The members of the gibberellin-regulated protein (GRP) family have been identified as allergens quite recently and they are currently attracting much attention. The GRPs are basic, small proteins (7 kDa) stabilized by six disulphide bridges. Similar to other allergens, such as non-specific lipid trasfer proteins (nsLTPs), they are stable for heating and gastrointestinal digestion (1). Peamaclein, the peach GRP, is the first member of this protein family to be identified and registered as an allergen (Pru p 7). Its detection was driven by the clinical history of a patient reporting anaphylaxis after peach consumption, but the patient had proved to be immunoglobulin E (IgE) negative to the isolated peach allergens available at that time (2). Then, Pru p 7 was associated with severe allergy symptoms and proposed as a marker allergen related to systemic reactions in peach allergy (3–5). In the meanwhile, the allergenic homologous proteins were identified in other fruits, such as Pru m 7 in Japanese apricot (4), Pun g 7 in pomegranate (6), Pru av 7 in cherry (WHO/IUIS), and Cit s 7 in orange (7). More recently, an allergenic GRP was identified in the pollen of Cupressus sempervirens (Cup s 7), and its IgE-binding competition with Pru p 7 and Pun g 7 was demonstrated (8). Cross-reaction between Cup s 7 and the homologous proteins from Cryptomeria japonica (Cry j 7) and Juniperus ashei (Jun a 7) pollens was also observed (9). In addition, a frequent association between the GRP-related food allergy and Cupressaceae pollen sensitization, especially occurring in the areas with high exposure to this type of tree pollen, was reported (9, 10). The cross-reactivity between the GRPs was suggested as a cause of multiple fruit allergies and the clinical cross-reactivity, at least among peach, Japanese apricot, orange, and pomegranate was proven (11). In addition to these fruits, the patients sensitized to GRP are frequently reported to experience allergic reactions against other fruits, such as apples (4, 12).

Apple is one of the most consumed fruits in the world. Nevertheless, in some subjects, it can cause different patterns of allergic reactions, from oral allergy syndrome (OAS) to severe systemic reactions. Sometimes, allergy to apple is associated with pollinosis, especially in the patients sensitized to Bet v 1 (13). Four allergens, Mal d 1–4, have been so far identified in this fruit and registered by the WHO/IUIS. Mal d 1 is a Bet v 1-like protein, Mal d 2 a thaumatin-like protein, Mal d 3 a 9 kDa nsLTP, and Mal d 4 a profilin (14, 15). The published observation that the patients sensitized to GRP frequently reported allergic symptoms against apple (4, 12), which, such as peach, taxonomically belongs to the Rosaceae family, suggested the presence in this fruit of a still uncharacterized allergen belonging to the GRP family.

In this context, this study aimed to investigate the presence of GRP in the peel and pulp of the apple fruit, the isolation of the protein, and the analysis of its structural and immunological properties. The specific objects of this study included the analysis of (i) the structural features by similarity search and molecular modeling, (ii) the frequency of sensitization to the apple GRP in a random population of suspected allergics, (iii) the frequency of sensitization in a subpopulation of the patients sensitized to GRP, (iv) reported allergy symptoms of patients following the apple ingestion and (v) a comparative analysis of the structural and immunological features of apple GRP with homologous allergenic proteins.

Applemaclein was found in quite similar amounts in the peel and pulp of both the analyzed cultivars, one (Annurca) is common in the South of Italy and the second (Golden Delicious) is a widespread commercial one. The protein shows a high sequence identity (94%) with Pru p 7, representing the prototype of allergenic GRP (2, 3, 5). Pru p 7 has mainly helical structure in the N-terminal region and the C-terminal appears flexible in solution by NMR analysis (1). The detected structural features suggest a possible shift of the molecule to a significantly more structured form following the interaction with a ligand, likely negatively charged surfaces, such as biological lipids/membranes (1). The two N-terminal helical stretches encompass residues 4–16 and 25–30. Compared with Pru p 7, the first amino acid substitution (residue 3) in an apple GRP is just before the first helical strech, the second substitution (residue 19) is placed between the first and the second helical stretch, whereas residues 59–60 are placed in the C-terminal loop region (Figures 1A,B). Therefore, the structural differences between the peach and apple GRP rely on these four residues that could be directly, or indirectly, involved in IgE-binding individual properties. Based on our modeling results, we suggest that most likely residues 59 and 60 are the most important in relation to the differences in IgE binding between applemaclein and Pru p 7. Our prediction is based on the fact that the neutral residues Asn59 and Pro60 of Pru p 7 are replaced with the two negatively charged residues Glu59 nd Asp60, respectively (Figure 1). Therefore, the surface properties of the corresponding C-terminal loop regions differ significantly. Additionaly, it was shown that the C-terminal loop region displays the highest mobility (27), therefore it is likely that a drastic change in the amino acid composition will also impact the conformation of this region and as a result, alter a surface region that potentially is recognized by IgE.

An apple GRP was recognized by specific IgE of the allergic subjects and efficiently competed with the peach and pomegranate GRP for IgE binding, producing a complete inhibition on both the homologous proteins. In a random and heterogeneous population of 4,721 Italian suspected allergics, the frequency of sensitization to the apple GRP was higher than that observed for the peach and pomegranate, but much more similar to that of Mediterranean cypress homologous protein. Nevertheless, in a population of 37 sensitized patients of Italy for which a clinical history was available, only one patient reported allergic reactions to apple, whereas some others did not report symptoms, but excluded this fruit from their diet. It is worth noting that the patient reporting allergic reactions, as well all the patients refusing the apple consumption, were not monosensitized to the GRP, but they were also IgE positive to nsLTP and/or Mal d 1. Therefore, it was not possible to associate the exclusion of apples from their diet to GRP and/or nsLTP and/or to other allergens. In addition, most of the 37 patients of Italy IgE positive to applemaclein, some of them at least apparently monosensitized to GRP, reported the consumption of apple without allergic symptoms connected to this fruit. In contrast, a Japanese study described the detection of patients monosenstized to GRP reporting allergic reactions against several fruits, frequently, such as apple (4), although peach was the fruit most frequently (92.3%) inducing allergic symptoms. Unfortunately, due to the insufficient amount of available sera, we could not investigate the possibility that the high levels of IgG/IgG4 blocking antibodies affected the results obtained in the analyzed population of Italy. Nevertheless, the mismatch between positive sensitization and the lack of reported clinical symptoms to apple claims further future investigations. Furthermore, in addition to the GRP allergens, it will be interesting to perform similar studies on other plant allergens and make a comparative analysis.

The researchers from France described the peach GRP as a major allergen in the patients allergic to peach from southern France, where the exposure to cypress pollen is high, with a prevalence of Pru p 7 sensitization exceeding 60% (9, 28). In contrast, a recent study reported that the allergy and sensitization to foods secondary to cypress pollen allergy appeared as rare phenomenon in the Italian population and that most of such patients react to peach, rather than other Rosaceae fruits (29). Therefore, discrepant results concerning the allergenic properties of GRP can sometimes be found in the literature. It seems that the same protein displays a different behavior, becoming a major or minor allergen, which may cause mainly localized or systemic allergic reactions, depending on the population/geographical area considered.

It is therefore conceivable that some factors, or a combination of them, can influence the GRP behavior and the response of our immune system. A great relevance has been so far given to the primary structure of allergenic proteins and certainly, it is of utmost importance and can explain the co-recognition and cross-reaction occurrence (30, 31). However, an increasing collection of data suggests that the final result, in the terms of sensitization and allergic reactions, could be affected also by other factors. For instance, a study performed in a Japanese population revealed that in more than half of the patients monosensitized to GRP (61.5%), exercise or aspirin intake acted as cofactors enhancing the allergic reaction onset (4). The influence of these factors, and also of additional ones, including alcohol, non-steroidal anti-inflammatory drugs, estrogens, antiacids, temperature, and compounds from Cannabis sativa, was reported for some allergens (32), especially for nsLTP (33, 34). nsLTP was described as the most frequent sensitizer in the Italian subjects with food-dependent exercise-induced anaphylaxis (FDEAIA) (35). It was reported that the allergenicity of nsLTP, as well as that of other allergens with the function of hydrophobic molecule carrier, could be influenced by the bound ligand which may act as an adjuvant and drives the immune system toward a type 2 (pro-allergenic) response (36, 37). For instance, the interaction of Pru p 3 with selected fatty acids was described to increase its IgE binding capacity, whereas the binding of some other lipids did not have the same effect (38). Different hydrophobic molecules have been suggested to bind nsLTP from several sources (39, 40). In fact, the nsLTP fold enables the formation of a hydrophobic tunnel-like cavity where hydrophobic ligands with different sizes can be accommodated (31, 37).

Similar to nsLTP, a GRP structure is predicted to form a cavity (12) where a ligand, such as biological lipids/membranes, can be bound (1). This is in agreement with the applemaclein modeling results (Figure 1C) showing that the surface of applemaclein is very irregular and there are some grooves/clefts that may potentially accommodate some small molecular ligands. However, we were not able to indentify such ligands. Due to its positive charge, the GRPs may interact with some negatively charged molecules, and therefore it is not surprising that they are associated with the membranes (41). It was also demonstrated that the GRPs have antimicrobial properties and are able to inhibit the growth of various bacterial and fungal pathogens (42, 43). In addition, the GRPs may be involved in the interactions with foreign proteins, such as bacterial adhesion protein from Pseudomaonas fluorescense (44), or interact with plant sucrose transporter SUT-1 (45). Generally, these proteins are implicated in many biological functions in plants, however, the molecular mechanism of their functions is still not understood. While the automated searches have not provided any clear hint on applemaclein molecular function, the available literature consistently reports that the GRPs are not susceptible to digestion and are quite stable (1). In line with the literature, applemaclein also proved to be resistant to trypsin and pepsin digestion. The combination of this information with the structural features of applemaclein (several disulfide bridges, small size and a presence of a loop that can protrude to an active site of an enzyme) hint that the protein may potentially act as a protease inhibitor or an enzyme inhibitor. For example, it was shown that Snakin-Z isolated from Zizphus jujuba fruits is able to inhibit human acetylcholinesterase and butyrylcholinesterease (46). The model of applemaclein (Figure 1B) suggests that one of the lysine residues (Lys57) is located on a C-terminal loop region of the protein and that the localization of this residue is somewhat similar to the position of a lysine residue present in bovine trypsin inhibitor, which blocks the active site of human trypsin (47). However, while there are some structural similarities between the applemaclein and small protein protease inhibitors, our experimental results have not confirmed the inhibition of bovine trypsin by the apple GRP, nor the formation of a stable complex between these two proteins.

In analogy with nsLTPs, which are described as the most common triggers and responsible for the main allergies in the Mediterranean area, we can speculate that the GRPs could represent important allergens frequently causing severe symptoms especially in the areas, such as Japan, but they could be less important in other geographical areas, such as Italy. Further studies are needed to identify the ligand(s) of applemaclein, and of GRPs in general, and to understand whether the bound molecules play a role in the modulation of IgE binding and the triggering of allergic reactions.

Additional data are available in Supplementary Material. Further inquiries can be directed to the corresponding author.

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

LT and RA: methodology and writing. CA and MT: data analysis and writing. IG: methodology. MC: conceptualization, funding acquisition, data analysis, and writing. AM: patient selection, data analysis, and funding. MAC: conceptualization, funding acquisition, data analysis, and writing. All authors have read and agreed to the published version of the manuscript.

This study was partially supported by ADL s.r.l., on the basis of an Agreement of Scientific Collaboration. MC was partially supported by the R01AI077653 grant from the National Institute of Allergy and Infectious Diseases. The content is solely the responsibility of the authors and does not necessarily represent the official views of ADL and of the National Institutes of Health.

IG and LT were employees of ADL s.r.l. when this study started, currently they are employees of IBBR-CNR. MT and MAC receive funding from ADL s.r.l. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.