranges for maize. As a result, the Committee regarded the differences not to be biologically significant. Further data was supplied in September 2002, which is described in paragraph 23.

29. The original study to determine the levels of the anti-nutritional substance phytic acid was conducted on silage produced from T25 maize. However, it is considered that the contribution of phytic acid from a processing maize variety to the human diet will be negligible. ACAF has recently reviewed the compositional analysis of maize grain, and was content that there were no important differences from other maize varieties in terms of essential nutrients and anti-nutrients.

30. Further compositional characterisation data (see paragraph 23) demonstrated that the level of phytic acid in the T25 maize line was at the bottom end of the range that is normally found for conventionally-bred maize.

Scientific Assessment
Toxicology

31. Toxicological assessments were based on the evaluation of any differences between the experimental lines and their parental counterparts i.e. the toxicity of the genes themselves, the introduced protein, and the consequences of any secondary effects arising from the modification.

32. The ACNFP considers each application individually and requests animal test data in the context of the overall toxicological profile of the food in question, where it is considered necessary. Since the products derived from T25 maize were demonstrated to be comparable to their conventional counterparts, no animal studies were required for the safety assessment.

33. The FAO/WHO joint report (Safety aspects of genetically modified foods of plant origin, 2000) notes that the practical constraints of obtaining meaningful information from animal-based toxicology studies on the safety of whole foods have been well recognised for many years. Foods are complex mixtures of compounds, which have a wide variation in their composition and nutritional value. They can only be fed to animals at low multiples of the amounts that might be present in the human diet. Identifying potential adverse effects and relating these to the food and not other factors can therefore be extremely difficult for a number of reasons.

34. The FAO/WHO joint report goes on to say that in certain cases, where insufficient data are presented for a safety assessment, animal testing may be deemed necessary.

35. Some additional animal studies (in broiler chickens and rats) were submitted to the French Competent Authority as part of the Part C evaluation under Directive 90/220. However, these studies were designed to support the wholesomeness and nutritional value of products derived from this maize line for use as animal feed ingredients and not as toxicity studies. As such, they would not have been relevant to a novel food application and would not have been considered as primary evidence in the food safety assessment. It is extremely important that studies of this nature should only be used for the purpose for which they were intended.

36. ACAF recently reviewed (September 2001) the broiler chicken and rat feeding studies. Although they were critical of some aspects of the information provided, Members saw nothing to indicate that T25 maize grain or its products pose any more risk to animals or humans if used in animal feed than non-GM maize varieties.

37. T25 maize was modified with the introduction of the pat gene that confers tolerance to glufosinate-ammonium based herbicides. The pat gene, which encodes for the enzyme phosphinothricin acetyl transferase (PAT), was originally isolated from a soil-borne bacteria, but the T25 maize line was produced using a synthetic version of this gene.

38. Evidence was provided to demonstrate that the PAT protein was degraded at temperatures above 35°C, whereas milling of the grain takes place at temperatures in excess of 100°C. In addition, the PAT enzyme was shown to be readily inactivated at pHs below 5. The food products obtained from T25 maize all undergo either wet milling, which involves steeping before milling and then processing by conventional means, using pressure, heat, and solvent extraction, or dry milling. The processes that are applied prior to consumption will either further degrade or remove protein. This, together with evidence of the rapid breakdown of the protein under the conditions experienced in the gut, satisfied the Committee that the PAT protein did not represent a food safety concern. The PAT protein was found to be absent from the maize oil and starch derived from this line, but was detected at 20–300 ppm of the crude protein of wet and dry milled hulls, grits, meal and flour.

39. The PAT protein is present at very low levels in T25 maize, and so, to be able to detect any potential toxicity, isolated protein was used in the rat feeding study. The protein used for this experiment was extracted from a strain of E.coli, modified to express the pat gene. There is a small, theoretical risk that the PAT protein derived from E.coli is different from the PAT protein from maize and so it is possible that the protein used for toxicological studies could differ from that produced by T25 plants. This would be due to a post-translational modification, resulting in changes in the properties of the target molecule. Such modifications would, however, affect the electrophoretic mobility of the PAT proteins, with the modified protein being retarded in a gel compared with the unmodified protein. Since the protein from bacteria and maize show the same electrophoretic mobility; the PAT protein is unlikely to have undergone significant post-translational modification. However, the rat feeding assessment is only part of the overall safety evaluation and evidence was supplied that demonstrated that the PAT protein was not present in processed food products. Therefore, any possible differences between the PAT protein from different sources is not relevant to the safety evaluation of the processed food products derived from T25 maize.

40. Post-translational modification is a term used to describe the changes that occur to proteins after peptide bond formation has occurred. Examples include glycosylation, acylation, limited proteolysis, phosphorylation, and isoprenylation. The most reliable way in which the effects of post-translational-modification can be assessed is by evaluating morphological and agronomic data. The gross phenotype and general health of a mature plant gives the best indication of how proteins function. As was stated in the executive summary of the ACNFP report on processed products derived from T25 maize, the data supplied to the Committee satisfied Members that no unintentional changes had taken place at the molecular level.

Allergenicity

41. The first decision tree for use to detect allergenicity in novel foods was published in 1996 and was further refined by the FAO/WHO in 2001. Searching for amino acid sequence homology with known allergens is a significant part of allergenicity detection, but this is supported by other methodologies, including the assessment of protein stability and digestibility.

42. The safety of products derived from T25 maize were assessed prior to the implementation of these decision trees. However, in the initial application searches of the EMBL and SWISSPROT databases were conducted and it was concluded that there was no evidence to suggest any homology between the PAT protein and known allergens. In addition, as mentioned above, there was evidence to demonstrate rapid breakdown of the PAT protein during both processing and digestion.

Genetic modification technology

43. The application from AgrEvo sought clearance of food products derived from the T25 maize line. The Committee was content that the processing that these products undergo would destroy any functional DNA present.

44. When the Committee first considered this application, there was considerable discussion regarding the presence of the antibiotic resistance marker. However, after Members had sought and interrogated further data supplied by AgrEvo, the Committee was satisfied that the GM line contains a single, truncated copy of an ampicillin resistance gene, and the data provided demonstrated that it was inactivate. This was therefore not a safety concern in relation to food products intended for human consumption.

45. In 2001, Aventis supplied the Committee with further characterisation data on the T25 maize line. The Committee was asked to consider these data and to discuss whether there were any issues that had implications for the food safety assessment of this line. The Committee was content that they did not. A letter from Aventis explaining these data is on the ACNFP pages of the Food Standards Agency website, together with the minutes from the ACNFP meetings when they were discussed.

46. To address the possibility of unexpected effects such as gene silencing, pleitropic effects or genetic instability, the assessment for the safety of GM foods addresses both intentional and unintentional effects that may occur as a result of the particular genetic modification.

47. The cauliflower mosaic virus is found world-wide in temperate regions and is common in commercial crops of cabbage, Brussels sprouts, and cauliflower. Therefore, plant material infected by this virus has been eaten for centuries and no ill effects have been reported.

48. There is limited published scientific evidence that suggests that the CaMV 35s promoter, which drives pat gene expression, is functional in E.coli. It was noted, however, that functional pat genes are found in bacteria in nature. The Committee was content that this is not a food safety concern.

49. The potential that honey may contain pollen collected from GM maize is not considered to be a cause for concern. AgrEvo provided analysis of the whole plant, leaves, roots, seeds and mature pollen from the GM maize line and its non-GM counterparts grown in Europe. The analysis was concerned with detection of the PAT protein. Although a small amount of the PAT protein was found in unprocessed seeds, leaves, stems and roots, none was found in mature pollen, and so there is no evidence that honey will contain this protein.

50. Research conducted by Laboratory of the Government Chemist into the pollen content of honey, estimated that consumers would ingest a maximum 1 part-per-billion of transgenic protein from this source. In December 1999 the ACNFP looked again at the advice it issued in 1991 to take account of the results from further research in this field. The Committee endorsed the advice given in 1991 and confirmed that it was still content that the presence of very small amounts of GM pollen in honey did not represent a risk to consumers.

Conclusion
51. The ACNFP has considered the information submitted to the Public Hearing on Chardon LL maize, and it is content that no new evidence has been presented that would question the safety of foods derived from Chardon LL maize.

ACNFP
December 2002