Science in Society Archive

Are Transgenic Proteins Allergenic?

Some two-thirds of all transgenic proteins have similarities to known allergens. Should we worry? Drs. Mae-Wan Ho, Arpad Pusztai, Susan Bardocz and Prof. Joe Cummins tell us why we should

Similarities to known allergens

A report published in 2002 [1] should raise concerns over the safety of the foreign proteins incorporated into GM crops that are commercially approved.

Researchers at the Institute of Food Safety in Wageningen, The Netherlands, screened transgenic proteins in GM food crops for the presence of short amino acid sequences identical to those in known allergens, and to find out if these are involved in binding IgE, the class of antibodies produced in allergic reactions.

They screened 33 transgenic proteins for continuous runs of at least 6 amino acids identical to known allergenic proteins. Twenty-two of the transgenic proteins showed positive results in runs of 6 or 7 amino acids; these include all the Bt toxins (Cry proteins), the CP4 EPSPS and GOX conferring glyphosate tolerance, the coat protein of the papaya ringspot virus, and even marker proteins such as GUS.

But on account of the limited data available, only a small number could be identified as linear epitopes (sites) that might bind to IgE antibodies. Although most identical stretches may be "false positives", the researchers said the results "warrant further clinical testing for potential allergenicity".

How reliable are current tests for potential allergenicity?

Potential allergenicity is one major aspect of safety assessment of GM crops. As many new proteins are introduced into GM food crops, it is important to find reliable methods of assessing their potential to cause allergic reactions, when eaten as food, through contact, or by inhaling (as pollen, for example).

One of the first steps in assessing if a protein is potentially allergenic is to compare its amino acid sequence with those of known allergenic proteins stored in computer databases, using available computer algorithms.

When such comparisons are made, identities of continuous runs of 8 or more amino acids are considered "immunologically relevant". But shorter stretches can also be relevant according to existing findings; for example, small sequences of four and six amino acids can be recognized and bound by IgE antibodies from allergic patients [2].

Apart from these continuous or linear epitopes, discontinuous epitopes may also be present, consisting of amino acids in different parts of the polypeptide chain that end up next to one-another when the polypeptide chain is folded up in its three-dimensional conformation. Thus, overall amino-acid similarity with an allergenic protein, i.e., 35% identity within a run of 80 amino acids, might be suspect. At the moment, it is difficult to predict which amino acids may form discontinuous epitopes, as we need to know the three-dimensional structure of the protein.

In addition to the linear and conformational peptide epitopes, glycans (carbohydrate chains linked to the protein) have also been shown to be major IgE binding sites in allergenic glycoproteins.

In a follow-up study published September 2004 [3], a new webtool was used to predict potential allergenicity of proteins and peptides according to the current recommendations of the FAO/WHO Expert Consultation, as outlined in the Codex Alimentarius [4, 5]. The Codex Alimentarius Commission was created by the United Nations FAO (Food and Agriculture Organization) and WHO (World Health Organization) to set international food standards.

The amino acid sequence of a protein is compared with all known allergenic proteins retrieved from the protein databases to identify stretches of 80 amino acids with more than 35% similarity, or small identical runs of at least 6 amino acids.

The ability of the procedure to predict allergens is evaluated by screening sets of known allergens and non-allergens. Apart from making correct predictions, both methods generated "false positive" and "false negative" hits. The number of false negatives decreases when a larger database of allergen sequences is used, whereas the number of false positives grows with the size of the database.

"False negatives", "false positives" and the need for precaution

The researchers point out that the number of false positives may be overestimated, because some of the 'non-allergens' used are related to and display similarities with their allergenic counterparts.

But that's precisely why we need to take any positive hits seriously. In fact, at least 5 of the 12 protein sequences used as 'non-allergens' were reported to react with other classes of antibodies, IgG and IgM, and are hence immunogenic, if not allergenic.

Another caveat, pointed out by the researchers, is that a protein belonging to a completely new group of allergens is likely to generate false negative results. This would apply to the majority of transgenic proteins that have never been part of our food chain.

As advised in the earlier publication, and also by the FAO/WHO, the outcomes should therefore be combined with other methods of assessing allergenicity, such as digestibility and binding of antisera from allergic patients, and possibly animal exposure tests. But that too, leaves a lot to be desired.

It is very difficult to assess the allergenicity of GM foods when the gene transferred into the plant is from an organism whose allergenic potential is unknown. Moreover, it is also possible that as a result of the gene transfer or insertion of the transgenic DNA, a new allergen is developed, or the expression of a minor allergen is elevated in the GM crop. The gene product can also have an allergenic adjuvant (helper) effect on a food component previously of low allergenic potential; or conversely, some component in the GM food may have an adjuvant effect on the allergenicity of the transgene product.

Unfortunately, while there are good animal models for nutritional/toxicological testing, no satisfactory animal models have so far been developed to test for allergenicity [6]. For the time being, only indirect methods are available for assessing the allergenic potential of GM foods derived from sources of unknown allergenicity. The screening tests described above are a useful preliminary step.

If the result is positive, then in vitro tests for IgE reaction need to be performed, especially as most epitopes are discontinuous. The absence of a positive in vitro reaction does not guarantee that the transgenic protein is not an allergen. In a decision-tree type of indirect approach, the next step is to consider the molecular size, glycosylation, stability, solubility and isoelecgtric point of the transgenic protein compared with known allergens [7]. Unfortunately, in most studies to-date, the all-important ability of the transgenic protein to resist breakdown in the gut is investigated in an in vitro simulated gastric/intestinal system [8, 9]; and this is fundamentally flawed. The results are therefore at best misleading and at worst erroneous. Reliance on the concept that most allergens are abundant proteins is probably also misleading because for example, Gadc1, the major allergen in codfish, is not a predominant protein [10].

In the absence of new and reliable methods for allergenicity testing, particularly the lack of good animal models, it is at present almost impossible to definitely establish
whether a new GM crop is allergenic or not in advance of its release into the human/animal food/feed chain.

In our view, with foods consumed by millions, any positive results should be assumed to be significant until fuller testing can definitively rule it out as a false positive. In North America and elsewhere, GM foods are not labelled and this may have led to the spread of allergens not identified as having originating with the GM foods that may in fact be the case.

Article first published 05/01/05


  1. Kleter GA and Peijnenburg Ad ACM. Screening of transgenic proteins expressed in transgenic food crops for the presence of short amino acid sequences identical to potential, IgE-binding linear epitopes of allergens. BMC Structural Biology 2002, 2:8
  2. Becker WM. Sequence homology and allergen structure (Topic 4). In Joint FAO/WHO Expert Consultation on Foods Derived from Biotechnology – Allergenicity of Genetically Modified Foods – Rome, 22 – 25 January 2001. Rome, Food and Agriculture Organisation of the United Nations, 2001.
  3. Fiers MWEJ, Kleter GA, Nijland H, Peijnenburg Ad ACM, Nap JP and van Ham R CHJ. Allermatch TM, a webtool for the prediction of potential allergenicity according to current FAO/WHO Codex alimentarius guidelines. BMC Bioinformatics 2004, 5:133
  4. FAO/WHO: Allergenicity of Genetically Modified Foods, 2001 [ ec_jan2001.pdf]. Rome, Italy, FAO/WHO
  5. FAO/WHO: CodexPrinciples and Guidelines on Foods Derived from Biotechnology, 2003 []. Rome, Italy, Joint FAO/WHO Food Standards Programme
  6. Helm RM and Burks AW. Mechanisms of food allergy. Current Opinion in Immunology 2000, 12, 647-53.
  7. O'Neil C, Reese G and Lehrer SB. Allergenic potential of recombinant food proteins. Allergy and Clinical Immunology International 1998, 10, 5-9.
  8. Astwood JD, Leach JN and Fuchs RL. Stability of food allergens to digestion in vitro. Nature Biotechnology 1996, 14, 1269-1273.
  9. Metcalf DD, Astwood JD, Townsend R, Sampson HA, Taylor SL and Fuchs RL. Assessment of the allergenic potential of foods derived from genetically engineered crop plants. In: Critical Reviews in Food Science and Nutrition 36(S) 1996, S165-86. CRC Press Inc. Boca Raton, USA.
  10. Bindslev-Jensen C and Poulsen LK. Hazards of unintentional/intentional introduction of allergens into foods. Allergy 1997, 52, 1184-6.

Got something to say about this page? Comment

Comment on this article

Comments may be published. All comments are moderated. Name and email details are required.

Email address:
Your comments:
Anti spam question:
How many legs on a spider?