Platelet endothelial aggregation receptor 1 (PEAR1), a novel epidermal growth factor repeat-containing transmembrane receptor, participates in platelet contact-induced activation.

The present study was designed to identify novel membrane proteins that signal during platelet aggregation. Because one putative mechanism for signaling by a membrane protein involves phosphorylation, we used oligonucleotide-based microarray analyses and mass spectrometric proteomics techniques to specifically discover membrane proteins and also identify those proteins that become phosphorylated on tyrosine, threonine, or serine residues upon platelet aggregation. Surprisingly, both techniques converged to identify a novel membrane protein we have termed PEAR1 (platelet endothelial aggregation receptor 1). Sequence analysis of PEAR1 predicts a type-1 membrane protein, 15 extracellular epidermal growth factor-like repeats, and multiple cytoplasmic tyrosines. Analysis of the tissue distribution of PEAR1 showed that it was most highly expressed in platelets and endothelial cells. Upon platelet aggregation induced by physiological agonists, PEAR1 became phosphorylated on tyrosine (Tyr-925), and serine (Ser-953 and Ser-1029) residues. PEAR1 tyrosine phosphorylation was blocked by eptifibatide, an alpha(IIb)beta(3) antagonist, which inhibits platelet aggregation. Immune clustering of PEAR1 resulted in PEAR1 phosphorylation. Aggregation-induced PEAR1 tyrosine phosphorylation lead to the subsequent association with the ShcB adaptor protein. Platelet proximity induced by centrifugation also induced PEAR1 tyrosine phosphorylation, a reaction not inhibited by eptifibatide. These data suggest that PEAR1 is a novel platelet receptor that signals secondary to alpha(IIb)beta(3)-mediated platelet-platelet contacts.

Platelet aggregation during arterial thrombosis results in ischemic complications precipitating in acute myocardial infarction and stroke. Platelet aggregation is known to be mediated by signaling events initiated by primary platelet agonists such as thrombin, ADP, and collagen, which induce a conformational change in the platelet integrin ␣ IIb ␤ 3 , allowing it to bind soluble fibrinogen and von Willebrand factor, resulting in platelet cross-linking. Platelet-platelet contacts during aggre-gation subsequently initiate secondary signaling events. Aggregation-induced signaling can result in multiple platelet secondary signaling events such as calcium mobilization, protein tyrosine phosphorylations, cytoskeletal rearrangements, and the release of platelet-dense bodies and ␣-granules. Aggregation-induced signaling is key to the formation of stable aggregates, particularly when aggregation is induced by low concentrations of one or more primary agonists. Platelet activation also causes the release of ADP from dense bodies and the generation of thromboxane A2, both of which induce further platelet stimulation.
Several mediators of aggregation-induced signals have been identified. One is ␣ IIb ␤ 3 itself, which becomes tyrosine-phosphorylated and also associates with numerous signaling and cytoskeletal proteins following platelet activation, allowing fibrinogen and/or von Willebrand factor binding and platelet aggregation. The importance of ␣ IIb ␤ 3 "outside-in" signaling in the enhancement of platelet aggregation was demonstrated by the generation of knock-in mice where tyrosine residues Tyr-747 and Tyr-759 were mutated to phenylalanine (1). These so-called DiYF mice displayed selective impairment of outside-in signaling, resulting in the formation of unstable aggregates. Other mediators are released from the activated platelets. One is the soluble CD40 ligand, a hydrolytic product produced by metalloprotease cleavage of the CD40 ligand on activated platelets that subsequently binds to ␣ IIb ␤ 3 . Another is GAS6, a protein released from ␣-granules that is involved in the stabilization of platelet-rich thrombi (2,3).
Although the importance of each of the platelet secondary signaling reactions described above is well documented, these reactions are dependent upon platelet activation. It has also been well documented, however, that platelet stimulation can be induced by platelet-platelet contact. Signaling of platelet receptors induced by platelet-platelet proximity, independent of platelet aggregation, have not been described. One exception may be the Eph kinases and ephrins, specifically EphA4 and ephrinB1, which, through receptor-ligand interactions on the platelet surface, enhance the binding of ␣ IIb ␤ 3 to immobilized fibrinogen in the presence of physiological agonists (4,5); however, the importance of this mechanism in platelet-platelet signaling on unstimulated platelets is unknown.
The present study was designed to identify novel platelet proteins involved in platelet proximity-induced activation. Reasoning that many signaling receptors become tyrosine-phosphorylated during signaling, we sought not only to identify novel receptors on platelets but to specifically identify those that become phosphorylated upon platelet-platelet interac-tions, independent of the activation state of the platelet. Initially, we used a bioinformatics approach and profiled platelet RNA on oligonucleotide-based microarrays to identify novel membrane proteins. We also used LC/MS/MS 1 to identify platelet proteins that become phosphorylated upon platelet aggregation. Surprisingly, both techniques converged on a novel 1034-amino acid protein we termed PEAR1 (platelet-endothelial aggregation receptor 1). Bioinformatic analyses revealed that PEAR1 is a type 1 membrane protein containing fifteen EGF-like repeats in its extracellular domain. The intracellular domain contains five proline rich domains, which may interact with Src homology 3 domain-containing proteins, as well as four potential tyrosine phosphorylation sites (Tyr-804, Tyr-925, Tyr-943, Tyr-979). The LC/MS/MS data demonstrated that PEAR1 is, in fact, tyrosine-phosphorylated at Tyr-925. We further demonstrated that PEAR1 becomes tyrosine phosphorylated in response to receptor clustering and during platelet aggregation. Finally, we showed that the PEAR1 signaling is induced through platelet-platelet contacts independent of platelet activation. These data are the first demonstration of a platelet receptor to signal through platelet-platelet contacts independent of platelet activation.
Purification of Platelet RNA-Platelets were collected from a healthy volunteer by apheresis, including a filtration step that removed lymphocyte contaminants following an Institutional Review Board-approved protocol. This procedure yielded ϳ250 ml of platelets (ϳ1 ϫ 10 9 platelets/ml). Platelets were diluted 1:4 with CGS (13 mM sodium citrate, 30 mM glucose, and 0.12 M sodium chloride) containing 0.1 g/ml prostacyclin (PGI 2 ). Platelets were pelleted by spinning for 20 min at 2200 rpm in a Beckman Coulter Allegra 6 centrifuge at room temperature. The supernatant was removed, and the platelet pellet was resuspended in RNA-STAT-60 (Tel-Test) using 1 ml of RNA-STAT-60 per 2.9 ϫ 10 8 platelets and immediately vortexed to facilitate platelet lysis. The platelet lysate was further processed according to RNA-STAT-60 protocol specifications. Total RNA was then treated with RNA Easy (Qiagen) for further purification.
Identification of PEAR1 cDNA and Full-length Cloning-5 g of DNase-treated platelet RNA was converted to cRNA, fragmented, and hybridized to the Millennium Pharmaceuticals cardiovascular custom 60-mer oligonucleotide array per the manufacturer's suggested protocol (Affymetrix, Santa Clara, CA). The probe sets contained on the custom array represented 25,000 genes that were selected based on publicly available sequence information and in-house sequencing efforts of endothelial and platelet cDNA libraries. Data were analyzed with MASv5 (Affymetrix) and Rosetta Resolver (Rosetta Biosoftware). cRNA hybridized to probe sets with intensity p values of Յ0.01 were identified. Platelet cRNA hybridized to a probe set for FLJ00193, and through bioinformatic analyses it was determined that the FLJ00193 cDNA encoded an incomplete open reading frame that contained two EGF repeats and a potential transmembrane domain. A proprietary in-house DNA sequence and a publicly available DNA sequence were used to assemble a 3.6-kb predicted cDNA. Platelet cDNA was synthesized by using Superscript II reverse transcriptase. PEAR1 cDNA was amplified using Platinum Pfx DNA polymerase and the primers 5Ј-GCAGGCT-TCATATCCTGAACGCTG-3Ј (forward) and 5Ј-GCTCTAGATTAACG-GTCCTGGCGTCGAAGTGGAGGTGATG-3Ј (reverse). The resulting 3.2-kb cDNA was cloned into pCR-BluntII-Topo vector.
In Situ Localization of PEAR1 mRNA Using Biotin-labeled Oligonucleotide Probes Detected by ABC Peroxidase-Formalin-fixed, paraffinembedded human and rodent tissues were used in this study. Digoxigenin-labeled riboprobes were used to localize PEAR1 mRNA. The details of the non-isotopic in situ hybridization and the tyramide-mediated signal amplification, as well as the methods used for acquiring digital images, have been described earlier (6,7). Following the manufacturer's protocol (Roche Applied Science), digoxigenin-labeled antisense and sense riboprobes were generated from a DNA template containing nucleotides 2369 -2896.
TaqMan Analysis-TaqMan experiments were performed using an ABI PRISM 7700 system (Applied Biosystems). Primers were designed using Primer Express software. 50 ng of RNA from a variety of human tissues and primary cell lines were used for analysis. PEAR1 and ␤-macroglobulin probes were labeled with different fluorescent dyes as per the manufacturer's protocol. ␤-Macroglobulin was used as an endogenous control to allow for normalization of the amount of RNA added to each reaction. The differential labeling of the target gene (PEAR1) and the reference gene (␤-macroglobulin) occurred in the same well.
Immunoprecipitations and Western Blotting from COS-7 Cells-Transfected cells were lysed with 0.4 ml of lysis buffer (40 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EGTA, 5 mM MgCl 2, 1% Triton X-100, 1 mM sodium orthovanadate, 10 mM NaF, 1 mM Pefabloc, 10 g/ml leupeptin, and 10 g/ml aprotinin). For immunoprecipitations, lysates were incubated with 2 g of the indicated antibodies. For Western analysis, polyvinylidene difluoride membranes were incubated with the indicated antibodies (1 g/ml) for 2 h at 4°C. Membranes were then incubated with 1:10,000 dilution of horseradish peroxidase-conjugated goat anti-mouse or goat anti-rabbit IgG for 1 h at room temperature and developed with the ECL™ chemiluminescent detection kit. Methodologies used for SDS-PAGE, Western blotting, and immunoprecipitations have been described previously (8).
Phosphopeptide Identification by In-gel Digestion and Mass Spectrometry-Washed human platelets were aggregated with 5 M human 1 The abbreviations used are: LC/MS/MS, liquid chromatography tandem mass spectrometry; aa, amino acid(s); ECD, extracellular domain; EGF, epidermal growth factor; FITC, fluorescein isothiocyanate; HUVEC, human umbilical vein endothelial cell; HRP, horseradish peroxidase; ICD, intracellular domain; PEAR1, platelet endothelial aggregation receptor 1; hPEAR1, human PEAR1; mPEAR1, mouse PEAR1; PTB, phosphotyrosine binding; SREC, scavenger receptor expressed by endothelial cells; TRAP, thrombin receptor activating peptide. TRAP as described below and solubilized in 2ϫ lysis buffer (20 mM Tris-HCl, pH 8, 2% Triton X-100, 4 mM EDTA, 250 mM NaCl, 20% glycerol, 1 mM phenylmethylsulfonyl fluoride, 2 mM Na 2 VO 4 , 2 g/ml leupeptin, and 2 g/ml aprotinin). Solubilized platelet-aggregated proteins from 6 ϫ 10 9 platelets were incubated with ␣-phosphotyrosine IgG beads (PY99; Santa Cruz Biotechnology) overnight at 4°C. The phosphoproteins were eluted with 100 mM phenyl phosphate, resolved by SDS-PAGE, and visualized by Coomassie staining. All Coomassiestained bands were excised and subjected to in-gel proteolytic digestion with trypsin and analyzed by LC/MS/MS (9). Data-dependent LC/ MS/MS was performed using electrospray ionization on an LCQ Deca XP ion trap mass spectrometer (Thermo Finnigan). An aliquot of each digest mixture was introduced to the mass spectrometer by reversedphase chromatographic separation with a 75-m inner diameter capillary column flowing at a rate of ϳ350 nl/min and eluted using a 30-min acetonitrile, 0.1% formic acid gradient. Chromatographic separation yielded ϳ30-s peak widths, and mass spectra were acquired in 9-s cycles. Each cycle was of the form consisting of one full mass spectrometry scan followed by four MS/MS scans on the most abundant precursor ions subjected to dynamic exclusion for a period of 1.5 min. The identity of each peptide sequenced was determined by interpreting the MS/MS spectra using the SpectrumMill software (Agilent Technologies).
Immunoprecipitiation and Cross-linking of Surface PEAR-1-Immunoprecipitations were performed with Protein G beads and 2.5 g/ml ␣-ICD PEAR1 IgG or isotype control overnight at 4°C. Immunoblotting of the complexes were performed with ␣-pTyr-4G10 (Upstate Cell Signaling Solutions) and ␣-pTyr-PY20 (Santa Cruz Biotechnology) mouse IgGs or ␣-ECD PEAR1 IgG rabbit IgG. For cross-linking experiments, 3 ϫ 10 8 platelets/ml were incubated with 5 g/ml ␣-ECD PEAR1 IgG in the presence of 25 g/ml IV.3 monoclonal antibody for 30 min at room temperature followed by cross-linking the bound IgG with ␣-rabbit secondary antibody for 10 min at 37°C. Post cross-linking, the reactions were stopped by the addition of ice-cold 2ϫ lysis buffer. Immunoprecipitation of PEAR1 was performed as described above.

RESULTS
Identification of PEAR1 through Gene Profiling-In an effort to identify novel receptors in human platelets, high throughput molecular profiling of RNA isolated from double-phoresed platelets was performed on a custom array that contained probe sets representing 25,000 genes. The sequences for the probe sets represented on the custom arrays were derived both from sequencing of platelet and HUVEC cDNA libraries as well as all the non-overlapping human genes present in the public gene data base. The protein coding sequences of the platelet cRNAs, which hybridized to the microarrays, were then classified based on PROSITE, Pfam, and SMART (11) sequencemotif searches. The selected probe sets were assigned struc-tural annotations and further subdivided into classes of interest. The following search terms were used to classify the protein sequences of the hybridized cRNAs: predicted signal sequences, potential transmembrane domains, GPCRs (G protein-coupled receptors, kinases, phosphatases, cadherins, EGF domains, fibronectins, Ig domains, SCR (short consensus repeat), leucine rich repeats, CUB (complement Clr/Cls, Uegf, and bone morphogenic protein-1), integrins, chemokines, thrombospondins, and proteases. A search for novel membrane proteins identified a 3282-bp cDNA that contained an open reading frame encoding a novel type 1 membrane protein of 1037 aa (Fig. 1). We termed the protein the platelet endothelial aggregation receptor 1 or PEAR1, based on its primary location on platelets and endothelial cells and its role in aggregationinduced signaling (see below). Homology searches revealed significant overall identity in the extracellular and intracellular domains of human PEAR1, mPEAR1, KIAA1780, and KIAA1781. PEAR1 also displays significant homology to the ECDs of CED-1 (12) and SREC (scavenger receptor expressed by endothelial cells) (13). A bioinformatics analysis of the PEAR1 ECD revealed the presence of 15 EGF-like repeats that vary in length from 39 to 42 aa and contain a consensus sequence of CX 1-2 GX 2 GX 2-4 CX 3 CX 1-3 CX 1-2 GX 1-2 CX 4 GX 1 CX 1 -CX 2 GX 2 GX 2 C. An alignment of the PEAR1 EGF-like repeats is presented in Fig. 2A. The chromosomal localizations and the degree of identities between hPEAR1 and mPEAR1, KIAA1780, KIAA1781, CED-1, and SREC-I are summarized in Table I. The domain structures present in PEAR1 and related proteins are shown in Fig. 2B. Putative signaling motifs present within the ICD of PEAR1 were identified by the Scansite program (14). Five potential Src homology 3-binding, prolinerich domains (Fig. 3) were identified in hPEAR1 and mPEAR1 but not in KIAA1780 or KIAA1781. Human PEAR1, mPEAR1, and KIAA1780 contain an NPXY motif (Fig. 3) that is known to act as interaction sites for PTB domain-containing proteins (15). The Scansite program also predicted that the tyrosine residues 804, 925, 943, and 979 present in hPEAR1 may be tyrosine-phosphorylated (Fig. 3). Similar tyrosine residues are also present in mPEAR1, KIAA1780, and KIAA1781. Based on overall homology, it is likely that PEAR1, KIAA1780, and KIAA1781 represent a novel family of EGF repeat-containing proteins. Because of the EGF repeats and their homology to SREC and CED-1, possible functions for this novel family of proteins include cell adhesion (16), the uptake of acetylated low density lipoprotein (13), and phagocytosis of apoptotic bodies (12).
Identification of hPEAR1 through Proteomics-We also employed a proteomics approach to identify novel membrane receptors and specifically targeted proteins that become tyrosinephosphorylated during aggregation. Platelet aggregates were detergent-solubilized, and phosphotyrosine proteins were captured on an ␣-phosphotyrosine affinity column (PY99). The bound proteins were eluted with phenylphosphate and subjected to SDS-PAGE. Eight bands were detected post Coomassie staining (Fig. 4A) PEAR1 and related proteins (B). A, the predicted EGF repeats present in PEAR1 were aligned based on a ClustalW analysis. 15 EGF-like domains contain a consensus sequence of CX 1-2 GX 2 GX 2-4 CX 3 CX 1-3 CX 1-2 GX 1-2 CX 4 GX 1 CX 1 CX 2 GX 2 GX 2 C and vary in length from 39 to 42 aa. The consensus EGF repeat contains six conserved glycine residues and eight conserved cysteine residues, suggesting four disulfide-bonded cysteine pairs in each EGF repeat. The exception is EGF repeat number 4 (EGF#4), which contains six cysteines and, therefore, three potential pairs of disulfide-bonded cysteines. The conserved cysteines and glycine residues are listed underneath the aligned EGF repeats. B, hPEAR1, mPEAR1, KIAA1780, KIAA1781, CED-1, and SREC are predicted to be type 1 membrane proteins. The single predicted transmembrane domains are identified as downward slanting striped boxes. The EGF repeats, which were identified by the Pfam search program (11), are presented as chains of solid black boxes. The intracellular domains of human and mPEAR1, KIAA1780, and KIAA1781 are shown as upward slanting striped boxes to indicate their sequence similarity. The intracellular domains of SREC and CED-1 do not share similarity with PEAR-1 and are represented as white (CED-1) and gray (SREC) boxes. responded to that of human PEAR1. The locations of the seven peptides are shown in Fig. 1. Most importantly, the mass spectra of three of these seven peptides were phosphorylated at Tyr-925, Ser-953, and Ser-1029 (Fig. 4B). The locations of the phosphorylated residues are shown in Figs. 1 and 3.
Tissue Distribution-To investigate the expression pattern of PEAR1, real time quantitative PCR (TaqMan) was performed on a variety of human tissues and primary cell lines (Fig. 5A). PEAR1 is most highly expressed in HUVECs, followed by megakaryocytes, osteoblasts, coronary smooth muscle cells, and erythroid cells. We also found low or no expression in peripheral blood leukocytes or macrophages. In normal human tissues, TaqMan analysis revealed expression in heart, kidney, skeletal muscle, pancreas, ovary, breast, lung, brain cortex, hypothalamus, spinal cord, and dorsal root ganglion. No expression was detected in liver, small intestine, and colon. A 5-kb transcript was detected by probing a Northern blot containing poly(A ϩ )-enriched mRNA from numerous human tissues; an expression pattern similar to that observed by Taq-Man analysis was obtained (data not shown).
In situ hybridizations revealed that PEAR1 is expressed in megakaryocytes present in rat adult femur marrow sections (Fig. 5B, section A), as well as in platelet aggregates present in human peripheral blood (Fig. 5B, section B). PEAR1 expression was also detected in endothelial cells of human umbilical cord artery and vein tissue sections (Fig. 5B, sections C and D) and also in microcapillaries and larger vessels of 16-day-old mouse embryo lung and heart tissues (Fig. 5B, sections E and F).
Western blotting with the ␣-ICD PEAR1 IgG detected a protein of ϳ150 kDa in COS-7 cells transfected with the fulllength hPEAR1 cDNA. The ␣-ICD IgG also recognized a 150-kDa protein in HUVEC and platelet lysates (Fig. 5C, lanes 2 and 5) along with an additional weaker band, migrating at ϳ180 kDa, cross-reacting in platelet lysates (lane 5). Only the 150-kDa band was immunoprecipitated from both HUVECs and platelets (Fig. 5C, lanes 4 and 7). The ␣-ICD PEAR1 IgG was also used to probe the expression of PEAR1 in sections of human umbilical cord, revealing expression in endothelial cells and aggregated platelets within a blood clot (data not shown).
Platelet Distribution of PEAR1-The expression of PEAR1 on the surface of control and stimulated platelets was detected by staining unpermeabilized, paraformaldehyde-fixed resting and TRAP-activated platelets with an ␣-rabbit IgG raised against the extracellular domain of PEAR1, ␣-ECD PEAR1 IgG  3. Alignment of the intracellular domains of human PEAR1, mPEAR1, KIAA1780, and KIAA1781 and identification of potential tyrosine phosphorylation sites and interaction motifs. The intracellular domains were aligned based on a ClustalW analysis. Amino acid identities are highlighted as boxed regions. The Scansite program (14) was used to identify potential tyrosine phosphorylation and protein-protein interaction motifs. Predicted tyrosine phosphorylation sites are marked by black stars. NPXY motifs are boxed by dashed lines, and proline-rich domains present in hPEAR1 are outlined by thick black boxes. Experimentally identified phosphorylation sites are marked by gray stars. The locations of the seven sequenced peptides are indicated by the dashed lines above the hPEAR1 sequence. (Fig. 6A). Fluorescence-activated cell sorter analyses indicated PEAR1 expression on the surface in resting platelets and, interestingly, there was no increase in ␣-ECD PEAR1 IgG binding upon agonist stimulation (Fig. 6A, top section). Platelet activation was confirmed by the increase in surface expression of P-selectin post TRAP activation (Fig. 6A, bottom section).
PEAR1 Tyrosine Phosphorylation-To determine the condition required for PEAR1 tyrosine phosphorylation, ␣-ECD PEAR1 immunocomplexes from resting or aggregated platelets were probed with ␣-phosphotyrosine-specific IgGs (Fig. 6B, top section). Phosphorylated PEAR1 was not detected in ␣-ECD PEAR1 immunocomplexes from resting platelets; however, upon platelet aggregation stimulated by collagen, TRAP, and thrombin, robust tyrosine-phosphorylation was observed. This phosphorylation was markedly reduced in the presence of the platelet ␣ IIb ␤ 3specific antagonist eptifibatide, which blocks platelet aggregation but not platelet activation (17). The data show that PEAR1 tyrosine phosphorylation is an aggregation-induced event.
We next examined the effect of immune clustering on PEAR1 tyrosine phosphorylation. Ligation of platelet-bound ␣-ECD PEAR1 IgG, but not nonspecific IgG crosslinked by a secondary IgG, resulted in tyrosine phosphorylation of PEAR1 (Fig. 6C,  top section). ⌻he ␣-ECD PEAR1 IgG clustering-induced phosphorylation was not inhibited in the presence of the FCR␥RIIa antibody IV.3, indicating that signaling was induced through PEAR1 and not through the platelet Fc receptor.

Non-receptor Tyrosine Kinase-induced Tyrosine Phosphorylation of PEAR1 and PEAR1 Association with the Shc Adaptor
Protein-Emerging data suggest that many signaling proteins in platelets are tyrosine-phosphorylated by Src kinases. PEAR1 contains four predicted tyrosine phosphorylation sites and an NPXY motif that may serve as an interaction site for PTB domain-containing proteins such as Shc (15). To determine whether Src kinases can phosphorylate PEAR1, Myctagged PEAR1 was co-expressed with v-Src in COS-7 cells. Panel 1 of Fig. 7A demonstrates that equal amounts of Myc-PEAR1 were immunoprecipitated when expressed in the absence or presence of v-Src. Immunoprecipitated Myc-PEAR1 cross-reacted weakly with the ␣-PY-IgG when expressed in the absence of v-Src (Fig. 7A, Panel 2, lane 2). However, upon co-expression with v-Src, Myc-PEAR1 was markedly tyrosinephosphorylated (Fig. 7A, Panel 2, lane 4). Because proteomics of the aggregated platelets showed that PEAR1 was tyrosinephosphorylated at position 925, we next assayed to determine whether phosphorylated PEAR1 bound PTB domain-containing signaling proteins. As shown in Fig. 7A, Panel 3, lane marked Lysate, COS-7 cells expressed the ShcA, ShcB, and ShcC isoforms. Immunoblotting Myc-PEAR1 immunoprecipitates demonstrated that ShcA and B co-immunoprecipitate with Myc-PEAR1 (Fig. 7A, Panel 3, lane 2). The association of ShcA and ShcB with Myc-PEAR1 is enhanced when Myc-PEAR1 is co-expressed with v-Src (Fig. 7A, Panel 3, lane 4). We next determined if Shc associated with tyrosine-phosphorylated PEAR1 in platelets. Equal amounts of PEAR1 were immunoprecipitated from unstimulated and aggregated platelet lysates (Fig. 7B, Panel 1). As shown in Fig. 7B, Panel 2, ShcB only coimmunoprecipitated with PEAR1 upon TRAP-induced platelet aggregation (Fig. 7B, Panel 2, lane 6).
Platelet-Platelet Contacts Induce PEAR1 Tyrosine Phosphorylation-Because the data above indicated that PEAR1 was tyrosine-phosphorylated upon platelet aggregation and that its surface expression was not increased by platelet activation, we investigated whether platelet-platelet contacts were sufficient in the absence of platelet stimulation to induced tyrosine phosphorylation of PEAR1. Platelet-platelet contacts were induced by centrifugation of human washed platelets; PEAR1 tyrosine phosphorylation (Fig. 8, Panel 1) was increased as compared with that occurring in resting platelets. The tyrosine phosphorylation occurred rapidly and was as apparent in samples solubilized immediately after centrifugation as in samples incubated for an additional 5 min (Fig. 8). Unlike the PEAR1 tyrosine phosphorylation observed in aggregation reactions, phosphorylation during centrifugation was not inhibited by the ␣ IIb ␤ 3 antagonist eptifibatide (Fig. 8, Panel 1). DISCUSSION It is well established that the platelet integrin ␣ IIb ␤ 3 is responsible for the primary interaction of platelets during thrombosis and hemostasis and that soluble stimuli (e.g. sCD40L, thromboxane A2, ADP, GAS6, and serotonin) are released from activated platelets to support this process. It is also established that aggregation-induced signaling also occurs. Numerous secondary receptors have been shown to be involved in this process, including outside-in signaling through ␣ IIb ␤ 3 , a reaction dependent in part upon tyrosine phosphorylation of ␤3 (1,18). Signaling reactions induced directly by platelet-platelet contact have not been previously described. Because aggregation-induced signaling is directly involved in aggregate stability, we sought to identify novel platelet receptors involved in platelet proximity-induced signaling.
The present study identified a novel platelet protein, termed PEAR1, that signals upon the formation of platelet-platelet contacts induced both by platelet aggregations or by platelet centrifugation. PEAR1 was identified using two independent techniques. First, by profiling platelet RNA on oligonucleotidebased microarrays, we identified a transcript encoding PEAR1. Second, experiments performed in parallel using a proteomics approach designed to identify platelet proteins phosphorylated upon platelet aggregation also led to the identification of PEAR1. In platelets, PEAR1 was shown to be a surface-expressed protein that, upon aggregation, was tyrosine-phosphorylated. This phosphorylation event was inhibited by the ␣ IIb ␤ 3 antagonist eptifibatide and, thus, demonstrated that PEAR1 tyrosine phosphorylation is dependent on surface contacts between activated platelets. Interestingly, we also demonstrated that, unlike other secondary signaling molecules that require platelet activation to signal, platelet-platelet contacts induced by centrifugation of washed platelets resulted in PEAR1 tyrosine phosphorylation. This phosphorylation was not inhibited by eptifibatide, implying that platelet-platelet contacts independent of platelet activation can induce PEAR1 phosphorylation.
Induction of PEAR1 phosphorylation by clustering through  7) were incubated with either nonspecific rabbit IgG (lanes 3 and 6) or ␣-PEAR1 ICD polyclonal IgG and precipitated with protein G-Sepharose (lanes 4 and 7). PEAR1 was detected by Western analysis with the anti-ICD PEAR1 polyclonal IgG and goat anti-rabbit IgG-HRP.
immune complexes also provided evidence that PEAR1 phosphorylation was an oligomerization-dependent event. Oligomerization of integrins and growth factor receptors is a well established mechanism for the promotion cell signaling (19,20). EGF and neuregulin induce oligomerization and signaling of ErbB receptors (21,22). In platelets, the forced clustering of Ephrin and Eph receptors leads to platelet activation and adhesion (4). Data presented in this paper suggest that, upon the formation of platelet-platelet surface contacts, an unidentified surface ligand binds to and induces clustering and phosphorylation of the intracellular tail of PEAR1.
PEAR1 displays extensive homology with KIAA1780 and KIAA1781 in the extracellular and intracellular domains; it is likely these proteins belong to a unique family of EGF repeatcontaining transmembrane proteins. The extracellular domain of PEAR1 is also highly similar to CED-1, SREC-I, and SREC-II. CED-1 is a Caenorhabditis elegans gene required for engulfment of apoptotic bodies (12). SREC-I was originally identified based on its ability to bind to and mediate uptake of anionic ligands such as acetylated low density lipoprotein and oxidized low density lipoprotein (13). SREC-II, which is 52% identical to SREC-I in the extracellular domain, does not mediate uptake of oxidized low density lipoprotein but appears to form transheterotypic interactions with SREC-I to promote cell-cell aggregation (23). By analogy, the activities of CED-1, SREC-I, and SREC-II suggest possible ligands for PEAR1. PEAR1 can be phosphorylated in an ␣ IIb ␤ 3 integrin-dependent manner on tyrosine (Tyr-925) and serine residues (Ser-953 and Ser-1029) and, potentially, at Tyr-804, Tyr-943, and Tyr-979. Src family kinases are known to transmit integrin-dependent signals that lead to platelet adhesion and aggregation (24,25). Co-transfection studies of v-Src with PEAR1 in COS-7 cells results in Src-promoted tyrosine phosphorylation of PEAR1. An NPXY motif, an interaction site for PTB domain-containing proteins, and five potential Src homology 3-binding domains were also identified in PEAR1. The Shc adaptor proteins ShcA, ShcB, and ShcC each contain a PTB domain as well as a phosphotyrosine-binding Src homology 2 domain (26). In platelets, outside-in signaling through integrin ␣ IIb ␤ 3 promotes the association of ShcB with tyrosine-phosphorylated ␤ 3 integrin tails (8,18), providing a mechanism to further promote platelet-platelet aggregation. In PEAR1-transfected COS-7 cells, we were able to detect the association of ShcA and ShcB, but not ShcC, with immunoprecipitated PEAR1; v-Src-promoted tyro-  1, 2, and 3) of 25 g/ml IV.3 IgG. IgG signaling was induced with 10 g/ml ␣-rabbit secondary whole IgG for 10 min at 37°C. Stimulated platelets were lysed and immunoprecipitated with ␣-PEAR1-ICD-IgG followed by Western blotting with ␣-phosphotyrosine antibodies PY20 and 4G10 (Panel A) or ␣-ICD PEAR1 (Panel B) IgG. sine phosphorylation of PEAR1 enhanced the association of ShcA and ShcB with PEAR1. ShcB is the only Shc isoform that is expressed in platelets (8) and, as demonstrated in this paper, associates with PEAR1 upon TRAP-induced platelet aggregation. The association of ShcB with PEAR1 may provide a mech-anism, in addition to its interaction with ␣ IIb ␤ 3 , to localize Shc to the plasma membrane where it may further enhance signaling pathways such as the activation of Ras.
In this study we have described the use of two techniques, genomics and proteomics, to identify a novel single transmem-  1, 4, 5, 6, and 7) in the absence (lanes 1, 2, 3, 4, and 5) or presence (lanes 6 and 7) of 2 M eptifibatide. Resting or spun platelets were lysed 0 or 5 min post spinning and immunoprecipitated (IP) with ␣-PEAR1-ICD-IgG followed by Western blotting with ␣-phosphotyrosine (␣-Ptyr) antibodies PY20 and 4G10 (Panel 1) or ␣-PEAR1-ICD (panel 2) polyclonal IgG. brane receptor expressed in platelets and endothelial cells that is designated PEAR1. This combination of these techniques may be applicable not only to characterizing proteins that signal during platelet aggregation but also proteins involved in other reactions important to platelet biology. Included in this list might be proteins involved in ␣-granule release and dense body secretion, proteins regulating cytoskeletal assembly and rearrangement, and proteins that signal in response to specific platelet agonists. We have also shown for the first time a platelet receptor that is activated directly by platelet-platelet contacts independent of platelet activation and secondary to platelet aggregation. Cell-cell contact-dependent mechanisms are responsible for a multitude of cellular processes. The cadherin/catenin adhesion system becomes activated upon cell-cell contact and has been implicated in contact inhibition of growth. Loss of these molecules has been implicated in tumorigenesis (27). Cell-contact dependent signaling events are also responsible for the immune synapse formation, where primary cellcell interactions are mediated by integrins, but secondary signaling by additional receptors is initiated by cell proximity. Although the function of PEAR1 signaling during plateletplatelet contact remains to be determined, the multitude of signaling events triggered by platelet aggregations is extensive and includes ␣-granule and dense body release, thromboxane A2 generation, cytoskeletal rearrangements, ␣ IIb ␤ 3 activation, and increase in cytosolic Ca 2ϩ concentration. Understanding the relationship of PEAR1 as a secondary adhesion receptor mediated by platelet cohesion in these processes is an important area for future studies.