Concomitant sensitization to legumin, Fag e 2 and Fag e 5 predicts buckwheat allergy

Summary Background Buckwheat (Fagopyrum esculentum) has become increasingly popular as a healthy food in Europe. However, for sensitized individuals, consumption can cause anaphylactic reactions. The aim of this study was to identify individual well‐characterized buckwheat allergens for component‐resolved diagnosis. Methods Patients were selected by positive skin prick test to buckwheat and divided into two groups: (1) sensitized to buckwheat without clinical symptoms and (2) buckwheat allergy. Buckwheat proteins were extracted from raw buckwheat seeds, purified applying a combination of protein precipitation and chromatographic methods, and analyzed by IgE immunoblotting and ELISA. Results Buckwheat‐allergic patients had a significantly larger median skin prick test weal diameter for buckwheat than the sensitized group and the positive control. Also, IgE immunoblotting clearly showed a distinct pattern in sera from allergic patients when compared to sensitized individuals. Several IgE‐reactive proteins were purified from crude buckwheat extract, namely legumin (Fag e 1 plus its large subunit), Fag e 2 (2S albumin), and newly identified Fag e 5 (vicilin‐like) as well as hevein‐like antimicrobial peptides, designated Fag e 4. All four allergens showed superior diagnostic precision compared to extract‐based ImmunoCAP with high sensitivity as well as high specificity. Conclusions Patients with clinical symptoms clearly show a distinct allergen recognition pattern. We characterized a buckwheat vicilin‐like protein as a new relevant marker allergen, designated Fag e 5. Additionally, another new allergen, Fag e 4, potentially important for cross‐reactivity to latex was added to the allergen panel of buckwheat. Further, our data show that the full‐length legumin comprising both, large and small subunit should be applied for component‐resolved diagnosis. Our data indicate that concomitant sensitization to legumin, Fag e 2 and Fag e 5, predicts buckwheat allergy.


| INTRODUCTION
In Asian countries, common buckwheat (Fagopyrum esculentum) is a popular, traditional food. Nowadays, it becomes increasingly popular also in Western countries as part of a healthy diet 1 due to its wellbalanced amino acid composition. 2 Buckwheat is also used as a substitute in gluten-free food, especially for people suffering from coeliac disease.
Although the prevalence of buckwheat allergy is relatively low (0.22% in Japan, 0.11% in Korea), 3,4 it is often associated with severe anaphylaxis. Similar to peanut allergy, even small amounts can cause severe, life-threatening reactions. 5 This is an important issue because buckwheat is often consumed as a hidden allergen in cakes, pancakes, and pastries. In Japan, it is estimated that 2.9%-3.4% of all reported anaphylactic events to foods are caused by buckwheat, 6 while in Korea, buckwheat has been identified as the leading cause of food allergy. 7 In Europe, data concerning the prevalence of buckwheat allergy are limited to date. Two Italian studies reported a sensitization prevalence of 3.6% and a prevalence of buckwheat anaphylaxis of 1%, respectively.
In France, the prevalence was 4.5% of cases of food anaphylaxis. [8][9][10] Currently, the widely used first-line diagnostic approach of buckwheat allergy is skin prick test (SPT) and in vitro tests such as Immu-noCAP, to detect and quantify buckwheat-specific IgE (sIgE).
However, despite high sensitivity, the specificity of these tests are low. 11 Thus, in many cases, food challenges are still the gold standard for a proper diagnosis.
Component-resolved diagnosis (CRD) allows to discriminate between clinically relevant and non-relevant sensitization for many food allergen sources (e.g. peanut, hazelnut, kiwifruit). 12 Although buckwheat is recognized as a major food allergen source, there is limited knowledge on the causative allergens. Up to now, three important buckwheat allergens have been identified, named Fag e 1-3. Urisu et al. 13 characterized a protein of approximately 24 kDa with high IgE-binding potential that was identified as the small subunit of buckwheat legumin (13S globulin) and tentatively designated Fag e 1. 14 Predominantly severe reactions are observed in patients sensitized to a 16 kDa protein that was identified as a member of the 2S albumin family. 15,16 This 16 kDa protein was officially designated Fag e 2 (www.allergen.org). In addition, Fag e 3, a 19 kDa N-terminal fragment of a vicilin-like protein was identified. 17 Most likely it originated from cleavage of a vicilin precursor protein as previously shown for a macadamia vicilin. 18 Choi et al. concluded that sIgE to Fag e 3 correlated with true buckwheat allergy and was less prevalent in individuals only sensitized to buckwheat without clinical symptoms. Other studies also reported IgE binding to 40-50 kDa proteins. 19,20 This study aimed to identify individual buckwheat allergens including potential marker allergens for CRD. Therefore, we evaluated and compared the allergen recognition pattern in sera from patients allergic to buckwheat and patients sensitized, but tolerant to ingestion of buckwheat, respectively. We further assessed the allergenic potential of the purified individual proteins including legumin, vicilin, and 2S albumin. A total of 389 patients were tested for buckwheat sensitization   by SPT at the Odense Research Center for Anaphylaxis (ORCA), Allergy Centre, Odense University Hospital, Denmark. Of these, 249 were evaluated due to suspicion of buckwheat allergy, food allergy of unknown cause or food-dependent exercise-induced anaphylaxis, and 140 were tested consecutively during a 4 months period. Based on a positive SPT to buckwheat, 52 patients were selected for further studies ( Figure S1). The diagnosis of buckwheat allergy was verified in 21% (11/

| Statistics
Results are given as medians and ranges. Statistical analyses were performed using the Mann-Whitney test for differences between independent samples and Chi square for differences between groups. P-values below .05 were considered significant. Analyses were performed with GraphPad Prism (GraphPad Software, La Jolla, USA) and STATA12 SE (Stata Corporation, College Station, TX, USA).  (Table S1). Among patients from the buckwheat-allergic group, eight patients were tested for buckwheat-sIgE and all were positive. From the SPT-sensitized subgroup, 18 were tested for buckwheat-sIgE and ten were positive (55.5%). Of the 10 in the sensitized subgroup, five had a negative food challenge, one patient had consumed buckwheat at home but without symptoms, and four had no clinical suspicion of buckwheat allergy ( Figure S1).

| Clinical data
Median sIgE for SPT-sensitized and allergic patients was 0.65 kU A / L (range: <0.35-14.2 kU A /L) and 4.7 kU A /L (range: 1.9-36.9 kU A /L), respectively. There was a significant difference between the groups (P < .002). Buckwheat-allergic patients had a significantly larger median SPT weal diameter for buckwheat than the sensitized group whereas the median diameter of the sensitized group was significantly smaller than that of the controls (P < .0001) (median: 4.5 mm vs 6.5 mm). Data are summarized in Table 1.
Of these, two were diagnosed with peanut allergy also having the highest buckwheat-specific IgE in the buckwheat-sensitized group, 14.2 and 6 kUA/L, respectively. Nine of 10 and 6 of 9 buckwheat-sensitized IgE positive patients were also cosensitized to grass pollen and birch pollen, respectively. Four of 11 and 0 of 10 allergic patients were cosensitized to grass pollen and birch pollen, respectively (Table S3).
To obtain a better overview of cosensitization patterns of the

| Purification and identification of several important buckwheat allergens
To identify and characterize the IgE-binding proteins, we purified the components using a combination of precipitation and chromatographic methods. The purified proteins were identified by N-terminal sequencing and/or tandem mass spectrometry or confirmed by MALDI-TOF mass spectrometry ( Figure S3A-D). Figure 2 shows all purified components at reducing and non-reducing conditions. The most abundant component was legumin (L-S, disulphide-linked small and large subunit; 55-70 kDa), which after reduction was separated into a small (S; Fag e 1; 18-22 kDa) and a large subunit (L; 33-38 kDa), providing the typical pattern of legumins. 22 Immunoblots using purified legumin were performed at non-reducing and reducing conditions ( Figure S4).
Binding of sIgE to the native protein (L-S) was observed when testing sera from buckwheat-allergic patients (5 of 7 sera). However, upon reduction, only 2 of 7 sera recognized the small subunit ( Figure S4). In addition, sera from 2 healthy donors as well as 3 from sensitized individuals also displayed IgE binding to the small subunit. The newly identified allergen Fag e 4 was recognized by 5 of 7 sera from buckwheat-allergic patients. Figure 4 gives an overview on the We showed that full-length buckwheat legumin (L-S; 55-70 kDa) from the seed storage protein family did not only provide a strong signal on the immunoblot, but also bound significantly more sIgE in sera from allergic patients as compared to the control group. Previous reports described the small subunit of the legumin (BW24KD) 13 as a major allergen, also designated Fag e 1. 14 In our study, fulllength legumin (L-S; 55-70 kDa) was recognized by 5 of 7 sera.
However, upon reduction, only 2 of 7 sera from allergic patients reacted with the small subunit (S; 18-22 kDa; Fag e 1) but also sera from the control group as well as from healthy donors ( Figure S4).  Serum S10 Serum S11 Serum S12 Serum S13 Serum S14 Serum S15 Serum S16 Serum S17 Serum S18 Serum S19 Serum S20