Summary

The allergy issue has raised many concerns about GM food, which have important implications. This policy brief attempts to put concerns into perspective.

Introduction

The possibility that genetically modified (GM) foods could be a source of novel allergens continues to concern consumers, and has contributed to the lack of acceptance of GM foods, certainly in Europe. A survey by the UK consumer organisation Which? found that 88 per cent of respondents were worried about the risk of new food allergies developing as a result of genetic modification. Such concerns are related to the undoubted increasing prevalence of conditions such as allergic asthma (often experienced together with other symptoms like a runny nose and skin rashes) in the developed world, and perceptions that this also extends to food allergies. The emergence of food allergies as a significant concern in developed countries is relatively recent, having become particularly problematic only in the last ten years. The number of allergenic foods and the frequency of severe reactions are also rising, with headlines relating to individuals dying from eating peanuts being unheard of 20 years ago. However, there are indications that many more individuals perceive themselves to suffer from a food allergy than actually suffer from this condition.

The issue of GM-related allergies flared up for the first time in 1996, when researchers showed transfer of a major allergen from Brazil nut into soybeans also transferred its ability to trigger allergic reactions in individuals with pre-existing Brazil nut allergies. The gene in question, coding for the 2S albumin (also known as the allergen Ber e 1), was being used to improve the nutritional quality of soybeans for animal feed. This discovery prompted consumer groups, as well as scientists, to push for comprehensive testing of the potential allergenicity of GM foods.

Given its public health, social and economic consequences, the possible introduction of new allergens through genetic modification must be rigorously considered, in both developed and developing countries. The allergy issue has raised many general concerns about GM food, which have important policy implications. This policy brief is an attempt to put concerns relating to potential allergenicity into perspective.

What is meant by the term 'allergy'?

The human immune system is designed to protect us from infection by many different kinds of organisms, including bacteria, fungi and parasites. One protective mechanism is the production of antibodies; for example, the immunoglobulin E (IgE) antibody is specifically produced in response to parasitic infections, such as malaria and river blindness. In some people IgE is also produced in response to other substances such as pollen and certain foods, resulting in allergies. 

Contact between these IgE antibodies and an allergen, triggers the release of the chemical histamine, which is responsible for symptoms such as rashes or wheezing. In some individuals these allergic reactions can be life-threatening, causing anaphylactic shock. Such severe reactions are rare but can be triggered by consuming very small amounts of the relevant food, for example eating just a single peanut.

Food allergy should not be confused with 'food intolerance', of which the best-defined with an immunological component is the gluten intolerance syndrome coeliac disease. In other instances the body's immune system is not involved. For example, lactose intolerance — when the body is unable to digest lactose (milk sugar) — is caused by low levels of the necessary enzyme. Food intolerance can also be used to describe a much less well-defined collection of symptoms that some people suffer from on consuming certain foods for which no clear basis is known.

Patterns of food allergy

One important factor in the development of food allergies is the extent of exposure to allergens. It follows that incidence of food allergies will be related to patterns of food consumption. For example, allergy to buckwheat is much more common in the Far East (where it is an ingredient of noodles) than in Europe. The incidence of pollen-associated allergies also differs geographically, and relates to the presence of specific types of pollen and consumption of fruits and vegetables. In Sweden, for example, where birch pollen allergy is common, so too is apple allergy with which it is associated; but the latter is uncommon in Spain where there are few birch trees. This is because the protein found in birch pollen to which people are allergic is very similar to proteins found in the flesh of apples. Consequently IgE molecules can bind to both the birch pollen allergens and the apple proteins triggering allergic reactions in sensitive individuals when they either breathe in birch pollen or eat an apple.

There are no objective data on the prevalence of food allergies for either the developed or developing countries. Factors such as the genetics of the individual, patterns of food consumption and the incidence of allergic disease in general are probably important.    It has been suggested that exposure to germs, including parasitic infections, in childhood (through infection, playing in dirt, close proximity to animals, and so on) helps to 're-programme' the immune system so that it becomes less prone to allergies — also known as the 'hygiene hypothesis'. Certainly allergic asthma in Africa is an urban rather than a rural disease, supporting the view that the incidence of allergies depends on environmental factors.

Why are GM foods potentially allergenic?

There is a risk that introducing a foreign gene into, for example, a plant may cause it to become allergenic. The product of the new gene may also induce an allergic response in individuals who have previously been sensitised to a closely related substance.

This happens with non-GM material; for example, a significant proportion of individuals who are allergic to substances such as pollen go on to develop allergies to fresh fruits and vegetables. In the example given above in relation to birch pollen and apple allergies, this is because closely related (homologous) proteins are present in both pollen and the edible plant tissues of apples, hazelnuts, potato, avocados and so on. The IgE produced in response to the pollen protein can bind to the homologues in these tissues, thereby triggering an allergic reaction.

Alternatively, an allergenic transgene might be able to make an individual become allergic who was not previously, so that they start to produce IgE. As known allergens are almost all proteins, this is unlikely to apply to non-protein products derived from GM organisms, such as highly refined oils.

It has been proposed by some that GM technology might actually offer a way to prevent food allergies. In the early 1990s a 'hypoallergenic' transgenic rice variety was developed in Japan in which the major rice allergen gene had been deleted. More recently the US biotech company Pioneer has developed a GM soybean with a major allergen deleted.

Given the fact that many allergenic plant foods contain a number of allergens — some of which are important seed storage proteins (as is the case with peanuts) — the likelihood of hypoallergenic GM versions of all known allergenic foods seems unlikely. However, such approaches may be realistic for plant foods where many individuals react to the presence of a single protein.

Allergenic risk assessment of GM foods

In many countries an assessment of the allergenic risks posed by a novel food (including those developed by genetic engineering) must be performed before it can be released into the market. Unlike analyses of chemical toxicity such assessments can be difficult, as we do not yet understand the cellular and molecular basis of food allergy, and there are no adequate animal models for it. Unlike a human being, it is difficult to make a small animal, such as a rat or mouse, allergic by simply feeding it the food, and scientists often have to give something extra, such as the cholera toxin, with the food. This is probably because the human gastrointestinal tract processes foods in a different way and the gut immune system responds in a different way.

In 1996 a 'decision-tree' based approach was developed by the International Life Sciences Institute  in conjunction with the International Food Biotechnology Council. In the original 1996 decision-tree transgenes with a high probability of being a food allergen could still be marketed providing they were appropriately labelled. It has since been shown that such an approach is untenable due to lack of consumer acceptance and now any transgene showing homology with a known allergen is not pursued as a commercial option.

This decision tree was revised in 2001 by the joint FAO (Food and Agriculture Organization)/WHO (World Health Organization) Expert Consultation on Foods Derived from Biotechnology in its report 'Evaluation of Allergenicity of Genetically Modified Foods'.

Since our present knowledge of what makes one particular food protein, and not another, become an allergen is far from complete. It is not possible to simply take the structure of an unknown protein and accurately predict its allergenicity. Consequently, the decision-tree approach is based primarily on direct and indirect comparisons with structures and allergenic properties of known allergens.

However, until we have a fuller understanding of the mechanisms of allergic disease, the predictive value of certain aspects of the risk assessment (such as the targeted serum screen) is uncertain.  Subsequently in March 2002, the FAO-WHO Codex Alimentarius Commission 's task force on Foods Derived from Biotechnology decided against elaborating the revised decision tree. Instead they recommended an integrated case-by-case approach, which takes account of the evidence derived from several types of information and data, as no single criterion is sufficiently predictive of allergenicity. The 26th Session of the commission then adopted the FAO-WHO guidelines in July 2003 (ALINORM 03/34A) as part of broader recommendations on food safety and GM food. This document and future modifications are likely to set the global standard for assessing allergenic risks of GM food.

Conclusion

The evidence to date supports the view that the allergenic risks posed by GM foods are generally no greater than those posed by new crops and foods developed by traditional methods or other 'novel foods' (which the European Union defines as any food that has not previously been used for human consumption to a significant degree within the Union).

The allergenic risk assessment process currently used to determine the safety of candidate transgenes makes it highly unlikely that a 'novel' food allergen would be introduced into the market. However, it is clear that such an assessment process will be even more effective once our understanding of the molecular basis of allergic disease has improved. Further research is undoubtedly needed to develop more effective model systems for predicting food allergenicty to improve the risk assessment process.

Furthermore, a report by the UK's Royal Society identified the need for post-market monitoring of allergies following the introduction of a new GM food. Against a background of increasing rates of allergic disease, however, such monitoring would be difficult to implement, and would need to be a part of much wider studies into the epidemiology of food allergy. 

Nevertheless, on the basis of our current understanding, the allergenic risks posed by foods produced from GM crops currently being grown are minimal. The challenge now is for governments — particularly those in the developing world — to ensure that adequate risk assessments strategies are put in place.

Dr Clare Mills leads the Physical Biochemistry Group at the Institute of Food Research, based at Norwich in the United Kingdom. She has a long-standing interest in the role of food protein properties in food functionality, particularly the way in which plant protein structure and component interactions affect food protein allergenicity.