Download pdf - allergenic potential OMG.pdf

7/27/2019 allergenic potential OMG.pdf

1/5

FORUM

Approaches to Assessment of the Allergenic Potential of NovelProteins in Food from Genetically Modied Crops

Ian Kimber 1 and Rebecca J. Dearman

Syngenta Central Toxicology Laboratory, Alderley Park, Maccleseld, Cheshire SK10 4TJ, United Kingdom

Received December 31, 2001; accepted February 13, 2002

The safety assessment of food derived from genetically modiedplants continues to attract considerable attention. Among theimportant issues that need to be considered is whether the prod-ucts of novel genes introduced into crop plants will have thepotential to induce allergic sensitization or to elicit allergic disease.Hierarchical approaches to allergenicity testing have been pro-posed, and these incorporate evaluation of the structural andsequence homology and serological identity of novel proteins withknown allergens, measurement of resistance to proteolytic diges-tion, and assessment of allergenic potential using animal models.Accounts of these approaches are available elsewhere, and it is notthe purpose of this article to provide a detailed critique of specicmethods. Our intention is rather to look more broadly at thestrategy for assessment of allergenic potential, the challenges suchassessments pose for the practicing toxicologist, and how some of

these might best be addressed.

The safety assessment of foods derived from geneticallymodied crops has been the subject of considerable discussionand debate (Goldman, 2000; Hodgson, 2001; Kuiper et al .,1999; Lachmann, 1999). Among the aspects that have attractedmost attention is allergenicity. The concern here is that geneticmodication of crop plants may in some instances be associ-ated with a risk of either inducing allergic sensitization insusceptible (but previously nonsensitized) individuals, or withthe elicitation of allergic reactions in those who are alreadysensitized. Specically, the concerns are that: (a) the product of a novel gene introduced into the plant may have the ability toinduce de novo sensitization amongst susceptible consumers,(b) that the product of a novel gene may be immunologicallycross-reactive with protein allergens to which consumers arealready sensitized, and will as a consequence have the potentialto provoke allergic reactions, or (c) that transformation will

result in the altered levels of endogenous protein allergensalready expressed by the host plant.

The requirement for allergenicity assessment of novel foodshas been reviewed previously (Kimber and Dearman, 2001a;Kimber et al ., 1997; Metcalfe et al ., 1996; Taylor, 1997;Taylor and Hee, 2001), and it is not our intention to cover thesame ground here. Nor is it the purpose of this article tocompare and contrast in detail the utility of specic methodsthat have been proposed for hazard identication and charac-terization. Our goal is rather to reect on some of the moregeneral aspects of protein allergenicity assessment, the chal-lenges they pose, and how some of these might best be ad-dressed.

Scientic Context

The concern that a novel gene product may have the poten-tial to induce sensitization is legitimate: food allergy is notuncommon, the prevalence being in the order of 12% in adultsand even higher among infants (Helm and Burks, 2000; Houri-hane, 1998), and a variety of plant proteins have been impli-cated as food allergens (Breiteneder and Ebner, 2001; Bushand Hee, 1996). However, it is important to appreciate that anormal diet will result in exposure to many thousands of proteins and that only a small proportion of these have beenshown to display sensitizing activity. The question that arisesfrom this is why some proteins are allergenic, whereas othersare not. This is of more than academic interestan understand-ing of the requirements for the initiation of sensitization willprovide the foundations for future developments in approachesto the identication and characterization of allergenic proteins,and may create opportunities for designing out sensitizingproperties.

In the context of food allergy, sensitization is dependent inmost instances on the induction of IgE antibody responses.This antibody associates with mast cells that are foundthroughout vascularized tissues, including the gastrointestinaltract. Following subsequent exposure, the inducing allergen

1 To whom correspondence should be addressed. Fax: 44 1625 590996.E-mail: [email protected].

TOXICOLOGICAL SCIENCES 68, 48 (2002)Copyright 2002 by the Society of Toxicology

4


2/5

cross-links membrane-bound IgE antibody, resulting in de-granulation and the release of a variety of inammatory me-diators, including histamine, serotonin, chemotactic factors,and prostaglandins. These factors act in concert to cause anincrease in vascular permeability, the contraction of smoothmuscle, and the accumulation of leukocytes that together pre-

cipitate an allergic reaction. The most frequent symptoms of food allergy include nausea and vomiting, abdominal pain, anddiarrhea. However, other organ systems (the skin and respira-tory tract) may be involved, and occasionally severe systemic(anaphylactic) reactions are provoked that may prove fatal.

It is generally acknowledged that a number of factors cancontribute to the expression by proteins of sensitizing activity,although not all of these may be inuential in each instance.Among these are the size and structure of proteins, resistanceto proteolytic digestion, glycosylation status, biologic function(such as enzymatic activity), overall immunogenicity (the abil-ity to provoke an immune response of any type), and the wayin which the protein is recognized, processed, and presented tothe immune system (Aalberse, 2000; Bredehorst and David,2001; Bufe, 1998; Huby et al ., 2000). It is assumed that these,and possibly other as yet unidentied factors, will collectivelydetermine the immunological outcome of exposure, andwhether and to what extent an allergic response is provoked. Inpractice, the assumption is that protein allergens have theproperties necessary to induce and sustain the quality of im-mune response required for IgE antibody production (a pref-erential type 2 response) (Stevens et al ., 1988) and/or toprohibit the genesis of type 1 immune responses that antago-nize the elaboration of IgE (Aalberse, 2000; Huby et al ., 2000;Kimber and Dearman, 2001a).

Safety Assessment: Available Approaches andGeneral Considerations

Perhaps the most important question to initially ask is whata safety assessment should have as an objective. The absenceof risk in this context is not achievable, and our view is that theaim, holistically, should be to ensure that a food deriving froma genetically modied crop is as safe as its traditional coun-terpart. With this in mind, the objective with regard to aller-genicity is to establish whether the novel food has an increasedpotential to induce sensitization or to elicit allergic reactions.

The rst structured approach to allergy safety assessmentresulted from a collaboration between the International FoodBiotechnology Council (IFBC) and the International Life Sci-ences (ILSI) Allergy and Immunology Institute. A tiered ap-proach was proposed in which the route taken was dictated bywhether or not the protein of interest derived from a source thathad previously been associated with allergic disease in humans(Metcalfe et al ., 1996). In such cases, the suggestion was thatthe identity of the protein of interest with proteins recognizedby patients allergic to the source of the transferred gene shouldbe investigated serologically (that is, reactivity of IgE antibody

from sensitized patients with the protein of interest) and pos-sibly by challenge of sensitized subjects. For proteins derivedfrom sources considered not to be allergenic, a different strat-egy was proposed. Here, the advice was to consider amino acidsequence homology with, and/or structural similarity to, knownprotein allergens (Gendel, 1998), and the resistance of the

protein of interest to digestion in a simulated gastric uid(Astwood et al ., 1996). Since the original description of thisdecision tree, modications have been proposed, among thesebeing more conservative criteria for establishing identity withknown allergens based on sequence homology, and the intro-duction of what was termed targeted serum screening toinvestigate the immunoactivity of the novel protein with seraprepared from subjects with allergy to materials that arebroadly related to the source material (FAO/WHO, 2001).More recently still, the Codex Ad Hoc Intergovernmental Task Force on foods derived from biotechnology has, through aworking group on allergenicity, given further consideration tosome of the recommendations made previously by IFBC/ILSIand by FAO/WHO.

For the purposes of this article, much of the detail is notdirectly relevant, and it is in any event available elsewhere(FAO/WHO, 2001; Metcalfe et al ., 1996). The overarchingquestion is whether the recommendations that resulted origi-nally from the IFBC/ILSI consultations, and/or some of themore recent suggested amendments to these, provide a suitableframework for ensuring that no new or increased risk of allergywill result from the introduction of novel foods. Our view isthat with certain provisos and caveats, these paradigms form asound, but as yet incomplete, basis for safety assessments.

One important issue is that both the initial and modied

recommendations for evaluation of allergenicity were dis-played, and have since been viewed, as a decision tree. Al-though a binary systemwherein the results of one investiga-tion dictate the route taken for further analyseshas someattraction, this hierarchical approach does not necessarily en-courage an holistic appreciation of the data available in reach-ing conclusions about the existence or otherwise of a potentialhazard. Components of the current decision trees do not nec-essarily provide an unambiguous indication of the presence orabsence of sensitizing hazard.

Thus, for instance, it is recognized that there are importantlimitations in relating sensitizing potential to a linear aminoacid sequence. Moreover, the apparent association betweenallergenic activity and resistance to digestion by pepsin or asimulated gastric uid is by no means absolute, and certainlynot all proteins that display resistance to proteolytic digestionhave the ability to induce allergic sensitization. The relation-ship between proteolytic stability and sensitizing potential isprobably worthy of further comment, as it serves to illustratethat our understanding of the factors that inuence allergenicactivity is as yet far from complete. It is generally accepted thatthe positive correlation between resistance to proteolytic di-gestion (measured frequently using a simulated gastric uid or

5FORUM


3/5


4/5

investigations it was found that when administered by gavage,ovalbumina known food allergenfailed to elicit IgE anti-body responses in BN rats and, compared with responsesprovoked by intraperitoneal exposure, elicited only low-gradeIgE responses in BALB/c mice (Dearman et al ., 2001). Al-though others have had greater success at eliciting IgE re-

sponses following gavage exposure of rats and mice to ovalbu-min (Akiyama et al ., 2001; Penninks and Knippels, 2001), ourview remains that, for the purposes of hazard identication,oral administration lacks sufcient sensitivity. Our preferredstrategy, therefore, is to administer test proteins by intraperi-toneal injection, under which conditions the opportunities forthe generation of vigorous immune responses are optimized.

Using this approach, the assessment of potential allergenic-ity is based on an evaluation of the ability of the test materialto provoke an IgE antibody response (as measured by homol-ogous passive cutaneous anaphylaxis). Importantly, however,such measurements are made within the context of the overallimmunogenicity of the test protein, and for this purpose spe-cic IgG antibody responses are also measured. The interpre-tation of such data is presently relatively unsophisticated, in-sofar as proteins with the inherent potential to induce allergicsensitization are dened as those that are able to stimulate ameasurable IgE antibody response. Conversely, proteins thatare clearly immunogenic in terms of eliciting an IgG antibodyresponse, as most foreign proteins will do, but which fail underthe same conditions of exposure to induce IgE antibody pro-duction, are viewed as having no (or at least a very limited)inherent potential to cause allergic sensitization.

It must be acknowledged that despite some considerableprogress, animal models have yet to be fully evaluated or

validated. However, work is continuing apace, and alreadythere are opportunities to combine the results of animal exper-iments with other sources of data about novel proteins toinform the safety evaluation process.

Towards an Holistic Approach to Safety Assessment

The question remaining is to what extent, with the toolscurrently available, is it possible to make informed judgmentsregarding the sensitizing potential of novel proteins. We areperhaps some little way from having robust paradigms, but it isrelevant to examine what may already be possible. Two hypo-thetical examples serve to illustrate this. Let us consider anovel protein that is found to be resistant to digestion by pepsinor in a simulated gastric uid, that displays some homologywith known human allergens, and that is able in mice or rats toprovoke a vigorous IgE antibody response. Collectively, thesedata would indicate that this protein has a signicant potentialto cause sensitization. At the other end of the spectrum, ahypothetical protein may be very labile and lack any sequenceor structural homology with known allergens. If, combinedwith this, the protein was found not to elicit the production of IgE antibody production under conditions of exposure where a

strong IgG antibody response was induced, then the conclusionmight be that the protein lacked any signicant inherent po-tential to stimulate the quality of immune response necessaryfor IgE production and the acquisition of sensitization.

Two points need to be made. First, the above paradigm israther conservative, insofar as it has to be recognized that

proteins identied as potential hazards, as a function of theirability to induce IgE antibody responses following intraperito-neal injection of mice, may not represent a risk of sensitizationto humans following dietary exposure. That is, the protein maylack the ability to stimulate IgE responses when encountered inthe gastrointestinal tract. Notwithstanding this, the currentvalue of this approach is that proteins that fail to stimulate IgEantibody production, in circumstances where vigorous IgGresponses are elicited, may be regarded as having little intrinsichazard. Second, while interpretation of polarized responsesmay appear relatively straightforward, intermediate responsespose more of a challenge. Increased experience will clearly beneeded to translate this into a robust paradigm that is applicablein a wide range of circumstances.

Concluding Comments

It will be apparent that much has still to be achieved.However, during the last ve years there has been a growingand substantial interest in the safety assessment of novel pro-teins generally, and in particular in the assessment of allergenicpotential. With continued investment we can anticipate accel-erated progress in the coming years. Such will be facilitated bythe following: (a) a detailed appreciation of the molecular andstructural characteristics that confer on proteins the ability to

induce allergic sensitization, (b) the further development of appropriate databases for consideration of the structural andsequence homology of novel proteins with known allergens,(c) standardization of methods for measuring the resistance of proteins to proteolytic digestion and the relationship of thischaracteristic with the genesis of allergic responses, and (d) thefurther calibration, renement, and evaluation of appropriateanimal models that, in conjunction with other data, are able toinform the safety assessment process.

REFERENCES

Aalberse, R. C. (2000). Structural biology of allergens. J. Allergy Clin. Immunol. 106, 228238.

Akiyama, H., Teshima, R., Sakushima, J.-I., Okunuki, H., Goda, Y., Sawada,J.-I., and Toyoda, M. (2001). Examination of oral sensitization with ovalbu-min in Brown-Norway rats and three strains of mice. Immunol. Lett . 78, 15.

Astwood, J. D., Leach, J. N., and Fuchs, R. L. (1996). Stability of foodallergens to digestion in vitro . Nat. Biotechnol . 14, 12691273.

Basketter, D. A., Gerberick, E. F., Kimber, I., and Willis, C. M. (1999).Toxicology of Contact Dermatitis. Allergy, Irritancy and Urticaria . JohnWiley and Sons, Chichester, U.K.

Bredehorst, R., and David, K. (2001). What establishes a protein as anallergen? J. Chromatogr. B Biomed. Sci. Appl . 756, 3340

7FORUM


5/5