Association of Fyn and Lyn with the proline-rich domain of glycoprotein VI regulates intracellular signaling.

The glycoprotein VI (GPVI)-Fc receptor (FcR) gamma-chain complex, a key activatory receptor for collagen on platelet surface membranes, is constitutively associated with the Src family kinases Fyn and Lyn. Molecular cloning of GPVI has revealed the presence of a proline-rich domain in the sequence of GPVI cytoplasmic tail which has the consensus for interaction with the Src homology 3 (SH3) domains of Fyn and Lyn. A series of in vitro experiments demonstrated the ability of the SH3 domains of both Src kinases to bind the proline-rich domain of GPVI. Furthermore, depletion of the proline-rich domain in GPVI (Pro(-)-GPVI) prevented binding of Fyn and Lyn and markedly reduced phosphorylation of FcR gamma-chain in transiently transfected COS-7 cells, but did not affect the association of the gamma-chain with GPVI. Jurkat cells stably transfected with wild type GPVI show robust increases in tyrosine phosphorylation and intracellular Ca2+ in response to the snake venom convulxin that targets GPVI. Importantly, convulxin is not able to activate cells transfected with Pro(-)-GPVI, even though the association with the immunoreceptor tyrosine-based activation motif-containing chains is maintained. These findings demonstrate that the proline-rich domain of GPVI mediates the association with Fyn/Lyn via their SH3 domain and that this interaction initiates activation signals through GPVI.

The adhesion and activation of platelets by subendothelial collagen fibers initiates aggregate formation at sites of vessel damage. Glycoprotein (GP) 1 VI plays a critical role in the activatory events induced by collagen as shown by the lack of response to collagen in human and mice platelets deficient in the glycoprotein (1,2). A collagen-related peptide and a snake venom toxin, convulxin, interact specifically with GPVI and mimic many of the responses to collagen (3)(4)(5).
Because of the physiological importance of GPVI, the mechanism of the GPVI-mediated signaling system has been extensively investigated (6 -8). GPVI is present as a complex with Fc receptor (FcR) ␥-chain in the platelet membrane (8 -10). The Src family kinases, Fyn and Lyn, are associated with GPVI-FcR ␥-chain complex in platelets and initiate activation through phosphorylation of the immunoreceptor tyrosinebased activation motif (ITAM) in the FcR ␥-chain leading to binding and activation of the tyrosine kinase Syk. A series of adapter molecules including LAT and SLP76 orchestrate a carefully regulated signaling network leading to activation of PLC␥2, phosphoinositol 3-kinase, and small molecular weight G proteins, leading to platelet activation (6,7).
The cloning of GPVI (11)(12)(13)(14) has revealed it to be a member of the immunoglobulin (Ig) superfamily, showing close homology to Fc␣RI. GPVI has a charged arginine residue in its transmembrane domain. This arginine, together with elements within the cytoplasmic domain, is crucial for association of GPVI with FcR ␥-chain and GPVI-mediated signal transduction (15,16). In addition, the cytoplasmic tail of GPVI has a cluster of 6 proline residues of unknown function (11)(12)(13)(14). This sequence of GPVI, RPLPPLPPLP, contains a consensus Src family kinase-SH3 recognition motif (RPLPPLP) (17,18), and provides a potential site of interaction with Fyn and Lyn via their SH3 domains.
In this study, we demonstrate that depletion of the prolinerich domain in GPVI abolishes the association with Fyn and Lyn and prevents tyrosine phosphorylation of FcR ␥-chain and downstream responses. From these findings, we suggest that Fyn/Lyn directly bind the proline-rich domain of GPVI and that this association is necessary for phosphorylation of the ITAM and downstream signals. described previously (19). Convulxin and anti-convulxin antibody were generous gifts from Drs. Mireille Leduc and Cassian Bon. Anti-phosphotyrosine monoclonal antibody 4G10, anti-FcR ␥-chain polyclonal Ab were purchased from Upstate Biotechnology, Inc. (TCS Biological Ltd., Botolph Claydon, UK). Anti-Fyn polyclonal antibody, anti-Lyn polyclonal Ab, normal mouse IgG, and anti--chain monoclonal antibody were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Dulbecco's modified Eagle's medium and trypsin-EDTA were from Invitrogen-RBL Life Technologies Ltd. (Paisley, UK). Other reagents were from previously described sources (16,20).
Preparation of MBP and GST Fusion Proteins-A maltose-binding protein (MBP)-GPVI fusion protein was prepared from cDNA encoding the GPVI cytoplasmic sequence Glu 266 -Ser 316 (11) subcloned into a pmalC2 vector (New England Biolabs, Beverly, MA) at EcoRI and XbaI sites. The construct encoded a fusion protein with the N-terminal region corresponding to Escherichia coli MBP and the GPVI sequence at the C terminus. The correct sequence was verified by sequencing. MBP and MBP-GPVI expressed in E. coli were purified on amylose-Sepharose according to the manufacturer's instructions (New England Biolabs). GST-SH3 domains of Fyn, Lyn, Src, Fgr, Btk, or PLC␥2 expressed in E. coli were purified on glutathione-agarose as described previously (19). Where appropriate, GST-Lyn-SH3 and GST-Src-SH3 were radioiodinated with sodium [ 125 I]iodide using the chloramine-T method, and separated from free label on Sephadex G-25 (Amersham Biosciences) washed with TS buffer (0.01 M Tris-HCl, 0.15 M NaCl, pH 7.4) as previously described (21,22).
Binding of GST-SH3 Domain Fusion Proteins to MBP-GPVI Fusion Protein or GPVI-related Peptide-Binding of GST-Src-SH3 or GST-Lyn-SH3 to MBP-GPVI cytoplamic domain fusion proteins or a synthetic peptide based on GPVI was assessed using a microtiter well assay as described elsewhere (22,23). Briefly, detachable microtiter wells (Immunlon-2 Remova-wells, Dynatech, Chantilly, VA) were coated with MBP or MBP-GPVI (5 g) in 50 l of TS buffer by incubation at 22°C overnight. A synthetic peptide corresponding to the GPVI cytoplasmic sequence His 269 -Ser 316 (Chiron Mimotopes, Clayton, Australia) was solubilized in distilled water at 0.5 mM, and wells coated overnight at 22°C with 50 l of a 1:10 dilution in 0.05 M NaHCO 3 , pH 9.2. MBP, MBP-GPVI, or His 269 -Ser 316 peptide-coated wells were incubated at 22°C with 5% (w/v) BSA in TS buffer for 2 h, then washed 4 times with 0.1 ml of TS buffer. To each well was added 200 l of 125 I-labeled GST-Lyn-SH3 or GST-Src-SH3 (final concentration, 1 g/ml) in TS buffer containing 1 mM CaCl 2 and 0.1% (w/v) BSA. Parallel assays included a 50-fold excess of unlabeled GST-Lyn-SH3 or GST-Src-SH3, or GPVI synthetic peptide Val 282 -Ser 298 (5-100 M). After 30 min at 22°C, the supernatant was aspirated, and bound radioactivity counted in a ␥-counter.
Preparation and Stimulation of Platelets-Platelets were obtained from drug-free volunteers on the day of the experiment and suspended in Tyrodes-Hepes buffer (134 mM NaCl, 0.34 mM Na 2 HPO 4 , 2.9 mM KCl, 12 mM NaHCO 3 , 20 mM Hepes, 5 mM glucose, and 1 mM MgCl 2 , pH 7.3). They were isolated and stimulated as described previously (20).
Ligand Blotting and Immunoprecipitation-For GPVI detection, membranes were incubated with 10 g/ml convulxin for 1 h at room temperature and then incubated with anti-convulxin antibody as described previously (24). GPVI was precipitated using its ligand, convulxin. Cell lysates were incubated with 10 g/ml convulxin for 2 h. 1:10,000 anti-convulxin antibody and 25 l of protein A-Sepharose were then added.
Cell Culture-COS-7 cells was grown in Dulbecco's modified Eagle's medium. Jurkat cells were grown in RPMI 1640 medium. Both medi-ums contain 100 units/ml penicillin, 100 g/ml streptomycin, and 10% heat-inactivated fetal bovine serum. Before the experiment, cells were washed with phosphate-buffered saline and resuspended in Tyrodes-Hepes buffer. Cells were detached by adding trypsin-EDTA and incubating for 5 min at 37°C before washing.
Constructs-The cDNA for human GPVI was subcloned into HindIII/ XbaI sites of pRc plasmid (FLAG-tagged). The construct encoded a fusion protein with the N-terminal region corresponding to human GPVI and FLAG sequence at the C terminus. The cDNA for the human FcR ␥-chain was subcloned into pMG (InvivoGen, San Diego, CA) and the cDNAs for Lyn or Fyn were subcloned into pcDNA3.1 (Invitrogen). The constructs for Lyn or Fyn encode a fusion protein with the Nterminal region corresponding to Lyn or Fyn and the C-terminal region corresponding to Myc. All sequences were verified by sequencing.
Transient and Stable Transfections-For stable transfections, 1 ϫ 10 7 Jurkat cells were washed once with serum-free medium and once with cytomix buffer (0.119 M KCl, 91 M CaCl 2 , 9.76 mM K 2 HPO 4 , 10.2 mM KH 2 PO 4 , 25 mM HEPES, 2.1 mM EGTA, 5 mM MgCl 2 , pH 7.6) which was supplemented on the day of experiment with 0.37 mg/ml glutathione. Cells were resuspended in 400-l cytomics and placed into an electroporation cuvette already containing 40 g of DNA. Cells were electroporated, added to 15 ml of completed medium, and incubated for 48 h before selection by 500 g/ml neomycin (G418). Surface expression of normal GPVI or proline-rich domain-deleted GPVI was confirmed by flow cytometry using convulxin and anti-convulxin antibody as described previously (24). For transient transfections, 5 g of each DNA (pRc alone or pRc containing GPVI, proline-rich domain-deleted GPVI, Fyn, Lyn, or FcR ␥-chain) was added to the buffer containing 252 mM CaCl 2 and 240 M chloroquine. Then, 500 l of 2 ϫ HBS buffer (280 mM NaCl, 10 mM KCl, 1.5 mM Na 2 HPO 4 , 50 mM HEPES, 12 mM dextrose, pH 7.5) was added to the DNA mixture drop by drop. Fifteen hours after adding this transfection mixture to COS-7 cells, the medium containing DNA was removed and completed medium was added. After 24 h, transfected cells were used for experimentation.
Flow Cytometry Studies-Jurkat cells were resuspended in Tyrodes-Hepes buffer containing 1 mg/ml bovine serum albumin. All incubation times were performed for 30 min unless otherwise indicated. For GPVI detection, Jurkat cells were incubated with 10 g/ml convulxin, washed, and incubated with 0.4 g/ml anticonvulxin antibody, washed again, and finally incubated with fluorescein isothiocyanate-conjugated anti-rabbit IgG secondary antibody diluted 1:500. Stained cells were analyzed immediately using a FACScalibur (Becton Dickinson). Data were recorded and analyzed using CellQuest software.

Src Kinases Bind to the Proline-rich Domain of GPVI in
Vitro-We used a MBP cytoplasmic tail of GPVI fusion protein to investigate whether Src kinases associate with the prolinerich region in GPVI via their SH3 domains. Binding of the 125 I-labeled GST fusion proteins of SH3 domains of Lyn and Src kinases to the MBP-GPVI tail fusion protein was displaced by an excess of the corresponding unlabeled fusion protein, whereas there was no specific binding to MBP itself (Fig. 1A). These findings demonstrate that Src kinases are able to bind the GPVI tail via their SH3 domains.
A proline-rich peptide (Val 282 -Ser 298 ) based on the GPVI tail was used to investigate whether this is able to displace the binding of the 125 I-labeled GST-Lyn-SH3 to a synthetic peptide corresponding to the GPVI cytoplasmic sequence His 269 -Ser 316 . Binding of 125 I-labeled GST-Lyn-SH3 to the GPVI cytoplasmic sequence was inhibited to a similar extent by an excess of the proline-rich peptide or the unlabeled GST fusion protein (Fig.  1B, i). The IC 50 for displacement by the peptide was ϳ6 M (Fig. 1B, ii). These results demonstrate that the SH3 domains of the Src family kinases are able to bind to the proline-rich domain of the GPVI tail.
The ability of the SH3 domains of Src kinases to bind GPVI was further investigated using platelet lysates. GST fusion proteins of the SH3 domains of the Src kinases Lyn, Fyn, Src, and Fgr bound similar levels of GPVI in control and convulxinstimulated platelets, as measured by ligand blotting using convulxin, whereas there was no specific binding to GST alone or the SH3 domains of PLC␥2 and Btk (Fig. 1C). These observations confirm the ability of the GPVI tail to selectively associate with the SH3 domains of Src kinases in vitro.
Fyn and Lyn Bind to the GPVI⅐FcR ␥-Chain Complex in Platelets-The association of Src kinases with the GPVI⅐FcR ␥-chain complex in platelets was investigated through immunoprecipitation of the glycoprotein receptor combined with Western blotting for Src kinases. In agreement with the results of Ezumi et al. (8), we observed a specific association of Fyn and Lyn ( Fig. 2A), but not Src (Fig. 2B) with this complex in resting platelets. We also confirmed the absence of Fyn or Lyn association with control mouse IgG immunoprecipitates (data not shown). A small increase (Ͻ30%) in association with both kinases was observed after a delay of ϳ20 s was seen in some but not all experiments. These findings demonstrate that Fyn and Lyn associate with GPVI⅐FcR ␥-chain complex in platelets and that an increase in this association is an early response to GPVI stimulation. However, these results do not provide information on the site of association of Src kinases within this receptor complex, bearing in mind that Src kinases have also been reported to bind non-phosphorylated and phosphorylated ITAMs (26 -29).
Lyn and Fyn Bind Directly to the Proline-rich Domain of GPVI in Vivo-To investigate whether Fyn and Lyn associate directly with the SH3 domain of GPVI in vivo, we used COS-7 cells reconstituted with wild type GPVI (WT-GPVI) and a receptor mutant lacking the proline-rich domain (Pro(Ϫ)-GPVI; see Fig. 3A). Both forms of GPVI were tagged with FLAG at the C terminus to facilitate analysis through precipitation. Western blotting studies using convulxin and an antibody to convulxin demonstrated that COS-7 cells express similar levels of WT-and Pro(Ϫ)-GPVI on their surface despite the absence of the FcR ␥-chain (data not shown). Immunoprecipitation of WT-GPVI with an anti-FLAG antibody demonstrated a direct association with co-transfected Fyn and Lyn (Fig. 3, B and C). In sharp contrast, there was only minimal association of the two Src kinases with Pro(Ϫ)-GPVI despite a similar level of expression of the glycoprotein and the two Src kinases in the transfected cells (Fig. 3, B and C). These findings demonstrate a direct association of Fyn and Lyn with GPVI in vivo and confirm that the proline-rich domain is critical for this interaction.

FIG. 1. Src kinases bind to the proline-rich domain of GPVI in vitro.
A, detachable microtiter wells were coated with MBP or MBP-GPVI, followed by blocking with 5% BSA. To each well was added 200 l of (i) 125 I-labeled GST-Lyn-SH3 or (ii) GST-Src-SH3. Parallel assays included a 50-fold excess of (i) unlabeled GST-Lyn-SH3 or (ii) GST-Src-SH3. B, detachable microtiter wells were coated with a synthetic peptide corresponding to the GPVI cytoplasmic sequence His 269 -Ser 316 , followed by incubation with 5% BSA. To each well was added 200 l of 125 I-labeled GST-Lyn-SH3. Parallel assays included a 50-fold excess of unlabeled GST-Lyn-SH3 or (i) 50 M GPVI synthetic peptide Val 282 -Ser 298 , or (ii) 5-100 M GPVI synthetic peptide Val 282 -Ser 298 . After 30 min at 22°C, bound radioactivity was counted in a ␥-counter. C, washed platelets stimulated with 10 g/ml convulxin for 20 s were lysed with lysis buffer. Proteins, precipitated with GST fusion proteins containing the SH3 domain of the indicated proteins, were resolved by 10% SDS-PAGE, transferred to polyvinylidene difluoride membranes. Precipitated GPVI was detected by convulxin ligand blotting. The data are representative of three experiments.

FIG. 2. Fyn and Lyn bind to GPVI in platelets.
Washed platelets stimulated with 10 g/ml convulxin for the indicated times were lysed with lysis buffer. Proteins precipitated with anti-GPVI antibody were dissolved in: A, reducing condition or B, non-reducing condition, resolved by 10% SDS-PAGE, and transferred to polyvinylidene difluoride membranes, and immunoblotted with antibodies against (A, i) Fyn or (A, ii) Lyn. A, iii; and B, ii, recruitment of GPVI was confirmed by convulxin ligand blotting. The data are representative of four experiments.

Inhibition of FcR ␥-Chain Phosphorylation by Deletion of GPVI Proline-rich Domain-We then used COS-7 cells express-
ing WT or Pro(Ϫ)-GPVI and co-transfected FcR ␥-chain to investigate whether the association with Lyn and Fyn is necessary for tyrosine phosphorylation of the ␥-chain ITAM. Tyrosine phosphorylation of ␥-chain was measured by Western blotting following precipitation with a GST fusion protein of the tandem SH2 domains of Syk which selectively binds tyrosinephosphorylated ITAMs (30). There was no phosphorylation of the FcR ␥-chain in the absence of Fyn or Lyn in either WT-or Pro(Ϫ)-GPVI expressing cells (not shown). In contrast, the FcR ␥-chain was tyrosine phosphorylated under non-stimulated conditions in COS-7 cells transfected with WT-GPVI and Lyn or Fyn but not in cells transfected with Pro(Ϫ)-GPVI and either Src kinase (Fig. 4, A and B). Furthermore, the level of tyrosine phosphorylation of the FcR ␥-chain was increased in WT-GPVI expressing cells stimulated by convulxin, whereas there was only a marginal change in cells expressing the Pro(Ϫ) mutant (Fig. 4, A and B). Importantly, the association of the FcR ␥-chain with GPVI was similar in both the WT-and Pro(Ϫ)expressing cells as revealed by FLAG precipitation and convulxin blotting for the FcR ␥-chain (Fig. 4C). This demonstrates that the reduced level of tyrosine phosphorylation of the FcR ␥-chain in Pro(Ϫ)-GPVI-transfected cells is not due to disruption of the GPVI-FcR ␥-chain complex. These results demonstrate that the proline-rich domain of GPVI is necessary for interaction with Fyn and Lyn and for subsequent tyrosine phosphorylation of FcR ␥-chain upon activation by convulxin.
Deletion of the Proline-rich Domain Abolishes Responses to GPVI following Stable Transfection into Jurkat Cells-Many of the proteins present in the GPVI signaling cascade are restricted to the hematopoietic cell lineage and are therefore absent in COS-7 cells. Because of this, it was important to extend these studies to cells of an hematopoeitic background. Available immortalized megakaryocytic-like cells, however, express low levels of endogenous GPVI and/or are not readily susceptible to transfection using standardized methodology. For these reasons, we chose to express WT-GPVI and Pro(Ϫ)-GPVI in human Jurkat T cells. It is not necessary to co-transfect the FcR ␥-chain to obtain functional responses as GPVI associates with the ITAM-containing -chains in this cell (16,32). In addition, Jurkat cells express the Src kinases Lck and FynT (31).
The snake toxin convulxin stimulated a rapid and sustained increase in intracellular Ca 2ϩ and whole tyrosine phosphorylation in Jurkat cells stably transfected with WT-GPVI but not with Pro(Ϫ)-GPVI (Fig. 5A), despite similar levels of receptor expression at the cell surfaces as demonstrated by flow cytometry (Fig. 5B). The G protein-coupled agonist thrombin stimulated a brisk increase of intracellular calcium in Jurkat cells transfected with Pro(Ϫ)-GPVI (Fig. 5B, ii), demonstrating that expression of the glycoprotein receptor mutant had not interfered with the Ca 2ϩ -releasing machinery. In addition, protein tyrosine phosphorylation increased in convulxin-stimulated Jurkat cells with WT-GPVI, but not in those with Pro(Ϫ)-GPVI (Fig. 5C). A similar level of association of the -chain was seen with WT-and Pro(Ϫ)-GPVI in transfected Jurkat cells, whereas the ITAM-containing protein was not precipitated in untransfected cells (Fig. 5D). These results demonstrate that the proline-rich domain in GPVI is required for ITAM phosphorylation and downstream responses in an hematopoietic cell line. DISCUSSION We have demonstrated that the proline-rich region in the GPVI tail associates with the SH3 domains of Fyn and Lyn and have presented evidence in support of a critical role for this interaction in mediating signaling by the receptor. These observations extend a number of previous studies proposing a role for the two Src family kinases in mediating GPVI-dependent phosphorylation of the FcR ␥-chain ITAM by providing a molecular basis for this interaction (12,25,33). The results suggest a model in which cross-linking of the glycoprotein receptor brings Fyn and Lyn to the FcR ␥-chain ITAM leading to tyrosine phosphorylation and initiation of downstream events through the tyrosine kinase Syk.
Several lines of evidence support a direct association of Src kinases with the proline-rich region in GPVI. Src kinases were found to bind to a chimera of MBP and the GPVI tail but not to MBP alone, and to a peptide containing the proline-rich region of GPVI. In addition, the SH3 domains of several Src kinases, but not those of Btk and PLC␥2, precipitated GPVI from control and stimulated platelets. Furthermore, the proline-rich domain of GPVI (RPLPPLPPLP) has the consensus sequence for binding to SH3 domains of the Src kinases (17,18). A direct association of the Lyn and Fyn kinases in a cellular environment with GPVI was confirmed by transient transfection studies in COS-7 cells, which lack the FcR ␥-chain. The functional significance of this interaction was demonstrated using COS-7 and Jurkat cells transfected with WT-and Pro(Ϫ)-GPVI. Cells transfected with WT-GPVI, but not Pro(Ϫ)-GPVI, exhibited a marked increase in tyrosine phosphorylation of the FcR ␥-chain ITAM and intracellular calcium mobilization. Importantly, the association of WT-and Pro(Ϫ)-GPVI with FcR ␥-chain or -chain was similar in COS-7 and Jurkat cells, respectively.
Src kinases are known to associate with a number of immune receptors and to play an important role in initiating signals. In Jurkat cells, Fyn and Lck associate with the T cell receptor (TCR)-CD3 complex and CD4, respectively, and it has been proposed that these associations mediate phosphorylation of the CD3-and -chain ITAMs (34 -36). Fyn and Lck interact directly with CD3 and CD4 complex via their N-terminal unique domains, respectively (34 -36). In addition, the N-terminal unique domains of the two Src kinases provide a signal for fatty acid acylation and specific plasma membrane localization, which may also serve to stabilize the interactions between the Fyn SH2 domain and phosphotyrosines in TCR -chain ITAMs (35). Interestingly, the CD3 ⑀-chain has a cluster of proline residues although, as yet, it is not known whether Fyn or Lck associates with this region via their SH3 domain. However, Denny et al. (37) reported that expression of SH3 domaindeleted Lck in an Lck-deficient T cell line inhibited activation of the mitogen-activated protein kinase pathway but not tyrosine phosphorylation of -chain. This demonstrates that the SH3 domain-proline-rich domain interaction is dispensable for T cell receptor signaling. In B cells, Lyn and Fyn are reported to associate with a short sequence, Asp-Cys-Ser-Met, within the Ig-␣ chain of the B cell antigen complex via their unique N terminals (26).
We have proposed a model in which the association of Src kinases with the proline-rich region of GPVI is necessary for ITAM phosphorylation. The SH3 domains of Src family kinases have been proposed to be autoinhibitory. X-ray crystallographic studies of Src and Hck revealed that the SH3 domain mediates an intramolecular interaction with an atypical binding site in the region linking the SH2 and kinase domains (38,39). Erpel et al. (40) have reported that inactivating the SH3 domain of Src induces an 8 -10-fold elevation of its kinase activity. Based on these findings, it seems likely that binding of the SH3 domain of Fyn and Lyn to the GPVI tail will increase their intrinsic activity. Since platelets need to be activated promptly upon platelet adhesion to collagen in subendothelium at the site of vessel damage, Src kinases may need to be "ready-to-go" even before stimulation. On the other hand, immune cells do not need to react so promptly to external stimuli, and, moreover, it may be favorable to have Src kinases in a "low" state of reactivity to avoid unwanted activation. Nevertheless, it is FIG. 5. Calcium mobilization and protein tyrosine phosphorylation are inhibited in Jurkat cells stably transfected with proline-rich domain-deleted GPVI. A, Jurkat cells stably expressed with: (i) wild type (WT) GPVI or (ii) proline-rich domain-deleted (Pro(Ϫ)) GPVI were stimulated with 10 or 20 g/ml convulxin or 2 units/ml thrombin. Intracellular calcium mobilization was measured by the ratio of Fura-2 emissions in a fluorometer. B, Jurkat cells transfected without (i, ii: shaded area) or with WT-GPVI (i: non-shaded area) or Pro(Ϫ)-GPVI (ii: non-shaded area) were incubated with convulxin, anti-convulxin, and fluorescein isothiocyanate-labeled anti-rabbit IgG. The fluorescence was analyzed by flow cytometry. C, cells were stimulated with 10 g/ml convulxin for the indicated times before addition of lysis buffer. D, -chain associated with GPVI immunoprecipitates was detected by immunoblotting with anti--chain antibody. The data are representative of three experiments. important to investigate whether proline-rich regions in other immune receptors are important for mediating activation signals.
In conclusion, we have shown that the proline-rich domain of the GPVI tail is necessary for the association with Src kinases via their SH3 domain and for mediating activation. This study therefore demonstrates a novel pathway of regulation of ITAM phosphorylation by an Ig domain containing receptor.