Polymeric α‐Hairpinin Allergens Induce a Functional Response via a Single Antibody

Vicilins are abundant seed storage proteins to which peanut and tree nut allergic patients are frequently sensitized. The large C-terminal cupin domain has been extensively characterized, but less well studied are the N-terminal α-hairpinin domains. These domains are characterized by two α-helices held together by two disulfide bonds. Additionally, a skewed amino acid distribution for Arg, Glu, and Gln, coupled with a high structural similarity, suggests α-hairpinins are prone to be cross-reactive across distantly related species and controversially with other protein folds [1, 2]. α-hairpinins occur in variable numbers in different seeds, and occasionally without a cupin domain, e.g., Pru du 8. Prior research suggests that α-hairpinins are cleaved to generate individual low-molecular-weight proteins whose allergenicity is independent of both each other and their cupin domain [3]. However, incomplete cleavage could have important consequences for allergies. In walnuts, Jug r 2.0102 has six α-hairpinins expressed with the central adjacent domains (α3 and α4) exhibiting the highest sequence identity of 90% (Figure 1A and Figure S1A,D). Using hybridoma-based technology we identified a human-derived IgE monoclonal antibody, 6D12, that binds specifically to both recombinant α3 and α4 allergens (Figure S1) [4]. Using X-ray crystallography, we determined a crystal structure that shows 6D12 bound to α4 at the turn between the helices (Figure 1E and Figure S1). Based on this structure, we tested site-directed mutants with ELISA to demonstrate that G30 and R34 of α4 were critical sites for the specificity of 6D12 (Figure S1). These residues are uniquely found in α3 and α4, suggesting why 6D12 is specific for these two α-hairpinins. Furthermore, western blot analysis of walnut extract revealed variable cleavage products of Jug r 2, ranging from one to three α-hairpinin(s) and larger bands still attached to the cupin domain (Figure 1C,D). Consequently, 6D12 antibody alone could potentially elicit an immune response to any cleavage products containing α3-α4 fragments via repeated cross-linking (Figure 2A). A similar scenario was suggested to enhance allergic reactivity to Phl p 5, which has two homologous domains tethered via a flexible linker [5]. To substantiate the functional response to a polymeric allergen with a single antibody, we recombinantly expressed an α3-α4 construct and conducted sandwich ELISA, basophil activation test (BAT), and a mouse model of passive systemic anaphylaxis (PSA). All three assays demonstrated a functional response with 6D12 (Figure 2B–D). However, walnut extract was unable to replicate this response likely due to the presence of cleavage products with individual α3 or α4 monomers. This is analogous to the known hook effect in BAT where high concentrations of antigen inhibit antibody cross-linking. To confirm this assumption, we demonstrated via sandwich ELISA and PSA that monomers of α3 could inhibit the functional response of the α3-α4 construct (Figure 2B,E). This suggests a feasible mechanism to inhibit or enhance activation. Consequently, this presents a uniquely different scenario to that of Phl p 5, whereby repeated cross-linking can be inhibited or activated due to the ratio of inhibitory molecules to activating polymers. The cleavage products of α-hairpinin allergens from other food sources raise concerns. For example, in hazelnut, Cor a 16 contains 12 α-hairpinin repeats with higher aggregate sequence identity (Figure S2A,B). Using single-cell sequencing technologies from allergic donors and ELISA-based screening, we identified a human IgE antibody, IGX-0705, specific for α-hairpinins of Cor a 16 in addition to several α-hairpinins from walnut (Figure S2C,D) [6]. First, this demonstrates that humans can make a cross-reactive antibody to hazelnut α-hairpinins. Second, western blot results indicate partial cleavage products in hazelnut with fewer α-hairpinin monomers (Figure S2E); this ratio could more readily induce a functional response. More research is needed as to the polymeric status of α-hairpinins in food products and the possible influence of repeated functional activation. All authors contributed to data acquisition and analysis. G.O.L. and G.A.M. wrote the manuscript. This research was funded in part by the Intramural Research Program of the National Institute of Environmental Health Sciences, NIH: 1ZIAES102906 (GAM), 1ZICES102645 (LCP), 1ZICES102487 (RMP). This work was supported by the National Institute of Environmental Health Sciences. Regarding 6D12: Human subjects with a diagnosis of food allergy were recruited during a routine visit with their allergist provider. Blood samples were collected for our study as approved by the Vanderbilt University Medical Center Institutional Review Board (IRB 141330 and 142,030). Regarding IGX-0705: In compliance with US regulations and policies governing human subject research, this study obtained institutional review board approval (no. 0000635, protocol reference Pro00049174) before proceeding with subject screening and patient consent. For BATs, blood samples were collected from adult volunteers with informed consent at the NIEHS Clinical Research Unit under the Environmental Exposure and Injury Protocol 000152 with all applicable institutional review board approvals. D.C., P.C., V.A., and J.W. are employees of and stakeholders in IgGenix. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Figure S1: 6D12 binding to Jug r 2. (A) SDS-PAGE of GST-tagged α-hairpinins and Western blot showing that 6D12 IgG binds to α3 and α4 of Jug r 2.0102, but not to a 4xCys-Ser mutation of α4. Sequence comparisons are shown below. G30 and R34 of α4 are highlighted. (B) ELISA data verifying results of Western blot that 6D12 is specific for GST-tagged α3 and α4. (C) Circular dichroism data showing that the 4xCys-Ser mutant lacks the helical structure typical of α-hairpinins. (D) Percent sequence identity matrix of Jug r 2.1012 α-hairpinins showing that α3 and α4 are the most similar overall and locally in the turn between the helices omitting gaps. (E) Crystal structure of rJug r 2 α4 bound to 6D12 zoomed in on the binding region highlighting the surface charge of the antibody. (F) ELISA data of mutant α4 proteins showing that the shallow pocket of G30, and the positive charge of R34 provide specificity for α3 and α4 versus other Jug r 2 α-hairpinins. Figure S2: Cor a 16 and IGX-0705 IgG. (A) Domain structure of Cor a 16. (B) Percent sequence identity matrix of α-hairpinin domains of Cor a 16 omitting gaps. (C) ELISA intensity data screening various α-hairpinins for binding to IGX-0705. (D) Western blot of GST-Cor a 16 α4 probed with IgG IGX-0705. (E) Western blot of 5 μg reduced hazelnut extract probed with IGX-0705. Table S1: Recombinant α-hairpinins. Table S2: Data collection and refinement statistics of the structure of 6D12 complexed with α4. Table S3: The kon, koff, and kD for 6D12 and IGX-0705 binding to GST-Jug r 2 α4. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.