Low Density Lipoprotein Phosphorylates the Focal Adhesion-associated Kinase p125FAK in Human Platelets Independent of Integrin αIIbβ3 *

Low density lipoprotein (LDL) is known to sensitize platelets to agonists via integrin mediated outside-in signaling (Hackeng, C. M., Huigsloot, M., Pladet, M. W., Nieuwenhuis, H. K., Rijn, H. J. M. v., and Akkerman, J. W. N. (1999) Arterioscler. Thromb. Vasc. Biol., in press). As outside in signaling is associated with phosphorylation of p125FAK, the effect of LDL on p125FAK phosphorylation in platelets was investigated. LDL induced p125FAK phosphorylation in a dose- and time- dependent manner. The phosphorylation was independent of ligand binding to integrin αIIbβ3 and aggregation, such in contrast to α-thrombin-induced p125FAK phosphorylation, that critically depended on platelet aggregation. Platelets from patients with Glanzmann’s thrombastenia showed the same LDL- induced phos- phorylation of p125FAK as control platelets, whereas α-thrombin completely failed to phosphorylate the kinase in the patients platelets. LDL signaling to p125FAK was independent of integrin α2β1, the FcγRII receptor, and the lysophosphatidic acid receptor and not affected by inhibitors of cyclooxygenase, protein kinase C, ERK1/2 or p38MAPK. Phosphorylation of p125FAK by LDL was strongly inhibited by cyclic AMP. These observations indicate that LDL is a unique platelet agonist, as it phosphorylates p125FAKin platelet suspensions, under unstirred conditions and independent of integrin αIIbβ3.

Focal adhesion kinase (p125 FAK ) is a nonreceptor tyrosine kinase implicated in signaling pathways mediated by integrins, G-protein coupled receptors, tyrosine kinase receptors, and the v-Src and v-Crk oncoproteins (1). After cell activation, p125 FAK translocates to the cytoskeleton at focal adhesions (2)(3)(4), where it serves as a docking site for signaling proteins (5,6). The protein contains six tyrosine phosphorylation sites. Autophosphorylation of Tyr 397 generates a high affinity binding site for the SH2 domain of Src family kinases. The association of Src subsequently initiates phosphorylation of Tyr 407,576,577 , inducing maximal kinase activity of p125 FAK (7). Association of Src also leads to the phosphorylation of Tyr 925 , thereby creating a docking site for the adaptor protein Grb2 that is known to mediate Ras activation by binding of the GDP/GTP exchange factor Sos, linking p125 FAK to the Ras/ MAP kinase pathway (8). The relevance of phosphorylation on Tyr 861 is not entirely clear, but it could serve as another site of p125 FAK interaction with Src-family kinases.
Phosphorylation of p125 FAK in blood platelets is different for platelets in suspension and platelets adherent to immobilized ligand. In platelet suspensions, p125 FAK phosphorylation only occurred under aggregating conditions (9). An antibody against integrin ␣ IIb ␤ 3 , that blocked fibrinogen binding and aggregation, totally abolished p125 FAK phosphorylation by ␣-thrombin and collagen in stirred suspensions. In the absence of stirring, ␣-thrombin (9) or the ␣ IIb ␤ 3 -activating antibody LIBS6 (10) failed to induce p125 FAK -phosphorylation. The role for ␣ IIb ␤ 3 in this signaling event was further supported by platelets from patients with Glanzmann's thrombastenia, that lack ␣ IIb ␤ 3 , in which neither ␣-thrombin nor collagen induced p125 FAK phosphorylation in stirred suspensions (9).
Platelets adherent to immobilized ligand show ␣ IIb ␤ 3 -dependent and -independent p125 FAK phosphorylation. Platelets bound to fibrinogen show p125 FAK phosphorylation via ␣ IIb ␤ 3 . Platelets from a patient with a truncated cytoplasmic domain of the ␤ 3 -subunit (11) and Chinese hamster ovary cells transfected with truncated forms of the ␤ 3 -subunit (12) bound to immobilized fibrinogen but failed to induce phosphorylation of p125 FAK . Expression of a constitutively active mutant of ␤ 3 together with ␣ IIb led to a slight degree of p125 FAK phosphorylation in suspended Chinese hamster ovary cells in the presence of fibrinogen. However, this phosphorylation was negligible compared with the same cells adherent to fibrinogen (13). Also ␣ IIb ␤ 3 -independent p125 FAK phosphorylation has been described. Collagen (14 -16) and immunoglobulins (15) immobilized on a surface induced p125 FAK phosphorylation in the presence of an anti-␣ IIb ␤ 3 antibody and in Glanzmann's platelets. Hence, in platelets p125 FAK may play a central role in signal transduction after ␣ IIb ␤ 3 ligation or in platelet adhesion, thereby strengthening ligand-receptor interaction and coordinating further signaling.
Low density lipoprotein (LDL) 1 is known to increase the sensitivity of human platelets to different agonists (17)(18)(19)(20), but the intracellular mechanisms involved remain largely unknown. Among the signal transducing elements that are activated by LDL are protein kinase C (PKC) (21,22), Ca 2ϩ mobi-lization (20,23,24) and phosphoinositide turnover (20,21,24,25). The faster collagen-induced secretion in LDL-treated platelets critically depends on ligand-induced outside-in signaling via integrin ␣ IIb ␤ 3 . This is illustrated in ␣ IIb ␤ 3 -deficient platelets, where pretreatment with LDL failed to increase dense granule secretion in response to collagen (22). Similar results were obtained when binding of released fibrinogen was blocked with the fibrinogen ␥-chain-derived peptide ␥ 400 -411 . As integrin signaling involves receptor ligation and clustering, and concentration of signaling elements in focal adhesions (reviewed in Refs. 5 and 6), we set to explore the involvement of p125 FAK in LDL-induced sensitization of platelets.
Anti-phosphotyrosine mAb 4G10 was from Upstate Biotechnology (Bucks, United Kingdom), anti-p125 FAK mAb was from Transduction Laboratories (Lexington, NY) and anti-p125 FAK polyclonal antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). The antibody IV-3 against the Fc␥RII receptor was purchased from Medarex (Annandale, NY), and monoclonal antibody against the integrin ␣ 2 -subunit (6F1) was a kind gift of Dr. Barry Coller, Mt. Sinai School of Medicine, New York. The fibrinogen-derived peptide HHLGGAKQAGDV (␥ 400 -411) was kindly provided by the department of Biochemistry, University Utrecht. All other chemicals used were of analytical grade.
LDL Isolation-Fresh, nonfrozen plasma from four healthy subjects each containing less than 100 mg lipoprotein(a)/liter was pooled, and LDL (density range 1.019 -1.063 kg/liter) was isolated by sequential flotation in a Beckman L-70 ultracentrifuge (26). To prevent lipid modification and bacterial contamination, 0.25 mM PMSF, 0.2 mM thimerosal, 2 mM NaN 3 , and 4 mM EDTA (final concentrations) were present during the first run (20 h, 175,000 ϫ g, 10°C). Subsequent centrifugations (20 h, 175,000 ϫ g, 10°C) were carried out in the absence of additives except for NaN 3 and EDTA. LDL was filtered through a 0.45-m filter (Millipore, Molsheim, France) and subsequently dialyzed against 10 3 volumes of 150 mM NaCl containing 1.5 mM NaN 3 and 1 mM EDTA. LDL was stored at 4°C under nitrogen for not longer than 14 days and before each experiment dialyzed overnight against 10 4 volumes 150 mM NaCl. ApoB100 and lipoprotein(a) concentrations were measured using the Behring Nephelometer 100. As described previously (22), these LDL preparations contained only minimal amounts of thiobarbituric acid-reactive substances (0.20 Ϯ 0.07 nmol/mg) and lipid peroxides (6.7 Ϯ 1.9 nmol/mg). The concentrations of these oxidation markers were below or within other reported values for native LDL (27)(28)(29)(30)(31). Lipoprotein(a) concentrations (Apotech, Organon Technika, Rockville, MD) were below 14 mg/liter. LDL preparations were analyzed for possible contamination by fibrinogen (Laurell technique), fibronectin (ELISA) or von Willebrand Factor (ELISA); all concentrations were below detection limits, which were Ͻ50 g fibrinogen, Ͻ50 ng fibronectin, and Ͻ 5 ng von Willebrand Factor per g of B100 protein.
The concentration of LDL was expressed as gram of apoB100 protein/liter.
Platelet Isolation-Freshly drawn venous blood from healthy volunteers was collected with informed consent into 0.1 volume of 130 mM trisodium citrate. The donors claimed not to have taken any medication during 2 weeks prior to blood collection. Platelet-rich plasma was prepared by centrifugation (200 ϫ g for 15 min at 22°C). Gel-filtered platelets were isolated by gel filtration through Sepharose 2B equilibrated in Ca 2ϩ -free Tyrode's solution (137 mM NaCl, 2.68 mM KCl, 0.42 mM NaH 2 PO 4 , 1.7 mM MgCl 2 , and 11.9 mM NaHCO 3 , pH 7.25) containing 0.2% BSA and 5 mM glucose. Platelets were adjusted to a final count of 2 ϫ 10 11 /liter.
Analysis of Phosphorylated and Total p125 FAK -Platelets (10 8 cells) were incubated with LDL or ␣-thrombin as indicated under "Results" and thereafter mixed with ice-cold lysis buffer (1:10 v/v) containing 10% Nonidet P-40, 5% N-octyl glucoside, 10 mM Na 3 VO 4 , 20 mM PMSF, 200 g/ml trypsin inhibitor, 50 mM N-ethylmaleimide, 10 mM EDTA, 1% SDS, and 100 mM benzamidine in Tyrode's solution. Tyrosine-phosphorylated proteins were precipitated using 1 g of 4G10 and protein A-Sepharose (100 l of a 1% suspension of protein A-Sepharose in lysis buffer, previously blocked in 1% BSA, 1 h, 22°C) for 5 h at 4°C. Precipitates were washed five times with lysis buffer and taken up in sample buffer. Proteins were separated by SDS-PAGE using a 7.5% gel and transferred to a nitrocellulose membrane. Phosphorylated p125 FAK was visualized by incubation with anti-p125 FAK mAb (0.25 g/ml, 15 h, 4°C), peroxidase-linked rabbit anti-mouse IgG (1:10,000 v/v, 1 h, 4°C), and enhanced chemiluminescence. In some experiments, as indicated, samples (2 ϫ 10 8 cells) were split in two equal aliquots and precipitated with polyclonal anti-p125 FAK (1.5 g) under the same conditions. Phosphorylated p125 FAK was visualized using 4G10 (0.5 g/ml) and total p125 FAK using anti-p125 FAK mAb. For semiquantitative determination of p125 FAK , the density of the bands was analyzed using Image Quant software.
Patients-Four unrelated patients with Glanzmann's thrombastenia (MAV, CPW, LYON1, and LYON2) were studied. The diagnosis was made on the basis of a markedly prolonged bleeding time (Simplate, Ͼ30 min; normal, Ͻ8 min) and a severe reduction in platelet ␣ IIb ␤ 3 . Patients MAV and CPW have been described previously (22). On a fluorescence-activated cell sorter, 0.3% and 0.4% ␣ IIb ␤ 3 -positive platelets were detected, respectively. Platelets from patients LYON1 and LYON2 (32) were obtained with the kind assistance of Dr. Claude Negrier, Hospital Edouard Herriot, Lyon, France. Fig. 1 illustrates that LDL induced a dosedependent increase in phosphotyrosine content of p125 FAK , in unstirred platelet suspensions. When precipitation was performed with an anti-phosphotyrosine antibody followed by SDS-PAGE, significant p125 FAK phosphorylation was observed at an LDL concentration of 0.5 g/liter apoB100 (Fig. 1, A and B). Inversely, when precipitation was performed with an anti-p125 FAK antibody, and blots were probed for phosphotyrosine containing proteins, similar results were obtained (Fig. 1C). Samples from Fig. 1C revealed equal amounts of p125 FAK protein in each lane when the blot was probed with anti-p125 FAK (Fig. 1D).

LDL Induces p125 FAK Phosphorylation in a Dose-and Timedependent Manner-
A rise in phosphorylation of p125 FAK after LDL incubation was observed as soon as 10 s after addition of LDL; activation was maximal at about 2 min. This maximum was sustained until 30 min (Fig. 2, A and B) and did not decline until 90 min after stimulation (not shown).
LDL-induced p125 FAK Phosphorylation Is Independent of Integrin ␣ IIb ␤ 3 -P125 FAK phosphorylation in platelet suspensions stimulated with ␣-thrombin and collagen is known to depend on ligand binding to integrin ␣ IIb ␤ 3 and aggregation (9,10,33). In concert with these reports, ␣-thrombin (1 unit/ml) failed to induce p125 FAK phosphorylation in unstirred platelet suspensions, where aggregation was absent (Fig. 3A). Under these conditions, prevention of fibrinogen binding to integrin ␣ IIb ␤ 3 by the fibrinogen ␥-chain-derived dodecapeptide (␥ 400 -411 ) (34) did not have any effect. Under stirred conditions, ␣-thrombin induced a strong phosphorylation of p125 FAK . The phosphorylation was completely abolished by ␥ 400 -411 , which inhibits ␣ IIb ␤ 3 occupancy by fibrinogen that is released by ␣-thrombin-stimulated platelets, which is in agreement with an earlier report (9). Again, LDL induced p125 FAK phosphorylation in the absence (Fig. 3) and presence (not shown) of stirring. Surprisingly, p125 FAK phosphorylation by LDL was insensitive to the presence of ␥ 400 -411 (Fig. 3A), although under these conditions LDL induced a slight but significant ␣-granule secretion (5-10% P-selectin (CD62P) expression compared with 20 M thrombin receptor activating peptide (SFLLRN)) (22). Thus, LDL induced p125 FAK phosphorylation independent of ligand binding to integrin ␣ IIb ␤ 3 and aggregation. To investi-gate whether the phosphorylation of p125 FAK by LDL involved ligand-independent signaling via ␣ IIb ␤ 3 , or completely bypassed the integrin, the experiments were repeated with platelets from patients with Glanzmann's thrombastenia that are deficient in ␣ IIb ␤ 3 . Fig. 3B shows the phosphorylation of p125 FAK by LDL in the absence and presence of ␥ 400 -411 in Glanzmann's platelets, which was the same as seen in normal subjects. In stirred suspensions of Glanzmann's platelets, ␣-thrombin completely failed to induce p125 FAK phosphorylation, both in the absence or presence of ␥ 400 -411 . From these experiments we conclude that LDL is a unique platelet agonist, as it initiates phosphorylation of p125 FAK in unstirred platelet suspensions, independent of integrin ␣ IIb ␤ 3 .
Characterization of Signaling Pathways Involved in LDLinduced p125 FAK Phosphorylation-To investigate whether LDL-induced p125 FAK phosphorylation involved the same sig-naling mechanisms as seen in surface-activated platelets, incubations were performed with agents that interfere with signal processing at the level of surface receptors (Fig. 4A) or intracellular signaling routes (Fig. 4B). As already shown in Fig. 3A, ␥ 400 -411 did not affect p125 FAK phosphorylation. Earlier work has shown that collagen-induced p125 FAK phosphorylation was mediated by integrin ␣ 2 ␤ 1 and was inhibited by the antibody 6F1 against this integrin (16). p125 FAK phosphorylation by immobilized IgG was inhibited by anti Fc␥RII antibody IV-3 (15). None of these antibodies affected LDL-induced FIG. 1. Activation of p125 FAK by different concentrations of LDL. A, platelets were incubated with LDL (10 min, 37°C) at the indicated concentrations. After lysis, phosphotyrosine-containing proteins were precipitated with anti-phosphotyrosine mAb 4G10. Identification of p125 FAK was performed by SDS-PAGE and Western blotting using a mAb against p125 FAK , followed by ECL. The blot shown is representative for four separate experiments. B, densitometric analysis of tyrosine-phosphorylated p125 FAK was performed using Image Quant software. Data were corrected for background intensities without LDL and were expressed as percentage of the intensity at 1 g/liter LDL (100%; open symbol). Data are means Ϯ S.D., n Ն 3; p Ͻ 0.05 for all concentrations at 0.5 g/liter and more. C, platelets were incubated with the indicated LDL concentrations (10 min, 37°C). After lysis, total p125 FAK was precipitated with anti-p125 FAK polyclonal antibody. Identification of tyrosine-phosphorylated proteins was performed by SDS-PAGE and Western blotting using anti-phosphotyrosine mAb 4G10. The blot shown is representative for three separate experiments. D, p125 FAK was precipitated from lysates described in the legend to C, and total p125 FAK was detected with anti-p125 FAK mAb, and bands were identified by ECL.

FIG. 2. LDL activates p125 FAK in a time-dependent manner.
A, platelets were incubated with LDL (1.0 g/liter, 37°C) for the indicated time periods. Tyrosine-phosphorylated proteins were precipitated using 4G10 mAb, and p125 FAK was identified as indicated in the legend to Fig.  1A. B, densitometric analysis of tyrosine-phosphorylated p125 FAK . For each time curve, the p125 FAK band intensity after 10-min incubation with LDL was set at 100% (open symbol). Data were corrected for background intensities without LDL and expressed as means Ϯ S.D., n Ն 3; p Ͻ 0.05 for all incubations at 30 s and more.

FIG. 3. LDL-induced phosphorylation of p125 FAK is independent of platelet integrin ␣ IIb ␤ 3 .
A, platelets were incubated with vehicle or ␥ 400 -411 (200 M, 2 min) and subsequently incubated with LDL (1 g/liter, 10 min, 37°C) or ␣-thrombin (1 unit/ml, 2 min, 37°C), under nonstirred or stirred (900 rpm) conditions as indicated. Phosphotyrosine-containing p125 FAK was identified using 4G10 as a precipitating antibody. The results are representative for seven similar experiments with LDL and three experiments with ␣-thrombin. B, platelets from patients with Glanzmann's thrombastenia were treated as in A. The shown blot is representative for results from four different patients. p125 FAK phosphorylation. Also inhibition of thromboxane A 2 (TxA 2 )-formation by indomethacin had no effect. As it was reported that p125 FAK phosphorylation is regulated via PKC in platelets adherent to immobilized fibrinogen (35) or collagen (15), the studies were repeated in the presence of the PKC inhibitor GF109203X, but p125 FAK phosphorylation remained the same. Also the mitogen-activated protein kinase (MAPK) ERK1/2 (p44/42 MAPK ) has been implicated in integrin-mediated signaling (36,37). Therefore, we tested the effect of the inhibitor PD98059 of MEK, the upstream activator of ERK1/2. The stress-activated p38 mitogen-activated protein kinase (p38 MAPK ) has been implicated in actin reorganization and cell migration in endothelial cells, both processes that are accompanied by focal adhesion formation (38). Inhibition of p38 MAPK was achieved using SB203580 (39). Both inhibitors did not change LDL-induced p125 FAK phosphorylation. In contrast, a strong inhibition was observed in the presence of agents that increase cyclic AMP (cAMP). PGI 2 (prostacyclin), its stable analogue iloprost and the cell-permeable cAMP analogue dibutyryl cAMP (Bt 2 cAMP) all strongly decreased p125 FAK phosphorylation induced by LDL.
Taken together, these data show that LDL-induced phospho-rylation of p125 FAK occurred upstream of or independent from signal transduction through TxA 2 formation, PKC, ERK1/2, and p38 MAPK . However, the strong inhibition by agents that raise cAMP reflects negative regulation by cAMP-dependent processes.
Previous studies have shown that LPA induces phosphorylation of p125 FAK in Swiss 3T3 cells (40) and rat hepatoma cells (41). As activated platelets release LPA in vitro (42) and possibly in vivo, we investigated whether the activating properties of LDL were caused by contamination with LPA. In a concentration that induces cell activation (1 mol/liter, Ref. 43) and 10-fold higher, LPA did not initiate the phosphorylation of p125 FAK . DISCUSSION The present findings show that LDL triggers the phosphorylation of p125 FAK in platelet suspensions independent of integrin ␣ IIb ␤ 3 and aggregation. This is in sharp contrast to phosphorylation by ␣-thrombin, collagen, and costimulation of epinephrine with ADP (9) or LIBS6 Fab fragments (33) in cell suspension, which requires ligand binding to ␣ IIb ␤ 3 and platelet-platelet contact. The requirement for ligand-induced outside-in signaling is illustrated by the absence of p125 FAK phosphorylation under conditions that prevent platelet spreading (10, 11) (reviewed in Ref. 5). LDL induces p125 FAK phosphorylation within seconds after stimulation, reaching a maximum at physiological concentrations of LDL (0.26 -1.23 g of apoB100/ liter) (44). LDL signaling to p125 FAK is not dependent on TxA 2 formation or activation of PKC, p38 MAPK , and ERK1/2. These observations indicate that p125 FAK phosphorylation occurs directly downstream of the LDL receptor.
P125 FAK is phosphorylated independent of ␣ IIb ␤ 3 when platelets adhere to immobilized collagen via integrin ␣ 2 ␤ 1 (14 -16) or to immobilized IgG via the Fc␥RII receptor (15). It is therefore possible that the LDL particle acts as an activating surface that phosphorylates p125 FAK by clustering of membrane receptors. However, antibodies against integrin ␣ 2 ␤ 1 and Fc␥RII, known to block further signal generation to p125 FAK , had no effect. Additionally, LPA did not induce p125 FAK phosphorylation. Another candidate for LDL-induced p125 FAK phosphorylation is the collagen receptor glycoprotein (GP)VI. Collagen signaling through this glycoprotein leads to phosphorylation of p125 FAK (45) and is inhibited by cAMP (46). However, GPVI-mediated p125 FAK phosphorylation is not observed in Glanzmann's platelets or in the presence of the RGDS peptide, such in contrast to the effect of LDL. The nature of the cAMP sensitivity of p125 FAK phosphorylation by LDL remains to be elucidated. A rise in cAMP leads to activation of protein kinase A, that in turn can activate vasodilator-stimulated protein (VASP), a 50-kDa protein that localizes to focal adhesions and regulates actin dynamics (47). Phosphorylation of VASP correlates with a decrease in ␣ IIb ␤ 3 activation and aggregation (48). Thus, cAMP might inhibit LDL signaling to p125 FAK via VASP by preventing cytoskeleton rearrangements.
It has been reported that phosphorylation of p125 FAK on Tyr 397 induces Src activity, leading to phosphorylation of Tyr 407,576,577 , inducing maximal kinase activity of the protein (7), but the importance of p125 FAK activity remains unclear. P125 FAK knockout mice were not viable, and embryonic cells of these mice had a reduced mobility. Surprisingly, the number of focal adhesions in these mice was increased. From these observations, it was proposed that p125 FAK regulates focal contact turnover (49), rather than their formation.
Two major signaling mechanisms downstream of p125 FAK are: (i) p130 CAS associates to one of the two C-terminal prolinerich regions of p125 FAK . This association results in phosphorylation of p130 CAS and subsequent binding of Crk via its SH2 . Phosphotyrosine-containing p125 FAK was identified using 4G10 as a precipitating antibody. Data were analyzed as indicated in the legend to Fig. 1B. Data were corrected for background intensities without LDL and expressed as means Ϯ S.D., n Ն 3. C, platelets were incubated with vehicle, LDL (1 g/liter), or LPA at the indicated concentrations and time periods. Phosphotyrosine-containing p125 FAK was identified using 4G10 as a precipitating antibody. domain (reviewed in Refs. 5 and 6). Subsequently, Crk can associate with the GDP/GTP exchange factors Sos (for Ras) or C3G (for Rap1) (50). (ii) The GTPase regulator associated with FAK (Graf) associates directly to p125 FAK via the proline-rich region of residues Pro 875 -Arg 880 in an SH3 domain-dependent manner. Activation of Graf was induced by phosphorylation mediated by ERK1/2 and was shown to stimulate the GTPase activity of CDC42 and Rho, but not Ras or Rac (51). The implications of CDC42 and Rho activity for stress fiber and focal adhesion assembly (52) led to the proposition that activation of Graf might down-regulate these CDC42 and Rho-mediated cytoskeletal changes (1).
Recently, we observed that LDL induced activation of the small GTPases Rap1 and Ral in platelets. 2 Activation of Rap1 and Ral critically depend on Ca 2ϩ (53,54), and Ral is a putative effector molecule of Rap1 (55). In turn, Ral is thought to be an upstream regulator of a member of the Rac/Rho family, CDC42. This small GTPase is involved in the rearrangement of the cytoskeleton. The strong phosphorylation of p125 FAK triggered by LDL described in the present report might therefore initiate these two pathways: (i) activation of Rap1 and Ral and CDC42 via C3G and (ii) cytoskeleton regulation via Ral-activated Rho, controlled by Graf.
In conclusion, via these two separate mechanisms, LDL might be controlling cytoskeleton rearrangements, thereby targeting several signal transducing proteins to an appropriate site of action, leading to an increased sensitivity to platelet agonists such as ␣-thrombin and collagen.