A Human Platelet Receptor Protein Microarray Identifies the High Affinity Immunoglobulin E Receptor Subunit α (FcεR1α) as an Activating Platelet Endothelium Aggregation Receptor 1 (PEAR1) Ligand

Abstract

Genome-wide association studies to identify loci responsible for platelet function and cardiovascular disease susceptibility have repeatedly identified polymorphisms linked to a gene encoding platelet endothelium aggregation receptor 1 (PEAR1), an "orphan" cell surface receptor that is activated to stabilize platelet aggregates. To investigate how PEAR1 signaling is initiated, we sought to identify its extracellular ligand by creating a protein microarray representing the secretome and receptor repertoire of the human platelet. Using an avid soluble recombinant PEAR1 protein and a systematic screening assay designed to detect extracellular interactions, we identified the high affinity immunoglobulin E (IgE) receptor subunit α (FcεR1α) as a PEAR1 ligand. FcεR1α and PEAR1 directly interacted through their membrane-proximal Ig-like and 13th epidermal growth factor domains with a relatively strong affinity (KD ∼ 30 nm). Precomplexing FcεR1α with IgE potently inhibited the FcεR1α-PEAR1 interaction, and this was relieved by the anti-IgE therapeutic omalizumab. Oligomerized FcεR1α potentiated platelet aggregation and led to PEAR1 phosphorylation, an effect that was also inhibited by IgE. These findings demonstrate how a protein microarray resource can be used to gain important insight into the function of platelet receptors and provide a mechanistic basis for the initiation of PEAR1 signaling in platelet aggregation.

Fig. 1.

Design and expression of a human platelet receptor library for common structural classes of cell surface and secreted proteins for AVEXIS. A, schematics show the design of ectodomain expression constructs. Type II and multispan proteins contained an exogenous signal peptide (Exg.SP), whereas the others retain their endogenous signal peptide (End.SP). Heteromeric complexes such as the integrins shown here were tagged on only one chain to ensure expression of only tagged complexes. *, stop codon. B, an anti-biotin Western blot of the 121 bait proteins organized into their structural categories. The majority of proteins were expressed at the expected size with little processing. Numbering is according to supplemental Table S2; note that the five expressed heteromeric complexes are not included because only one chain would have been detected. GPI, glycosylphosphatidylinositol; Bio, biotin; 6H, His₆; ITG, integrin; TM, transmembrane.

Fig. 2.

A human platelet secretome and receptor protein microarray identifies FcεR1α as a ligand for PEAR1. Soluble recombinant biotinylated proteins representing the secretome and receptor repertoire of the human platelet were purified and arrayed in six 3-fold dilutions on streptavidin-coated slides. A, the array was screened with a control (rat Cd200) pentamerized FLAG-tagged prey that bound the background baits FcγR2α (red box) and P4HB (white box). B, PEAR1 prey additionally interacted with the FcεR1α bait (orange box) when compared with the control in A; fluorescence intensities are quantified in supplemental Fig. S4. C, FcεR1α prey additionally interacted with the PEAR1 bait (blue box) in comparison with the control. Note that the nine regularly spaced markers are orientation markers and that the boxed areas marked on the array enclose the location of the six spots containing the dilutions of the named immobilized baits (see supplemental Fig. S3).

Fig. 3.

FcεR1α and PEAR1 directly and specifically interact with a relatively high affinity. A, purified monomeric FcεR1α-Cd4-His₆ was serially diluted and injected over immobilized PEAR1 until equilibrium was achieved (inset). Binding data that had been reference-subtracted were plotted as a binding curve, and a K_D of 27.4 ± 1.3 nm was calculated. B, association and dissociation rate constants derived from an independent kinetic analysis of the FcεR1α-PEAR1 interaction were consistent with the equilibrium analysis. Seven serial dilutions of purified, soluble FcεR1α-Cd4-His₆ were injected over immobilized PEAR1 (black lines), and kinetic parameters for the interaction derived from a 1:1 binding model were fitted to the family of sensorgrams (red lines). 6H, His₆; RU, response units.

Fig. 4.

The FcεR1α-PEAR1 interaction is mediated by membrane-proximal domains and can be specifically inhibited by IgE. A, a schematic illustrating the domain organization of the PEAR1 receptor in the membrane. Expressed fragments of the PEAR1 protein are represented by orange bars, and the ticks and crosses indicate the ability of the fragments to bind the full-length (FL) ectodomain of FcεR1α as determined by AVEXIS; a similar summary for FcεR1α but tested for binding to PEAR1 is shown in B. C, binding data using AVEXIS showing that the 13th epidermal growth factor (EGF) domain of PEAR1 and the second Ig-like domain of FcεR1α are necessary and sufficient for binding. Bars represent means ± S.E. (n ≥ 3). D, the FcεR1α-PEAR1 interaction detected by AVEXIS using PEAR1 as a plate-immobilized bait was completely inhibited by low (IC₅₀ ∼ 0.5 ng/ml) concentrations of IgE (filled circles) but not control IgG (open circles). A control interaction, rat Cd200-Cd200 receptor (Cd200R) (squares), was not inhibited by either antibody. E, the indicated concentrations of purified full-length (FL) and both smaller fragments of the IgE constant heavy chain (C2–4) and (C2–3) were preincubated with the FcεR1α prey before being added to the PEAR1 bait, and the interaction was detected using AVEXIS. Data points are mean ± S.E. (n = 3). F, the FcεR1α-PEAR1 interaction was detected using AVEXIS, and the inhibition by IgE (IgE alone; filled circles) could be relieved by the addition of 2.5 μg of omalizumab (open circles). Data points are mean ± S.E. (n ≥ 3). Error bars represent S.E. EMI, EMILIN-family domain.

Fig. 5.

Oligomerized FcεR1α promotes platelet aggregation and phosphorylates PEAR1. A, soluble recombinant pentamerized (s5) PEAR1 ectodomains modestly inhibited whereas a similar s5-FcεR1α protein strongly promoted platelet aggregation relative to a control (rat s5-Cd200) when added prior to collagen-induced platelet aggregation. B, precomplexing s5-FcεR1α with IgE completely inhibited s5-FcεR1α-potentiated aggregation of collagen-activated platelets. C, oligomeric FcεR1α, but not a control protein, triggered the phosphorylation of PEAR1 (top panel) in human platelets with similar potency to an anti-PEAR1 antibody (Ab) (lower panel); total PEAR1 protein was detected as a loading control. D, oligomeric FcεR1α induced tyrosine phosphorylation in human platelets as shown by anti-phosphotyrosine (P-Tyr) Western blotting of lysates. Phosphorylation of PEAR1 and AKT but not PLCγ2 (a mediator of FcεR1α signaling) was observed. Total PLCγ2 protein was used as a loading control. Aggregation data are representative from at least 10 independent experiments.