INTRODUCTION

Platelets express a single-chain low affinity receptor for immunoglobulin, FcRII (1, 2). Fc receptors expression is restricted to cells of hematopoietic lineage (3). Fc receptor activation has been linked to diverse functions that include activation of tyrosine kinases, elevation of intracellular calcium, and regulation of transcription of genes encoding cytokines (3). In platelets, soluble immune complexes or cross-linking of the FcRII receptor with secondary antibodies trigger a robust activation response that includes changes in intracellular calcium concentration, phosphatidic acid metabolism, and thromboxane production (4, 5). Additional activation responses in suspended platelets, including granule secretion and aggregation, are dependent on thromboxane production (4, 5). FcRII ligation also triggers the induction of tyrosine phosphorylation of multiple cellular proteins (6). These include the FcRII receptor itself (2, 6), a 40-kDa sialoglycoprotein that does not have an intrinsic kinase activity, and the FcRII-associated protein-tyrosine kinase, p72^SYK (2, 6). p72^SYK, a homologue of the T cell-associated protein-tyrosine kinase ZAP-70, is a non-receptor tyrosine kinase containing two SH2 domains but no SH3 domain (7). Clustering of chimeric transmembrane proteins bearing intracellular SYK or its homologue ZAP-70 sequences in T cells is sufficient to trigger calcium mobilization and cytolytic effector functions (8, 9). Similarly, clustering of p72^SYK chimera introduced into rat basophilic leukemia (RBL-2H3) cells is sufficient to trigger cellular responses that include protein tyrosine phosphorylation and synthesis and release of allergic mediators (10). p72^SYK activation may thus provide an essential trigger for multiple downstream signaling events.

The mechanism by which Fc receptors trigger SYK phosphorylation and activation is not fully understood. Based on model(s) proposed for multichain immune recognition receptors such as the T cell antigen receptor, TCR (3, 9, 11, 12, 13), FcRII ligation most likely affects several sequential events. First among them is an association between FcRII with a Src family member that triggers transient FcRII receptor phosphorylation. An association between FcRII and the Src family member p56/53^lyn was detected in monocytic THP-1 cells (14) and human B lymphocytes (15, 16), whereas in natural killer cells Fc receptor ligation initiated activation of Lck followed by Syk phosphorylation (17). In platelets activated by immune complexes, FcRII appeared associated with Src (6) but since FcRII phosphorylation was observed in Src-deficient mice (18) another kinase may also affect FcRII phosphorylation. Phosphorylation may enable the FcRII receptor to recruit the p72^SYK kinase, through an interaction between the FcRII-ARAM motif and the p72^SYK SH2 binding domain, as a result of which p72^SYK will become phosphorylated and active (3).

Another yet unresolved issue is the identification of downstream effector targets of p72^SYK. This issue is rather complex because distinct cellular functions are likely to be defined by substrate specificity. We are interested in substrates that are potentially involved in the regulation of cell shape and spreading. Both of these functions are frequently regulated by members of the integrin adhesion receptor family. Integrins are transmembrane heterodimers that interact both with extracellular matrix and cytoskeletal proteins (for reviews see Refs. 19, 20, 21). Recent studies have implicated p72^SYK in signal transduction downstream from integrins. Induction of p72^SYK phosphorylation was detected upon adhesion of THP-1 cells to fibronectin (22). Platelet interaction with fibrinogen or collagen, mediated by the integrin receptors _IIb3 (GP IIb-IIIa) and ₂₁ respectively, similarly triggered p72^SYK phosphorylation (23). In most cell systems studied thus far, integrin receptor ligation triggers the induction of tyrosine phosphorylation of a 125-kDa protein, itself a tyrosine kinase localized in focal adhesion plaques, pp125^FAK (for review see Ref. 24). It is not clear at present whether there is a direct connection between p72^SYK and pp125^FAK phosphorylation and activation. In the THF-1 cells, inhibition of actin polymerization by cytochalasin D prevented pp125^FAK phosphorylation but not p72^SYK phosphorylation (22). Treatment of the platelets with cytochalasin D similarly prevented pp125^FAK (23) but not p72^SYK phosphorylation (2, 23). In addition, both in THF-1 cells and platelets, integrin receptor ligation effectively triggered p72^SYK phosphorylation but not pp125^FAK phosphorylation (18, 23). These data suggest that if the two events are sequentially linked, p72^SYK activation may be an earlier event than pp125^FAK phosphorylation.

In the present study we examined whether FcRII receptor-mediated signals in platelets involve pp125^FAK phosphorylation. Our data demonstrate that FcRII mediates platelet adhesion to immobilized IgG and the induction of pp125^FAK phosphorylation independent of the integrin _IIb3. We have also found that protein kinase C regulates pp125^FAK phosphorylation, but not p72^SYK phosphorylation downstream from FcRII.

MATERIALS AND METHODS

Rabbit polyclonal antisera BC3 and BR12 were used to immunoprecipitate pp125^FAK and p72^SYK, respectively. A rabbit polyclonal antiserum to p72^SYK was used for immunoblotting. Monoclonal antibody IV.3 (1) was from Merdex Inc. Monoclonal antibody 4G10 was from Upstate Biotechnology Inc. Bisindolylmaleimide (BIS¹; GF 1092003X) (25) and thapsigargin were from Calbiochem. Fibrinogen, apyrase, PGI₂, indomethacin, and BSA were from Sigma. Human IgG was purchased from Sigma and Pierce. BAPTA-AM and Fura-2-AM were from Molecular Probes. Cytochalasin D (CD) was from Aldrich. Bisindolylmaleimide, BAPTA-AM, thapsigargin, and CD were dissolved in Me₂SO. The final Me₂SO concentration was 0.5%. Untreated polystyrene tissue culture plates (Corning) were from Fisher. BCA reagents were purchased from Pierce.

Human platelets were isolated by gel filtration from freshly drawn blood anticoagulated with 0.15 volume of NIH formula A acid-citrate-dextrose solution supplemented with 1 µM PGE₁ and 1 unit/ml apyrase as described previously (26). Platelet concentration was adjusted to 2-5 × 10⁸ platelets per ml in an incubation buffer containing 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl₂, 5.6 mM glucose, 1 mg/ml BSA, 3.3 mM NaH₂PO₄, and 20 mM HEPES, pH 7.4. Platelet adhesion to IgG (50 µg/ml), Fc-IgG (50 µg/ml), or FBGN (100 µg/ml) was studied in polystyrene plates precoated with the specific protein and blocked with BSA as described previously (26). Platelets were added to the IgG-, FBGN-, Fc-IgG-, or BSA-coated plates for 60 min at room temperature. Adherent cell morphology was examined by scanning electron microscopy (27).

To examine the effect of the inhibitor on platelets, gel-filtered platelets were incubated with 12 µM BIS (1 h), 12 µM CD (10 min), 1 µM prostaglandin I₂ (PGI₂) (1 min), 10 µM indomethacin (10 min), or with 0.5% Me₂SO alone (Me₂SO vehicle) (1 h). When the effect of BAPTA-AM was studied, platelets in plasma were incubated with 100 µM BAPTA-AM at 37 °C for 30 min and gel-filtered. To examine the effect of BAPTA-AM on intracellular Ca²⁺, platelets in plasma were loaded in some experiments with 100 µM Fura-2-AM for 30 min at 37 °C followed by BAPTA-AM or Me₂SO alone for 30 min before gel filtration (28). Changes in Ca²⁺ were monitored at a wavelength pair of 340 nm/510 nm in a Perkin-Elmer LS-5B Luminescence spectrofluorimeter. To specifically effect the [Ca²⁺]_i concentration, the endomembranous Ca²⁺-ATPase, thapsigargin which rapidly induces calcium mobilization from the platelet intracellular stores (29) and 1 mM CaCl₂ were sequentially added to the platelet suspension, and recording was continued after each addition. Treatment of the platelets with thapsigargin triggered an abrupt increase in [Ca²⁺]_i that was further augmented with the addition of CaCl₂. This biphasic response was not detected in the BAPTA-AM-loaded platelets demonstrating that BAPTA was able to prevent changes in [Ca²⁺]_i (data not shown).

Protein tyrosine phosphorylation was studied as described previously (26). Briefly, adherent platelets were washed with phosphate-buffered saline containing 1 mM sodium vanadate and scraped into 100 µl of sample buffer (66 mM Tris, pH 7.4, 2% SDS) at 90 °C. Lysates were heated at 100 °C for 10 min. After centrifugation for 10 min at 12,000 × g, supernatants were analyzed for protein content using the BCA method, and lysates containing equal amounts of protein were subjected to immunoblotting analyses.

For immunoprecipitation studies, adherent platelets were lysed at 4 °C for 30 min in radioimmune precipitation (RIPA) buffer containing 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 158 mM NaCl, 10 mM Tris-HCl, pH 7.2, 1 mM sodium EGTA, 1 mM phenylmethylsulfonyl fluoride, and 1 mM sodium vanadate. After clarification at 15,000 × g for 30 min, lysates containing equal amounts of protein were immunoprecipitated with rabbit polyclonal antiserum BC3 to pp125^FAK or polyclonal antiserum BR12 to p72^SYK. Immunoprecipitated proteins were examined by immunoblotting with mAb 4G10.

Platelets in plasma were loaded with 50-100 µCi/ml of ⁵¹Cr (ICN Biomedical Inc.) at 37 °C for 1 h and gel-filtered. Counts from samples labeled by this method were >95% cell associated. One hundred-microliter aliquots of untreated and inhibitor- or mAb-pretreated platelets were incubated in IgG- or FBGN-coated 96-well plates for 1 h at room temperature. Nonadherent cells were aspirated and wells were rinsed three times with phosphate-buffered saline. Adherent cells were lysed in 2% SDS. Cell adhesion was determined by measuring  emission. Binding is expressed as a ratio relative to the radioactivity present in an aliquot of untreated platelets. Each experiment was done in triplicate. Data represent the mean ± S.D. of at least four experiments. Data were analyzed using analysis of variance and the Fisher's protected least significant difference post hoc analysis method using a confidence level of p < 0.05.

Platelets in plasma were loaded with [¹⁴C]serotonin (2 µCi/ml; DuPont NEN) at 37 °C for 1 h and gel-filtered. After addition of 1 mM imipramine to inhibit serotonin re-uptake, the platelets were incubated on the protein-coated surfaces for 1 h. After addition of 1% formalin and 5 mM EDTA, incubation mixtures were collected, centrifuged for 10 min at 15,000 × g, and supernatants were assayed for [¹⁴C]serotonin in a scintillation counter. Serotonin release was expressed as a percent of the radioactivity present in an identical platelet aliquot exposed to a BSA-coated surface. Each measurement was done in triplicate. Data represent the mean ± S.D. of at least four experiments.

Untreated gel-filtered platelets or platelets pretreated for 10 min with 10 µM indomethacin, or for 1 min with 1 unit/ml thrombin, were incubated for 1 h on BSA-, FBGN-, or IgG-coated surfaces. After addition of 10 µM indomethacin for 10 min, supernatants were recovered and spun for 3 s at 15,000 × g to remove residual nonadherent platelets. Release of thromboxane B₂ was measured by an enzyme immunoassay kit (Cayman Chemical) (30). Thromboxane production, defined as pg/ml, was expressed as a ratio relative to thromboxane present in an identical platelet aliquot pretreated for 10 min with indomethacin and exposed to a BSA-coated surface. Each measurement was done in triplicate. Data represent the mean ± S.E. of at least four experiments.

RESULTS

Unstimulated gel-filtered platelets were exposed to IgG coated on a polystyrene surface. Suspended platelet aggregates of various sizes were observed within 10 min of platelet exposure to the IgG-coated surfaces. Adherent platelet aggregates were observed shortly thereafter (Fig. 1A). Binding to IgG was prevented by the FcRII receptor mAb, IV.3 (1). Adherent platelet aggregates were also observed on surfaces coated with Fc-IgG at the same concentration (data not shown).

We assumed that platelet aggregation on IgG was indicative of granule secretion and/or activation of the cyclooxygenase pathway leading to thromboxane production. To assess dense granule secretion, [¹⁴C]serotonin-loaded platelets were exposed to IgG-, BSA-, or FBGN-coated surfaces, and serotonin release into the medium was monitored after 1 h. Platelets adherent to IgG released 4.1 ± 1.33 times more [¹⁴C]serotonin than platelets exposed to a BSA-coated surface but, as previously shown (26), platelet adherence to FBGN did not cause detectable serotonin release (1.26 ± 0.34 relative to platelets exposed to BSA-coated surfaces). When compared with the release triggered by pretreatment of the platelets with 1 unit/ml thrombin for 1 min before exposure to the IgG-coated surface, the untreated IgG-adherent platelets released 59% ± 10% (n = 4) of their dense granule content.

To measure thromboxane production supernatants of untreated platelets exposed for 1 h to IgG-, FBGN-, or BSA-coated surfaces were collected and assayed for thromboxane B₂, a stable analogue of thromboxane A₂ (30). Platelets pretreated with indomethacin, a cyclooxygenase inhibitor, and exposed to the BSA-coated surfaces for 1 h served as controls. Untreated platelet interaction with BSA or adherence to FBGN caused no detectable thromboxane release. In contrast, platelet adhesion to IgG-coated surfaces triggered massive thromboxane production (308 ± 143 relative to platelets exposed to BSA-coated surfaces) that was completely inhibited by indomethacin (1 relative to platelets exposed to BSA-coated surfaces). Platelets adherent to IgG released on the average 10 times more thromboxane than FBGN-adherent platelets prestimulated with 1 unit/ml thrombin for 1 min. These results indicate that immobilized IgG cannot only support platelet adhesion but also is a potent stimulant of granule secretion and thromboxane production.

For these studies, platelets were pretreated with several specific and well characterized inhibitors that include bisindolylmaleimide (BIS), a specific protein kinase C inhibitor (25), indomethacin, PGI₂, a prostacyclin that increases cAMP and blocks activation in response to a variety of agonists (31), and cytochalasin D (CD) which blocks actin polymerization in agonist-stimulated platelets and spreading on FBGN (26). As shown in Fig. 2, pretreatment of the platelets with indomethacin, PGI₂, or BIS caused a partial yet statistically significant decrease in platelet binding to IgG (65% ± 5% and 69% ± 4%, and 56% ± 4%, respectively). Bisindolylmaleimide, but not indomethacin or PGI₂, blocked aggregation (Fig. 1B) and reduced secretion by more than 50%. These results are consistent with the previously observed effects of BIS on platelet aggregation and secretion (25). Treatment of the platelets with CD caused an even greater decrease in platelet adhesion to IgG (24% ± 6%, p < 0.0001). The results with CD were not statistically different from the results obtained with the Fc receptor-specific mAb, IV.3 (17% ± 4%). Despite its inhibitory effect on platelet adhesion to the IgG-coated surface, treatment of the platelets with CD did not prevent [¹⁴C]serotonin release (data not shown). Neither one of these inhibitors affected _IIb3-mediated platelet binding to FBGN. These data suggested that FcRII and _IIb3 mediate adhesion and/or organization of the cytoskeleton via distinct mechanisms. Furthermore, these results suggested that platelet adhesion to the IgG-coated surface is greatly dependent on an intact cytoskeleton and/or a cytoskeleton-dependent function(s).

To examine whether pp125^FAK was phosphorylated in IgG-adherent platelets, pp125^FAK immunoprecipitates were analyzed by immunoblotting with mAb 4G10. As shown in Fig. 3A, a tyrosine-phosphorylated pp125^FAK kinase was immunoprecipitated from lysates of IgG- and FBGN-adherent platelets. A tyrosine-phosphorylated p72^SYK kinase was also immunoprecipitated from lysates of platelets adherent both to FBGN and IgG (Fig. 3, A-C). A higher intensity of the tyrosine-phosphorylated p72^SYK band was consistently detected in the IgG-adherent as compared to the FBGN-adherent platelet lysates (Fig. 3, A and B) although the p72^SYK signal itself was unchanged (Fig. 3C).

The integrin _IIb3 mediates platelet aggregation and pp125^FAK phosphorylation (26, 32). Since platelets adherent to IgG aggregate, we next examined whether _IIb3 contributed to the induction of pp125^FAK phosphorylation in the IgG-adherent platelets. For these studies, platelets from two patients with Glanzmann's thrombasthenia, previously shown to contain <5% of the normal amount of _IIb3 (26), were exposed to IgG-coated surfaces. These platelets formed a confluent monolayer on IgG indistinguishable from the one observed on a collagen-coated surface (data not shown). In addition, pp125^FAK and p72^SYK were phosphorylated to a similar degree in the normal as compared to _IIb3-deficient platelets adherent to IgG or collagen (Fig. 4). Immobilized collagen served as a positive control in this study since pp125^FAK phosphorylation in Glanzmann's thrombasthenia platelets adherent to this surface has been previously demonstrated (26). These data indicate that _IIb3 is not required for the induction of tyrosine phosphorylation of this particular protein-tyrosine kinase substrate or for platelet adhesion to IgG.

in normal and

Recent studies have indicated that protein kinase C regulates pp125^FAK phosphorylation upon integrin ligation in suspended (28) and in FBGN-adherent platelets (27). Others have shown that tyrosine phosphorylation of p72^SYK is negatively regulated through calcium mobilization in thrombin-stimulated platelets (33). To examine the intracellular signaling pathways that regulate tyrosine phosphorylation of pp125^FAK and p72^SYK in the IgG-adherent platelets, these specific proteins were immunoprecipitated from BIS- or BAPTA-AM-pretreated platelets. Treatment with BIS prevented the detection of a phosphorylated pp125^FAK protein but did not affect p72^SYK phosphorylation while BAPTA-AM had no effect on tyrosine phosphorylation of either pp125^FAK or p72^SYK (Fig. 5). Treatment of the platelets with indomethacin or PGI₂ also had no effect on pp125^FAK phosphorylation. These data indicated that p72^SYK phosphorylation may be necessary but it is certainly not sufficient to trigger pp125^FAK phosphorylation.

Platelet adhesion to the IgG-coated surface stimulated tyrosine phosphorylation of multiple proteins, most of which displayed an electrophoretic mobility similar to that of proteins detected in lysates of FBGN-adherent platelets (Fig. 6). However, tyrosine phosphorylation of a 72-kDa protein, migrating with an electrophoretic mobility similar to p72^SYK, a 47- and 44-kDa doublet, and a broad 39-42-kDa band was predominantly observed in the IgG-adherent platelets (Fig. 6). Tyrosine phosphorylation of proteins of 105 and 101 kDa and of proteins smaller than 47 kDa was sensitive to pretreatment of the platelets with BIS (Fig. 6, lanes 3 and 7). Pretreatment of the platelets with indomethacin or PGI₂ had no detectable effects on the induction of protein tyrosine phosphorylation (Fig. 6, lanes 6 and 7, respectively). These data suggested that adhesion to IgG triggers protein tyrosine phosphorylation partially dependent on protein kinase C activation but independent of thromboxane production.

DISCUSSION

Platelet adhesion to IgG-coated surfaces triggered tyrosine phosphorylation of p72^SYK and pp125^FAK. These specific proteins were also tyrosine-phosphorylated in IgG-adherent platelets from patients with Glanzmann's thrombasthenia consistent with an _IIb3-independent phosphorylation mechanism. Evidence is accumulating that several distinct types of receptors signal pp125^FAK phosphorylation. These include the G-protein-linked neuropeptide receptors for bombesin, vasopressin, and endothelin (34), as well as the platelet-derived growth factor receptor, a transmembrane protein with intrinsic tyrosine kinase activity (35). Enhanced pp125^FAK phosphorylation was also observed following treatment of fibronectin-adherent RBL-2H3 cells with FcRI-divalent antibodies. No phosphorylation was induced, however, by the aggregation of the FcRI receptor in suspended cells (36, 37). In mouse macrophages, FcRIII and FcRI receptor-mediated phagocytosis was associated with induction of p72^SYK, but not pp125^FAK, tyrosine phosphorylation (38). The present study is thus the first to demonstrate that an immunoglobulin receptor can directly signal pp125^FAK phosphorylation.

Both in platelets and in neutrophils the cytoskeleton assembly is a required component of the Fc receptor-mediated signaling pathways. Treatment of the neutrophils with CD prevented the actin filaments assembly and phagocytosis suggesting that these events may be functionally linked (39). Kang et al. (5) have similarly shown that pretreatment of platelets with cytochalasin B blocked phosphatidic acid accumulation, intracellular Ca²⁺ increase, p47 and p20 phosphorylation, serotonin release, and aggregation, induced by heat-aggregated IgG. Consistent with these data we observed that CD prevents platelet binding to IgG. Integrin-mediated tyrosine phosphorylation of pp125^FAK in platelets is sensitive to CD treatment (26, 32). pp125^FAK may regulate the cytoskeleton assembly through its interaction with actin-binding proteins or indirectly by phosphorylation of other regulatory proteins (20, 21, 24). If pp125^FAK phosphorylation is essential for platelet adhesion to an IgG-coated surface, inhibitors of pp125^FAK phosphorylation are expected to prevent adhesion. Our studies with BIS, however, indicate that inhibition of pp125^FAK phosphorylation does not prevent platelet binding to IgG. Preliminary studies with human erythroleukemia (HEL) cells similarly suggest that pp125^FAK phosphorylation is not required for cell binding to IgG. HEL cells express both the FcRII and _IIb3 receptor (40, 41). pp125^FAK phosphorylation was detected in phorbol 12-myristate 13-acetate-pretreated HEL cells adherent to FBGN.² In contrast, pp125^FAK phosphorylation was not observed in either untreated or phorbol 12-myristate 13-acetate-treated IgG-adherent HEL cells. A tyrosine-phosphorylated p72^SYK protein was, however, detected in the same lysates. Taken together, these data suggest that adhesion to IgG is regulated in a manner independent of pp125^FAK phosphorylation.

Tyrosine phosphorylation of p72^SYK has been demonstrated in thrombin-stimulated platelets (33, 42), following FcRII receptor clustering (2), or when the platelet _IIb3 or ₂₁ integrin receptors were engaged (23). This phosphorylation event was insensitive to the presence of CD or EGTA (2). In contrast, translocation of the tyrosine-phosphorylated p72^SYK kinase to the Triton X-100-insoluble fraction that is stimulated in platelets by thrombin was affected by CD (23, 42). These data raise the possibility that p72^SYK plays a role in linking the Fc receptor to the actin cytoskeleton and that severing this link with CD interferes with many of the Fc receptor-mediated functions. A second possibility, not exclusive of the former, is that once the p72^SYK and Fc receptor complex is formed, additional signaling components may become activated. In mouse macrophages (43) or the human monocytic cell line THP-1 (44), for example, FcRIIA ligation triggers tyrosine phosphorylation of several proteins including Shc, GTPase-activating protein, phospholipase C-1, phospholipase C-2 and Vav. Shc and GTPase-activating protein activation may initiate activation of Ras-dependent pathway(s) while phospholipase C-1 activation can trigger diacylglycerol and inositol 1,4,5-trisphosphate formation from phosphatidylinositol 4,5-biphosphate, leading to protein kinase C activation and release of intracellular calcium (45, 46, 47, 48). In concert, these secondary signaling mediators may affect the cytoskeleton assembly. Similar mechanisms may explain the Fc receptor ability to respond to immobilized IgG as an adhesion receptor would to a matrix ligand.