From Allergen Molecules to Molecular Immunotherapy of Nut Allergy: A Hard Nut to Crack

Peanuts and tree nuts are two of the most common elicitors of immunoglobulin E (IgE)-mediated food allergy. Nut allergy is frequently associated with systemic reactions and can lead to potentially life-threatening respiratory and circulatory symptoms. Furthermore, nut allergy usually persists throughout life. Whether sensitized patients exhibit severe and life-threatening reactions (e.g., anaphylaxis), mild and/or local reactions (e.g., pollen-food allergy syndrome) or no relevant symptoms depends much on IgE recognition of digestion-resistant class I food allergens, IgE cross-reactivity of class II food allergens with respiratory allergens and clinically not relevant plant-derived carbohydrate epitopes, respectively. Accordingly, molecular allergy diagnosis based on the measurement of allergen-specific IgE levels to allergen molecules provides important information in addition to provocation testing in the diagnosis of food allergy. Molecular allergy diagnosis helps identifying the genuinely sensitizing nuts, it determines IgE sensitization to class I and II food allergen molecules and hence provides a basis for personalized forms of treatment such as precise prescription of diet and allergen-specific immunotherapy (AIT). Currently available forms of nut-specific AIT are based only on allergen extracts, have been mainly developed for peanut but not for other nuts and, unlike AIT for respiratory allergies which utilize often subcutaneous administration, are given preferentially by the oral route. Here we review prevalence of allergy to peanut and tree nuts in different populations of the world, summarize knowledge regarding the involved nut allergen molecules and current AIT approaches for nut allergy. We argue that nut-specific AIT may benefit from molecular subcutaneous AIT (SCIT) approaches but identify also possible hurdles for such an approach and explain why molecular SCIT may be a hard nut to crack.


INTRODUCTION
Nuts are nutrient-dense foods that receive increasing attention due to reports regarding their possible health-promoting properties and their pleasant taste (1,2). At the same time, tree nuts and peanuts are among the most common elicitors of anaphylaxis, a severe, potentially life-threatening hypersensitivity reaction mediated by allergen-specific IgE antibody-induced mast cell and basophil activation (3)(4)(5)(6). The possibility of accidental nut ingestion and the associated fear of experiencing severe allergic reactions is particularly challenging for nut-allergic children and their parents and results in a considerable reduction in quality of life (7)(8)(9)(10).
In allergology, a distinction is made between tree nuts and peanuts by defining a nut according to what is considered a nut in the culinary sense and less according to the botanical definition. Generally, "real" botanical nuts like the hazelnut, but also several seeds and drupes that grow on trees are considered tree nuts. Peanuts, which grow underground, are classified as legumes (11). Walnut, pistachio, pecan, hazelnut, almond, cashew, Brazil nut and macadamia are responsible for most allergic reactions to tree nuts and therefore included in this review under the umbrella of "tree nuts" (11) and the term "nut" used herein generally refers to peanuts and tree nuts unless otherwise specified.
True food allergy (class I) is characterized by the primary sensitization to the allergy-causing food via the gastrointestinal tract (12,13) (Figure 1). Therefore, class I food allergens have usually higher stability to gastric digestion than other allergens (14). Immediate allergic reactions to nuts in sensitized patients occur within minutes after nut ingestion. It has been also speculated that IgE sensitization to class I food allergens may occur by epicutaneous sensitization (15) but on the other hand it was found that epicutaneous allergen application does not induce or boost allergen-specific IgE responses (16).
The severity of the allergic reaction depends on the amount of allergen to which the patient is exposed and on other factors such as barrier function and allergen-specific sensitivity which often is associated with specific IgE levels. Class I food allergens often contain sequential IgE epitopes in addition to conformational IgE epitopes which indicates that sensitization occurs also to allergen fragments emerging through digestion in the gastrointestinal tract (17)(18)(19). Allergic reactions to nuts are typically IgE-mediated (type I reactions) and might cause symptoms affecting the gastrointestinal tract (abdominal pain, vomiting), the skin (urticaria, angioedema), the respiratory tract (rhino-conjunctivitis, wheezing) and, in severe cases, the cardiovascular system (loss of consciousness, low blood pressure) (Figure 1). Anaphylactic shock characterized by drop in blood pressure and cardiovascular failure involves several organ systems and requires immediate treatment with epinephrine (20). Several factors such as mast cell activation and/or load, existing co-allergies or asthma might enhance the risk of anaphylactic reactions to tree nuts (21).
Class II food allergy is associated with sensitization to pollen allergens. Patients are usually first sensitized to a pollen allergen and produce IgE antibodies which cross-react with allergens present in food. Examples include the major birch pollen allergen, Bet v 1 and the panallergen, profilin which were discovered first in birch pollen (22)(23)(24)(25)(26). IgE sensitization to class II food allergens is usually associated with mild allergic reactions known as pollen-food allergy syndrome (PFAS) or oral allergy syndrome (OAS) (20,27). Clinical characteristics of PFAS include mainly oropharyngeal symptoms (27). Interestingly, it has been indicated that ingestion of birch pollen-related allergens from food sources such as Cor a 1 from hazelnut, could activate allergen-specific T cells independent of IgE, leading to late-phase and chronic allergic inflammation and this might further cause disorders such as atopic dermatitis in sensitized patients (28,29). Moreover, pollen-related nut allergens causing PFAS might be associated with eosinophilic esophagitis, although they seem to be of less relevance than homologues from fruits and vegetables (30). However, eosinophilic esophagitis can be caused also by class I food allergens from milk, egg and wheat, while peanut and tree nuts seem to be of minor relevance (31). Major features of class II food allergens are that they contain mainly conformational but not sequential IgE epitopes which are sensitive to digestion and heating (32)(33)(34). The sensitization to class II food allergens is initially caused by pollen allergens and results in IgE and T cell cross-reactivity with the related food allergens (35,36). IgE sensitization to class II food allergens is highly prevalent in countries with high exposure to the crossreactive pollen allergens (37,38). Accordingly, diagnostics including the measurement of IgE against the originally sensitizing pollen allergens (39) and allergen-specific immunotherapy to the cross-reactive pollen allergens can improve not only pollen allergy but also the associated food allergy to some extent (40,41).
Diagnosis of nut allergy usually starts with a thorough evaluation of the patient's history. Allergic sensitization can be detected by skin prick tests (SPT) and in vitro diagnostics with allergen extracts. However, sensitization determined by measurement of specific IgE antibodies and SPT does not always indicate clinical food allergy, which can only be confirmed by the occurrence of specific allergic symptoms after food exposure. Double-blind placebo-controlled food challenges (DBPCFC) are still considered the "gold standard" of food allergy testing, although patients are at risk of anaphylaxis during the procedure (42,43). Lip dose challenge (LDC) is another possibility of testing which has a good predictive value for nut allergy (44). However, in recent years, molecular diagnosis with defined and mainly recombinant allergens by IgE serology has turned out to be very helpful in diagnosing nut allergy, in particular when it is combined with a thorough medical history (45). Another key problem in therapy of nut allergy is the lack of modern and effective allergen-specific treatment options. At present, avoidance of the disease-causing allergens is a possible option but there is also evidence that early introduction of for example of peanuts in the diet of sensitized but not yet symptomatic children may have beneficial effects (46). Accordingly, there are different opinions whether avoidance or rather intake should be recommended for sensitized children. Another major problem is that there is currently little progress regarding the development of modern molecular immunotherapy forms for nut allergy. Hypoallergenic allergen derivatives have been described (47) but no clinical studies have been performed so far. Sensitization to different nut allergens varies in different parts of the world due to dietary habits in diverse populations and varying allergen exposure in different areas but this is undergoing changes due to globalization and migration.

IMPORTANCE OF VARIOUS NUTS AS ALLERGEN SOURCES IN DIFFERENT PARTS OF THE WORLD
The prevalence of nut allergies among children and adults has been investigated in several studies (11,(48)(49)(50)(51). However, there are large variations regarding methodology and study design which make it difficult to compare the studies and to understand the true nut allergy rates. It seems that reports on nut allergy prevalence do not provide accurate information regarding actual prevalences in the different populations due to several reasons. First of all, most studies that include a representative study population are limited to self-reports and do not contain a detailed clinical evaluation of patients. Moreover, several studies do not distinguish between sensitization to class I and class II food allergens. In this context, it must be considered that allergic reactions to nuts might be due to cross-reactivity with pollen allergens and are not caused by primary nut sensitization (11). Especially in studies from Europe, hazelnut allergy prevalence might thus be overestimated and sensitization should therefore be evaluated by molecular diagnosis to clearly distinguish between birch pollen allergic patients and those with true hazelnut allergy. This applies also to several other nuts that contain cross-reactive panallergens and cross-reactive carbohydrate determinants (CCDs). For example, many subjects who were tested positive by IgE serology using peanut allergen extracts in Zimbabwe were found to be sensitized to CCDs which usually do not cause allergic reactions (52). Table 1 provides an overview of nut allergy prevalence studies, in particular from Europe, Northern America, Asia, Australia and Africa (38,(48)(49)(50).
Importantly, the worldwide prevalence of nuts causing allergy correlates strongly with the nuts that are consumed in this FIGURE 1 | Sensitization to class I and class II nut allergens is associated with different clinical symptoms. Sensitization to class II nut allergens usually occurs by respiratory sensitization to cross-reactive respiratory allergens (e.g., pollen allergens) and is associated with mild symptoms such as oral allergy syndrome, local reactions in the oropharynx, esophagus and may trigger atopic dermatitis and/or urticaria. Sensitization to class I digestion-resistant nut allergens usually occurs via the gastrointestinal tract and eventually via the skin and is associated with systemic and severe manifestations such as anaphylaxis but also mild symptoms are possible.   were tested for at least one allergen mix. 4220 were tested for all five food allergen mixes. Participants had been previously included in the "random sample" group during the second phase of the European Community Respiratory Health Survey.

Singapore Philippines
Administration of a questionnaire to assess prevalence of peanut and tree nut allergy in Singapore (local and expatriate) and Philippine schoolchildren of different age groups. Allergy diagnosis was based on convincing history which was defined by reports on the appearance of specific allergic symptoms within two hours after food ingestion.   region. However, for improved nut allergy management it is more relevant to consider the sensitization profile of nut allergic patients on a molecular level. As an example, sensitization to allergens of the family of pathogenesis-related class 10 (PR-10) proteins is widespread in northern countries, while IgE reactivity to non-specific lipid transfer proteins (nsLTPs) is predominant in the Mediterranean region. Molecular diagnostics significantly helps to distinguish between cross-reactive allergens and those that are a true indicator of sensitization to a particular nut.
In Europe, regional as well as ethnical differences in the sensitization profile of nut allergic patients have been observed (48,50,56). Generally, self-reported prevalence is significantly higher than food challenge-confirmed nut allergy (58). Several studies that investigated peanut allergy prevalence in Europe revealed varying prevalence rates (53)(54)(55)59). In Russia, peanut allergy does not seem to play a major role in food allergy (38). Peanuts and cashew nuts are among the most common elicitors of anaphylaxis (85). Co-sensitization to different nuts correlates Tree nut tolerance was defined by negative OFC result, SPT wheal size of 0-2 mm, SPT response of 3-7 mm and reported food ingestion, or lack of reaction since age 4 years without food avoidance.
allergic to at least one tree nut. Cashew was the most common tree nut causing allergy (2.7%), followed by hazelnut (0.9%) and almond (0.3%).

Zimbabwe Sweden
Sera from peanut-sensitized and peanut-allergic patients were analyzed for IgE to Ara h 1-3, 6, 8 and 9 using an allergen microarray. IgE levels were considered low (0. 35 Listed are studies investigating prevalences of allergy to different nuts as determined by different methodologies in different populations with the corresponding references. strongest between nuts of the same botanical family such as cashew and pistachio or pecan and walnut (60). In the US, peanut is one of the most common foods causing allergy (76)(77)(78). Among tree nuts, walnut and cashew cause most of the allergic reactions, followed by almond, pistachio, Brazil nut, hazelnut and macadamia (76). Similar results were seen in a Canadian study with peanut allergy being most prevalent, predominantly in children (81).
In Central and South America, few studies reported sensitization of allergic patients to peanut and almond, although in this region, allergy to nuts seems to be low in general (79,80,86,87). In most Latin American countries, frequent foods that cause allergy include fish, seafood, milk, egg, vegetables and fruits (87,88).
In Asia, peanut allergy prevalence seems to be low compared to US and certain western countries (76,(89)(90)(91)(92). Cashew nut is one of the most common reported tree nuts causing allergy in Asia (67,70,71,74). However, tree nut allergy prevalence varies significantly across Asia especially between East and Southeast Asia and the Middle East (62,63,66,70,74). It can be assumed that the availability of nuts in certain regions contributes to the prevalence of allergies to these nuts, as can be seen by the increased frequency of pistachio allergy in pistachio cultivation regions (64).
In Australia, peanut allergy is one of the most frequent elicitors of IgE-mediated food allergy (49,93). Tree nut allergy in Australia is less common than peanut allergy and prevalence rates of individual tree nut allergies vary significantly between studies (82,83,93).
Peanut allergens are the most frequently recognized nut allergens in South Africa (84) as determined in allergic children whereas IgE recognition of peanut allergens seems to be often asymptomatic as reported for Zimbabwe (52) but data regarding the prevalence of nut allergies in Africa are rare. Figure 2 provides an overview of the role of different nuts as allergen sources for different regions of the world. Peanut allergy seems to be most frequent in most parts of the world whereas in Europe hazel nut allergy seems to be more important. Interestingly, different molecular IgE sensitization patterns can be observed in different geographic regions depending on birch pollen exposure involving IgE reactivity to Ara h 8, sensitization to lipid transfer proteins in southern Europe with sensitization to Ara h 9, and the classical peanut sensitization involving storage proteins such as Ara h 1, Ara h 2, Ara h 3 and Ara h 6 (94-96). In South America, nut allergy seems to be less common than in other parts of the world. Only few data are available for Africa indicating a need for further studies. It seems that early introduction of peanut in the diet as it occurs in Zimbabwe results in a low rate of symptomatic peanut allergy (52).
Notably, reports on the prevalence of nut allergies among adults are rare and most studies have been conducted in children. More studies taking into account the molecular IgE sensitization profiles and symptoms verified by highly indicative case history and/or provocation testing in children and adults are needed to obtain a more complete picture of the dominating nut allergies in different parts of the world.

CLINICAL RELEVANCE OF NUT ALLERGEN MOLECULES
Peanut allergy is a good example for the importance of molecular diagnosis for identifying the culprit sensitizing allergen source. Patients may be allergic to peanut due to primary sensitization to Specific Immunotherapy for Nut Allergy birch pollen and cross-reactivity of PR-10 allergen (i.e., crossreactivity between Bet v 1 and Ara h 8), some are sensitized to lipid transfer proteins from fruits and eventually certain pollen (e.g., cross-reactivity between Pru p 3 and Ara h 9), others may be genuinely sensitized to peanut and the corresponding peanutspecific marker allergens (Ara h 1, 2, 3 and 6) and there can be mixed sensitizations (94)(95)(96). The deconvolution of the molecular IgE sensitization profiles is therefore of high importance for identifying the genuinely sensitizing allergen source, predicting clinical manifestations (mild or severe forms of allergy), prevention and treatment based on avoidance/diet and AIT (13). New approaches for the diagnosis and therapy of nut allergies will be increasingly based on individual nut allergen molecules. The clinical relevance of different allergens significantly varies by region and age. In the overview of nut allergen molecules in  (94,99). Valcour et al. showed that in the US, patients with reported peanut allergy most frequently recognized Ara h 2 but IgE reactivity to Ara h 1 and Ara h 3 was also highly prevalent in the tested patients (104). Kleber-Janke et al. reported IgE reactivity to Ara h 1 in 65% and to Ara h 2 in 85% of sera from patients (n = 40) with reported peanut allergy (100). Koppelman et al. compared the IgE reactivity of 32 peanut-allergic patients to Ara h 1, Ara h 2 and Ara h 3 and showed that of these three allergens, Ara h 2 was most frequently recognized (26/32) (102). Importantly, sensitization to Ara h 2 is associated with severe allergic reactions (103). Ara h 2 further has the potential to crossreact with other 2S albumins such as Ara h 6 and Ara h 7, with Ara h 2 possibly representing the primary sensitizing agent (108,162). However, in rare cases, monosensitization to Ara h 6 and Ara h 7 might be observed and thus must be considered for accurate diagnosis (108,163). It has been shown that detection of IgE reactivity to peanut extract together with reactivity to rAra h 2 and rAra h 6 allows reliable peanut allergy diagnosis and Ara h 2 could significantly increase diagnostic specificity (164).
In comparison to Ara h 1 and Ara h 2, sensitization to Ara h 3 is less frequently observed (94,102,105).
In the Mediterranean region, sensitization to the nsLTP Ara h 9 is common and has high cross-reactive potential with homologous allergens of the Rosaceae family, in particular the peach nsLTP Pru p 3 (94,110,111,165).
Schwager et al. reported sensitization to peanut oleosins in patients with a history of severe allergic reactions (113). According to the authors, roasting of peanuts seemed to increase the IgE-binding capacity of oleosins. Previously, several studies have reported that roasting might enhance the allergenic activity of peanut allergens (166)(167)(168)(169).
So far, little is known regarding the clinical relevance of peanut defensins and the nsLTPs Ara h 16 and Ara h 17 as well as the currently approved cyclophilin-peptidyl-prolyl cistrans isomerase Ara h 18 which may be cross-reactive with corresponding pollen and respiratory allergens.

Walnut
For the English walnut (Juglans regia), which belongs to the Juglandaceae family, 8 allergens have been officially approved by the allergen nomenclature (Jug r 1 to 8), making it the clinically most relevant walnut species (97, 116) ( Table 2). For the black walnut (Juglans nigra) 3 allergens have been identified (Jug n 1, 2, 4) (97). However, their clinical relevance is not yet well described in the literature.
Teuber et al. reported that 12 out of 16 walnut-allergic patients showed IgE reactivity to a 2S albumin from English walnut, designated Jug r 1, thus identifying it as a major walnut allergen (115).
IgE reactivity to another major walnut allergen, the vicilin Jug r 2, was detected in 9 out of 15 patients from the US (117). In a study by Pastorello et al., IgE reactivity to vicilin-like protein precursors and vicilin precursors of 9 kD was observed in 10 out of 46 sera from Italian patients, suggesting a minor role of vicilins in allergic patients in the Mediterranean region (118).
Pastorello et al. further reported that 37 out of 46 sera showed IgE binding to the walnut nsLTP Jug r 3, leading to the conclusion that in southern Europe, Jug r 3 represents a major allergen of walnut (118). Notably, peach LTP (Pru p 3) completely inhibited IgE binding to Jug r 3, indicating strong cross-reactivity between walnut and peach.
In 2003, Teuber et al. observed IgE sensitization of patients who experienced life-threatening systemic reactions after walnut consumption to a walnut protein of the legumin group, designated Jug r 4 (119). IgE binding to a recombinant Jug r 4 fusion protein was observed in 15 out of 23 tested sera, suggesting major importance of Jug r 4 in patients with confirmed symptoms. Another study showed IgE reactivity to recombinant Jug r 4 in 21 out of 37 sera from walnut-allergic patients (120).
Jug r 6, like Jug r 2 and Jug r 4, is a member of the cupin superfamily. Although Jug r 2 and Jug r 6 belong to the same protein family, they share only 44% identity (122). In comparison to Jug r 2, which was identified as a major walnut allergen by Teuber et al., Jug r 6 showed IgE reactivity in 20 of 77 walnut-allergic patients,    indicating it is of minor clinical relevance (117,122). Interestingly, cross-reactivity has been shown between Jug r 6 and homologues from pistachio, sesame and hazelnut, which, however, did not apply for Jug r 2 (122).

Hazelnut
So far, 11 allergens from common hazel (Corylus avellana) are registered in the WHO/IUIS database (97) ( Table 2). Sensitization to the nsLTP, Cor a 8 predominantly occurs in patients from the Mediterranean region and has been associated with severe allergic reactions (128,130). However, also in birchendemic regions, sensitization to Cor a 8 was found in children who had objective reactions during DBPCFC (129). Pastorello et al. reported IgE reactivity to Cor a 8 in patients with a history of anaphylactic reactions to hazelnuts and demonstrated inhibition of IgE binding to Cor a 8 by the purified Pru p 3 (124).
Severe allergic reactions unrelated to pollen allergy have also been reported from patients with sensitization to the 11S globulin Cor a 9 and the 7S globulin Cor a 11 (132). IgE reactivity to Cor a 9 was detected in 12 of 14 patients with a history of systemic reactions to hazelnuts (131). In hazelnutallergic patients from birch-endemic regions, age-related differences regarding the sensitization to Cor a 9 were observed (126). In total, 65% of pre-school children and 50% of schoolchildren, but only 17% of adults with systemic reactions were sensitized to Cor a 9. In a study by Lauer et al., IgE sensitization to Cor a 11 was observed in less than 50% of 65 hazelnut-allergic patients and the allergen demonstrated significantly lower biological activity in comparison to Cor a 1, suggesting that Cor a 11 is a less relevant hazelnut allergen (134). Similar to Cor a 9, in birch-endemic regions, sensitization to Cor a 11 is age-dependent and is recognized predominantly by children with objective symptoms (135).
The 2S albumin Cor a 14 was first identified in 2010 (137). In a study by Faber et al., IgE reactivity of hazelnut-allergic patients to Cor a 14 was analyzed in different age groups, revealing that Cor a 14 was predominantly recognized in pre-school (18/20) and school-aged children (8/10) (139). In Dutch patients with hazelnut allergy, sensitization to Cor a 14 and Cor a 9 was shown to be highly specific for predicting more severe hazelnut allergy (138). Similar results were obtained in another study that examined the role of component resolved diagnostics for the prediction of clinical allergy in hazelnut-allergic children (170). Specific IgE to Cor a 14 was found to be reliable for the discrimination between patients with clinical reactivity and those that were nonreactive.
The hazelnut oleosins Cor a 12, Cor a 13 and Cor a 15 might be associated with severe allergic reactions (136,171). However, more studies are needed to establish their clinical relevance. In Europe, sensitization to Cor a 12 in patients with reported reactions to hazelnuts ranged from 10 to 25% and appeared to be more frequent in children than adults (172). The clinical relevance of Cor a 6, a isoflavone reductase-related protein, and Cor a 10 a luminal binding protein with possible pollen crossreactivity remains to be determined.

Pistachio
Five allergens from Pistacia vera (Pis v 1, Pis v 2, Pis v 3, Pis v 4 and Pis v 5) have been officially approved (97) ( Table 2). The sensitization profile of patients with pistachio allergy varies significantly across Europe, indicating age-related, demographic and ethnic differences among the population (56,60,63). The clinical relevance of individual pistachio allergens has not been investigated in detail, but it has been shown that pistachio allergy can lead to severe allergic reactions (173).
Ahn et al. reported IgE reactivity in the serum of 19 out of 28 pistachio-allergic patients to a 7 kDa 2S albumin, which was designated Pis v 1. Moreover, 14 out of 28 patients showed IgE binding to the legumin-like protein Pis v 2 (140). These allergens were further identified as homologous of the cashew allergens Ana o 3 and Ana o 2, respectively. The cashew tree belongs just like pistachio to the Anacardiaceae family, which explains the high structural similarity of the proteins and indicates cross-reactivity.
IgE sensitization to the 7S globulin Pis v 3 was shown in 7 of 19 patients who had a history of allergic reactions to pistachio and/or cashew (141). The patients with IgE reactivity to rPis v 3 also reacted to rAna o 1 from cashew nut.
In 16 out of 27 sera from pistachio-allergic patients, IgE reactivity to a manganese superoxide dismutase (MnSOD)-like protein, designated Pis v 4, from pistachio was detected (142). MnSOD-like proteins are known as cross-reactive respiratory allergens (174) and hence Pis v 4 may be considered as a class II

Cashew
Currently, three cashew (Anacardium occidentale) allergens are registered in the database of the WHO/IUIS (97) ( Table 2
Pru du 3 belongs to the nsLTP family, which is usually associated with high allergenic activity and cross-reactivity between members of the Rosaceae family, mainly in the Mediterranean region (176,177). However, large clinical studies evaluating the prevalence of IgE sensitization to Pru du 3 in almond-allergic patients from different regions are needed.
The 60s acidic ribosomal protein P2 has been identified as Pru du 5, and IgE reactivity to a recombinant variant of the protein was shown in 4 of 8 almond-sensitized patients (149). Acid ribosomal proteins have been identified in molds as allergens and it may therefore be considered that this allergen may represent a class II food allergen (178).
Reactivity to recombinant variants of the amandin Pru du 6, Pru du 6.01 and Pru du 6.02, was seen in 9 of 18 and 5 of 18 almond-allergic patients, respectively, while only 4 of the tested patients showed IgE reactivity to both isoforms (151). Kabasser et al. suggested that Pru du 6 might be a specific marker for almond allergy since 16 of 18 almond-allergic patients showed IgE reactivity to the allergen (152). Moreover, positive sIgE to Pru du 6 provided a specificity of 78% and a sensitivity of 83% for almond allergy, while at the same threshold level, the detection of sIgE to almond extract significantly lacked specificity (33%). In comparison, Pru du 8 and Pru du 10 had specificities of 100% and 61% but were less sensitive (41% and 67%) (152). The antigenicity of almond amandin does not seem to be influenced by roasting, blanching or autoclaving, indicating high protein stability (179,180). In

Brazil Nut
To this date, the 2S albumin Ber e 1 and the 11S globulin Ber e 2 from Brazil nut (Bertholletia excelsa) have been registered in the allergen data base (97) ( Table 2).
Pastorello et al. reported that each of 11 patients with a history of anaphylaxis after the consumption of Brazil nut, showed IgE reactivity to a 2S albumin, implying that it represents a major allergen from Brazil nut (154). Rayes

Pecan
Three proteins from Carya illinoinensis, the 2S albumin Car i 1, the vicilin Car i 2 and the legumin Car i 4 have been officially approved as allergens (97) ( Table 2).
In 2011, the 2S albumin Car i 1 was characterized and IgE binding to recombinant Car i 1 was detected in 22 of 28 patients with pecan allergy (158). The same study showed that pecan and walnut extracts inhibited IgE binding to recombinant Car i 1, indicating strong cross-reactivity with homologous proteins from these nuts. In 2016, Zhang et al. reported that 6 out of 25 patients with DBPCFC-confirmed pecan allergy, showed IgE reactivity to pecan vicilin Car i 2 (159). In a study by Sharma et al., an 11S globulin from pecan, designated Cari i 4, was recognized by IgE from 16 out of 28 patients with pecan allergy (160). Furthermore, extracts from pecan as well as walnut inhibited IgE binding to rCar i 4, suggesting cross-reactivity with legumins from other tree nuts.

Macadamia
To date, 2 allergens from macadamia nut (Macadamia integrifolia), the vicilin Mac i 1 and the legumin Mac i 2, are included in the allergen list of the WHO/IUIS Allergen Nomenclature Sub-committee (97) (  (184). According to available data, measurement of specific IgE to macadamia nut does not always predict clinical allergy and might lead to false-negative results (185,186). However, single allergen molecules of macadamia nut for component resolved diagnosis are lacking and it must be considered that macadamia extracts might not contain all relevant allergens and thus provide low diagnostic sensitivity (186). Therefore, the identification and characterization of macadamia proteins with established allergenic potential is urgently needed. Possible cross-reactivity between macadamia and hazelnut has been suggested (182, 183).

Clinically Relevant Panallergens to Be Considered as Class II Food Allergens
In peanuts, one of the most relevant panallergens is the Bet v 1like homologue Ara h 8, which is of major importance in patients from birch-endemic regions where allergic reactions to peanuts can be strongly associated with sensitization to birch pollen (94,104,109). Similarly, IgE reactivity to the profilin Ara h 5 is associated with previous sensitization to pollen (106). In walnut, the pathogenesis-related protein (PR-10) Jug r 5 is associated with IgE cross-reactivity between homologous allergens from different plant sources and of minor relevance for patients with primary walnut allergy (121). The Bet v 1-like Cor a 1 and the profilin Cor a 2 are cross-reactive allergens of hazelnut and sensitization to these allergens is typically seen in birch-endemic regions (50,125,127). Both allergens are expressed in hazelnut as well as in hazel pollen. The profilin Pru du 4 is a minor allergen of almond and cross-reactivity with profilins from grass pollen was reported (148). It is quite likely that additional "food allergens" ( Table 2, light blue) will be identified for which sensitization occurs by respiratory allergen sources and symptoms of food allergy will be low because the allergens are not heat stable and/or become easily digested and then lose their allergenic activity. Ara h 18, Cor a 6, Pis v 4 and Pru du 5 are possible candidates and there may be more discovered in the future ( Table 2, dark blue). IgE reactivity to the class II nut allergens is not due to genuine nut sensitization and symptoms caused by these allergens may be treated by AIT directed to the originally sensitizing respiratory allergens.

DIAGNOSIS OF NUT ALLERGY
Diagnosis of nut allergies usually starts with the evaluation of the medical history of the patient. While in the past, diagnosis was mainly achieved by allergen extract-based tests (SPT, OFC), these are increasingly being replaced by modern molecular techniques using specific allergen molecules (Figure 3) (187). Figure 3 compares traditional allergen extract-based diagnosis for nut allergy with modern molecular allergy diagnosis. Traditional extract-based diagnosis uses allergen extracts prepared from the allergen sources for serology and provocation testing in conjunction with the clinical history to determine food which can elicit allergic reactions. Molecular allergy diagnosis is based on IgE serology to a broad panel of defined allergen molecules in combination with the clinical history. In this pathway provocation testing is reduced and usually only performed if necessary to confirm clinically relevant allergy if this cannot be determined by molecular testing and medical history. Molecular testing offers high precision regarding the identification of the culprit allergen molecules is fast and helps to reduce provocation testing which can give rise to severe reactions (187).

Food Challenges
Double-blind, placebo-controlled oral food challenge is still a common procedure for food allergy diagnosis, although in the case of strong clinical suspicion, this is usually avoided. Generally, it is recommended that DBPCFC is performed in a standardized procedure under consideration of several patientrelated and procedure-related parameters (188, 189). Nevertheless, it must be taken into account that oral food challenges (OFC) bear the risk of potentially fatal anaphylaxis during the procedure (43). This applies particularly to nuts, Specific Immunotherapy for Nut Allergy which are among the most common foods causing anaphylaxis (5). In recent studies, lip dose challenges (LDC), using fresh nuts or nut paste, were suggested as a supplement for oral challenges for nut allergy diagnosis (44,190). LDC might be performed as a preliminary test to an OFC but currently cannot replace the latter. However, LDC, in combination with modern molecular diagnostic, might reduce the need for OFC in the future.

Skin Tests
In principle, two types of skin tests can be performed for diagnostics purposes. Skin prick testing measures the induction of mast cell degranulation caused by cross-linking of IgE bound to the high affinity IgE receptor (FcϵRI) (191) whereas atopy patch testing (APT) detects allergen-specific T cell activation even in the absence of IgE-mediated effects (191,192). Accordingly, SPT may be considered as surrogate test for IgEmediated immediate allergic inflammation and APT as surrogate test for chronic, T cell-mediated allergic inflammation. SPT and the detection of food-specific serum IgE with allergen extracts have been traditionally used for allergy diagnosis but have major weaknesses. First of all, these tests are performed with poorly defined allergen extracts and hence do not identify the sensitizing allergen molecules (193). Second, both methods cannot be used to predict clinical sensitivity with certainty because the extent to which digestion affects allergenic activity cannot be measured with these methods. Several authors suggested that the use of fresh food might increase test sensitivity (194,195). Therefore, food challenge tests are still recommended despite the associated risk factors.

Molecular Allergy Diagnosis
Molecular allergy diagnosis is based on the use of purified allergen molecules, mainly recombinant allergens, to determine the IgE sensitization profile of allergic patients (45). There are also attempts to improve the diagnosis of nut allergy by combining different forms of allergen extracts-based diagnosis. For example, it has been shown that prediction of clinical reactivity to pistachio and cashew was improved by SPT in combination with measurement of sIgE (196). However, nowadays native purified or recombinant single allergen molecules are increasingly replacing conventional extracts in in vitro diagnostics. Molecular tests that allow the detection of specific IgE antibodies to individual allergen molecules are also known under the term component-resolved diagnostics (CRD) (197). For peanut allergy, it was demonstrated that by measuring Ara h 2specific IgE, the diagnostic accuracy could be considerably improved (198)(199)(200)(201). When measured together, sIgE reactivity to Ara h 6 and Ara h 2 was shown to be predictive for severe peanut allergy (103). For the prediction of positive outcomes of food challenges in children, it was demonstrated that Ara h 2-specific IgE levels of 14.4 kU A /L and Cor a 14-specific IgE levels of 47.8 kU A / L had an estimated probability of 90% for predicting a positive peanut or hazelnut challenge (202). In another study, Cor a 14specific IgE levels of 0.5 and 1.0 kU A /L had a probability of 50% and 95% to predict clinical reactivity to hazelnut in sensitized patients, respectively (170). Moreover, it was shown that measurement of sIgE levels for Cor a 9 in hazelnut-sensitized patients might improve the diagnostic accuracy for the prediction of hazelnut allergy in Japanese children (203). For cashew it was found that sIgE to individual allergen molecules from cashew nut had a predictive value for the diagnosis of clinical allergy (204)(205)(206). Measurement of Jug r 1-specific IgE was suggested for the prediction of walnut allergy in children due to improved clinical specificity in comparison with IgE to walnut extracts (207). Several assays have been developed for the detection of serum IgE to either a single allergen analyte (singleplex assay) or various allergens at a time (multiplex assay) (187,208,209). The availabilities of single allergens and advanced microarray technology have made it possible to obtain a quick insight into the sensitization profile of a patient (210). In order to enable quantitative conversion between different multiplex IgE testplatforms for nut allergens, statistical models have been established recently (211). For the European MeDALL research project, an allergen chip with 170 allergen molecules, including natural purified and recombinant allergens from almond, cashew, pistachio and peanut, was developed which could be used even for dried blood samples (212). Recently, a study showed moderate agreement of microarray-based analysis in comparison with clinical diagnosis but high sensitivity of the microarray was seen for tree nuts (213). Moreover, the microarray results for tree nuts correlated with SPT results, promising a superior role of component resolved diagnostic for nut allergies in the future.
Another interesting approach for in vitro allergy diagnosis of nut allergy is the basophil activation test (BAT). Since the early description of allergen-induced histamine release from basophils (214) and the demonstration of the applicability of basophil activation testing for recombinant allergens (215), basophil activation testing has continuously developed (216). Importantly, basophil activation can discriminate between IgE-reactive antigens with no or poor ability to induce IgE-mediated receptor aggregation from potent allergens which induce basophil activation already at low doses (32,217). Thus basophil activation testing is useful to address a major problem of in vitro allergy diagnostics, i.e., the possibility of false-positive results due to the presence of crossreactive carbohydrate determinants (218). In plants, these IgEbinding carbohydrate structures are usually N-glycans with a core a-1,3-linked fucose residue. It is well established that CCDs are responsible for IgE cross-reactivity between a wide range of plant allergens and other unrelated allergen sources (219). Furthermore, the presence of N-glycans in cellulose-based ImmunoCap assays could lead to false-positive results in patients with high levels of CCD-reactive IgE antibodies (220). Possibilities to overcome IgE reactivity to CCDs are the production of non-glycosylated recombinant allergen molecules or the use of specific CCD inhibitors (221). CCDdirected IgE antibodies seem to have poor biological activity and are not associated with clinical symptoms (222)(223)(224). In basophil activation tests, flow cytometry can be used to analyze basophil activation, which, for example can be defined by the upregulation of the lineage-specific basophil marker CD203c together with the degranulation marker CD63 (225) as has been shown for hazelnut allergy (226). Alternatively, rat basophil cell lines transfected with human FcϵRI can be loaded with serum IgE and then stimulated with allergens (227). Basophil activation was found useful for predicting clinical reactions in peanut allergic patients. Glaumann et al. reported that negative basophil allergen threshold sensitivity correlated with negative DBPCFC in children with peanut allergy (228). Moreover, 92% with positive DBPCFC had positive threshold sensitivity results and increased levels of IgE antibodies to the major peanut allergens Ara h 1, Ara h 2 and Ara h 3. More recently, basophil activation testing was reported to have high accuracy for the diagnosis of peanut and tree nut allergy but it has not been studied if it can be used to differentiate between sensitization to class I and class II food allergens, causing mild and severe systemic anaphylactic reactions, respectively (229).
Basophil activation testing is also a useful tool to investigate the efficacy of AIT for nut allergy by demonstrating the ability of allergen-specific immunoglobulin G (IgG) antibodies to block IgE-mediated immediate allergic reactions (230, 231).

ALLERGEN-SPECIFIC IMMUNOTHERAPIES FOR NUT ALLERGIES
Most of the strategies for treatment and prevention of food allergy and in particular of nut allergy (e.g., allergen avoidance, diet, use of hypoallergenic food products, AIT) are tightly connected with the accurate identification of the culprit allergens. However, some measures like the management of severe acute and chronic inflammation may be achieved by drugs such as epinephrine injection for treatment of acute anaphylactic reactions, immunosuppressive drugs and anti-IgE treatment (232). Besides diet, AIT is the most important form of allergen-specific treatment. The immunological mechanisms underlying AIT include a modified allergen-specific antibody, cellular and cytokine response (233). Besides complex alterations of the cellular and cytokine responses it has become clear that the induction of allergenspecific IgG and perhaps of allergen-specific IgA antibodies which block IgE binding to the allergen and accordingly the IgE antibodymediate pathology is a key mechanism of AIT (234)(235)(236). This has been evidenced in clinical studies using molecular approaches for AIT (237,238) and by the demonstration that passive immunization with allergen-specific blocking IgG antibodies is clinically effective (239-241).

Current Forms of AIT For Nut Allergy Are Mainly Based on Allergen Extracts and Subcutaneous AIT Is Rarely Used
Regarding the treatment of respiratory allergy by AIT subcutaneous injection immunotherapy remains to be the most frequently used and effective form of AIT as documented by a large number of clinical studies although a huge effort has been done to promote sublingual immunotherapy (SLIT) in multiple studies (235,242). However, SCIT is more effective than SLIT and patients adherence to SCIT is much better than to SLIT (235,243). Regarding AIT of food allergy it is of note, that there are only few early studies regarding SCIT (244,245) and it seems that due to unfavorable side effect profiles SCIT has not been further pursued for food allergy. Instead, oral immunotherapy (OIT) has been developed for class I food allergens which are resistant to digestion whereas OIT studies for respiratory allergens and class II food allergens which are sensitive to digestion have not been successful (246)(247)(248). Another important aspect is that only few attempts were made to introduce molecular forms of AIT for food allergy whereas different forms of molecular AIT have been evaluated for respiratory allergy (235). One possible reason for this could be that many more patients suffer from respiratory allergy than from food allergy and usually new forms of treatment are mainly evaluated for frequently occurring forms of allergy because the costs for the preclinical and clinical development of novel vaccines are high. Accordingly, the majority of AIT trials for food allergy have been performed with allergen extracts and by using the OIT approach.

Oral Immunotherapy
OIT is based on the controlled ingestion of the allergen-causing food, intending to achieve sustained desensitization in the patients. It has been shown that similar as for SCIT, the success of treatment is associated with the development of allergen-specific IgG blocking antibodies which have actually been measured in many of the OIT studies. Table 3 provides and overview of OIT studies  informing about the number of participants, the study design, clinical and immunological outcomes, side effects and references and/or trial registration numbers which allow to track the studies in the Clinical Trials data base (https://clinicaltrials.gov/). Most of the studies were conducted for peanut allergy whereas OIT studies for tree nut allergies are scarce ( Table 3). A study by Andorf et al. (280) is one of the few studies providing evidence for effects of OIT to several different nuts when OIT was combined with anti-IgE treatment.
There are methods available for determining major peanut allergens in natural allergen extracts (281) but the precise concentrations of the individual peanut allergens in the natural extracts is not known. Currently, there is no standardized procedure for OIT neither regarding the study design nor are there defined vaccines with known composition. Usually, OIT starts with a dose-escalation day, followed by a buildup phase during which increasing amounts of the allergen are ingested until the maintenance dose is reached. DBPCFC might be performed after a defined food avoidance period to confirm sustained desensitization in the treated subjects. Already in 2009, Jones et al. reported a clinical trial of peanut OIT (249). Since then, the efficacy and safety of peanut OIT have been extensively studied. OIT studies demonstrated successful desensitization and the production of protective IgG4 antibodies but reports of adverse reactions raised safety concerns (267,269). Adverse reactions affecting the gastrointestinal and respiratory tract during peanut OIT are common (282). To reduce the risk of side effects and to accelerate the desensitization process, the supplementation of OIT with omalizumab, an anti-IgE monoclonal antibody, has been suggested (283)(284)(285). The optimal time point to start OIT, treatment duration and length of the maintenance phase are still a matter of debate. With exception of few studies (261,265,(267)(268)(269)(270), most studies involved less than 100 patients and the achieved clinical       Of the initial 37 eligible study participants, 3 withdrew due to treatmentrelated adverse reactions. Overall, 95% of the participants were affected by adverse events which occurred more frequently during the build-up phase. Most adverse events were mild (85%), 15% were considered moderate and no severe reaction was reported.

•
In the peanut-OIT group a significant reduction in IL-4, IL-5, IL-10 and IL-2 production, a significant increase in median peanutspecific IgG4 levels and a decrease in the peanutspecific IgE/IgG4 ratio were observed after treatment when compared between the randomized arms.
Two patients in each group withdrew due to experiencing adverse events, which were severe in one of the subjects of each group. All patients experienced adverse events at some point, however, only 1.2% of placebo doses and 4.3% of peanut OIT doses were associated with treatment-related adverse reactions. 40% in the placebo group and 45% in the active OIT group needed at least 1 dose reduction because of reported side effects.         (272) Initial study was described above (272). An additional cohort of patients that followed identical protocol were also included in the extension study. During the long-term extension study, subjects received SLIT with maintenance daily dose of 2 mg peanut protein (up to 5 years). After the final day of SLIT, sensitization was assessed by DBPCFC with 5 g peanut protein.   Design of studies, numbers of participants, outcomes, side effects and references are listed.
benefits were relatively modest when put into context with side effects. Accordingly there are different opinions about OIT. One metanalysis (286) concluded: "In patients with peanut allergy, high-certainty evidence shows that available peanut oral immunotherapy regimens considerably increase allergic and anaphylactic reactions over avoidance or placebo, despite effectively inducing desensitization. Safer peanut allergy treatment approaches and rigorous randomized controlled trials that evaluate patient-important outcomes are needed." whereas another opinion was more optimistic (287). Nevertheless, Aimmune's peanut OIT has been approved by FDA in the USA and is now marketed as "Palforzia" (https:// www.fda.gov/vaccines-blood-biologics/allergenics/palforzia).

Sublingual Immunotherapy
Another possible form of immunotherapy for nut allergy is sublingual immunotherapy, which is given in the form of allergen-containing tablets or drops that must be kept under the tongue. One intention for the development of SLIT was the reduction of side effects and its simplified application for selfadministration by the patients. However, clinical effects of SLIT are less pronounced than for SCIT for respiratory allergens (235) and there are only few studies, most of them performed in few patients for nut allergy (Table 3) (255,(271)(272)(273)(274)(275). Although few studies showed desensitization in some of the participants by the end of the therapy, the results regarding sustained unresponsiveness and long-term compliance are not encouraging (255,274).

Epicutaneous Immunotherapy
Epicutaneous immunotherapy (EPIT) is a more recent approach which has been developed originally for AIT of respiratory allergy (288) but has now been evaluated also for AIT of peanut allergy (235,289). EPIT is based on the direct application of an allergen-containing patch on the patient´s skin, similar as it is performed in APT. In theory, EPIT promises a reduced risk of systemic reactions and an uncomplicated application, also for children, due to its non-invasive nature. Table 3 provides an overview of current EPIT studies for nut allergy (276)(277)(278), which, however, is currently limited exclusively to peanut. Moderate success for the treatment of peanut allergy was reported, with one study showing some efficacy in children between 6 and 11 years (277). A review of available data states that "EPIT might induce desensitization in peanut allergy and an increased risk of local adverse events (AEs). These findings should be interpreted with caution owing to the limited study and heterogeneity. More data in the older (children ≥ 12 years and adults) and other allergic diseases are needed" (289). The analysis of systemic peanut allergen-specific IgG responses has shown that epicutaneous allergen administration induces only a very modest production of allergen-specific IgG and mainly specific T cell activation (16).

Molecular Immunotherapy via the Subcutaneous Route
As already mentioned above, SCIT has not been developed for AIT of allergy to class I food allergens, most likely because of the risk of inducing anaphylactic side effects when natural allergen extracts are used (244,245).
Regarding molecular AIT we found only one published study in which peanut allergic subjects had been treated by a molecular form of AIT using recombinant modified Ara h 1, 2 and 3 encapsulated in inactivated Escherichia coli (279) ( Table 3) but half of the subjects (5/10) in this trial experienced adverse reactions, and two of them had anaphylactic reactions.
For AIT of respiratory allergy, several molecular AIT approaches have been evaluated already in clinical trials (Figure 4), yielding encouraging results in terms of inducing protective IgG responses, alterations of cellular immune responses and evidence for clinical efficacy (235). These approaches include SCIT with recombinant or purified major allergen molecules (290), SCIT based on recombinant hypoallergenic allergen derivatives with (291) and without allergen-specific T cell epitopes (237,292). For the latter approaches the induction of allergen-specific blocking IgG antibodies has been demonstrated and evidence for clinical efficacy has been obtained. SCIT with allergen-derived T cell epitope-containing peptides has not been successful and an induction of allergen-specific IgG has only been demonstrated when relatively long peptides had been used [reviewed in (235)].
Regarding the development of molecular AIT approaches for treatment of allergy to class I food allergens, important and promising results have been collected for the major fish allergen parvalbumin which such as the major nut allergens represents a digestion-resistant and highly allergenic molecule (293). Within the European Union-funded research program FAST, a hypoallergenic recombinant mutant protein of the major carp allergen Cyp c 1 (294) has been produced, characterized and shown to be hypoallergenic in vivo (295)(296)(297). Furthermore, safety and ability to induce protective specific IgG responses has been demonstrated in first clinical trials for this molecular vaccine (https://clinicaltrials. gov/: NCT02017626; NCT02382718). Thus it has been proven that it is possible to develop recombinant hypoallergens for SCIT of class I allergens. First recombinant hypoallergenic derivatives of peanut allergens have been characterized in preclinical studies. In fact, several studies reported the production of modified allergen variants of the peanut allergens Ara h 1, Ara h 2 and Ara h 3 and demonstrated reduced IgE reactivity by immunoblotting using patient's sera (298)(299)(300). More recently, the generation of hypoallergenic variants of Ara h 2 and Ara h 6 with decreased allergenic activity but preserved T-cell proliferation capacity has been described (301). Similarly, Tscheppe et al. reported the production of a novel Ara h 2 hypoallergen lacking linear and conformational IgE epitopes (47). IgE reactivity to the unfolded mutant was tested using sera from Ara h 2-sensitized patients and showed reduced IgE-binding capacity compared to natural Ara h 2. The Ara h 2 mutant exhibited low basophil activation ability but still induced T-cell proliferation.
It is known that for allergy to class II food allergens beneficial effects can be obtained by SCIT with the genuinely sensitizing cross-reactive respiratory allergens (40,302) but the effects on food allergy seem to be lower due to limited cross-reactivity of the induced IgG antibodies (303).
Likewise, molecular AIT with recombinant hypoallergenic birch pollen allergen derivatives was found to induce also cross-protective IgG antibodies to cross-reactive food allergens (41, 291) but similar as for natural allergen extracts, there seems to be limited crossreactivity of therapy-induced IgG with the cross-reactive food allergens. This has been observed in the clinical trials but also in preclinical studies investigating the cross-protective potential of antibodies induced with molecular vaccines made for the treatment of respiratory allergy (304,305). Accordingly, it has been suggested to develop recombinant hypoallergens which incorporate also epitopes of the cross-reactive food allergen molecules (306).

Future Molecular Forms of AITs for Nut
Allergy: How to Crack the Nut Originally, recombinant hypoallergenic allergen derivatives have been made to incorporate allergen-specific T cell epitopes but it has been realized that also non-IgE reactive T cell epitopes can cause side effects by activating allergen-specific T cells leading to late phase side effects (192,307,308). The more recently developed technology of replacing allergen-specific T cell epitopes by unrelated carrier proteins (309) seems to reduce T cell-mediated side effects and has been shown to yield promising clinical data with approximately 25% improvement of symptoms over placebo when tested for SCIT of grass pollen allergy (237). One may therefore consider the development of carrier-based B cell epitope-containing vaccines by combining peptides derived from the IgE binding sites of the respiratory allergens and the corresponding cross-reactive class II food allergens to obtain combination vaccines for treatment of pollen allergy and the associated oral allergy syndrome (Figure 4, lower, right).
The technology of producing fusion proteins consisting of hypoallergenic peptides derived from IgE binding sites of allergens and allergen-unrelated carrier proteins may be applicable also for class I food allergens. However, it needs to be born in mind that it may be more difficult to identify hypoallergenic peptides in class I food allergens because they may harbor not only conformational IgE epitopes which can be easily disrupted but also sequential IgE epitopes of which some may be cryptic (i.e., hidden in the intact allergen structure and exposed only after digestion). It may therefore be difficult to identify non-allergenic peptides derived from the IgE binding sites of class I food allergens which are needed for the construction of the carrier-bound B cell epitope-containing vaccines. SCIT with recombinant purified class I food allergens is in principle possible but vaccines based on purified wild-type allergens may cause severe side effects. SCIT with recombinant T cell epitope-containing hypoallergens derived from class I food allergens seems possible and effective if the vaccines induce allergen-specific protective IgG antibodies but late phase, T cell-mediated side effects may occur. Treatment with T cell epitopecontaining peptides from class I food allergens will likely not be successful because short peptides fail to induce protective IgG antibodies but tolerogenic peptides may be considered for preventive approaches (Figure 4).
If one performs an analysis of strengths and weaknesses of current allergen extract-based AIT approaches for nut allergy and future molecular AIT vaccines several aspects need to be considered   Figure 5). Without doubt, advances have been made regarding the development of allergen extract-based AIT for nut allergy and experience has been collected in several clinical trials ( Figure 5 and Table 3). However, the major limitation for allergen extractbased forms of treatment resides in the fact that allergen extracts represent natural products which have major limitations regarding quality, allergen composition, purity and allergenic activity which only can be overcome by introducing molecular approaches for treatment ( Figure 5) (193). It seems to be due to side effects that SCIT approaches with natural allergen extracts for treating allergy to class I food allergens were not pursued. Instead, mainly OIT approaches have been investigated in larger trials whereas SLIT and EPIT are still in an experimental stage. Side effects are still a concern in OIT with allergen extracts and may be overcome with molecular AIT technologies using hypoallergenic allergen derivatives ( Figure 5). Studies performed with molecular AIT approaches indicate high potential but more efforts are needed to advance this treatment into clinical trials and into clinical use. Accordingly, hypoallergenic derivatives need to be developed for the most important allergens, and thus a thorough preclinical and clinical characterization needs to be performed which will require large efforts and investment into the development ( Figure 5). Most of the experiences have been collected for AIT of respiratory allergy but experience from preclinical and clinical trials in food allergy suggest a common mode of action indicating that SCIT with recombinant nut hypoallergens should be safe, induce protective IgG responses and exhibit clinical efficacy but clinical studies are lacking. Clear advantages of molecular AIT forms are the defined mode of production which satisfies Good Manufacturing Practice requirements needed for clinical studies. A major possible advantage is that molecular design will allow to develop safe and effective forms of AIT for allergy to class I food allergens. Furthermore, molecular AIT can be ideally combined with the already established forms of molecular diagnosis allowing the adequate selection of patients for treatment and also the monitoring of the treatment using molecular biomarkers (209,236,310). Figure 5 provides a summary of the SWOT analysis of existing allergen extract-based forms of AIT for nut allergy and future forms of molecular AIT but much more needs to be done regarding the preclinical and clinical development of molecular AIT forms for food allergy.

SUMMARY AND CONCLUSION
Nut allergies might lead to severe allergic reactions or even death, and yet the only current treatment option is avoidance of the allergen source. For AIT as well as in nut allergy diagnosis, extract-based methods are still used. Molecular diagnosis is an alternative to traditional allergen-extract based diagnosis and molecular AIT is a promising future perspective. Molecular AIT approaches require knowledge of molecular sensitization profiles in the population intended to treat. It is evident that currently available studies regarding prevalence of sensitization and allergy to nuts are highly heterogeneous regarding design and only few contain information about molecular sensitization profiles. Therefore, there is a need for molecular studies to obtain comparable data regarding the prevalence of allergy to certain nuts. Molecular IgE-based diagnosis for nut allergy diagnosis may reduce the risk of side effects by reducing the need for provocation tests and promises more comprehensive results. At the moment mainly oral forms of allergen-specific immunotherapy are studied which suffer from poor patients compliance and severe side effects. Molecular AIT is not yet well investigated for treatment of nut allergy although it promises a reduction of side effects through the use of recombinant hypoallergens.