Epitopes with similar physicochemical properties contribute to cross reactivity between peanut and tree nuts

Highlights

• Allergen peptides with similar PhysicoChemical Properties (PCP) can bind IgE.

• An IgE epitope PCP-consensus identified similar repeats in diverse plant allergens.

• Synthetic consensus peptide (cp) inhibited patient IgE binding to nut extracts.

• cp antibodies preferentially bound to allergens containing multiple PCP repeats.

• These repeats may contribute to cross reactions and the severity of nut allergies.

Abstract

Many individuals with peanut (PN) allergy have severe reactions to tree nuts (TN) such as walnuts or cashews. Although allergenic proteins in TN and PN have overall low identity, they share discrete sequences similar in physicochemical properties (PCP) to known IgE epitopes. Here, PCP-consensus peptides (cp, 13 aa and 31 aa) were identified from an alignment of epitope rich regions of walnut vicilin, Jug r 2, leader sequence (J2LS) and cross-reactive epitopes in the 2S albumins of peanut and synthesized. A peptide similarity search in the Structural Database of Allergenic Proteins (SDAP) revealed a network of peptides similar (low property distance, PD) to the 13 aa cp (13cp) in many different plant allergens. Peptides similar to the 13cp in PN and TN allergens bound IgE from sera of patients allergic to PN and TN in peptide microarray analysis. The 13cp was used to produce a rabbit consensus peptide antibody (cpAB) that detected proteins containing repeats similar to the 13cp in western blots of various nut extracts, in which reactive proteins were identified by mass spectrometry. The cpAB bound more specifically to allergens and nut extracts containing multiple repeats similar to the 13 cp, such as almond (Pru du 6), peanut (Ara h 2) and walnut (Jug r 2). IgE binding to various nut extracts is inhibited by recombinant J2LS sequence and synthetic 31cp. Thus, several repeated sequences similar to the 13cp are bound by IgE. Multiple similar repeats in several allergens could account for reaction severity and clinically relevant cross-reactivity to PN and TN. These findings may help improve detection, diagnostic, and therapeutic tools.

Introduction

Up to 30 % of individuals with a clinically significant peanut (PN) allergy are also clinically allergic (not just sensitized) to at least one tree nut (TN) species (Bublin and Breiteneder, 2014; de Leon et al., 2007, 2003; Glaspole et al., 2011; Kulis et al., 2009; Maloney et al., 2008; Sicherer et al., 2003; Uotila et al., 2018). While the major TN and PN allergens can be grouped into similar protein families (PFAM) (Ivanciuc et al., 2009a; Schein et al., 2010), the overall sequence identity even within the same PFAM is quite low. For example, the peanut 2S albumin allergen, Ara h 2, is <35 % identical to the related walnut allergen, Jug r 1 (Smeekens et al., 2018). This suggests their allergenicity lies in discrete sequence areas. Previously, the peptide similarity search program, in the Structural Database of Allergic Proteins (SDAP) (Ivanciuc et al., 2009a, a; Ivanciuc et al., 2009b, 2002) was used to identify sequences with similar physicochemical properies (PCPs) to an immunodominant epitope of the 2S albumin, Ara h 2 and predict potential epitopes in other allergens (Maleki et al., 2011, 2012). The PD search revealed similar IgE epitopes in the leader sequence (LS) of the walnut allergen, the 7S globulin Jug r 2 (J2LS). These repeated sequences were recognized by serum IgE of both walnut and peanut allergic individuals. Therefore the walnut allergen Jug r 2, whose mature region has a vicilin fold similar to Ara h 1, has several IgE reactive sequences in its LS that are similar (low property distance, PD) to an immunodominant epitope of Ara h 2, a 2S albumin. Here we address the question of whether repeats with similar PCPs to a consensus sequence could play a role in IgE binding, and cross reactivity of peanuts and nuts.

Plant proteins are known to contain repeated sequences, due to the action of transposons and the polyploidy of their genomes. For example, repeated sequences account for as much as 67 % of cotton genomes (Li et al., 2015) and microsatellite repeats are used to classify cultivars of plants such as peanuts (Wang et al., 2016), rice (Kaur et al., 2015) flowering cabbage (Chen et al., 2017) and many others. Plant selection can also contribute to gene duplications if the sequences enhance characteristics such as stability to temperature fluctuation, growth rate or resistance to environmental stresses. One would expect that the genes of rapidly growing, short lived plants such as peanuts would evolve at a faster rate than the genome of TN (Luo et al., 2015) and therefore contain more repeats.

To determine whether other plant allergens contain peptides with similar potential epitopes (or repeats), the corresponding areas in peanut 2S albumin allergens (Ara h 2, h 6, and h 7) were aligned with the three long repeats of the J2LS (Fig. 1). The PCPcon program (Danecek et al., 2010; Danecek and Schein, 2010; Schein et al., 2012) was then used to calculate a PCP-consensus peptide (cp) for a 31 consensus peptide (31cp) and a shorter 13cp from the most conserved region of the trimeric repeat. Extending the search for peptides similar to the 13cp in SDAP revealed a network of similar peptide repeats in many other allergens, from different protein families. Further, rabbit polyclonal antibodies to the 13cp (cpAB) detected allergens predicted to contain repeated sequences in nut extracts. We empirically determined the binding of a specific antibody and IgE from patient sera to similar peptides from PN and TN allergens synthesized on microarrays. We found that the most reactive peptides were indeed those with the highest similarity (lowest PD) (Ivanciuc et al., 2004) to the 13cp.

The high similarity of the cross-reactive, repeated sequences in allergens from strikingly different plants and trees indicates they must have arisen from an ancient common ancestor. These results provide evidence that the 13cp represents an evolutionarily conserved region that may account for clinically significant cross-reactivity. This knowledge may be useful in anticipating proteins with high allergenic potential and improving diagnosis of cross-reactivity.