Biochemical properties of peanut allergens

Food allergens and their corresponding IgE-binding epitopes are thought to possess physicochemical properties that confer resistance to digestive enzymes and thermal processing. This increases their allergenicity by enhancing their ability to reach the intestinal mucosa. Ara h 1 is capable of forming stable dimers, trimers and larger complexes upon heating without affecting IgE reactivity (Ref. Reference Koppelman32). Treatment of Ara h 1 with gastrointestinal enzymes such as pepsin, trypsin and chymotrypsin produces large proteolytic fragments that retain binding affinity for serum IgE from peanut-allergic individuals (Ref. Reference Maleki33). The resistance of Ara h 1 to degradation following heating and treatment with digestive enzymes may be related to its stable, homotrimeric structure (Ref. Reference Maleki33). This monomer–monomer interaction reduces access to the catalytic sites within the protein, allowing Ara h 1 to survive intact during food processing or passage along the digestive tract, thus contributing to its potency as an allergen.

Similarly, Ara h 2 and Ara h 6 retain allergenicity following proteolytic digestion as well as heat treatment (Refs Reference Suhr34, Reference Lehmann35). Structural studies revealed that following digestion, both allergens yielded immunologically active core structures that were able to induce release of inflammatory mediators even in the context of decreased serum IgE binding (Ref. Reference Lehmann35). Ara h 2 shares sequence homology with trypsin inhibitors and can act as a weak trypsin inhibitor with increased activity following roasting, thus protecting itself as well as Ara h 1 from trypsin digestion (Ref. Reference Maleki36). This function of Ara h 2 may contribute to the overall allergenic properties of peanut proteins through increased resistance to digestive enzymes. By contrast, Ara h 8 has low stability to heat and gastric digestion (Ref. Reference Mittag30).

The Maillard reaction is also a contributing factor to the allergenicity of peanuts (Refs Reference Chung and Champagne37, Reference Maleki38, Reference Chung39). This is a nonenzymatic reaction between a protein and a reducing sugar that occurs during thermal processing and cooking (Ref. Reference Namiki40). The amino groups of proteins become glycosylated to form Amadori products, which degrade into dicarbonyl intermediates. These intermediary compounds react with amino groups of proteins to form stable end products known as advanced glycation end products. Chung and Champagne (Ref. Reference Chung and Champagne37) demonstrated that advanced glycation end products, formed by heating a previously nonallergenic peanut lectin protein in the presence of sugars, can effectively compete with untreated peanut allergens for serum IgE antibodies, suggesting that the Maillard reaction can convert a nonallergenic protein into a potentially allergenic protein. Similarly, Maleki and colleagues observed that roasted peanut proteins inhibited IgE binding to raw peanut proteins more effectively (90-fold higher) than raw peanut proteins (Ref. Reference Maleki38). Both studies concluded that the presence of advanced glycation end products in heat-treated peanuts contributed to their overall heightened allergenicity.

T-cell epitopes

By contrast to B-cell epitopes, data on the T-cell epitopes of peanut allergens are extremely limited. To date, dominant sites of T-cell reactivity have been reported for only one peanut allergen, Ara h 2, and the core T-cell epitopes within these sites have not yet been identified. Glaspole et al. identified two highly immunogenic T-cell-reactive regions: Ara h 2 (19–47) and Ara h 2 (73–119). Peptides spanning these two regions induced strong T-cell proliferation associated with a Th2-type cytokine response (Ref. Reference Glaspole47). Knowledge of the dominant T-cell epitopes of allergens is critical information for the development of a T-cell-targeted vaccine for peanut-specific allergen immunotherapy.

Crossreactivity between peanut and tree nuts

Clinical studies indicate that peanut-allergic subjects, particularly adults, usually exhibit multiple sensitivities to both peanut and tree nuts. In specialist allergy clinics, the prevalence of co-allergy to peanut and tree nuts was reported as 35–40% (Refs Reference Ewan48, Reference Sicherer, Burks and Sampson49), although in our experience up to 80% of adult peanut-allergic patients are clinically allergic to one or more tree nuts (R. O'Hehir, unpublished). Sicherer and colleagues examined the serology of peanut-allergic children and found significant correlations between the level of peanut-specific IgE and tree-nut-specific IgE antibodies. In particular, peanut-specific IgE levels correlated with IgE levels for hazelnut, Brazil nut and almond (Ref. Reference Sicherer, Burks and Sampson49). Given this, several studies have investigated the presence of crossreactive allergens in peanut and various tree nuts. IgE crossreactivity was demonstrated between peanut and the tree-nuts pistachio, macadamia, almond, Brazil nut and hazelnut (Refs Reference Parra50, Reference Sutherland51, Reference de Leon52). In the majority of these cases the identity of the crossreacting allergen(s) was not established, with the exception of Ara h 2, which shares IgE-binding epitopes with almond and Brazil nut allergens (Ref. Reference de Leon53).

Although peanut is a legume and is taxonomically distantly related to tree nuts (Fig. 3), they are both defined as ‘edible’ seeds and perform similar functions in plant development. Indeed, several known peanut and tree-nut allergens belong to the same protein families. The major peanut allergen, Ara h 1, is a member of the vicilin family of seed storage proteins (Ref. Reference Burks24) together with allergens from walnut, cashew and hazelnut (Refs Reference Teuber54, Reference Wang55, Reference Pastorello56). Also, 2S albumin seed storage proteins, implicated as allergens in almond, Brazil nut, hazelnut and walnut (Refs Reference Pastorello56, Reference Poltronieri57, Reference Pastorello58, Reference Teuber59), are related to conglutin seed storage proteins, which include allergens from almond (Ref. Reference Poltronieri57) and peanut – namely Ara h 2, Ara h 6 and Ara h 7 (Refs Reference Burks25, Reference Kleber-Janke29). Legumins or 11S globulins are seed storage proteins shown to be allergenic in peanut (Ara h 3), cashew, hazelnut and walnut (Refs Reference Rabjohn27, Reference Pastorello56, Reference Wang60, Reference Beyer61, Reference Teuber62). Therefore, it is highly likely that a number of peanut allergens are responsible for the observed IgE crossreactivity between peanut and certain tree nuts.

Figure 3

Taxonomic classification of peanut and tree nuts. Many people suffering from peanut allergy are also allergic to one or more tree nuts. Although peanut and a number of tree nuts belong to different plant families, they arise from the same class, namely Magnoliopsida, and therefore may contain homologous proteins that play a role in allergenic crossreactivity. Tree nuts that have been shown to share IgE-binding epitopes with peanuts include almond, Brazil nut, hazelnut, macadamia and pistachio. The classification shown is based on information at the website of the US Department of Agriculture (http://plants.usda.gov/).

Clinical and biological relevance of IgE crossreactivity

Whether or not in vitro allergen crossreactivity translates into clinical hypersensitivity is controversial. In addition to peanut, the legume family includes soybean, pea, lima bean and green bean; however, although these foods are closely related taxonomically, clinical hypersensitivity to more than one legume is rare (Refs Reference Bernhisel-Broadbent and Sampson63, Reference Bernhisel-Broadbent, Taylor and Sampson64). van der Veen and colleagues observed that a third of patients sensitised to grass pollen (including timothy and orchard grass as confirmed by serum specific IgE test) had significant serum levels of peanut-specific IgE antibodies but no clinical symptoms (Ref. Reference van der Veen65). This phenomenon was attributed to crossreactive carbohydrate determinants (CCDs) – N-linked carbohydrate groups of glycoprotein allergens that induce the production of crossreactive IgE antibodies to food and grass pollen allergens (Refs Reference Batanero66, Reference Petersen67). Subjects with true peanut allergy had minimal levels of IgE antibodies specific for CCDs, in contrast to those with false-positive skin prick tests to peanut (Ref. Reference van der Veen65).

IgE antibodies to CCDs also appear to have negligible biological activity. In the same study by van der Veen and colleagues, higher concentrations of peanut extract (∼1000-fold) were required for basophil histamine release in patients with high levels of anti-CCD IgE antibodies with no peanut allergy compared with peanut-allergic patients, further validating the assertion that crossreactive IgE antibodies directed at CCDs are not indicative of clinical sensitivity. However, biological activity of peanut-specific IgE antibodies that crossreact with tree-nut allergens was demonstrated by basophil-activation assays (Ref. Reference de Leon68). Human basophils passively resensitised with peanut-specific serum IgE antibodies from peanut- and tree-nut-allergic subjects become activated following incubation with almond and Brazil-nut extracts, as indicated by detection of CD63 surface expression by flow cytometry (Ref. Reference de Leon68). Although the threshold doses of peanut, almond and Brazil-nut extracts required for basophil activation were not compared, it is highly likely that crosslinking of peanut-specific IgE antibodies present on effector cells by tree-nut allergens is clinically relevant given the high incidence of cosensitisation to peanut and tree nuts in peanut-allergic subjects.

Modified peanut allergens with reduced IgE reactivity

Given the high-affinity IgE binding of peanut allergens, there is a need to develop suitable allergen preparations that do not crosslink IgE antibodies bound to effector cells. One approach is the engineering of recombinant allergens with conserved T-cell reactivity and decreased IgE binding. Knowledge of the IgE-binding epitopes of peanut allergens has allowed preparation of hypoallergenic variants with decreased IgE binding (Refs Reference Burks, King and Bannon69, Reference King70).

PCR mutagenesis targeting linear IgE-binding epitopes was utilised to produce a hypoallergenic Ara h 2 mutant; however, only 75% efficacy was achieved, with serum IgE from 12/16 patients demonstrating a considerable decrease in reactivity to the Ara h 2 mutant in comparison with the wild-type allergen (Ref. Reference Burks, King and Bannon69). Similarly, another hypoallergenic Ara h 2 mutant was produced but this variant also retained IgE reactivity in a small cohort of peanut-allergic subjects and induced significant mediator release from basophils passively sensitised by peanut-allergic sera (Ref. Reference King70).

Site-directed mutagenesis was utilised in the development of a hypoallergenic variant of Ara h 3. A 40  kDa acidic subunit of Ara h 3 containing the four identified Ara h 3 linear IgE-binding epitopes was expressed using a bacterial system, with the critical residues for IgE binding targeted for point mutations by substitution with an alanine residue (Ref. Reference Rabjohn71). The resulting modified protein was still IgE-reactive although binding was substantially decreased in comparison with the wild-type protein (Ref. Reference Rabjohn71). A hypoallergenic form of Ara h 1 with T-cell reactivity has also been produced (Ref. Reference Bannon72) but the characterisation of this modified allergen has been limited.

The retention of IgE reactivity with the Ara h 2 and Ara h 3 variants following mutation of the linear IgE-binding epitopes suggests that conformational B-cell epitopes should also be disrupted to avoid IgE-mediated adverse reactions in patients. In view of the multiple sensitivities to peanut allergens in most patients (Ref. Reference Lewis73), it is likely that a cocktail of hypoallergenic mutants would be required in a vaccine.

T-cell-epitope peptide vaccines

An alternative approach to specific immunotherapy uses short T-cell-epitope peptides that lack IgE binding. This strategy shows promise in the treatment of certain allergies and is particularly attractive because the peptides are too short to crosslink effector-cell-bound IgE antibodies. In preclinical studies, the administration of peptides based on dominant T-cell epitopes of allergens from cat and house dust mites induced specific T-cell tolerance in mice (Refs Reference Briner74, Reference Hoyne75). This approach has been encouraging in clinical trials using allergen peptides for cat allergy (Refs Reference Oldfield, Larche and Kay76, Reference Alexander77) and bee venom allergy (Ref. Reference Muller78), and experimental data with a potent allergen of natural rubber latex have shown some promise (Ref. Reference Drew79). Treatment with multiple doses of cat allergen (Fel d 1) peptides inhibited early- and late-phase allergic reactions to the whole allergen, with an associated decrease in Th1 and Th2 responses and an increase in IL-10 production by peripheral blood mononuclear cells (Ref. Reference Oldfield, Larche and Kay76). Furthermore, this treatment improved nasal symptom scores in subjects with allergic rhinitis (Ref. Reference Alexander77). In the case of allergy to bee venom, immunotherapy using T-cell-epitope peptides of the major bee venom allergen, phospholipase A₂ (PLA₂), protected patients from challenge with PLA₂ (Ref. Reference Muller78). Concurrent decreases in PLA₂-allergen-induced T-cell proliferation and cytokine secretion were detected, consistent with T-cell anergy.

A similar therapeutic approach for peanut allergy requires full characterisation of the dominant T-cell epitopes of the major peanut allergens. While dominant T-cell epitopes of Ara h 2 have been reported (Ref. Reference Glaspole47), there is no information available on Ara h 1 T-cell epitopes. In addition, if Ara h 3 is also confirmed as a major peanut allergen, T-cell epitopes of Ara h 3 will also need definition to ensure that the necessary dominant T-cell-reactive sites are included in a potential vaccine.

Nonspecific immunotherapeutic strategies

Non-allergen-specific strategies have also been considered for the treatment of peanut allergy. Anti-human-IgE therapy showed promise in studies of asthma and allergic rhinitis. One preparation used in such therapy is TNX-901, a humanised IgG₁ monoclonal antibody specific for an epitope in the CH3 region of IgE – the region that binds to the high-affinity FcɛRI receptor on mast cells and basophils (Ref. Reference Leung80). In one study on peanut-allergic patients, TNX-901 was administered subcutaneously every 4 weeks for 16 weeks and subsequent peanut food challenges demonstrated a significant increase in the threshold dose of peanut flour eliciting symptoms (the new threshold dose was equivalent to approximately six to eight peanuts) (Ref. Reference Leung80). Although this treatment was not curative, the risk of a severe or fatal reaction after accidental ingestion was diminished. The nonspecific nature of this form of treatment may be advantageous in patients with food allergies where a strict diet is difficult to manage and specific immunotherapy is not available, as is the case with peanut allergy. Conceptually, it has appeal for patients who are extremely sensitive to more than one allergen from any given source. However, high patient commitment and compliance would be needed and the cost of such therapy is considerable. Currently, an alternative humanised anti-IgE antibody is being tested for peanut allergy for future therapeutic use (Ref. Reference Teuber and Beyer81). Other nonspecific therapies reported to suppress peanut-induced anaphylaxis in murine studies include the administration of IL-12 (Ref. Reference Lee82) and treatment with a Chinese herbal formula, FAHF-1 (Ref. Reference Li83), but efficacy and safety in humans have not been shown.

Humanised anti-IgE antibody treatment might be useful as an adjunct during the induction phase of conventional injection allergen-specific immunotherapy. Intuitively, the high potency of crude peanut allergens suggests that hypoallergenic extracts or T-cell-epitope peptide preparations would still be preferable for safety because of the difficulty in delineating the optimal dose of anti-IgE therapy that would give a satisfactory level of protection from adverse reactions for every peanut-allergic individual. Nevertheless, the possibilities of sublingual routes (see next section) with their apparently decreased adverse events remain to be fully explored.

Sublingual immunotherapy – an alternative route?

Currently, much of the research on peanut allergy is being driven by the need to provide an effective immunotherapy for this life-threatening food allergy. As discussed above, several immunotherapeutic interventions have been proposed for peanut allergy, with the ultimate goal of inducing tolerance to peanuts. In general, the route of administration for conventional immunotherapy is subcutaneous; however, local routes, most notably sublingual, are being investigated as an alternative. Sublingual immunotherapy (SLIT) has raised substantial interest because it is noninvasive and is reported to be efficacious with minimal adverse events. In this immunotherapy regimen, allergen extract is placed under the tongue for 1–2  min and then either swallowed or spat out (Ref. Reference Moingeon84). While SLIT has been utilised successfully for environmental allergens (Refs Reference Dahl85, Reference Bousquet86), it is conceptually difficult to imagine that it can be used for peanut-allergic individuals, given that accidental ingestion of peanut extract can lead to life-threatening anaphylactic reactions. However, SLIT using standardised hazelnut extract to desensitise patients who were allergic to hazelnut showed promising results. Following treatment, there was a significant increase in the threshold dose of hazelnut that elicited objective symptoms, together with an increase in the level of hazelnut-specific IgG₄ (Ref. Reference Enrique87). A low rate of side effects was reported in the study (three reactions out of 1466 doses), which included subjects who previously experienced both local and systemic reactions after hazelnut ingestion (Ref. Reference Enrique87). Although complete tolerance to hazelnut was not achieved, the higher threshold dose after treatment would nevertheless provide a level of protection in cases of accidental ingestion. It is yet to be seen if similar results can be achieved with SLIT using peanut extract, although the primary concern would be the risk of extreme side effects or the development of apparent initial benefit followed by tachyphylaxis (resistance to benefit) with the risk of a subsequent severe reaction.

An alternative strategy to avoid IgE-mediated reactions is to use hypoallergenic preparations for SLIT. However, whether IgE-mediated uptake of the allergen at the oral mucosa is important for tolerance induction remains to be resolved (Ref. Reference Moingeon84). The need for IgE-reactive allergen preparations for successful SLIT could pose difficulties in the treatment of peanut allergy, especially in extremely sensitive patients. A high level of caution would be needed during clinical investigation with perhaps an emphasis on the desirability of good animal models for proof of concept studies.

Burks and colleagues are currently investigating oral immunotherapy in children with peanut allergy (http://www.foodallergyproject.org/Duke-Burks.pdf). The challenge will be to determine the duration of treatment required (3 years, 5 years or lifelong) and a safe dosing interval. Again, compliance concerns with chronic therapy and risks of tachyphylaxis will need careful consideration with the use of crude extracts, because a lapse of regimen could result in a heightened risk of a catastrophic allergic reaction.

Identification of crossreactive peanut and tree-nut allergens

With the observation that cosensitisation to peanuts and tree nuts might be due to allergenic crossreactivity, there is a need to identify the crossreactive components. Such information will improve the diagnosis and management of these allergies. The seed storage proteins, which are conserved throughout different plant families, are strong candidates for mediating the observed crossreactivity. Membership of the same protein family, however, does not necessarily translate into allergenic crossreactivity. For example, no crossreactivity was demonstrated between Ara h 1 and another member of the vicilin family in walnut, Jug r 2 (Ref. Reference Teuber54). The cashew allergen Ana o 1, another vicilin protein, also showed no common IgE-binding epitopes despite sharing 45% amino acid sequence similarity with Ara h 1 (Ref. Reference Wang55). It is likely that 50–70% amino acid sequence homology is required for immunological B-cell crossreactivity to occur between proteins (Ref. Reference Aalberse88). Alternatively, homology between two proteins limited to a small stretch of amino acids might result in crossreactivity if there are similarities in the tertiary structure (Ref. Reference Aalberse, Akkerdaas and van Ree89). Further characterisation of potential crossreactive peanut and tree-nut allergens at the level of secondary and tertiary molecular structure would provide valuable insight into the molecular characteristics required for immunologically relevant crossreactivity to occur. Evaluation using functional assays such as basophil-activation assays will provide convincing evidence for clinically relevant crossreactivity.

Transgenic plants for the production of hypoallergenic peanuts

A novel strategy for decreasing the incidence of peanut allergy and, most significantly, avoiding the risk of exposure to peanut allergens, is the engineering of transgenic plants to produce hypoallergenic peanuts. This may be achieved by introducing antisense RNA copies of the allergen gene into the plant to suppress allergen gene expression (Ref. Reference Moseley90). Using this approach, expression of a 16 kDa allergen was decreased in rice seeds (Ref. Reference Tada91) and expression of the major allergen Gly m Bd 30  K was silenced in soybean seeds (Ref. Reference Herman92). Importantly, suppression of the allergen in transgenic soybeans did not interfere with the normal plant life cycle, with no obvious differences in growth, development, reproduction and seed maturation compared with wild-type soybeans (Ref. Reference Herman92).

An alternative strategy, post-transcriptional gene silencing (PTGS) by RNA degradation, is being trialled for the production of hypoallergenic peanuts (Ref. Reference Dodo, Konan and Viquez93). For this, an extra copy of an endogenous gene is introduced into plants, leading to the degradation of RNA encoded by both the transgene and homologous endogenous gene (Ref. Reference Vaucheret, Beclin and Fagard94). Preliminary studies showed that transformation of peanut embryos with a fragment of Ara h 2 mRNA results in the integration of the Ara h 2 transgene into the peanut genome followed by stable expression throughout plant development (Ref. Reference Dodo, Konan and Viquez93). It is hypothesised that the truncated Ara h 2 mRNA will be synthesised in the peanut seeds and subsequently trigger the specific degradation of endogenous Ara h 2 mRNA (Ref. Reference Dodo, Konan and Viquez93). The success of this strategy is yet to be seen as no results have been reported on the production of peanuts lacking the Ara h 2 protein, if it is indeed possible. For clinical effect, it is likely that peanuts also lacking Ara h 1 and Ara h 3 would need to be engineered as minimum requirements for decreased allergenicity in the peanut-allergic population. However, the deletion of at least three genes from the peanut genome might have deleterious effects on plant growth and development and thus hamper the production of a hypoallergenic peanut.