Fine resolution of the N-terminal IgE-binding epitope of Ara h 2: Discovery of variants with enhanced IgE binding

Abstract

Background: IgE binding to linear peptides from the N-terminal region of Ara h 2 (epitope 1) is associated with the achievement of sustained unresponsiveness in young children receiving oral immunotherapy and may be important for cross-reactivity between peanuts and tree nuts. This region is also part of the binding site for neutralizing IgG monoclonal antibodies associated with sustained unresponsiveness following oral immunotherapy.

Objective: We sought to perform alanine scanning of this epitope to determine the importance of individual amino acids and then amino acid scanning to screen for sequences with enhanced binding of IgE.

Methods: A streptavidin IgE ELISA with biotinylated peptides was used to measure the binding of IgE to full-length and truncated peptides to identify a core sequence (DRRCQSQLERAN, amino acids 30-41 in the Ara h 2 sequence). Peptide microarrays were used to screen multiple peptides and quantitate binding of IgE. Statistical analysis included one-way ANOVA followed by the Dunnett multiple comparison test.

Results: IgE binding was greatly reduced when alanine was substituted for arginine at positions 31, 32, and 39 (R31, P < .001; R32, P < .01; R39, P < .001); glutamine at positions 34 and 36 (Q34, P < .01; Q36, P < .001); and glutamate at position 38 (E38, P < .01). Substitution of aspartate with asparagine at position D30 in conjunction with substitution of asparagine at position N41 with either leucine or lysine gave enhanced binding (P < .0001). Molecular modeling of these data suggests a conformational basis for recognition by polyclonal IgE.

Conclusions: IgE binding assays using pooled and individual sera demonstrated the importance of amino acids throughout the sequence of epitope 1 for immune recognition. The results of alanine scanning indicated residues that could be changed as part of a larger strategy to generate hypoallergenic forms of Ara h 2, whereas sequence variants with enhanced binding were identified that may be useful for improving diagnostics.

Keywords: 2S albumins; Ara h 2; IgE; epitope; food allergy; mimotope; peanuts; peptides.

FIG 1.

IgE binding to epitope 1 and alanine scanned peptides of epitope 1. IgE microarray assays of alanine scanned peptides were performed as described in the Methods section. Background (100–300 density units) with empty spots was subtracted. Data from 4 individual assays were pooled. Data were normalized so that the signal with native sequence was set to 100%. Substitution of 5 of the 12 amino acids (R31, R32, Q34, E37, R38) individually with alanine showed significantly reduced binding of polyclonal IgE. The data reported as density units are shown in Fig E2. ***P < .001, **** P < .0001.

FIG 2.

Residues in the N-terminal epitope 1 of Ara h 2 with reduced IgE binding in the alanine scan have substantial SASAs in the protein. IgE microarray assays were performed as described in the Methods section. (A) Residues that, when changed to A, have significantly reduced IgE binding are shown in red on this region of the Ara h 2 structure. (B) SASAs were calculated by the GetArea program, which measures quantitatively the area of a peptide that is accessible by a water molecule. The differences in SASAs for the residues of epitope 1 in the isolated peptide (SASA Peptide) and in the folded 3D structure of Ara h 2 (SASA Fold) are shown. The difference between these values (SASA Diff) shows that these critical residues are surface exposed in both forms, indicating that IgE recognition of the epitope 1 peptides reflects the structure of Ara h 2.

FIG 3.

IgE binding to novel variants of epitope 1. IgE microarray assays of novel variants of epitope 1 were performed with the serum pool in 3 independent assays as described in the Methods section. Background (100–300 density units) with empty spots was subtracted. Data were then normalized so that the signal with the native sequence was set to 100%. Data are shown as percent of the signal seen with the native sequence. The data reported as density units are shown in Fig E5. ***P < .001, ****P < .0001.

FIG 4.

Dose response of IgE binding to epitope 1, D30N/N41K, epitope 1, D30N/41F and epitope 1, D30N/N41L, compared with the native sequence. Three independent IgE microarray assays were performed as described in the Methods section with varying dilutions of the serum pool to yield the Ara h 2–specific IgE concentrations noted. Circles indicate native sequence; triangles indicate epitope 1, D30N/N41K; upside-down triangles indicate epitope 1, D30N/N41F; and squares indicate epitope 1, D30N/N41L. *P < .05, **P < .01, ***P < .001, ****P < .0001.

FIG 5.

Proposed structural changes in epitope 1, D30N/N41K, and epitope 1, D30N/N41L, peptides with higher affinity for IgE, indicates that the altered residues have similar high surface exposure. The critical sidechains of epitope 1 for IgE binding are illustrated in red on the 3D model of Ara h 2 (A) and orange for epitope 1, D30N/N41K (B), and epitope 1, D30N/N41L (C). The amino acid changes at positions 1 and 12 in the variants are colored red.