Single B-cell deconvolution of peanut-specific antibody responses in allergic patients

Abstract

Background: The frequencies, cellular phenotypes, epitope specificity, and clonal diversity of allergen-specific B cells in patients with food allergy are not fully understood but are of major pathogenic and therapeutic significance.

Objective: We sought to characterize peanut allergen-specific B-cell populations and the sequences and binding activities of their antibodies before and during immunotherapy.

Methods: B cells binding fluorescently labeled Ara h 1 or Ara h 2 were phenotyped and isolated by means of flow cytometric sorting from 18 patients at baseline and 13 patients during therapy. Fifty-seven mAbs derived from allergen-binding single B cells were evaluated by using ELISA, Western blotting, and peptide epitope mapping. Deep sequencing of the B-cell repertoires identified additional members of the allergen-specific B-cell clones.

Results: Median allergen-binding B-cell frequencies were 0.0097% (Ara h 1) or 0.029% (Ara h 2) of B cells in baseline blood from allergic patients and approximately 3-fold higher during immunotherapy. Five of 57 allergen-specific cells belonged to clones containing IgE-expressing members. Almost all allergen-specific antibodies were mutated, and binding to both conformational and linear allergen epitopes was detected. Increasing somatic mutation of IgG4 members of a clone was seen in immunotherapy, whereas IgE mutation levels in the clone did not increase.

Conclusion: Most peanut allergen-binding B cells isolated by means of antigen-specific flow sorting express mutated and isotype-switched antibodies. Immunotherapy increases their frequency in the blood, and even narrowly defined allergen epitopes are recognized by numerous distinct B-cell clones in a patient. The results also suggest that oral immunotherapy can stimulate somatic mutation of allergen-specific IgG4.

Keywords: B cell; IgE; IgG(4); Peanut; allergen specific; allergy; antibody; antigen specific; high-throughput DNA sequencing; immunotherapy; repertoire; somatic mutation.

Fig 1

Representative flow cytometry plots of Ara h 1 and Ara h 2-specific B cell responses in peanut-allergic individuals undergoing peanut oral immunotherapy. Samples are shown from a healthy donor (Patient 27, left panels), a peanut-allergic individual at baseline (Patient 11, middle panels) and a peanut-allergic individual after 9 months of OIT (Patient 5, right panels). (A) Ara h 1-specific (top panels) and Ara h 2-specific (bottom panels) B cells (CD19+/CD3-/CD14-/CD16-/CD235a-) were double positive for labeled allergen. Boxplots show frequencies of Ara h 1-specific (top panels) and Ara h 2-specific (bottom panels) total B cells (B), class-switched memory B cells (C), non-class-switched memory B cells (D) and plasmablasts (E) for healthy (n=9), baseline allergic (n=18), and OIT (n=18) samples analyzed by flow cytometry. Frequencies of allergen binding B cells within each subset are plotted as the fraction of total B cells within each subset. P-values were calculated by Wilcoxon tests (*p<0.05, **p<0.01, and ***p<0.001).

Fig 2

Venn diagrams summarizing the results of Ara h 1 and Ara h 2 binding assays with monoclonal antibodies generated from allergic patient B cells isolated by allergen-specific FACS. 21 Ara h 1-specific monoclonal antibodies and 36 Ara h 2-specific monoclonal antibodies were tested. Shown are results of ELISA, Western blotting and phage display experiments with Ara h 1 (A) and Ara h 2 (B) proteins, and results of Western blotting experiments with reduced Ara h 1 (C) and Ara h 2 (D) proteins.

Fig 3

Peptide epitope mapping of patient-derived Ara h 1 and Ara h 2-specific mAbs. 22 monoclonal antibodies derived from single cells isolated from five patients undergoing oral immunotherapy bound 24-mer peptides from Ara h 1 and/or Ara h 2 expressed in a phage-display library. Shown are averaged data from at least two independent experiments, and the plotted fold enrichment is relative to a mock control containing no antibody (row 1). Antibodies from each person are presented together. 11 of these 22 antibodies from three patients bound the same Ara h 2 region (peptides 3 to 5). Positive control anti-Ara h 1 and anti-Ara h 2 antibodies were polyclonal rabbit antibody preparations.

Fig 4

V, D and J gene usage, and isotype frequencies of allergen-specific antibody heavy chains. Data were derived from 21 Ara h 1-specific antibodies (top panels) and 36 Ara h 2-specific antibodies (bottom panels) generated from 6 allergic individuals. V, D and J gene frequencies (A–C) and V and D gene family frequencies (D–E) are graphed, as well as isotype frequencies (F) and IGHV somatic mutation levels for each isotype (G).

Fig 5

Identification of peanut-specific lineages in deeply sequenced patient antibody heavy chain repertoires. (A) Representative heavy chain sequences from a lineage clonally related to AbID 70 were obtained from Patient 7 at 9, 21, 25, 34, 36 and 42 months during OIT. The sequence derived from the single cell AbID70 is also shown. The germline sequences of the gene segments used in this antibody gene rearrangement are shown for reference. Darker shading indicates a higher percent identity. CDRs are underlined. Average distance trees were computed by BLOSUM62 on sequences trimmed to the length of the shortest read. (B) Boxplots show IGHV somatic mutation levels for members of IgE-containing lineages clonally related to five peanut-specific monoclonal antibodies. (C) IGHV mutation levels for IgG4 (left panel) and IgE (right panel) reads in the AbID 70 lineage are plotted over time. Arrows indicate the time points sampled. Points are plotted with horizontal jitter to prevent superposition. The adjusted R-squared values are indicated in each panel.