Thrombin Induces Proteinase-activated Receptor-1 Gene Expression in Endothelial Cells via Activation of Gi-linked Ras/Mitogen-activated Protein Kinase Pathway*

We addressed the mechanisms of restoration of cell surface proteinase-activated receptor-1 (PAR-1) by investigating thrombin-activated signaling pathways involved in PAR-1 re-expression in endothelial cells. Exposure of endothelial cells transfected with PAR-1 promoter-luciferase reporter construct to either thrombin or PAR-1 activating peptide increased the steady-state PAR-1 mRNA and reporter activity, respectively. Pretreatment of reporter-transfected endothelial cells with pertussis toxin or co-expression of a minigene encoding 11-amino acid sequence of COOH-terminal Gαi prevented the thrombin-induced increase in reporter activity. Pertussis toxin treatment also prevented thrombin-induced MAPK phosphorylation, indicating a role of Gαi in activating the downstream MAPK pathway. Expression of constitutively active Gαi2 mutant or Gβ1γ2 subunits increased reporter activity 3–4-fold in the absence of thrombin stimulation. Co-expression of dominant negative mutants of either Ras or MEK1 with the reporter construct inhibited the thrombin-induced PAR-1 expression, whereas constitutively active forms of either Ras or MEK1 activated PAR-1 expression in the absence of thrombin stimulation. Expression of dominant negative Src kinase or inhibitors of phosphoinositide 3-kinase also prevented the MAPK activation and PAR-1 expression. We conclude that thrombin-induced activation of PAR-1 mediates PAR-1 expression by signaling through Gi1/2 coupled to Src and phosphoinositide 3-kinase, and thereby activating the downstream Ras/MAPK cascade.

Endothelial Cell Cultures-Human microvessel endothelial cells (a human dermal microvascular endothelial cell line) were obtained from Dr. Edwin W. Ades (National Center for Infectious Diseases, Center for Disease Control, Atlanta, GA) (24). The cells were grown in endothelial basal medium MCDB-131 supplemented with 10% FBS, 2 mM L-glutamine, and 1 mg/ml hydrocortisone. Human pulmonary artery endothelial cells (HPAEC) obtained from Clonetics Corp. (San Diego, CA) were grown in EBM-2 medium supplemented 10% FBS. Cells were cultured on tissue culture dishes coated with 0.1% gelatin. Cells were used between passages 4 and 8.
Northern Blot Analysis-Endothelial cells were incubated in serumfree DMEM for 2 h prior to agonist exposure for the indicated times. Total cellular RNA was extracted using TRIzol reagent per manufacturer's instructions. Briefly, following treatment, cells were lysed using 2 ml of TRIzol reagent/100-mm 2 tissue culture dish. Chloroform (0.2 volume) was added to cell lysate, and the aqueous phase was collected by centrifugation. The total cellular RNA from the aqueous phase was precipitated with an equal volume of isopropanol for 1-2 h at Ϫ70°C. After centrifugation, the RNA pellet was collected, washed with 75% ethanol, dried, and dissolved in diethyl pyrocarbonate-treated water containing 0.5% SDS. RNA was quantified by spectrophotometry. Approximately equal amounts of total cellular RNA was separated on a denaturing 1% formaldehyde-agarose gel. After separation, the RNA was transferred to Duralose membranes and immobilized. The membranes were prehybridized for 1 h at 68°C in Quickhyb solution and hybridized for 2 h at 68°C with a 1.36-kb thrombin receptor cDNA probe (provided by Dr. S. R. Coughlin, University of California, San Francisco) and 1.1-kb glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The cDNA probes were labeled with [ 32 P]dCTP using a Prime-It II kit (Stratagene). The membranes were washed and visualized by autoradiography. Autoradiograms were scanned into a personal computer, and levels of PAR-1 and GAPDH mRNA were quantified by NIH Image version 1.6. -Fold increase was determined as PAR-1/ GAPDH ratio) ϫ (untreated PAR-1/GAPDH ratio) Ϫ1 .
Construction of PAR-1 Reporter Plasmid-A 1.82-kb PAR-1 promoter sequence was made by PCR amplification of directly upstream of the initiation codon from an ϳ3.0-kb exon I containing PAR-1 gene 5Јuntranslated region (19). The proximal primer incorporated a HindIII restriction site into the sequence for subcloning purposes (proximal primer 3Ј-aagcttttgtcccgggctctgcgcggcgctgc-5Ј; distal primer 5Ј-aagctttatttaactgggtacttcc-3Ј). The amplified fragment (PCR-1) was inserted into pUC118 for amplification and restriction analysis. This plasmid, PCR-1/pUC118, was digested with KpnI/BglII to release PCR-1 with cohesive ends compatible with KpnI/BamHI-digested pGL2 to form PCR-1/Luc. The sea pansy (R. reniformis) luciferase gene driven by herpes simplex virus-thymidine kinase promoter (TK/pRL) served as an internal control to correct for experimental variation.
Transfections-HMEC, grown in six-well plates to 50 -70% confluence, were incubated with LipofectAMINE-DNA complexes for 2-4 h. LipofectAMINE-DNA complexes were made by incubating 4 g of Li-pofectAMINE with 0.1-1.0 g of plasmid DNA in 0.2 ml of Opti-MEM I for 45 min at 22°C. LipofectAMINE-DNA complexes were diluted with 0.8 ml of Opti-MEM I before being added to HMEC, prewashed two times with Opti-MEM I, for 2-4 h. To end transfection, 2 ml of MCDB 131 medium supplemented with 10% FBS was added to each well. We determined the transfection efficiency in HMEC using green fluorescent protein (GFP) plasmid, which encodes a mutated form of the GFP gene driven by a cytomegalovirus immediate early promoter. HMEC transfected with GFP plasmid were visualized by fluorescence microscopy with a blue filter (peak excitation 490 nm). Approximately 30% of HMEC appeared green 24 h after being transfected with GFP plasmid DNA.
Dual Luciferase Reporter Assay-After transfection, the cells were incubated in growth medium for 20 -24 h. Cells were then incubated in serum-free DMEM for 2 h prior to addition of agonist. Following stimulation, cells were lysed (according to manufacturer's instructions) and 20 l of lysate (500 l total) was assayed for reporter gene expression. Firefly (P. pyralis) and sea pansy (R. reniformis) luciferase activity were assessed by the Dual Luciferase Reagent Assay System. Protein concentrations were determined using Bio-Rad reagents.
p44/42 MAPK Assay-MAPK activity was measured by using the MAPK assay kit from New England Biolabs, Inc. (Beverly, MA). Endothelial cell monolayers in 100-mm 2 culture dishes were treated with or without agonist for the indicated times at 37°C. Cells were washed three times with phosphate-buffered saline and lysed with lysis buffer (20 mM Tris-HCl, pH 7.5, containing 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1.0% Triton X-100, 1 mM Na 3 VO 4 , 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerol phosphate, 2 g/ml leupeptin, 2 g/ml pepstatin A, 2 g/ml aprotinin, and 44 g/ml PMSF) for 10 min at 4°C. Cells were scraped from dishes, and lysates were transferred to microcentrifuge tubes and incubated 4°C for 30 min. Insoluble material was removed by centrifugation at 13,000 ϫ g for 15 min, and precleared supernatants were used for immunoprecipitation and in vitro kinase assay according to the manufacturer's instructions. Briefly, lysates were immunoprecipitated with phosphospecific p42/44 MAPK monoclonal antibody at 4°C for 1 h. Protein A-Sepharose beads were added and incubated at 4°C for 2-4 h with gentle rotation. Immunoprecipitates were pelleted, washed twice in lysis buffer, and twice in kinase assay buffer (25 mM Tris-HCl, pH 7.5, 1 mM Na 3 VO 4 , 1 mM ␤-glycerol phosphate, 2 mM dithiothreitol, 10 mM MgCl 2 ). Immune complexes were resuspended in 50 l of kinase assay buffer containing 200 M ATP and 2 g of Elk-1 fusion protein and incubated at 30°C for 30 min. Reactions were terminated by addition of 30 l of SDS sample buffer. Samples were resolved on 12% SDS-polyacrylamide gel, transferred to Duralose membrane, and subsequently incubated with phospho-Elk-1 antibodies. Phosphorylated Elk-1 was determined by the ECL (enhanced chemiluminescence) method.
Immunoprecipitation-Endothelial cell monolayers in 100-mm 2 culture dishes were treated with agonist for the indicated times at 37°C. Cells were washed three times with phosphate-buffered saline and lysed in lysis buffer (50 mM Tris-HCl, pH 7.5, containing 150 mM NaCl, 1 mM EDTA, 0.25% sodium deoxycholate, 1.0% Nonidet P-40, 0.1% SDS, 1 mM Na 3 VO 4 , 1 mM NaF, 2 g/ml leupeptin, 2 g/ml pepstatin A, 2 g/ml aprotinin, and 44 g/ml PMSF) for 30 min at 4°C. Insoluble material was removed by centrifugation (13,000 ϫ g for 15 min) prior to overnight immunoprecipitation with 1 g/ml antibody (as indicated) at 4°C. Protein A-or G-agarose beads were added to each sample and incubated for 1 h at 4°C. Immunoprecipitates were gently washed three times with wash buffer (Tris-buffered saline containing 0.05% Triton X-100, 1 mM Na 3 VO 4 , 1 mM NaF, 2 g/ml leupeptin, 2 g/ml pepstatin A, 2 g/ml aprotinin, and 44 g/ml PMSF). Immunoprecipitated proteins were resolved on SDS-PAGE and transferred to Duralose membrane. The membrane was utilized for immunoblotting with indicated antibodies (described below).
Western Blot Analysis-Endothelial cell lysates or immunoprecipitates were resolved by SDS-PAGE on a 12% separating gel under reducing conditions and transferred to Duralose membrane. Membranes were blocked with (5% dry milk in 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.05% Tween 20) for 2 h at 22°C. Membranes were incubated with indicated primary antibody (diluted in blocking buffer) at 22°C for at least 1 h. Following washes, membranes were incubated at 22°C with horseradish peroxidase-conjugated goat anti-rabbit or mouse antibody. Protein bands were detected by ECL method.
Endothelial Cell Shape Change Assay-The time course of endothelial cell shape change (measured in real time) in response to thrombin was determined according to procedures described by us (25). In brief, HPAEC were grown to confluence on small gold electrodes (4.9 ϫ 10 Ϫ4 cm 2 ). Prior to experiments, monolayers were washed two times with serum-free medium and incubated for 2 h in 1% serum-supplemented culture medium. The small and large electrodes were connected to a phase-sensitive lock-in amplifier. A constant current of 1 A was applied by a 1-V, 4000-Hz AC connected serially to a 1-megohm resistor between the small electrode and the larger counter electrode. The voltage change between small electrode and larger counter electrode was continuously monitored by lock-in amplifier, stored, and processed on a computer. The data are presented as change in resistive (in-phase) portion of the impedance normalized to its initial value at time zero as described (25,26).
Statistical Analysis-Statistical comparisons were made using twotailed Student's t test. Experimental values were reported as mean Ϯ S.E. Differences in mean values were considered significant at p Ͻ 0.05.

Thrombin-induced PAR-1 Activation Increases Steady-state
PAR-1 mRNA-To determine if PAR-1 activation increases the PAR-1 mRNA levels in endothelial cells, Northern blot analysis was performed on total RNA isolated from agonist-stimulated cells (see "Experimental Procedures"). Confluent endothelial cell monolayers were stimulated for the indicated times with PAR-1 agonists, thrombin, or the 14-amino acid PAR-1 tethered ligand, TRP-14 (12). The expression of GAPDH was unaffected by the experimental conditions and was thus used an internal control. In primary cells, the HPAEC, PAR-1/GAPDH ratio increased 2-fold (over untreated controls) after thrombin treatment (Fig. 1A). This increase was sustained through 6 h of thrombin treatment. A similar response was observed with thrombin treatment of HMEC (Fig. 1B). Stimulation of HPAEC with TRP-14 also increased PAR-1/GAPDH 1.9-fold by 6 h (Fig. 1C).
Thrombin-mediated Transactivation of PAR-1 Promoter-To explore the induction of PAR-1 gene, the PAR-1 promoter re-porter plasmid (PAR-1/Luc) and TK/pRL (see "Experimental Procedures") were transiently expressed in HMEC. Cells were challenged with either thrombin or TRP-14 for various times and lysed, and luciferase activity was measured (see "Experimental Procedures"). Luciferase activity significantly increased over basal at 2 h after thrombin stimulation. By 4 -6 h of thrombin exposure, luciferase levels were ϳ3-fold greater than in untreated cells. After 10 h of thrombin exposure, luciferase activity showed a downward trend (Fig. 2). Exposure of reporter plasmid-transfected HMEC with different concentrations of thrombin ranging from 10 nM to 300 nM did not alter the expression of reporter (data not shown). Luciferase activity (Fig. 2) and PAR-1 mRNA levels ( Fig. 1) were increased over a similar time course with maximal gene expression occurring between 3 and 6 h. To determine whether PAR-1 activation itself is sufficient to induce the PAR-1 gene, HMEC transfected with the reporter plasmid were exposed to TRP-14. Luciferase FIG. 3. Thrombin-mediated induction of PAR-1-driven luciferase activity is pertussis toxin-sensitive. HMEC were transfected with luciferase-reporter plasmids (described under "Experimental Procedures") and pretreated overnight with pertussis toxin (100 ng/ml). HMEC were then stimulated with 100 nM thrombin for 4 h and lysed, and luciferase activity was measured. RLU were corrected for milligrams of total protein. Luciferase activity (mean Ϯ S.E.) from HMEC -fold increase is [(RLU/mg protein treated)/(RLU/mg protein untreated)]. Black columns, 100 nM thrombin treatment; white columns, without thrombin treatment. The mean Ϯ S.E. from four experiments repeated in triplicate are shown. Asterisk (*) indicates the difference from thrombin-stimulated control (p Ͻ 0.005). activity increased Ͼ2.5-fold over basal after 4 h of 100 mM TRP-14 stimulation, indicating that activation of endothelial cell surface PAR-1 is responsible for the induction of PAR-1 gene expression.
Pertussis Toxin-sensitive G Proteins Mediate Thrombininduced PAR-1 Gene Expression-To determine the role of G i in thrombin-mediated PAR-1 gene induction, HMEC expressing the reporter construct were pretreated with pertussis toxin (PTX) to uncouple PAR-1 from PTX-sensitive G proteins. PTX pretreatment prevented the increase in luciferase activity by 90%, following thrombin stimulation (Fig. 3). Basal luciferase levels were not altered by PTX pretreatment. Thus, PTX-sensitive G proteins are critically involved in mediating thrombininduced PAR-1 gene expression.
The interface between the G␣ subunits of heterotrimeric G proteins and cytosolic domains of heptahelical receptors are potential targets for interrupting agonist-mediated signaling events (23). 2 To address further the role of G i , we determined whether a G␣ i -peptide antagonist interferes with thrombinmediated PAR-1 expression. We transfected plasmids encoding a peptide corresponding to the COOH terminus of G␣ i to disrupt the G protein-receptor interaction (23). 2 Expression of the G␣ i antagonist peptide inhibited thrombin-induced PAR-1 expression (Fig. 4A). Overexpression of a scrambled peptide of the same amino acid composition did not alter luciferase expression following thrombin treatment. These results combined with the inhibitory effect of PTX on PAR-1 gene expression indicate the involvement of G i -PAR-1 coupling in mediating thrombin-induced PAR-1 gene induction.
Receptors coupled to PTX-sensitive G i proteins have been shown to activate downstream signal transduction pathways in both ␣ and ␤␥ subunit-dependent manner (14,15,27). To address if ␣ or ␤␥ subunits coupling to PAR-1 is responsible for downstream signaling, we co-expressed a constitutively active G␣ i2 mutant (G i2 -Q205L) or ␤␥ subunits with reporter construct in HMEC. Expression of mutant G␣ i2 increased reporter activity 3-4-fold over basal values (Fig. 4B), suggesting the involvement of PTX-sensitive G proteins in the thrombin-induced PAR-1 gene induction. Co-expression of plasmids encoding the G␤1-and G␥2-isoforms along with luciferase-reporter construct induced PAR-1 gene expression in the absence of thrombin treatment (Fig. 4C). However, neither G␤1 nor G␥2 alone induced PAR-1 gene expression. Further, we co-expressed the G␤␥ sequesters such as G protein-coupled receptor kinase 2 (GRK2) and transducin with the luciferase-reporter construct (28,29). Co-expression of either GRK2 or transducin FIG. 4. A, overexpression of peptide antagonists specific for G␣ i -receptor interface prevents thrombin-induced increase in PAR-1 expression. HMEC were co-transfected with the reporter plasmids and 100 ng of pCDNA 3.1 (vector), G-scrambled minigene (G␣scr), or G␣ i minigene (G␣ i ). Cells were stimulated with 100 nM thrombin for 4 h, lysed, and assayed for luciferase activity. Black columns, 4 h of 100 nM thrombin treatment; white columns, without thrombin treatment. Results are mean Ϯ S.E. from five experiments. Asterisk (*) indicates the difference from thrombin-stimulated control (p Ͻ 0.005). B, constitutively active G␣ i2 mutant (G␣ i2 -Q205L) induces PAR-1 expression. HMEC were cotransfected with the reporter plasmids (as described in Fig. 2) and 100 ng of G␣ i2 -Q205L for 4 h. Cells were lysed and assayed for luciferase activity. with the PAR-1 reporter construct markedly reduced thrombininduced PAR-1 expression (Fig. 4D). These results indicate that G␣ i activation mediates thrombin-induced PAR-1 expression by G␤␥ dissociation from the ligand-activated G␣ i.
PTX Prevents p42/44 MAPK Activation-Because MAPK regulates gene transcription (30), we determined if MAPK contributes to the downstream signaling required for PAR-1 gene expression. Thrombin exposure in HMEC caused a rapid and transient phosphorylation of MAPK, which was maximal at 2 min (Fig. 5A). Pretreatment of HMEC with PTX prevented the thrombin-induced MAPK phosphorylation at 2 min (Fig. 5B). However, PTX pretreatment did not inhibit the phosphorylation of MAPK by 12-O-tetradecanoylphorbol-13-acetate. Activation of MAPK was also measured by its ability to phosphorylate Elk-1 in vitro (see details under Experimental Procedures") as the phosphorylation of Ser-383 on Elk-1 by MAPK is required for Elk-1-dependent gene transcription (31). We showed that PTX pretreatment prevented the MAPK-mediated Elk-1 phosphorylation after thrombin stimulation. In contrast, basic fibroblast growth factor-mediated Elk-1 phosphorylation was PTX-insensitive (Fig. 5C). Thus, activation of G i/o mediated the thrombin-induced MAPK phosphorylation and activation of MAPK.
Since an immediate downstream target of Ras is the serine/ threonine kinase Raf, we tested the ability of CAAX-Raf, which constitutively localizes to the cytosolic face of plasma membrane, to induce PAR-1 gene transcription. In this experiment, the luciferase activity increased Ͼ2-fold in HMEC overexpressing CAAX-Raf (Fig. 6A). As a known substrate for Raf kinase is the MAPK kinase, MEK1, the upstream activator of MAPK, we also determined the effects of overexpression of dominant negative MEK1. These results indicated that the dominant negative MEK1 abolished the thrombin-induced PAR-1 gene expression (Fig. 6B). However, the oncogenic MEK1 induced the PAR-1-driven luciferase activity (1.8-fold over untreated) inde- Blots from A and B were stripped and reprobed with anti-Erk1 antibodies. C, to assess MAPK activity, phosphorylated MAPK was immunoprecipitated from HMEC lysates (see "Experimental Procedures") and in vitro kinase assay was performed using an Elk-1 GST fusion protein as a MAPK substrate. Samples were resolved by SDS-PAGE and immunoblotted with a phosphospecific Elk-1 polyclonal antibody. HMEC were stimulated with 50 ng/ml basic fibroblast growth factor for 10 min where indicated. Each immunoblot is representative of three experiments.

FIG. 6. Blocking Ras/Raf/MAP kinase signal transduction cascade prevents thrombin-induced PAR-1 expression.
HMEC were co-transfected with 100 ng of plasmid DNA encoding (A) dominantnegative Ras (Ras S17N), constitutively active Ras (Ras G12V), constitutively active Raf-1 (CAAX-Raf) or (B) transfected with 300 ng of plasmid DNA encoding dominant-negative MEK 1 (MKK1 K97M), constitutively active MEK 1 (MKK1S218E/S222E), and the luciferasereporter plasmids. Cells were grown to confluence, stimulated with thrombin (where indicated) for 4 h, lysed, and luciferase activity was measured. Black columns, 4 h of 100 nM thrombin treatment; white columns, without thrombin treatment. Results are the mean Ϯ S.E. from four experiments of three wells from each group. Inset, lysates from HMEC, transfected with increasing amounts of MKK1 plasmids were probed for MKK1 expression using chicken anti-HA antibodies. Each MKK1 plasmid encodes MKK1-HA chimera. In A, asterisk (*) indicates the significant difference from thrombin stimulated control reporter expression (p Ͻ 0.05). In B, asterisk (*) indicates difference from basal reporter expression (p Ͻ 0.05). pendent of thrombin stimulation, a response similar to the constitutively active mutants of Ras and Raf. Western blot analysis using an anti-hemagglutinin polyclonal antibody demonstrated the relative expression of mutant MEK1 isoforms in HMEC (Fig. 6B, inset). Thus, the Ras/Raf/MAPK signaling pathway is critical in mediating the thrombin-induced expression of PAR-1 gene in endothelial cells.

Inhibition of Protein-tyrosine Kinase Signaling Pathways Prevents Thrombin-induced PAR-1 Gene Expression-Because
tyrosine kinase signaling activates Ras (32), we assessed the role of protein-tyrosine kinase (PTK)-dependent signaling in mediating PAR-1 gene expression in HMEC. Genistein, a PTK inhibitor, was used to investigate PTK-dependent signal transduction. Thrombin-activated MAPK phosphorylation was measured after pretreating HMEC with 30, 100, or 300 M genistein. We showed that genistein decreased the thrombininduced MAPK phosphorylation in a dose-dependent manner (Fig. 7A).
To assess the role of PTK-dependent signaling in PAR-1 gene transcription, HMEC transfected with luciferase-reporter plasmids were pretreated with 100 M genistein for 2 h prior to stimulation with 100 nM thrombin. These results showed that genistein pretreatment inhibited luciferase activity by 80% (Fig. 7B), whereas basal luciferase activity was unaffected by PTK inhibition. Genistein also prevented the thrombin-induced increase in PAR-1 mRNA levels at 3 and 6 h (Fig. 7C), indicating the dependence of PTK signaling in the induction of PAR-1 gene.
Following functional depletion of endothelial cell PAR-1, cellular resensitization to thrombin required 18 h and the response was de novo PAR-1 synthesis-dependent (5). Therefore, we examined a functional consequence of genistein on the protein synthesis-dependent resensitization to thrombin at 18 h. HPAEC were pretreated with 100 M genistein prior to measuring thrombin-induced cell shape change by the electrical resistance method (25) (see "Experimental Procedures"). The drop in transendothelial electrical resistance (a measure of cell retraction) in response to the initial thrombin concentration, which was sufficient to deplete PAR-1 in HPAEC, was genistein-insensitive (Fig. 8, inset); however, the protein synthesis-dependent cell retraction (i.e. the cycloheximide-sensitive response) occurring at 18 h after thrombin challenge (5) was inhibited 70% by genistein (Fig. 8). Therefore, genistein did not prevent the initial thrombin-induced HPAEC shape change, but it prevented the delayed, cycloheximide-sensitive response following PAR-1 depletion.
Shc Phosphorylation following Thrombin Treatment-The activation of Ras by cell surface receptors involves formation of a Shc-Grb2-SOS1 complex that depends on Src homology 2-mediated recognition of phosphotyrosine (33). Src-mediated tyrosine phosphorylation of Shc is essential for this complex formation (34). As phosphorylation of Shc is critical in such proteinprotein interactions (35), we determined the phosphorylation of Shc adapter proteins in HMEC. Phosphotyrosine-enriched proteins were immunoprecipitated from lysates of thrombin-stimulated cells. Phosphorylation of p46, p52, and p66 Shc as detected by anti-Shc immunoblotting increased following 1-2 min of thrombin exposure (Fig. 9A). Pretreatment of HMEC with Src kinase inhibitor PP1 markedly reduced the thrombin-induced tyrosine phosphorylation of Shc isoforms. Fig. 9B shows the expression of all three isoforms of Shc in HMEC. These results indicate that Src-mediated Shc phosphorylation serves as an upstream signaling intermediate contributing to thrombin-activated Src homology 2-dependent protein interactions. Inhibition of Src Kinase Activity Prevents Thrombin-induced MAPK Phosphorylation and Luciferase Expression-Src ki-nase is an upstream regulator of Ras (36), and stimulation of Src kinases by G protein-coupled receptors results in tyrosine phosphorylation of Shc and subsequent p21 ras activation (34,37). We used two distinct inhibitors, herbimycin A and PP1, to assess the role of Src family of PTK in thrombin-induced intracellular signaling. Phosphorylation of MAPK in response to 2 min of thrombin exposure was measured in HMEC pretreated with 1, 10, or 30 nM herbimycin A (Fig. 10A). Herbimycin A concentrations as low as 1 nM inhibited thrombin-induced MAPK phosphorylation to levels comparable to unstimulated cells. PP1 (10 M) also inhibited the thrombin-induced MAPK phosphorylation (data not shown). In addition, pretreatment of HMEC with 10 M PP1 prevented thrombin-mediated increases in luciferase activity (Fig. 10B). Moreover, co-expression of dominant negative Src (c-Src K295M/Y527F) with the FIG. 7. A, genistein prevents thrombin-induced MAPK phosphorylation. HMEC were pretreated with 30, 100, or 300 M genistein for 2 h prior to 2 min stimulation with 100 nM thrombin. Approximately 30 g of total cellular protein was resolved by SDS-PAGE, and Western blot analysis was performed using phosphospecific MAPK and anti-Erk1 antibodies (see "Experimental Procedures"). Results are representative of three experiments. B and C, genistein inhibits thrombin-induced PAR-1 gene expression. B, HMEC were transfected with the luciferasereporter plasmids (as described in Fig. 2). Cells were pretreated with 100 M genistein for 2 h prior to 4 h stimulation with 100 nM thrombin. Cells were lysed, and luciferase activity was measured. Values are shown as mean Ϯ S.E. from four experiments of three wells. Black columns, 4 h of 100 nM thrombin treatment; white columns, without thrombin treatment. Asterisk (*) indicates difference from thrombinstimulated PAR-1 expression (p Ͻ 0.05). C, HPAEC were pretreated with 100 M genistein (or vehicle; 0.05% dimethyl sulfoxide) 2 h prior to thrombin stimulation (25 nM) for indicated times. Total RNA was isolated, and Northern blot analysis was performed. Band intensities were quantified as described in Fig. 1. Ratios of PAR-1 to GAPDH plotted for each autoradiogram are shown. The results shown are mean of three experiments.
reporter plasmid prevented the induction of luciferase expression (Fig. 11A). Western blot analysis demonstrated the relative expression of c-Src in each group (Fig. 11B). Thus, Src kinase phosphorylates Shc to activate downstream signals, and signals the thrombin-induced PAR-1 gene expression.
Phosphoinositide 3-Kinase Activity Is Necessary for Thrombin-induced PAR-1 Gene Expression-PI3Ks, lipid kinases existing in heterodimeric (PI3K␣) and monomeric (PI3K␥) forms, have been implicated in G protein-coupled receptor-mediated signaling (38,39). One isoform, PI3K␥, is stimulated by G␤␥ derived from ligand-activated G protein-coupled receptor (34,(37)(38)(39)(40). This activation recruits PI3K␥ to plasma membrane, which enhances Src-like kinase activity and facilitates activation of Shc-Grb2-Sos-Ras pathway (39,40). We used pharmacologically distinct inhibitors, wortmannin and LY294002, to address the role of PI3K in thrombin-mediated signaling. Both inhibitors prevented thrombin-induced MAPK phosphorylation at 2 min of thrombin stimulation (Fig. 12A). Moreover, wortmannin (100 nM) and LY294002 (100 nM) pretreatment inhibited thrombin-induced increase in reporter activity (Fig. 12B), indicating the involvement of PI3K in PAR-1 gene induction. Thus, the PI3K activity upstream of MAPK is necessary for thrombin-induced signaling of PAR-1 gene expression. DISCUSSION Irreversible activation and degradation of PAR-1 precludes receptor recycling as a means of restoring responsivity to thrombin via the cell surface PAR-1 (10). As PAR-1 is cleaved and targeted to lysosomes for degradation (10), it is necessary that PAR-1 be re-synthesized to replenish functional cell surface receptor. Although endothelial cells contain a preformed cytosolic pool of PAR-1 that translocates within minutes to the cell surface in response thrombin (10,11), this response is independent of de novo receptor synthesis and is rapidly lost once the pre-formed PAR-1 pool has been depleted (5,10). Inhibition of protein synthesis by cycloheximide prevented the full recovery of cell surface PAR-1 after thrombin-induced PAR-1 depletion (5), indicating de novo protein synthesis was required for the response. In the present study, we demonstrate that activation of cell surface PAR-1 by thrombin is itself a critical stimulus for PAR-1 gene expression in vascular endothelial cells. We showed in HPAEC and HMEC that PAR-1 activation increased PAR-1 mRNA and PAR-1 promoter-driven luciferase-reporter activity. Activation of PAR-1 was sufficient to increase PAR-1 mRNA since endothelial cells challenged with PAR-1 activating peptide (TRP-14), also induced PAR-1 gene expression. Therefore, the protein synthesis-dependent PAR-1 resensitization in vascular endothelial cells is attributed to the activation of PAR-1 gene expression.
Studies on the effects of thrombin on PAR-1 mRNA in human glomerular mesangial and human erythroleukemia (HEL) cells (41,42) showed that PAR-1 mRNA was unaffected by thrombin in mesangial cells, whereas PAR-1 mRNA increased within 6 h of thrombin exposure in HEL cells. HEL cells undergo prolonged PAR-1 desensitization following thrombin exposure (43); however, HEL cells, unlike endothelial cells, lack a cytosolic PAR-1 pool (43). Recovery of thrombin sensitivity in HEL cells was associated with new receptor synthesis since cycloheximide prevented the resensitization (43). Results of the present study showing that thrombin activation of PAR-1 in endothelial cells induced PAR-1 gene expression within 3-6 h after thrombin exposure are consistent with the HEL cell experiments (43).
Proteolytic cleavage of the surface receptors PAR-1, PAR-2, PAR-3, and PAR-4 by proteases (i.e. thrombin in the case of PAR-1, PAR-3, and PAR-4 and tryptase or trypsin in the case of PAR-2) unmasks a cryptic ligand that in PAR-1 interacts with the extracellular loop 2 (12, 44 -47). Binding of the tethered ligand to extracellular loop 2 domain stimulates PAR-1, and generates intracellular second messengers (47). PAR-2 and PAR-3 have been shown to be expressed in human umbilical vein endothelial cells (48,49). PAR-2 is trypsin-specific receptor, but it can also be activated by PAR-1 activating peptide (TRP-14) (50). PAR-3 and PAR-4 cannot be activated by TRP-14 (45,46). In the present study, we were unable to show a functionally active PAR-2 in HPAEC since the PAR-2-specific activating peptide (SLIGKV) did not induce either cell retraction or increase in [Ca 2ϩ ] i (data not shown). Therefore, the activation of TRP-14-induced PAR-1 gene expression can only be ascribed to stimulation of PAR-1.
Thrombin activates signaling in endothelial cells by coupling to PAR-1 to heterotrimeric G i and G q (13,51). Our results using PTX to address the role of heterotrimeric G i in activating PAR-1 gene showed that thrombin-induced PAR-1 gene transcription was PTX-sensitive. To explore further the role of PTX-sensitive heterotrimeric G proteins, a minigene encoding a peptide antagonist homologous to the G␣ COOH terminus of G i was co-expressed with the reporter construct in HMEC. We showed that overexpression of this G␣ i peptide inhibited thrombin-induced PAR-1 gene expression. Expression of constitutively active mutant of G␣ i2 or G␤␥ subunits in HMEC was sufficient to transactivate the PAR-1 promoter in absence of thrombin. Further, we showed that co-expression of G␤␥ subunits sequesters such as GRK2 and transducin with PAR-1 promoter construct prevented thrombin-induced expression of PAR-1 in HMEC (28,29). These data demonstrate that G␣ i activation mediates thrombin-induced PAR-1 expression by G␤␥ dissociation from the ligand-activated G␣ i .
We measured the activation of MAPK to ascertain MAPK involvement in the signal transduction pathway downstream of FIG. 12. A, wortmannin and LY294002 inhibit thrombin-induced MAPK phosphorylation. HMEC pretreated for 2 h with wortmannin (30, 100, and 300 nM) (upper) or LY294002 (30, 100, and 300 nM) (lower) were assayed for MAPK phosphorylation following 2 min of 100 nM thrombin stimulation. Blots were stripped and reprobed with anti-Erk1 antibodies. Phospho-MAPK positive control (ϩ). B, PI3K activity is required for thrombin-induced PAR-1 gene expression. HMEC transfected with luciferase-reporter constructs were pretreated with 100 nM wortmannin or 100 nM LY294002 for 2 h prior to 4 h of thrombin exposure. Cells were lysed, and luciferase activity was measured. Black columns, 4 h of 100 nM thrombin treatment; white columns, without thrombin treatment. Values are shown as mean Ϯ S.E. from four experiments of three to six wells in each group. Asterisk (*) indicates difference from the thrombin-stimulated control (p Ͻ 0.05). G i . Phosphorylation of MAPK on threonine and tyrosine residues by MEK is known to stimulate MAPK activity (52). We showed that thrombin maximally activated MAPK phosphorylation within 2 min, and pretreatment of HMEC with PTX prevented the thrombin-induced MAPK phosphorylation. PTX also abolished MAPK-dependent phosphorylation of Elk-1 in an in vitro kinase assay. To delineate the signaling mechanisms conveying the thrombin signal to PAR-1 promoter, we introduced dominant negative (dn) mutants of Ras and MEK1 into HMEC. We showed that overexpressing dn-mutant of Ras or MEK1 interfered with activation of downstream signaling. Expression of dn-Ras (N17Ras) in contrast to dn-MEK1 inhibited PAR-1 gene induction by 50% in response to thrombin challenge. This may be the result of incomplete inhibition of thrombin-mediated Ras activation by N17 Ras expression or alternatively to a Ras-independent signaling mechanism (53). The finding that dn-MEK1 mutant fully inhibited PAR-1 gene transcription following thrombin stimulation implies that MAPK activity is necessary for PAR-1 gene induction since MEK1 is situated directly upstream of MAPK in the Ras/Raf/ MAPK signaling cascade. We also showed that overexpressing the constitutively active mutants of Ras (V12Ras), Raf (CAAX-Raf), and MEK1 (S218E/S222E) resulted in transactivation of the PAR-1 promoter independent of thrombin stimulation.
Genistein, as in the case of PTX, prevents thrombin-induced p21 ras and MAPK activity (22), suggesting the involvement of protein-tyrosine kinase (PTK) activity in the G i -transduced signaling of PAR-1 gene expression. In thrombin-stimulated endothelial cells, we showed that genistein pretreatment prevented MAPK phosphorylation and PAR-1 gene induction, as demonstrated by both reporter assay and mRNA blotting. Genistein has been shown to prevent the de novo protein synthesis-dependent resensitization to thrombin following PAR-1 depletion in endothelial cells (5). Thus, thrombin-induced activation of PAR-1 promoter relies on the PTK-dependent signaling pathway.
PI3K and Src-like tyrosine kinases lie between G i and the downstream Ras/MAPK signaling cascade (14,15). We showed that inhibition of Src activity by PP1 prevented the induction of PAR-1 gene in response to thrombin. In addition, overexpression of dn-Src abrogated PAR-1 gene induction, whereas wild type Src had no effect on thrombin signaling. Previous studies have linked Src-like PTK activity to the formation of Shc-Grb2-SOS1, which in turn activates p21 ras (14,15), and phosphorylation of Shc on tyrosine residues by Src kinase is a prerequisite for Grb2 association and subsequent SOS1 recruitment (34). We showed that thrombin stimulated the rapid tyrosine phosphorylation of Shc adapter proteins and that this was inhibited by Src kinase inhibitor PP1, suggesting that Src can activate Ras/MAPK via this mechanism, and thus signal PAR-1 expression.
Thrombin-induced MAPK phosphorylation and PAR-1 gene expression were sensitive to PI3K inhibition by both wortmannin and LY294002. The p85 subunit of PI3K␣ contains a Src homology 2 domain interacting with Shc-Grb2 complex (39,40), suggesting that activation of PI3K␣ can lead to activation of the downstream PTK pathway, and thereby to PAR-1 gene expression. The present data are consistent with the hypothesis that G␤␥ released from G␣ i activates Src-like PTK and PI3K, and this in turn activates Ras/MAPK signal transduction, and induces PAR-1 gene expression in endothelial cells.
In summary, the present data indicate that thrombin-mediated PAR-1 gene expression in endothelial cells requires the heterotrimeric G i -activated Ras/MAPK signaling pathway. The pathway linking G i to Ras relies on PI3K and Src-like PTK activity as well as Shc-Grb2-SOS1 complex formation. We pro-pose that thrombin-induced PAR-1 gene expression activates the G i -linked Ras/MAPK cascade resulting in transactivation of the PAR-1 promoter and PAR-1 expression.