Role of a JAK3-dependent Biochemical Signaling Pathway in Platelet Activation and Aggregation*

Here we provide experimental evidence that identifies JAK3 as one of the regulators of platelet function. Treatment of platelets with thrombin induced tyrosine phosphorylation of the JAK3 target substrates STAT1 and STAT3. Platelets from JAK3-deficient mice displayed a decrease in tyrosine phosphorylation of STAT1 and STAT3. In accordance with these data, pretreatment of human platelets with the JAK3 inhibitor WHI-P131 markedly decreased the base-line enzymatic activity of constitutively active JAK3 and abolished the thrombin-induced tyrosine phosphorylation of STAT1 and STAT3. Following thrombin stimulation, WHI-P131-treated platelets did not undergo shape changes indicative of activation such as pseudopod formation. WHI-P131 inhibited thrombin-induced degranulation/serotonin release as well as platelet aggregation. Highly effective platelet inhibitory plasma concentrations of WHI-P131 were achieved in mice without toxicity. WHI-P131 prolonged the bleeding time of mice in a dose-dependent manner and improved event-free survival in a mouse model of thromboplastin-induced generalized and invariably fatal thromboembolism. To our knowledge, WHI-P131 is the first anti-thrombotic agent that prevents platelet aggregation by inhibiting JAK3.

Here we provide experimental evidence that identifies JAK3 as one of the regulators of platelet function. Treatment of platelets with thrombin induced tyrosine phosphorylation of the JAK3 target substrates STAT1 and STAT3. Platelets from JAK3-deficient mice displayed a decrease in tyrosine phosphorylation of STAT1 and STAT3. In accordance with these data, pretreatment of human platelets with the JAK3 inhibitor WHI-P131 markedly decreased the base-line enzymatic activity of constitutively active JAK3 and abolished the thrombin-induced tyrosine phosphorylation of STAT1 and STAT3. Following thrombin stimulation, WHI-P131treated platelets did not undergo shape changes indicative of activation such as pseudopod formation. WHI-P131 inhibited thrombin-induced degranulation/serotonin release as well as platelet aggregation. Highly effective platelet inhibitory plasma concentrations of WHI-P131 were achieved in mice without toxicity. WHI-P131 prolonged the bleeding time of mice in a dose-dependent manner and improved event-free survival in a mouse model of thromboplastin-induced generalized and invariably fatal thromboembolism. To our knowledge, WHI-P131 is the first anti-thrombotic agent that prevents platelet aggregation by inhibiting JAK3.
Platelet activation via cleavage of the protease-activated receptors by thrombin leads to formation of a platelet-rich thrombus, which can severely impair blood flow to vital organs, including the brain, heart, lungs, and kidneys (1). Since the serine protease thrombin is the most potent activator of platelet-mediated coagulation, a better understanding of the signaling pathways regulating thrombin-induced platelet activation may provide the basis for new and effective strategies for prevention and/or treatment of thromboembolism.
Recent studies have revealed important roles for several protein-tyrosine kinases in platelet physiology (2)(3)(4)(5). Notably, JAK3, 1 a member of the Janus family of protein-tyrosine kinases (6 -8), was shown to be constitutively active in human platelets, but its potential physiologic role in agonist-induced platelet activation or aggregation remains unknown. The purpose of this study was to examine the role of JAK3 in thrombininduced platelet activation and aggregation. Here we show genetic and biochemical evidence that implicates JAK3 as one of the regulators of platelet function. Furthermore, our study uniquely identifies a small molecule chemical inhibitor of JAK3 as a novel antiplatelet agent for prevention of potentially fatal thromboembolic events.
Platelet Aggregation Assays-Platelet-rich plasma (PRP) was purchased from the Memorial Blood Bank (Minneapolis, MN) and used according to the guidelines of the Parker Hughes Institute Human Subjects Committee. The PRP samples were treated with varying concentrations of WHI-P131 for 20 min at 37°C. Control PRP samples were treated with vehicle alone. The treated PRP samples were diluted 1:4 with sterile normal saline, and platelets were stimulated with thrombin (0.1 unit/ml; Chronolog Inc., Philadelphia, PA) under stirring. Platelet aggregation was monitored using the Born method of turbidimetric aggregation in a Whole Blood Platelet Aggregometer (Model 560 dual chamber instrument, Chronolog Inc.) for 5 min. When the aggregating agent is added to the platelet-rich plasma according to this method, the formation of the large platelet aggregate is accompanied by a clearing of the plasma. The IC 50 values for WHI-P131-mediated inhibition of agonist-induced platelet aggregation were calculated by nonlinear regression analysis using GraphPAD Prism Version 2.0 software. In impedance aggregation studies, the increase in electrical impedance caused by adherence and aggregation of platelets on an electrode is measured. For these studies, blood was extracted from Jak3 knockout and control C57BL/6 mice by eye bleeds into tubes containing 15% (v/v) buffer containing 0.8% (w/v) citric acid, 2.2% (w/v) trisodium citrate, and 2.45% (w/v) dextrose and mixed gently to prevent coagulation. Citrated blood was diluted with an equal volume of saline and prewarmed at 37°C for 5 min. The platelet agonist thrombin (0.1 unit/ml) was added at 1 min to induce aggregation. Thrombin-induced platelet aggregation was measured in wild-type and knockout mice (n ϭ 3 for each type) in the whole blood platelet aggregometer.
Immunoprecipitations, immune complex protein kinase assays, and immunoblotting on polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA) using the ECL chemiluminescence detection system (Amersham Pharmacia Biotech) were conducted as described previously (9, 14 -17). For immunoblotting, we used antibodies against phosphotyrosine, JAK3, STAT1, STAT3, phospho-STAT1 (recognizes both STAT1␣ and STAT1␤), and phospho-STAT3 (recognizes both STAT3␣ and STAT3␤) (New England Biolabs Inc., Beverly, MA). Horseradish peroxidase-conjugated sheep anti-mouse and donkey anti-rabbit secondary antibodies were purchased from Transduction Laboratories. Horseradish peroxidase-conjugated sheep anti-goat antibodies were purchased from Santa Cruz Biotechnology. Following electrophoresis, kinase gels were dried onto Whatman No. 3 Filter Paper and subjected to phosphorimaging on a molecular imager (Bio-Rad) as well as autoradiography on film. Similarly, all chemiluminescent JAK3 Western blots were subjected to three-dimensional densitometric scanning with the molecular imager and an imaging densitometer using Molecular Analyst/Macintosh Version 2.1 software following the specifications of the manufacturer (Bio-Rad). A JAK3 kinase activity index was determined by comparing the ratios of the kinase activity in PhosphorImager units (PIU) and density of the protein bands in densitometric scanning units (DSU) with those of the base-line sample using the following FIG. 1. Jak3 knockout mice. The homozygous wild-type Jak3 ϩ/ϩ genotype was documented by detection of a single 720-bp multiplex PCR product, and the homozygous knockout Jak3 Ϫ/Ϫ genotype was documented by detection of a single 620-bp multiplex PCR product. Neg Con, negative control.
formulas: activity index ϭ (PIU of kinase band/DSU of JAK3 protein band) test sample ; and stimulation index ϭ (PIU of kinase band/DSU of JAK3 protein band) test sample /(PIU of kinase band/DSU of JAK3 protein band) base-line control sample .
Serotonin Release-Platelet samples were prepared as previously described (20). Release of serotonin from thrombin (0.1 unit/ml)-stimulated platelets was measured using a serotonin detection kit (Immunotech, Marseilles, France) according to the manufacturer's specifications. Sonicated platelets were used for measurement of the total serotonin content of platelets.
High-resolution Low-voltage Scanning Electron Microscopy (HR-LVSEM)-HR-LVSEM was utilized for topographical imaging of the platelet surface membrane as previously reported (21). Aliquots of human platelets were incubated with 100 M WHI-P131 or vehicle alone for 30 min. Treated platelets were then stimulated with thrombin (0.1 unit/ml) for 10 s. 3% glutaraldehyde was added to stop the reaction. Samples were prepared for HR-LVSEM as previously described (21) and analyzed using a Hitachi S-900 SEM instrument at an accelerating voltage of 2 kV.
Transmission Electron Microscopy (TEM)-Aliquots of human platelets were incubated with 100 M WHI-P131 or vehicle alone for 30 min and then stimulated with thrombin (0.1 unit/ml) for 10 s. Samples were prepared for TEM as previously described (22). Briefly, 0.1% glutaraldehyde was added to stop the reaction. Following a brief centrifugation, the sample pellets were layered with 3% glutaraldehyde for 40 min at room temperature. The samples were then post-fixed in 1% OsO 4 for 1 h at 4°C, rinsed three times in distilled water at room temperature, and dehydrated in a graded ethanol series (25, 50, 75, 90, 95, and 100%) and 100% propylene oxide. The samples were embedded in Embed 812 (Electron Microscopy Sciences, Fort Washington, PA). Silver sections were picked up on mesh grids and stained for 10 min in 1% uranyl acetate and 70% ethanol and for 10 min in Reynold's lead citrate. Sections were viewed in a Jeol 100ϫ electron microscope at 60 kV. True magnifications were determined by photographing a calibration grid at each magnification step on the microscope and using this scale to determine final print enlargements.  STAT3 (B, upper panels), and STAT3 (B, lower panels). C, STAT1 was immunoprecipitated from human platelets stimulated with 0.1 unit/ml thrombin. The immunoprecipitates were subjected to Western blot analysis utilizing antibodies raised against phosphotyrosine (PO 4 Y; upper panel) and STAT1 (lower panel). D, STAT3 was immunoprecipitated from human platelets that were stimulated with 0.1 unit/ml thrombin. The immunoprecipitates were subjected to Western blot analysis utilizing antibodies raised against phosphorylated STAT3 (upper panel) and STAT3 (lower panel). E, JAK3 was immunoprecipitated from platelets that were stimulated with thrombin (0.1 unit/ml) after treatment with vehicle (1% dimethyl sulfoxide (DMSO) in phosphate-buffered saline) or WHI-P131 (100 M). The immunoprecipitates were subjected to quantitative kinase assays (upper panel) and immunoblotting with an anti-JAK3 antibody (lower panel) as described under "Experimental Procedures." F and G, human platelets were pretreated with vehicle or WHI-P131 (100 M) prior to thrombin stimulation. In F, STAT1 was immunoprecipitated from platelets that were stimulated with 0.1 unit/ml thrombin. The immunoprecipitates were subjected to Western blot analysis utilizing antibodies raised against phosphotyrosine (upper panel) and STAT1 (lower panel). In G, whole cell lysates from platelets stimulated with 0.1 unit/ml thrombin were subjected to Western blot analysis utilizing antibodies raised against phosphorylated STAT3 (upper panel) and STAT3 (lower panel). PIU, PhosphorImager units; DSU, densitometric scanning units.

Measurement of Bleeding and Clotting Times in
ments. Mice were placed in a tube holder, and tail bleeding was performed with a 2-mm cut from the protruding tail tip. The tail was placed vertically into 10 ml of normal saline in a 37°C water bath, and bleeding times were determined at 30 min post-intraperitoneal injection of WHI-P131 as previously described (1).
Thromboplastin-induced Thromboembolism Model-4 -6-week-old male ICR mice were treated intravenously with 200 l of vehicle (PBS supplemented with 10% Me 2 SO) or varying doses of WHI-P131 in 200 l of vehicle administered in two bolus injections 30 min (intraperitoneal bolus) and 5 min (intravenous bolus) prior to the thromboplastin challenge. The mice were challenged with 25 mg/kg thromboplastin (Sigma) via an intravenous bolus injection into the tail vein as previously described (23). At the time of thromboembolism-related death after the thromboplastin injection or elective sacrifice at 48 h using ketamine/xylazine, all mice were perfused with PBS, followed by 4% phosphate-buffered Formalin. PBS and Formalin were pumped through the left ventricle of the heart and allowed to exit through a 3-mm incision through the anterior wall of the right ventricle. During necropsy, several selected tissues (brain, heart, liver, and lungs) were harvested, fixed in 10% neutral buffered Formalin, dehydrated, and embedded in paraffin by routine methods for histopathological examination. Glass slides with affixed 6-m tissue sections were prepared and stained with hematoxylin and eosin or Masson's trichrome.

JAK3-dependent Tyrosine Phosphorylation of STAT1 and STAT3 Proteins in Thrombin-stimulated Platelets-
We first set out to examine the effects of thrombin stimulation on the phosphorylation status of STAT1 and STAT3 proteins in platelets from wild-type C57BL/6 mice. Notably, treatment of platelets with 0.1 unit/ml thrombin induced tyrosine phosphorylation of the ␣ isoform (p91) of STAT1 ( Fig. 2A) and the ␤ isoform (p83) of STAT3 (Fig. 2B). Thrombin-induced tyrosine phosphorylation of STAT1 and STAT3 was JAK3-dependent because thrombin stimulation failed to induce tyrosine phosphorylation of these STAT proteins in JAK3-deficient platelets from Jak3 knockout mice (Jak3 Ϫ/Ϫ ). Similarly, stimulation of human platelets with 0.1 unit/ml thrombin enhanced the tyrosine phosphorylation of STAT1 and STAT3 proteins (Fig. 2, C and  D). Pretreatment of human platelets with the JAK3 inhibitor WHI-P131 (100 M) decreased the base-line enzymatic activity of constitutively active JAK3 by 81% as measured by autophosphorylation (Fig. 2E) and abolished the thrombin-induced ty-rosine phosphorylation of STAT1 and STAT3 (F and G). It is noteworthy that at 60 s after thrombin stimulation, the enzymatic activity of JAK3 was reduced by 41% (activity index: 0.35 ϭ 59% of the activity index before thrombin stimulation) (Fig. 2E, first and second lanes). The significance of this observation is currently unknown.
Effects of the JAK3 Inhibitor WHI-P131 on Thrombin-induced Platelet Activation-Activation of platelets after exposure to thrombin is associated with actin polymerization and rapid translocation of the tyrosine kinase SYK (24,25) as well as tubulin to the Triton X-100-insoluble fraction that is associated with the actin filament network. As shown in Fig. 3A, Western blot analysis of the cytoplasmic and Triton X-100soluble and -insoluble fractions from unstimulated platelets confirmed the presence of abundant amounts of actin in the Triton X-100-insoluble fraction and of SYK as well as tubulin in the Triton X-100-soluble (but not -insoluble) fraction. Within 60 s after thrombin stimulation, a significant amount of SYK and tubulin translocated to the membrane-associated cytoskeleton as evidenced by the Western blot detection of SYK and tubulin in the actin-containing Triton X-100-insoluble fractions. Notably, thrombin stimulation also induced the translocation of JAK3, STAT1␣/␤, and STAT3␤ proteins to the Triton X-100-insoluble fraction. As shown in Fig. 3B, pretreatment of platelets with the JAK3 inhibitor WHI-P131 prevented the thrombin-induced relocalization of SYK, tubulin, JAK3, STAT1, and STAT3 to the Triton X-100-insoluble fractions.
Platelet activation after thrombin stimulation was accompanied by marked changes in platelet shape and ultrastructural organization. Topographical imaging of the surface membrane of thrombin (0.1 unit/ml)-stimulated human platelets by HR-LVSEM at 40ϫ magnification showed development of pseudopodious extensions indicative of activation (Fig. 4, A and B). WHI-P131 (100 M) inhibited thrombin-induced pseudopod formation (Fig. 4, C and D). Examination of thrombin-stimulated platelets by TEM at 40,000ϫ magnification showed a rapid shape change from discoidal cells to spheres with pseudopods extending from the surface and coalescence of granules as well as canalicular cisternae in the center of the platelet as a prel- ude to degranulation (Fig. 5, A and B). In contrast, no pseudopods were observed, and the granules remained uniformly dispersed after thrombin stimulation of WHI-P131-treated platelets (Fig. 5, C and D). In accordance with its inhibitory effects on activation-associated shape changes and granule migration in thrombin-stimulated platelets, WHI-P131 inhibited platelet degranulation after thrombin stimulation as evidenced by a markedly reduced amount of serotonin secreted from WHI-P131-treated platelets after thrombin challenge (Fig. 5E). The measured serotonin values in platelet supernatants were 157 Ϯ 26 nM for vehicle-treated control platelets (n ϭ 4), 907 Ϯ 20 nM for vehicle-treated, thrombin-stimulated platelets (n ϭ 4), and 313 Ϯ 19 nM for WHI-P131-treated, thrombin-stimulated platelets (n ϭ 4). Taken together, these results provide unprecedented evidence that JAK3 plays a critical role during the earliest events of thrombin-induced platelet activation.
Role of JAK3 in Thrombin-induced Platelet Aggregation-We next sought to examine the role of JAK3 in thrombininduced platelet aggregation. To this end, we first compared the thrombin-induced aggregatory responses of platelets from wild-type and Jak3 knockout mice. As shown in Fig. 6, the magnitude of the thrombin (0.1 unit/ml)-induced aggregatory response of Jak3 ϩ/ϩ platelets from wild-type mice was greater than that of Jak3 Ϫ/Ϫ platelets from Jak3 knockout mice. In accordance with these results, pretreatment of human platelets with the JAK3 inhibitor WHI-P131 for 30 min inhibited throm-bin (0.1 unit/ml)-induced platelet aggregation in a concentrationdependent fashion, with an average IC 50 value of 1.5 M (Fig. 7,  A and B). By comparison, WHI-P258, a structurally similar compound that does not inhibit JAK3, did not affect the thrombin-induced aggregation of platelets even at 100 M (Fig. 7, A  and C).
Notably, WHI-P131 also improved the survival outcome in a mouse model of thromboplastin-induced generalized and invariably fatal thromboembolism (Fig. 8). In this model, 100% of the challenged mice develop dyspnea, ataxia, and seizures and die within 10 min after the thromboplastin challenge from widespread thrombosis in multiple organs and massive pulmonary thromboembolism. All of the 20 vehicle-treated mice died after the thromboplastin challenge, with a median survival time of 2.5 min. WHI-P131 more than doubled the median survival time and produced an event-free survival outcome of FIG. 4. Effects of WHI-P131 on thrombin-induced shape changes in platelets. HR-LVSEM was utilized for topographical imaging of the platelet surface membrane as previously reported (21). Aliquots of human platelets were incubated with 100 M WHI-P131 or vehicle alone for 30 min and then stimulated with thrombin (0.1 unit/ml). Samples were prepared for HR-LVSEM as previously described (21) and analyzed at an accelerating voltage of 2 kV. A, resting platelets with a discoid appearance and smooth contours; B, vehicle-pretreated control platelets stimulated with thrombin; C, WHI-P131-pretreated unstimulated platelets; D, WHI-P131-pretreated platelets stimulated with thrombin.

FIG. 5. Effects of WHI-P131 on thrombin-induced ultrastructural changes and degranulation in platelets.
A-C, aliquots of human platelets were incubated with 100 M WHI-P131 or vehicle alone for 30 min and then stimulated with thrombin (0.1 unit/ml) for 10 s. Samples were prepared for TEM as described under "Experimental Procedures." Sections were viewed in a Jeol 100ϫ electron microscope at 60 kV. A, TEM images of untreated unstimulated control (CON) platelets with a typical discoid appearance and disperse distribution of granules; B, TEM images of vehicle-treated, thrombin-stimulated platelets with spike-like pseudopodia and coalescence of granules in the center; C, TEM images of WHI-P131-treated unstimulated platelets; D, TEM images of WHI-P131-treated, thrombin-stimulated platelets with the largely discoid appearance of resting platelets; E, serotonin release from platelets stimulated with 1 unit/ml thrombin for 60 s measured using the serotonin detection kit according to the manufacturer's specifications. 30 Ϯ 15% (Fig. 8). The cause of death in WHI-P131-pretreated thromboplastin-challenged mice was generalized thromboembolism. No drug-related toxic lesions were detected in any of the organs of these mice. All of the 20 control mice treated with 80 mg/kg WHI-P131 without a subsequent thromboplastin challenge survived beyond the 48-h observation period without any evidence of impaired health status or bleeding. DISCUSSION In summary, our findings reveal an essential role for JAK3 in thrombin-induced platelet activation and aggregation. As a serine protease, thrombin activates protease-activated receptors 1 and 4 (26) by cleaving the N-terminal portion of the receptor. The cleaved peptide then acts as a tethered ligand that activates the G-protein-coupled receptor independent of receptor cleavage (27). JAK3 may bind to the cytoplasmic Cterminal portion of the protease-activated receptor(s) and play a pivotal role in transduction of the thrombin-induced biochemical signal once the receptor is cleaved. Further studies are needed to decipher the molecular mechanism of JAK3-mediated regulation of platelet function.
WHI-P131 inhibited thrombin-induced tyrosine phosphorylation of STAT1 and STAT3 proteins as well as activationassociated translocation of SYK and tubulin to the Triton X-100-insoluble fraction. In agreement with these results, platelets from JAK3-deficient mice displayed a decrease in thrombin-induced platelet aggregation and tyrosine phosphorylation of STAT1 and STAT3. Following thrombin stimula-FIG. 6. Role of JAK3 in thrombin-induced platelet aggregation. Shown are representative traces of aggregation curves of platelets from Jak3 knockout mice and wild-type (WT) C57BL/6 mice. Thrombin (0.1 unit/ml)-induced platelet aggregation in citrated whole blood was measured by electrical impedance.

FIG. 7. Effects of the JAK3 inhibitor WHI-P131
on thrombin-induced platelet aggregation. A, composite concentration effect curve of WHI-P131. In two independent experiments, triplicate platelet-rich plasma samples were treated with varying concentrations of the JAK3 inhibitor WHI-P131, the parent compound WHI-P258, or vehicle (1% Me 2 SO in phosphate-buffered saline) and then stimulated with thrombin (0.1 unit/ ml). Platelet aggregation was monitored in a platelet aggregometer. Results are expressed as the percent control of thrombin-induced maximum platelet aggregation as a function of the applied WHI-P131 concentration. B and C, representative traces of aggregation curves of platelets treated with WHI-P131 (100 M; shown in B), WHI-P258 (shown in C), or vehicle and then stimulated with thrombin (0.1 unit/ml). Platelet aggregation was monitored in a platelet aggregometer. tion, WHI-P131-treated platelets did not undergo shape changes indicative of activation such as pseudopod formation. WHI-P131 inhibited thrombin-induced degranulation/serotonin release as well as platelet aggregation. Highly effective platelet inhibitory plasma concentrations (Ն10 M) of WHI-P131 were achieved in mice without toxicity. WHI-P131 prolonged the bleeding time of mice in a dose-dependent manner and improved event-free survival in a mouse model of thromboplastin-induced generalized and fatal thromboembolism, involving the lungs, liver, heart, and central nervous system. Thus, this study uniquely identifies WHI-P131 as a novel antiplatelet agent targeting JAK3 for prevention of potentially fatal thromboembolic events. To our knowledge, WHI-P131 is the first anti-thrombotic agent that prevents platelet aggregation by inhibiting JAK3. WHI-P131 is also being developed as an apoptosis-promoting anticancer agent (28). JAK3 inhibitors such as WHI-P131 may be useful as a new class of anticoagulants for treatment of hypercoagulable metastatic cancer patients as well as patients with a primary cardiovascular, cerebrovascular, or hematologic disease at risk for thromboembolic complications.