This Opinion follows a request from the European Food Safety Authority (EFSA) to the Panel on Genetically Modified Organisms (GMO Panel) to establish a Working Group on “The Assessment of allergenicity of genetically modified foods”.
The strategy summarised in this report for assessing the allergenicity of GM food and feed considers the allergenicity of the newly expressed proteins, the whole GM food and feed, and also other aspects, such as exposure. Particularly with regard to newly expressed proteins, it is based on a weight-of-evidence, case-by-case approach, in line with the approach followed in the EFSA guidance document and the Codex Alimentarius guideline. Different aspects of the allergenicity risk assessment of GMOs are discussed in this report, and where relevant, additional recommendations are provided to further strengthen the weight-of-evidence approach. These include recommendations with regard to the search for sequence homology and structural similarities, IgE binding tests and testing of the digestibility of newly expressed proteins. The recommendations also pertain to the comparative analysis of endogenous allergens and their structural characteristics in whole GM food and feed. In addition, proposals have been made with regard to other additional testing that may improve the assessment, e.g. animal models, as well as suggestions for further exploring new promising methods that are as yet in an early phase of development, particularly on 3-dimensional structure of allergens, profiling technology, exposure assessment and post-market monitoring.
The background information, scientific evidence and rationale on which the conclusions and recommendations are based, can be found in the annexes.
Annex 1 (Clinical aspects of food allergy) deals with the mechanisms, the prevalence and clinical pictures of food allergy, as well as the diagnostic procedures. Within the assessment of allergenicity of GMOs, attention is paid to the risk for populations with a particular or altered digestive physiology such as infants. In addition, the possible role of GMOs as adjuvants, i.e. substances that, when co-administered with a protein increases its immunogenicity and therefore might increase as well its allergenicity, is considered within the allergenicity assessment of GMOs. Where possible, the use of surveillance programmes such as post-market monitoring and survey of occurrence of occupational allergic reactions through different routes of exposure is proposed.
Annex 2 (Structural aspects of food allergens: conformation, in planta processing and food matrix interactions) addresses the structure and biological properties of a newly expressed protein in an integrative approach for assessing its possible allergenicity. Current understanding of how the allergenic potential relates to protein structure, biological properties, post-translational modifications and plant processing, and how it may be affected by food processing and interactions with the food matrix is presented. This is discussed within the context of the natural variability and taking into account that genetics, environmental factors and post-harvest conditions may affect the expression of allergens in plants. Therefore, the considerations developed in this Annex do also pertain to the assessment of allergenicity of the whole plant.
Annex 3 (Bioinformatics for the assessment of allergenicity of newly expressed proteins in GMOs) covers various bioinformatics methods available including allergen databases and algorithms for search of sequence identity of newly expressed proteins with known allergens and the assessment of the relevance of alignments observed. A recommendation is made that allergen databases are regularly updated and checked for accuracy (e.g. absence of irrelevant allergens and presence of minor true allergens with low frequency of sensitisation). Criteria for inclusion of proteins into allergen databases and for sequence identity searches are discussed as to improve the accuracy, sensitivity and specificity of current bioinformatics approaches. Bioinformatics analyses are not appropriate for the assessment of the de novo sensitisation potential of a newly expressed protein but provide useful information on the possible IgE-cross-reactivity with known allergens. Such information should be confirmed by other tests to conclude on the likelihood of allergenicity of the newly expressed protein(s).
Annex 4 (Assessment of allergenicity of newly expressed proteins in GMOs using in vitro and cell-based tests) describes in vitro methods that can be used for assessing the allergenicity of newly expressed protein(s). This includes a discussion of different IgE binding assays and criteria for serum selection, and an overview of experimental conditions for performing and interpreting the outcome of in vitro digestibility tests. IgE-binding assays aim to test for possible IgE-cross reactivity of the newly expressed protein with known allergens. Specific serum screening is required if there is any indication of relationship or structure similarity with known allergens as evidenced in the previous step of the assessment, or if the source of the transgenic protein is considered allergenic, i.e. is known to produce one or more allergen(s). In view of the problems associated with the use of sera of human origin, the possibility of using well-characterised antibodies raised in animals for a pre-screening may be envisaged. Stability towards in vitro digestibility can provide additional information about the possible allergenic potential of the newly expressed protein. The conditions that should be observed for performing those tests and correctly interpreting the outcome are discussed in relation with the clinical considerations presented in Annex 1. Cell-based assays that employ either cells isolated from human or animal tissues or propagated from immortal cell lines are also considered.
Annex 5 (Analytical and profiling technology/in vitro protein analysis and proteomics methods for the allergenicity assessment of the whole GM plant) addresses analytical methods and profiling techniques for assessing the potential increase of the intrinsic allergenicity of the whole plant and derived products as an unintended effect of the genetic modification. When the recipient of the introduced gene is known to be allergenic, a potential quantitative and/or qualitative change in the allergenicity of the whole GM food should be tested by comparing the allergen repertoire of the GM plant with that of appropriate comparator(s) taking into account the natural variability in the endogenous allergen expression. Allergens in whole plants can be analysed based on their immunochemical and biological properties with gel-electrophoresis followed by immunoblotting. Alternative proteomic methods using, e.g. mass spectrometry for identifying and quantifying allergens are also reviewed. High throughput analytical techniques are a promising tool for non-targeted profiling of differences in protein expression in the GM plant compared with its appropriate comparator(s). It is proposed that they should be thoroughly assessed for accuracy, sensitivity, specificity and feasibility before being routinely used for allergenicity assessment.
In Annex 6 (Animal models) a review and a critical appraisal is made of the numerous animal models (including transgenic animals) that are currently available to test the capacity of a newly expressed protein to elicit an allergic reaction in allergic consumers already sensitised to a cross-reactive allergen or its potential to de novo sensitise predisposed individuals. Advantages and pitfalls of the different models are analysed. It is noted that none of these models fully reproduce either the diversity and variability of the IgE response in heterogeneous populations of allergic humans or the conditions of sensitisation that occur in the real life upon given conditions of exposure and environment. In addition, none of the current animal models has both enough sensitivity and specificity in order to guarantee the absence of false negative and false positive results. Animal models are, therefore, in general considered not validated and inconclusive for the assessment of the sensitising potential of a novel protein. However, they can provide useful information on the different mechanisms underlying the induction and development of an allergic reaction when there are indications of a sensitising or adjuvant potential of the newly expressed proteins.