Expression and Function of the Mouse Collagen Receptor Glycoprotein VI Is Strictly Dependent on Its Association with the FcRγ Chain*

Platelet glycoprotein (GP) VI has been proposed as the major collagen receptor for activation of human platelets. Human GPVI belongs to the immunoglobulin superfamily and is noncovalently associated with the FcRγ chain that is involved in signaling through the receptor. In mice, similar mechanisms seem to exist as platelets from FcRγ chain-deficient mice do not aggregate in response to collagen. However, the activating collagen receptor on mouse platelets has not been definitively identified. In the current study we examined the function and in vivo expression of GPVI in control and FcRγ chain-deficient mice with the first monoclonal antibody against GPVI (JAQ1). On wild type platelets, JAQ1 inhibited platelet aggregation induced by collagen but not PMA or thrombin. Cross-linking of bound JAQ1, on the other hand, induced aggregation of wild type but not FcRγ chain-deficient platelets. JAQ1 stained platelets and megakaryocytes from wild type but not FcRγ chain-deficient mice. Furthermore, JAQ1 recognized GPVI (approximately 60 kDa) in immunoprecipitation and Western blot experiments with wild type but not FcRγ chain-deficient platelets. These results strongly suggest that GPVI is the collagen receptor responsible for platelet activation in mice and demonstrate that the association with the FcRγ chain is critical for its expression and function.

Platelet glycoprotein (GP) VI has been proposed as the major collagen receptor for activation of human platelets. Human GPVI belongs to the immunoglobulin superfamily and is noncovalently associated with the FcR␥ chain that is involved in signaling through the receptor. In mice, similar mechanisms seem to exist as platelets from FcR␥ chain-deficient mice do not aggregate in response to collagen. However, the activating collagen receptor on mouse platelets has not been definitively identified. In the current study we examined the function and in vivo expression of GPVI in control and FcR␥ chain-deficient mice with the first monoclonal antibody against GPVI (JAQ1). On wild type platelets, JAQ1 inhibited platelet aggregation induced by collagen but not PMA or thrombin. Cross-linking of bound JAQ1, on the other hand, induced aggregation of wild type but not FcR␥ chain-deficient platelets. JAQ1 stained platelets and megakaryocytes from wild type but not FcR␥ chaindeficient mice. Furthermore, JAQ1 recognized GPVI (approximately 60 kDa) in immunoprecipitation and Western blot experiments with wild type but not FcR␥ chain-deficient platelets. These results strongly suggest that GPVI is the collagen receptor responsible for platelet activation in mice and demonstrate that the association with the FcR␥ chain is critical for its expression and function.
Collagen is one of the major components of the vessel wall responsible for platelet adhesion and activation at sites of vascular injury (1). A variety of collagens have been identified, seven of which are found in the subendothelial layer. The interaction between platelets and collagen can either occur indirectly via intermediary proteins like von Willebrand factor, which complexes to collagen(s) in the vessel wall and concomitantly binds to the platelet receptors glycoprotein (GP) 1 Ib-V-IX and activated GPIIb/IIIa, or by direct recognition of collagen by specific receptors. Several different receptors for collagen have been identified on platelets including CD36 (2), a p65 collagen type I-specific receptor (3), integrin ␣ 2 ␤ 1 (4), and the non-integrin GPVI (5). In humans, there is growing evidence for GPVI as the major collagen receptor for platelet activation (6). Very recently, molecular cloning of GPVI cDNA demonstrated this receptor as a type I transmembrane protein belonging to the immunoglobulin (Ig) superfamily (7). Its sequence is most closely related to human Fc␣R and natural killer cell receptors as well as to polymorphic mouse receptors known as paired Ig-like receptors (PIR) (8,9). GPVI (62-65 kDa) contains two Ig-C2-like extracellular domains formed by disulfide bonds and a 51-amino acid cytoplasmic tail that is free of any consensus sequence motifs for activation. Instead, GPVI harbors a positively charged arginine in its transmembrane region. The presence of transmembrane-charged residues is a typical feature seen in a variety of stimulatory receptors, such as Fc receptors for IgG (Fc␥RI and Fc␥RIII) (10), IgE (Fc⑀RI) (10), and IgA (Fc␣R) (11), as well as PIR-A (8) and natural killer receptor-P1 (CD161) (12). All of them lack amino acid activation motifs but can associate with the FcR␥ signaling subunit to trigger activation in response to receptor aggregation. According to in vitro studies on Fc␣R and mouse PIR-A, it has been suggested that the transmembrane arginine of GPVI is similarly required for its function and association with the FcR␥ chain (7,8,11). Findings by Gibbins and co-workers (13,14) support a model where human GPVI couples collagen stimulation of platelets to phosphorylation of the FcR␥ chain leading to activation of Syk and phospholipase C␥2.
Activation of mouse platelets by collagen is also found to be associated with tyrosine phosphorylation of multiple proteins including the FcR␥ chain, Syk, and phospholipase C␥2 (15). However, identification of the mouse collagen receptor coupled to the FcR␥ chain has not been established due to the lack of antibodies recognizing GPVI. Here we report that GPVI expression on mouse platelets is strictly dependent on the presence of the FcR␥ chain as demonstrated by flow cytometry, immunoprecipitation, and Western blot analysis with a newly developed mAb (JAQ1) against GPVI in FcR␥ chain-deficient and control mice. Moreover, JAQ1 completely inhibited collagen-induced aggregation of mouse platelets. Therefore, it is concluded that GPVI represents the major collagen receptor for platelet activation in both mice and humans.

EXPERIMENTAL PROCEDURES
Animals-Specific pathogen-free mice (NMRI, C57Bl/6) 6 -10 weeks of age were obtained from Charles River Breeding Laboratories, Sulzfeld, Germany, and kept in our animal facilities. C57Bl/6 mice deficient in the FcR␥ chain (16) were obtained from Taconics, Germantown, NY, and kept under dry barrier conditions in the animal facilities at the Hannover Medical School (Hannover, Germany) until usage at 2-4 months of age.
Antibodies-The rat anti-mouse P-selectin mAb RB40.34 was kindly provided by D. Vestweber, Mü nster, and modified in our laboratories. FITC-labeled rabbit anti-rat Ig, HRP-labeled rabbit anti-FITC, and rabbit anti-fibrinogen were purchased from Dako. All the following antibodies were generated, produced, and modified in our laboratories: Platelet Preparation-Mice were bled under ether anesthesia from the retro-orbital plexus. Blood was collected in a tube containing 10% (v/v) 0.1 mol/liter sodium citrate or 7.5 units/ml heparin, and plateletrich plasma (prp) was obtained by centrifugation at 300 ϫ g for 10 min at room temperature. The platelets were washed twice with Tris buffer (TB, 20 mmol/liter Tris-HCl, pH 7.3, 0.9% NaCl) by centrifugation at 1,300 ϫ g for 10 min and used immediately. Isolated platelets did not show any signs of activation as shown by flow cytometry (staining for P-selectin and surface-expressed fibrinogen).
Production of Monoclonal Antibodies (mAbs)-Female Wistar rats, 6 -8 weeks of age, were immunized repeatedly with mouse platelets. The rat spleen cells were then fused with mouse myeloma cells (Ag8.653), and hybridomas were selected in HAT medium. Hybridomas secreting mAbs directed against platelet receptors were identified by flow cytometry. Briefly, a 1:1 mixture of resting and thrombin-activated FIG. 1. Platelet aggregation by collagen, PMA, and ␣-thrombin in the presence of JAQ1. Heparinized prp was incubated with JAQ1 (20 g/ml) or irrelevant rat IgG2a (20 g/ml) before addition of collagen (5 g/ml) or PMA (50 ng/ ml). Thrombin-induced aggregation (0.1 unit/ml) was performed with washed platelets.
FIG. 2. JAQ1 binds to mouse GPVI. a, flow cytometric detection of GPVI on mouse platelets. Platelets were incubated with JAQ1 (10 g/ml, solid line) or irrelevant rat IgG2a (10 g/ml, shaded area) for 30 min at room temperature. Bound JAQ1 was detected with rabbit anti-rat IgG-FITC. b, immunoprecipitation of GP-VI from surface-biotinylated mouse platelets. Nonidet P-40 lysates were incubated with 10 g/ml nonimmune rat IgG2a (control, contr.) or JAQ1, followed by protein G-Sepharose. Proteins were separated on a 9 -15% gradient SDS-PAGE gel under reducing (red.) conditions, transferred onto a PVDF membrane, and detected by streptavidin-HRP and ECL. c, platelet proteins were separated by SDS-PAGE and immunoblotted with FITC-labeled JAQ1. Bound JAQ1 was detected by HRPlabeled rabbit anti-FITC. d, heparinized prp was incubated with JAQ1 (20 g/ml) or irrelevant rat IgG2a (20 g/ml) followed by addition of polyclonal rabbit anti-rat IgG antibodies (10 g/ml). nonred., non-reduced. platelets (10 6 ) was incubated with 100 L supernatant for 30 min at room temperature, washed with PBS (1,300 ϫ g, 10 min), and stained with FITC-labeled rabbit anti-rat Ig (Dako) for 15 min. The samples were analyzed on a FACScan (Becton Dickinson, Heidelberg, Germany) in the set-up mode. Platelets were gated by FSC/SSC characteristics. Positive hybridomas were subcloned twice prior to large scale production.
Immunoprecipitation and Immunoblotting-Immunoprecipitation was performed as described previously (19). Briefly, 10 8 washed platelets were surface-labeled with EZ-Link sulfo-NHS-LC-biotin (Pierce, 100 g/ml in PBS) and subsequently solubilized in 1 ml of lysis buffer (Tris-buffered saline containing 20 mmol/liter Tris/HCl, pH 8.0, 150 mmol/liter NaCl, 1 mmol/liter EDTA, 1 mmol/liter phenylmethylsulfonyl fluoride, 2 g/ml aprotinin, 0.5 g/ml leupeptin, and 0.5% Nonidet P-40, all from Roche Molecular Biochemicals). Cell debris was removed by centrifugation (15,000 ϫ g, 10 min). Following preclearing (8 h), 10 g of mAb was added together with 25 l of protein G-Sepharose (Amersham Pharmacia Biotech), and precipitation took place overnight at 4°C. Samples were separated on a 9 -15% gradient SDS-polyacrylamide gel (PAGE) along with a molecular weight marker and transferred onto a polyvinylidene difluoride (PVDF) membrane. The membrane was incubated with streptavidin/horseradish peroxidase (1 g/ ml) for 1 h after blocking. After extensive washing, biotinylated proteins were visualized by ECL (Amersham Pharmacia Biotech). For immunoblotting, platelets were not surface-labeled. After lysis, whole cell extract was run on an SDS-polyacrylamide gel and transferred onto a PVDF membrane. The membrane was first incubated with 5 g/ml FITC-labeled mAb followed by rabbit anti-FITC horseradish peroxidase (1 g/ml). Proteins were visualized by ECL.
Immunohistochemistry-Acetone-fixed cryosections (6 m) were blocked (5% normal goat serum, 5 mg/ml bovine serum albumin in PBS) for 30 min at room temperature. JAQ1 was added at a final concentration of 2 g/ml. After 90 min, the sections were washed three times with PBS and subsequently incubated with the HRP-labeled rabbit anti-rat IgG antibodies at a final concentration of 2 g/ml for 60 min at room temperature. The 3-amino-9-ethylcarbazole substrate was added after the three washing steps, and the sections were then counterstained with hematoxylin.

RESULTS AND DISCUSSION
A new mAb against mouse GPVI (JAQ1, rat IgG2a) was generated. JAQ1 blocked collagen-induced platelet aggregation in prp and on washed platelets in a dose-dependent manner, whereas it had no effect on aggregation induced by PMA (in prp) or ␣-thrombin (washed platelets) (Fig. 1). Incubation of prp with varying concentrations (3-30 g) of JAQ1 under stirring (1000 rpm, 37°C) or static conditions did not induce any signs of platelet activation (surface expression of P-selectin or fibrinogen). Consequently, no aggregation was induced by the mAb itself. JAQ1 bound to mouse platelets (Fig. 2a) and precipitated a single chain protein of an apparent molecular weight of approximately 60 kDa under nonreducing conditions (Fig. 2b). The molecular weight slightly shifted to approximately 65 kDa under reducing conditions, demonstrating that the apparent molecular weight of mouse GPVI is very similar to its human homolog (7,20). Furthermore, JAQ1 recognized GPVI as a single band of approximately 60 kDa under nonreducing conditions in a Western blot analysis of platelet lysates (Fig. 2c), whereas no band was detected under reducing conditions (not shown) indicating that JAQ1 binds to a conformational epitope on the receptor. On human platelets, polyclonal antibodies against GPVI have been reported to induce platelet aggregation when used as IgG or F(abЈ) 2 fragments (7, 21), whereas monovalent Fab fragments had no activator effect, suggesting that clustering of GPVI is required for signal transduction. To induce clustering of GPVI, we cross-linked surface-bound JAQ1 by the addition of polyclonal rabbit anti-rat IgG antibodies (10 g/ml). As shown in Fig. 2D, such treatment rapidly resulted in irreversible platelet aggregation. In contrast, addition of irrelevant rat IgG2a followed by polyclonal rabbit anti-rat IgG antibodies (10 g/ml) had no effect. To our knowledge, JAQ1 is the first mAb directed against GPVI and therefore the only divalent anti-GPVI reagent described to date. This may explain why JAQ1, in contrast to polyclonal antibodies against human GPVI, did not induce platelet activation/aggregation by itself but only upon cross-linking. The finding that blockage of GPVI with JAQ1 completely inhibited collagen-induced platelet activation strongly suggests that GPVI is the principal collagen receptor for platelet activation in mice. Like in the human system, receptor clustering seems to be required for signaling through GPVI.
GPVI Is Not Detectable on Platelets and Megakaryocytes from FcR␥ Chain-deficient Mice-Human GPVI has been described to be associated physically and functionally with the FcR␥ chain which seems to be critical for platelet activation through FIG. 3. JAQ1 has no effect on platelets from FcR ␥ chain-deficient mice. Heparinized prp from FcR␥ chain-deficient (␥ Ϫ/Ϫ) and control (␥ ϩ/ϩ) mice was incubated with JAQ1 (20 g/ml) followed by addition of polyclonal rabbit anti-rat IgG antibodies (10 g/ml). the receptor (13,22). Furthermore, it is well recognized that platelets from FcR␥ chain-deficient mice do not aggregate in response to collagen (15). Based on these findings, we expected JAQ1 not to induce platelet aggregation upon cross-linking on platelets from FcR␥ chain-deficient mice. Indeed, addition of increasing concentrations of JAQ1 (3-30 g/ml) followed by polyclonal rabbit anti-rat IgG antibodies (10 g/ml) had no effect on FcR␥ chain-deficient platelets (Fig. 3). In parallel experiments with platelets from C57Bl/6 control mice, the same procedure always resulted in platelet aggregation. This finding again demonstrated that the antigen recognized by JAQ1 is mouse GPVI and further suggested that function through mouse GPVI requires the association with the FcR␥ chain.
Therefore, we assessed the expression of GPVI in platelets and megakaryocytes from C57Bl/6 mice in comparison to FcR␥ chain-deficient mice. JAQ1 did not bind to FcR␥ chain-deficient platelets as determined by flow cytometry (Fig. 4a). Immunohistochemical examination of splenic megakaryocytes from FcR␥ chain-deficient mice demonstrated no specific binding of JAQ1, whereas megakaryocytes from C57Bl/6 control mice were clearly stained (Fig. 4b).
Moreover, in the absence of the FcR␥ chain, GPVI-specific bands were not detectable in JAQ1 immunoprecipitates from surface-biotinylated platelets nor in a Western blot analysis of whole platelet lysate. These data strongly suggest that GPVI is not expressed in vivo in the absence of the FcR␥ chain, although we cannot definitively rule out the possibility that the protein is expressed but fails to undergo the folding necessary for JAQ1 to bind. In contrast to GPVI, similar amounts of GPIIb/IIIa were immunoprecipitated from control and FcR␥ chain-deficient platelets, and GPIIIa was detected in both platelet preparations by Western blot analysis (Fig. 4, c and d).
Furthermore, flow cytometric studies demonstrated that the expression of GPIIb/IIIa, the collagen receptor GPIaIIa, and the GPIb-IX-V complex was indistinguishable between platelets from C57Bl/6 and FcR␥ chain-deficient mice (data not shown).
Taken together, our results give strong evidence that GPVI is the dominant receptor for collagen-induced activation of platelets in mice. We have used FcR␥ chain-deficient mice in combination with the first monoclonal antibody against mouse GPVI to demonstrate that correct expression and function of mouse GPVI is strictly dependent on the presence of the FcR␥ Platelets were incubated with JAQ1 (10 g/ml, solid line) or irrelevant rat IgG2a (10 g/ml, shaded area) for 30 min at room temperature, followed by rabbit anti-rat IgG-FITC. b, immunohistochemical detection of GPVI on splenic megakaryocytes on acetone-fixed frozen sections. Unstained megakaryocytes are indicated by arrows. c, immunoprecipitation of GPVI and GPIIb/IIIa from surface-biotinylated control and FcR␥ chain-deficient platelets. Nonidet P-40 lysates were incubated with 10 g/ml JAQ1 or MWReg30 (anti-GPIIb/IIIa), followed by protein G-Sepharose (Amersham Pharmacia Biotech). Proteins were separated on a 9 -15% gradient SDS-PAGE gel under non-reducing conditions, transferred onto a PVDF membrane, and detected by streptavidin-HRP and ECL. d, whole platelet proteins were separated by SDS-PAGE and immunoblotted with FITC-labeled JAQ1 or EDL1 (anti-GPIIIa). Bound mAb was detected by HRP-labeled rabbit anti-FITC.
subunit. These observations, combined with reports that human GPVI is also physically and functionally associated with the FcR␥ chain (13,14), indicate that the mechanisms leading to collagen-induced platelet activation are similar in mice and humans.