Effector cell protease receptor-1 is a vascular receptor for coagulation factor Xa.

The binding and assembly of the coagulation proteases on the endothelial cell surface are important steps not only in the generation of thrombin and thrombogenesis, but also in vascular cell signaling. Effector cell protease receptor (EPR-1) was identified as a novel leukocyte cell surface receptor recognizing the coagulation serine protease Factor Xa but not the precursor Factor X. We now demonstrate that EPR-1 is expressed on vascular endothelial cells and smooth muscle cells. Northern blots of endothelial and smooth muscle cells demonstrated three abundant mRNA bands of 3.0, 1.8, and 1.3 kDa. 125I-Labeled Factor Xa bound to endothelial cells in a dose-dependent saturable manner, and the binding was inhibited by antibody to EPR-1. No specific binding was observed with a recombinant mutant Factor X in which the activation site was substituted by Arg196 → Gln to prevent the proteolytic conversion to Xa. EPR-1 was identified immunohistochemically on microvascular endothelial and smooth muscle cells. Functionally, exposure of smooth muscle cells or endothelial cells to Factor Xa induced a 3-fold and a 2-fold increase in [3H]thymidine uptake, respectively. However, receptor occupancy alone is insufficient for mitogenic signaling because the active site of the enzyme is required for mitogenesis. Thus, EPR-1 represents a site of specific protease-receptor complex assembly, which during local initiation of the coagulation cascade could mediate cellular signaling and responses of the vessel wall.

The ordered assembly of proteins of the coagulation cascade on cell surfaces results in greatly enhanced kinetic efficiency of protease function and protease generation as well as protection of proteases from their extracellular inhibitors. Endothelial cells, which constitute the cellular barrier between blood and the smooth muscle cells of the vessel wall, may initiate and assemble functional thrombin-generating cascades following vessel injury. In addition to thrombus generation, proteins in the cascade can directly initiate receptor-coupled cellular signaling and functional responses. In addition to its role in clotting and in the anticoagulation pathways mediated by specific proteolytic conversion of fibrinogen to fibrin and thrombomodu-lin-dependent activation of protein C, thrombin binds to a G-protein coupled cell surface receptor and initiates intracellular signal transduction and cell activation events. In contrast to the thrombin receptor, which is activated by cleavage, EPR-1, 1 the receptor recognizing the protease Factor Xa, possesses high affinity (1), is not cleaved (1), can elicit responses in lymphocytes (2), and can serve as a functional cofactor for the bound protease (1). As a result, the association of the generated proteases with cell surface receptors can initiate such diverse cellular functions as proliferation (3,4), chemotaxis (5), growth factor gene expression and synthesis (6,7), and adhesion molecule expression (8).
EPR-1 was identified as a novel leukocyte cell surface receptor for the coagulation protease Factor Xa but not the precursor zymogen Factor X (1). The cDNA of this cell surface transmembrane receptor has been cloned and found to encode a novel protein (9). Occupancy of EPR-1 on B and T cell lymphocyte subsets by Xa or surrogate monoclonal antibody ligands increased cytosolic free [Ca 2ϩ ] i and enhanced CD3-dependent mononuclear cell proliferation (2). Whether this is the result of direct receptor occupancy has not been clearly delineated. In the present study we addressed the hypothesis that EPR-1 may be expressed on vascular cells and if so might participate in vascular cell signaling following receptor occupancy by Factor Xa generated locally as a consequence of vascular injury, tissue factor expression, and the local thrombogenic response (10,11).

EXPERIMENTAL PROCEDURES
Materials-Tissue culture reagents were obtained from Life Technologies, Inc. The porcine heparin used to culture human umbilical vein endothelial cells was from Sigma. Endothelial cell growth supplement was from Organon Technika Corp. (Rockville, MD). The human arterial smooth muscle cells were from Clonetics Corporation (San Diego, CA). THP-1 cells were from American Type Culture Collection (Rockville, MD). ITS (insulin, transferrin, and selenious acid) was from Collaborative Biomedical Products (Bedford, MA). Recombinant tick anticoagulant protein (TAP) was kindly provided by Dr. George Valsuk (Corvas International, San Diego, CA). Factor Xa, which was inactivated at the active site by reacting with the covalent inhibitor glutamyl-glycylarginyl-chloromethylketone, was obtained from Hematologic Technologies Inc. (Essex Junction, VT).
The radioisotopes 125 I and [ 32 P]dCTP were obtained from Amersham Corp. The [ 3 H]thymidine was from ICN (Costa Mesa, CA). Biospin chromatography columns and Zetaprobe membranes were from Bio-Rad, the random priming kit was from Boehringer Mannheim, and the UV cross-linker was from Stratagene (La Jolla, CA).
Purification of Factors X and Xa-Factor X was highly purified from human plasma as described (18) followed by an additional final ultra purification using a Factor X-specific monoclonal antibody as follows. It was considered necessary to remove traces of Factor VII/VIIa, which were uniformly found in all conventionally purified Factor X preparations. The contamination was readily detected when the rate of Factor Xa generation was analyzed in the presence of relipidated recombinant human TF. Typically, 10 -50 pM Factor VII was detected in 100 nM Factor X, corresponding to contamination of 1-5:10,000. When further purified, as follows, the contamination with Factor VII was Ͻ1:50,000. The pool of Factor X was bound to anti-Factor X monoclonal antibody f21-4.2 immobilized on Ultralink Hydrazide, and the column was washed extensively with 1 M NaCl, 5 mM EDTA, pH 8.0. The bound Factor X was eluted with 2 M guanidine hydrochloride and immediately dialyzed against 10 mM Tris, 140 mM NaCl, pH 7.4. Affinity purified Factor X was homogeneous by SDS-PAGE. Factor Xa was produced from the highly purified Factor X by limited proteolytic activation with Russel's viper venom Factor X activator. The Factor Xa product was again purified on benzamidine-Sepharose according to Krishnaswamy et al. (19).
Expression and Purification of Factor X Arg 196 3 Gln-The initial full-length cDNA for Factor X was kindly provided by Dr. W. Church (20). The mutation was introduced by oligonucleotide directed mutagenesis according to Kunkel (21). The coding sequence DNA was subcloned into pED4 (22) and sequenced in its entirety to detect potential unrecognized mutations in the expression plasmid. The plasmid was transfected into the Chinese hamster ovary cell line DG44. Stable transfectants were selected and grown in large scale serum-free cultures as described previously in detail for the expression of the homologous protein factor VII (23).
Factor X Arg 196 3 Gln was isolated from the serum-free culture supernatant by affinity chromatography on immobilized monoclonal antibody f21-4.2, described above. The protein obtained from this purification was homogenous when analyzed by SDS-PAGE under nonreducing conditions, but under reducing conditions the preparation showed a small amount (5-10%) of unprocessed single chain Factor X. Amino-terminal sequencing of the light chain yielded the sequence predicted from the cDNA, demonstrating proper processing of the signal peptide. The mutant Factor X, upon prolonged incubation with the enzymatic complex of Factor VIIa and TF (TF⅐VIIa), was not converted to Factor Xa, as determined by chromogenic substrate hydrolysis. The mutant Factor X (Arg 196 3 Gln) protein competitively inhibited the activation of wild-type plasma derived Factor X consistent with recognition by the enzymatically active TF⅐VIIa complex.
Cell Culture-Rat smooth muscle cells were isolated from thoracic aorta explants. Cells were cultivated for up to 12 passages only and were routinely grown in medium 199, 10% fetal calf serum, 50 g/ml each of penicillin and streptomycin, 2.5 g/ml Fungizone® (Amphotericin B), and 2 mM glutamine. Human vascular smooth muscle cells were cultivated up to the third passage only and grown in medium 199, 20% fetal calf serum, 50 g/ml each of penicillin and streptomycin, 2 mM glutamine, and 20 g/ml endothelial cell growth supplement. Primary cultures of human umbilical vein endothelial cells were prepared by standard methods (24). Cells were cultivated up to passage 3 only and grown in medium 199, 20% fetal calf serum, penicillin/streptomycin, 2 mM glutamine, Fungizone®, 90 g/ml heparin, and 20 g/ml endothelial cell growth supplement.
Binding Experiments-Human umbilical vein endothelial cells were plated into gelatinized 96-well plates at a density of 60 ϫ 10 4 cells/well and cultured for 24 h to allow for maximum adherence. The cell monolayers were washed twice with PBS (0.14 M NaCl, 0.01 M sodium phosphate, pH 7.3), 5 mM EDTA at 4°C. 125 I-Labeled Factor Xa (IODO-BEAD method of labeling) (1) was added to cells at serial concentrations from 15 to 120 g/ml in the presence of medium 199, 0.5% BSA at 4°C for 1 h. The cell monolayers were then washed four times in PBS, 0.2% BSA, solubilized in 0.2 N NaOH, and counted in a ␥ counter. For the antibody inhibition of binding experiments, a similar protocol was followed with the respective antibodies, which were added in various dilutions for 1.5 h at 4°C at the same time as the addition of the labeled Factor Xa.
Western Analysis-Endothelial and smooth muscle cell monolayers were washed, scraped into PBS, and homogenized in lysis buffer (1% Triton, 0.1% Nonidet P-40, 20 mM Tris, pH 7.4, 150 mM NaCl, 20 g/ml aprotinin, 10 g/ml soybean trypsin inhibitor, 2 mM phenylmethylsulfonyl fluoride, 10 mM benzamidine). Lysis continued at 4°C for 30 min, after which the lysate was clarified by centrifugation and protein concentrations were determined. 5 ϫ sample buffer was diluted to 1 ϫ; samples were run on 7% SDS-PAGE gels applying 220 g of protein/ lane. The protein bands were transferred to Immobilon-P and EPR-1 protein was detected using the ECL protocol and reagents from Amersham Corp. with anti-EPR-1 mAb 9D4 at 4 g/ml.
RNA Analyses-Cell monolayers were washed once in PBS and lysed in guanidinium isothiocyanate. Total RNA was isolated using ultracentrifugation in a cesium chloride gradient. RNA was quantified spectrophotometrically and loaded on a 1% agarose-formaldehyde gel at 20 g/lane. RNA was transferred to Zetaprobe membranes in 10 ϫ SSC buffer (1.5 M NaCl, 0.1 M NaH 2 PO 4, 0.01 M EDTA) by capillary action. RNA was UV cross-linked to the membrane. The cDNA for EPR-1 was labeled with [ 32 P]dCTP using random priming and then purified by centrifugation through a Sephadex G-50 spin column. Blots were prehybridized for 10 min and then hybridized for 16 h at 43°C in 50% formamide, 7% SDS, 120 mM Na 2 HPO 4 , 250 mM NaCl, and 1 mM EDTA. Blots were washed at room temperature for 20 min in 2 ϫ SSC, 0.1% SDS, then for 20 min at 65°C in 0.5 ϫ SSC, 0.1% SDS, and finally for 20 min at 65°C in 0.1 ϫ SSC, 0.1% SDS. The blots were analyzed by exposure to film for 24 h.
Proliferation Assays-Smooth muscle cells were plated overnight in 96-well plates in medium with 10% serum. The cells were then washed in PBS and allowed to become quiescent in medium 199, 1% ITS for 24 h. Cells were then incubated with serum-free medium with or without reagents. After 18 h, 1 Ci of [ 3 H]thymidine was added to each well. After 4 h, cells were washed twice in PBS, trypsinized, harvested, and counted in a scintillation counter. Endothelial cell proliferation assays were similar with the following exceptions. Cells were plated in 5% serum for 24 h. The following day, they were placed in medium with 1% serum, with or without reagents.
Fluorescence-activated Cell Sorter-Flow Cytometry-Human endothelial cells were grown to confluence in T-75 cm 2 tissue culture flasks, washed once in PBS, and then aliquoted in 2 ϫ 10 6 cells/100-ml aliquot. Cells were incubated with anti-EPR-1 mAb B6, 2E1, or normal mouse serum for 30 min at 4°C. Cells were then pelleted by centrifugation, washed twice in PBS, 1% BSA, and resuspended in PBS, 1% BSA. The fluorescein isothiocyanate-labeled second antibody was then added (1: 100, v/v), and cells were covered in aluminum foil and rotated at 4°C for 30 min The cells were washed twice in PBS, 1% BSA and then resuspended in 2% paraformaldehyde, PBS. The samples were reconstituted to 1 ml with PBS, 1% BSA, and analyzed on an EPICS XL flow cytometer. Sizing gates were set to include all nucleated cells, and for each sample, at least 10 4 cells were analyzed.
Statistical Analyses-The data were analyzed by ANOVA. Differences with p Ͻ 0.05 were considered to be significant.

Characteristics of Factor Xa Binding to Vascular Cells-
Radiolabeled Factor Xa was reproducibly (n ϭ 6) observed to bind to human endothelial cells in a dose-dependent, saturable, and specific manner (Fig. 1). Zero-order kinetics were achieved at 20 nM Factor Xa. The K d for the binding was 12.5 nM, and the number of binding sites were calculated (B max ) to be ϳ220,000/ endothelial cell (Fig. 1, inset). Binding of Factor Xa was also observed on smooth muscle cells with a K d of 32 nM and 143,000 binding sites/cell (not shown). A Factor X mutant (Arg l96 3 Gln), which is not converted to the active serine protease Factor Xa by TF⅐VIIa, did not exhibit significant binding to endothelial or smooth muscle cells (Fig. 1), thereby providing supplemental data to support the specificity for Factor Xa.
Receptor Identity-In order to identify the endothelial cell receptor for Factor Xa, monoclonal antibodies recognizing the Factor Xa receptor previously identified on leukocytes, namely EPR-1 (1), were tested for their ability to inhibit binding of 125 I-Factor Xa to endothelial cell monolayers (Fig. 2). Anti-EPR-1 monoclonal antibody B6 diminished specific binding to endothelial cells by approximately 60%, whereas characterized antibodies to annexin II (12,13), TF (14,15), and PECAM (17) failed to block binding of Factor Xa (not shown). In addition, no significant inhibition of Xa binding to endothelial cells was observed with monoclonal antibody OKM1 to Mac-1, which blocks binding of Factor X to leukocyte surface Mac-1 (16) (not shown). To further substantiate endothelial cell surface expression, we demonstrated mAb B6 reactivity with the endothelial cell surface by flow cytometry (Fig. 3).
To determine whether vascular cells can synthesize and express EPR-1, we first analyzed mRNA transcript levels for EPR-1 in endothelial and smooth muscle cells by Northern blot hybridization (Fig. 4). Three EPR-1 transcripts of approximately 1.3, 1.9, and 3.0 kilobases, in agreement with previous observations in THP-1 cells (9), were identified in both cell types. These are in accord with the known differentially spliced transcripts of EPR-1 mRNA (25). Western blot analysis demonstrated a ϳ65-kDa protein present in arterial smooth muscle cell lysates (using anti-EPR-1 mAb B6). In endothelial cells a doublet of ϳ55 kDa (Fig. 5) was found.
To further substantiate the functional expression of EPR-1 by these vascular cells in vivo, immunohistochemical analysis for EPR-1 was carried out (Fig. 6). Anti-EPR antibody strongly reacted with blood vessels in frozen sections of human tonsil (Fig. 6A). Both vascular smooth muscle cells and endothelial cells were immunoreactive. In contrast control mouse IgG 1 did not react (data not shown).
To explore the potential functional impact of docking of Factor Xa with vascular cell EPR-1, in vitro mitogenesis experiments were performed. These experiments were designed to exclude proliferative responses that might result from local generation of thrombin. In Fig. 7, we show a representative experiment that demonstrates the ability of Factor Xa at 50 nM, slightly in excess of the 32 nM K d , to induce 3-fold increased proliferation of arterial smooth muscle cells relative to cells grown in medium alone in an 18-h assay. Factor Xa-induced proliferation was inhibited by a monoclonal antibody (B6) against EPR-1. Notably, addition of the thrombin inhibitor, hirudin, did not diminish Factor Xa-induced mitogenesis, thereby excluding cellular signaling by thrombin via the thrombin receptor. Also, neither very highly repurified Factor X, with Factor VII contamination at Ͻ1:50,000, nor a nonactivable mutant Factor X, neither of which bind or react with EPR-1, influenced these cell proliferation assays.
To determine if the active site of Factor Xa was required for Factor Xa-induced smooth muscle cell mitogenesis, we performed proliferation assays in the presence of recombinant TAP, which blocks Xa conversion of prothrombin to thrombin, or with Factor Xa, which was inactivated at the active site by reacting with the covalent inhibitor glutamyl-glycyl-arginylchloromethylketone. In Fig. 8 we show that Factor Xa (50 nM) increased smooth muscle cell mitogenesis by more than 4-fold relative to untreated cells. However, Factor Xa in the presence of TAP (50 g/ml) or glutamyl-glycyl-arginyl-chloromethylketone (50 nM) did not significantly increase smooth muscle cell mitogenesis, implying that the intact catalytic active site of Factor Xa is required for EPR-1⅐Xa vascular cell signaling. DISCUSSION The macromolecular assembly of the proteins of the coagulation cascade on cell surfaces can result in a diverse set of effects including conversion of zymogens to highly specific proteases, generation of proteolytic fragments of various substrate proteins, platelet and endothelial cell activation via the thrombin receptor, and formation of the fibrin gel. Assembly of these proteases on cell surface receptors results in receptor-protease complexes with greatly enhanced function such as for assembly of the TF⅐VIIa initiation complex (10), the thrombin-thrombomodulin complex (26), the prothrombinase complex Factor Xa-Factor Va (27), and the intrinsic complex factor IXa-Factor VIIIa (28). Less well appreciated has been the docking of proteases on cognate cell surface receptors, a molecular mechanism well recognized for the adhesive proteins and their receptors but also evidenced by the docking of Factor X on the integrin Mac-1 (16), the assembly of protein S and homologous proteins on Tyro 3 (29), and the identification and cloning of EPR-1, a novel high affinity receptor for Factor Xa (1,9). These receptors as well as the assembly of Factor VIIa on tissue factor not only mediate functional molecular assemblies on the cell surface but possess the capability of cellular signaling via their FIG. 5. Western analysis of EPR-1 in human EC and smooth muscle cells. Endothelial and smooth muscle cells were washed, solubilized in homogenization (lysis) buffer, and centrifuged. Supernatants (220 g of protein/lane) were analyzed on 7% SDS-PAGE gels. Following transfer to Immobilon-P, the membrane was reacted with anti-EPR-1 monoclonal antibody B6 at 4 g/ml. Samples were visualized using the ECL detection kit. cytoplasmic domains (2,29,30).
Here, we demonstrate that vascular endothelial cells and smooth muscle cells express EPR-1, a Factor Xa receptor previously identified only on leukocytes. Further, Factor Xa binding to EPR-1 induced arterial smooth muscle cell activation, analogous to previously described engagement of EPR-1 on lymphocytes (2), but only if the active site of the enzyme is intact and functional. This implies that receptor occupancy alone is not sufficient for cellular activation.
Quantitative characterization of the binding of Factor Xa to endothelial cells and smooth muscle cells demonstrated that it is a receptor-mediated event with a K d of ϳ12.5 nM and ϳ220,000 binding sites/cell on endothelial cells (Fig. 1) and 32 nM and ϳ143,000 binding sites/cell on smooth muscle cells (Fig.  1). An inhibitory EPR-1 antibody significantly inhibited specific binding (Fig. 2). Western immunoblotting experiments demonstrated that an antibody to EPR-1 reacted with a ϳ55 K d band (Fig. 5). In contrast, the EPR-1 of vascular smooth muscle cells was ϳ65 kDa. The difference between endothelial and smooth muscle cell EPR-1 may reflect heterogeneity of glycosylation, different cell type-specific homologous isoforms of the receptor, or alternative splicing of EPR-1 mRNA in endothelial cells consistent with previous observations for other cells (31). Consistent with this, Northern blot hybridization detected one major and two minor transcripts of EPR-1 mRNA in both endothelial and smooth muscle cells (Fig. 4). We have also demonstrated EPR-1 immunostaining of vessels within human tissue, introducing the potential for EPR-1 to participate in cellular responses in association with local activation of coagulation cascades, which might therefore elicit smooth muscle cell mitogenesis in vivo.
Recent data suggest that in addition to participating in protease generating molecular pathways, some of these proteins may directly initiate receptor-coupled intracellular signaling and cell activation. Thrombin, in addition to cleaving fibrinogen to fibrin, also proteolytically activates the cell surface thrombin receptor resulting in intracellular signaling via a G-protein-coupled mechanism (32,33). Protein S, an homologous ␥-carboxylated protein, but not a protease, binds to the receptor Tyro 3 resulting in intracellular signals (29), at least under some experimental conditions (34). Similarly, TF initiates a cytosolic Ca 2ϩ signal following binding of the coagulation protease factor VIIa (30). In addition, the product of the gas6 gene, which is related to protein S, a negative co-regulator of the coagulation cascade, potentiates smooth muscle cell proliferation mediated by Ca 2ϩ mobilizing receptors (35).
EPR-1 was originally identified on monocytes and various leukocyte subsets as a binding protein for the protease Factor Xa (1, 2). On these cell types, the assembled EPR-1⅐Xa complex can facilitate activation of prothrombin to thrombin, though weakly (1,36). In addition, occupancy of EPR-1 by Factor Xa or by certain surrogate mAb ligands, increases cytosolic free [Ca 2ϩ ] i in single adherent T cells and co-stimulates lymphocyte proliferation initiated by suboptimal stimulation through the CD3/T cell receptor complex (2). Whether this is a direct result of receptor ligation or signaling is initiated by other means (i.e. patching the receptor) has not been established. Because mitogenesis has been implicated as one of the responses triggered by Factor Xa binding to vascular cells (6,37), we initially hypothesized that a molecular mechanism for the action of Factor Xa might involve direct receptor-ligand assembly of Factor Xa and EPR-1. Factor Xa induced a 3-fold increase in thymidine incorporation, which was inhibited by an antibody to EPR-1. Importantly, very highly purified Factor X and mutant recombinant Factor X, which cannot be proteolytically cleaved to Factor Xa, did not stimulate arterial smooth muscle cell proliferation. This supports our concept that Factor X must be cleaved to Factor Xa before binding to EPR-1 occurs, and the cellular response is evoked by EPR-1⅐Xa formation. Although Gasic et al. (37) demonstrated that Factor X was mitogenic for smooth muscle cells, our results suggest that proteolytic activation of Factor X to Factor Xa is necessary to elicit the cellular response. We have found functionally significant Factor VII contamination of conventional highly purified Factor X. These experiments were conducted with additionally purified Factor X, which has significantly reduced Factor VII levels (Ͻ1:50,000). We attribute differences in experimental results of prior studies to the quality of protein preparations. This interpretation is further supported by the lack of mitogenic effect of the mutant recombinant Factor X where the activation of the sessile bond is was mutated by the replacement of Arg 196 with Gln.
The mitogenicity observed in our studies was directly mediated by Factor Xa but not via thrombin generation. This was confirmed by the inability of the thrombin inhibitor, hirudin, to decrease Factor Xa-induced response-coupled mitogenic activity (Fig. 7). However the intact catalytic active site of Factor Xa is required for EPR-1⅐Xa smooth muscle cell mitogenic signaling because glutamyl-glycyl-arginyl-chloromethylketone or Factor Xa in the presence of TAP did not induce smooth muscle cell mitogenesis (Fig. 8). Binding of Xa to EPR-1 did not require the active site because both Factor Xa and Factor Xa in the presence of TAP bind equivalantly to vascular smooth muscle cells (data not shown). Thus, post-receptor binding signaling events but not EPR-1⅐Xa binding require the catalytic active site of Xa to be functional. It will now be necessary to define the proteolytic events leading to this cellular response. Factor Xa was also mitogenic for endothelial cells, inducing a 2-fold increase in [ 3 H]thymidine incorporation (data not shown). However, the mitogenic response was not blocked by a purified monoclonal antibody directed against EPR-1. Therefore, the mechanism(s) of mitogenic signaling by factor Xa in endothelial cells is not clear and is the subject of further investigation.
Previous reports by Stern and others have documented the ability of factor Xa to bind to vascular cells (6,38,39). Although endothelial cells synthesize Factor V (40), it is unclear whether they express Va on their surface to support the association of factor Xa. Colman and his colleagues were unable to detect Factor V on the surface of adherent monolayers of human endothelial cells; however, they found that injury induced expression of Factor V (41). Endothelial cells grown in human serum, devoid of factor V coagulant activity, expressed approximately 15,000 molecules/cell as measured by a monoclonal antibody directed against factor V/Va (42). The source of this Va, i.e. endothelial cell or serum derived, was unclear; however, the reported number of Factor Va molecules per cell are entirely inadequate to account for the ϳ220,000 sites for Factor Xa binding found in the present studies. Others have shown that although prothrombin activation by endothelial cells is inhibited by anti-Factor V antibodies, the binding of Factor Xa does not appear to be affected (39). The binding of Factor Xa to HepG2 and J82 tumor cells has also been shown to be Factor Va-independent (43).
The specific role played by EPR-1 on endothelial cell prothrombinase complex assembly and prothrombin formation is under investigation. Vascular mechanisms of prothrombin activation have been extensively characterized in the platelet model, in which membrane assembly of Factor V/Va functions as a nonenzymatic co-factor to increase the catalytic activity of factor Xa by 300,000-fold (27). Although the platelet is probably the primary site of coagulation complex assembly in vivo, the endothelium, in vitro, is also capable of forming a prothrombinase complex (44). The catalytic efficiency of prothrombin activation by the endothelium (in cell culture) is similar to platelets (44). Synthetic EPR-1 peptides inhibit prothrombinase activity on endothelial cells (45) and monoclonal anti-EPR-1 antibodies inhibit prothrombinase activity on platelets (46) in a dose-dependent manner (each in the absence of exogenous V/Va). However, the relationship of EPR-1 and factor Va and the relative importance of each to prothrombin formation on the endothelial cell surface in vitro remains to be determined. More importantly, the contribution of the vascular endothelium (relative to the platelet) to prothrombinase complex assembly and coagulation in vivo is also unclear, but the platelet is likely to play a more important role in this regard.
In summary, our data support the hypothesis that EPR-1, a receptor on endothelial cells and vascular smooth muscle cells, can specifically mediate a Factor Xa interaction with a concomitant cellular response. Further experiments are needed to determine the nature of signaling events, which are initiated by the binding of Factor Xa to EPR-1, but because the active site of Xa is required for mitogenic signaling, it is likely that the enzymatic activity of the protease and not receptor occupancy per se is involved in the signaling cascade. In addition, it is recognized that the coagulation and related protease cascades are generated in vivo during inflammatory and hemostatic responses. Binding of locally generated Factor Xa to EPR-1 on vascular cells may generate an "independent accessory signal" to increase arterial smooth muscle cell proliferation and contribute to the early molecular events associated with vascular cell activation that accompanies vascular disease.