Networking in the Hemostatic System

Prothrombin activation is a pivotal event in thrombosis and hemostasis because thrombin can mediate fibrin formation and can activate and aggregate platelets. Platelet aggregation depends upon the binding of adhesive proteins to integrin αIIbβ3 on the platelet surface. In the present study, a novel interface between the blood coagulation system and platelets is demonstrated by showing that 1) prothrombin binds to αIIbβ3 and 2) this interaction accelerates prothrombin activation. Prothrombin bound to purified αIIbβ3 in a specific, saturable, and divalent cation-dependent manner. This interaction was inhibited by certain monoclonal antibodies to αIIbβ3, by the αIIbβ3 ligands fibrinogen and RGD peptides, but not by thrombin or unrelated proteins. Prothrombin also interacted with αIIbβ3 on resting and stimulated platelets as demonstrated by soluble ligand binding and platelet adhesion assays. Activation of prothrombin by Factor Xa alone or Factor Xa-Va was accelerated by αIIbβ3, and this enhancement was blocked by a monoclonal antibody that inhibited prothrombin binding to the receptor. Taken together, these data identify a previously unrecognized linkage between platelets and the blood coagulation system that may have a significant regulatory consequence.

The blood coagulation cascade and platelets collaborate to mediate thrombus formation and maintain homeostasis. Both arms of the hemostatic system must be functional to prevent excessive blood loss at sites of vascular injury. Linkages between the platelet and coagulation pathways are multifaceted and necessary to achieve efficient thrombus formation. Several examples of such networking are relevant to the present study. Thrombin, formed in the blood coagulation system, converts fibrinogen to fibrin, the major protein component of the hemostatic plug, and thrombin also is a major physiological and potent activator of platelets. Fibrinogen not only forms fibrin but also aggregates platelets by engaging its receptor, integrin ␣ IIb ␤ 3 . Platelets influence thrombin generation by providing a surface for assembly of the prothrombinase complex, consisting of the zymogen, prothrombin, Factor Xa, Factor Va, and phospholipid.
Recently, Reverter et al. (1) identified a previously unrecognized linkage between the blood coagulation system and plate-lets by showing that certain antiplatelet agents could inhibit thrombin generation. These agents included a chimeric monoclonal antibody (mAb) 1 7E3, which binds to ␣ IIb ␤ 3 (and ␣ V ␤ 3 ), blocks ligand binding to this receptor and is currently being used as an antithrombotic drug in humans (2,3), and peptides containing an Arg-Gly-Asp (RGD) sequence that interact with ␣ IIb ␤ 3 (4) and inhibit the binding of adhesive proteins to the receptor (5)(6)(7). Moreover, thrombin generation in the presence of platelets from patients with Glanzmann's thrombasthenia, a congenital deficiency of ␣ IIb ␤ 3 , was also reduced compared with normal platelets (1,8). These observations suggest that ␣ IIb ␤ 3 can influence prothrombin activation, but the molecular mechanism for this effect was undetermined.
In the present study, we provide a potential basis for the influence of ␣ IIb ␤ 3 on thrombin formation. Noting that prothrombin contains an RGD sequence within its catalytic domain (9), we considered whether the zymogen could be a ligand for the receptor. A specific and saturable interaction of the zymogen but not thrombin itself is demonstrated. Furthermore, binding of prothrombin to the receptor is shown to accelerate thrombin formation. These data identify a novel and potentially important linkage between platelets and the blood coagulation system.

EXPERIMENTAL PROCEDURES
Peptides, Proteins, and Antibodies-GRGDSP peptide was prepared as described (10). The cyclic RGD peptide, KYGC(s-s)RGDWPC(s-s), was synthesized by the same method and cyclized with potassium ferricyanide (11). The peptides were purified to homogeneity by high pressure liquid chromatography using a C18 Vydac column and characterized by amino acid composition.
Bovine serum albumin (BSA; fraction V, crystalline) was purchased from Calbiochem-Novabiochem Corp. (La Jolla, CA). Human prothrombin, purchased from Alexis Corp (San Diego, CA), was more 99% pure as assessed by SDS-PAGE (12). The preparations used contained only one major Coomassie Blue staining band, and this protein reacted with mAb to prothrombin (Biodesign International, Kennebunk, ME) in Western blots. Human thrombin with a specific activity of Ͼ1000 NIH units/mg protein was from Alexis Corp. Factor Xa with specific activity of 1029 plasma equivalent units/mg protein and Spectrozyme TH, the chromogenic substrate used to measure the amidolytic activity of thrombin, were purchased from American Diagnostica Inc. (Greenwich, CT). mAb 7E3 (13) was kindly provided by Dr. Barry S. Coller, (Mt. Sinai School of Medicine, New York, NY); mAb CRC64 was described previously (14); mAb 4F10 was from Dr. Virgil Woods (University of California, San Diego, CA); and mAb LM609 was from Dr. David Cheresh, Scripps Research Institute (La Jolla, CA). FITC goat antimouse IgG was purchased from Zymed Laboratories (South San Francisco, CA).
Purification of Proteins-Fibrinogen was purified from fresh human plasma by differential ethanol precipitation (15). ␣ IIb ␤ 3 was purified from octylglucoside extracts of human platelets by affinity chromatography using a KYGRGDP-Sepharose column followed by elution with excess RGD-containing peptide (4). The preparations used exhibited only two major bands corresponding to the ␣ IIb and ␤ 3 subunits and did not react with mAb LM609, indicative of negligible ␣ V ␤ 3 contamination. For some experiments, thrombin was inactivated by treatment with a 10-fold molar excess of D-phenylalanyl-L-prolyl-arginine chloromethylketone (PPACK). All proteins were Ͼ99% pure as judged by SDS-PAGE and protein staining with Coomassie Brilliant Blue.
Radioiodination-Na 125 I (specific activity ϭ 15-17 mCi 125 I/mg of iodine) from Amersham Life Science, Inc. was used for radioiodination. Fibrinogen and prothrombin were radiolabeled using a modified chloramine-T method (15). The labeled prothrombin was not distinguishable from the unlabeled form upon SDS-PAGE under reducing and nonreducing conditions. Iodinated prothrombin was stored at 4°C and used within 3 days of labeling.
Solid Phase Ligand Binding Assays-The binding of prothrombin and fibrinogen to immobilized ␣ IIb ␤ 3 was performed as described by Charo et al. (16,17) with minor modifications. ␣ IIb ␤ 3 (125 g/ml) was diluted 1:25 in a buffer containing 10 mM Tris, 150 mM NaCl, pH 7.4 (Buffer A) and immobilized onto 96-well microtiter plates (Corning Costar Corp., Cambridge, MA) at 500 ng/well for 18 h at 4°C. The plates were then washed and post-coated with 40 mg/ml BSA overnight at 4°C.
Radiolabeled mAb CRC64 specific for ␣ IIb ␤ 3 (14) was used to quantitate the amount of immobilized receptor per well. At saturation, 5.6 ϫ 10 10 Ϯ 0.4 ϫ 10 10 molecules of CRC64 bound to the ␣ IIb ␤ 3 -coated wells. The functional activity of the immobilized ␣ IIb ␤ 3 was assessed on the basis of 125 I-fibrinogen binding. At saturation, 4 ϫ 10 10 Ϯ 0.32 ϫ 10 10 fibrinogen molecules bound per well. Thus, most of the bound receptor was capable of interacting with fibrinogen. 125 I-Prothrombin was added in Buffer A containing 2 mg/ml BSA and the selected divalent cations (1 mM Ca 2ϩ unless otherwise specified). After incubation for selected times (150 -180 min) at 37°C, wells were washed four or five times with Buffer A, and bound prothrombin was quantitated by counting the bound radioactivity in a ␥-counter. The bound prothrombin was eluted and analyzed by SDS-PAGE, autoradiography, and Western blotting using prothrombin specific mAbs. No difference between nonlabeled prothrombin, labeled prothrombin and eluted prothrombin was found. In some experiments, ␣ IIb ␤ 3 -coated wells were preincubated for 20 min with mAbs, peptides, or proteins prior to the addition of 125 I-prothrombin. Nonspecific binding was measured in the presence of a 40-fold excess of unlabeled ligand. Prothrombin binding to wells, coated by BSA, served as an additional negative control. Data were determined as the means of triplicate or quadruplicate measurements at each experimental point.
Binding of Prothrombin to Washed Platelets-Platelets, isolated by differential centrifugation followed by gel filtration (18), were suspended at 1 ϫ 10 8 /ml in modified Tyrode's buffer, containing 0.1% BSA, 1 mM CaCl 2 or 10 mM EDTA. Platelets were preincubated with mAb 7E3 (20 mg/ml), nonimmune immunoglobulins (20 g/ml), or fibrinogen (1 M) for 20 min, and 125 I-prothrombin was then added at selected concentrations. After 1.5 h at room temperature, platelet-bound ligand was separated by centrifugation through 20% sucrose for 2.5 min at 22°C in Beckman microfuge, and the cell-bound radioactivity was measured in a ␥-counter. For some experiments, the platelets were activated with phorbol myristate acetate (PMA) at 400 nM or with ADP at 10 M for 15 min. Thrombin was used to activate platelets at 0.1 unit/ml, and 3 mM PPACK was added 10 min later.
Flow Cytometry-Isolated platelets were suspended at 1 ϫ 10 8 /ml in modified Tyrode's buffer containing 0.1% BSA, 1 mM CaCl 2 , 3 g/ml prostaglandin I 2 , and 2 mM PPACK (Buffer B) and incubated with prothrombin (100 g/ml), cyclic RGD peptide (50 M), or with no additions for 60 min at room temperature. LIBS I mAb (19) or control mouse IgG was added to a final concentration of 50 g/ml. After 30 min, the platelets were washed by centrifugation in Buffer B, incubated with FITC-goat anti-mouse IgG on ice for 20 min, and then analyzed by flow cytometry. Flow cytometry was performed using a FACScan instrument (Becton Dickinson, San Jose, CA); 10,000 events were recorded; and the data were analyzed using the CellQuest software program (version 1.2).
Platelet Adhesion Assays-Platelets were labeled in platelet-rich plasma with 51 Cr (specific activity 100 cpm/1000 platelets) as described (20). The radiolabeled platelets were then isolated as described above and suspended at 5 ϫ 10 7 /ml in modified Tyrode's buffer containing 0.2% BSA and 1 mM CaCl 2 . The platelet suspension (100 l) was added into microtiter wells, which had been coated with prothrombin at a concentration of 10 g/ml and post-coated with BSA (30 mg/ml). After 90 min at 37°C, the wells were washed four or five times with Tyrode's buffer, and adherent cells were quantitated by counting the bound radioactivity in a ␤-counter. Nonspecific platelet adhesion was defined as the radioactivity bound to wells coated only with BSA, and this value was subtracted in the reported results.
Thrombin Generation Assays-␣ IIb ␤ 3 was immobilized onto 96-well microtiter plates as described above. After post-coating with BSA, prothrombin was added at concentrations of 20 -50 mg/ml. mAb 7E3 (20 g/ml) was used to inhibit prothrombin binding to the receptor, and nonspecific mouse IgG at the same concentration served as control. After selected incubation times in the presence of 1 mM Ca 2ϩ to allow binding to the immobilized receptor, prothrombin activation was initiated by addition of 2 g/ml Factor Xa or a mixture of 2 g/ml Factor Xa and 2 g/ml Factor Va. Thrombin generation was assayed at 37°C by measuring the amidolytic activity of thrombin in samples removed after 12 min for Factor Xa-induced activation or 6 min for Factors Xa-Vainduced activation. These samples (10 l) were added to the prewarmed (37°C) thrombin chromogenic substrate, Spectrozyme TH (100 l), at a final concentration of 0.5 mM in 0.1 M NaCl, 0.05 M Tris/HCl, 0.05% BSA (w/v), 20 mM EDTA, pH 7.9. The release of free chromophore (p-nitroanilide-diacetate) from substrate was monitored as a function of time at 405 nm in a Molecular Devices microplate reader. The means of V max obtained using the Softmax program were used to calculate the thrombin concentration. Thrombin of a known specific activity was used as a calibrator. The means Ϯ S.D. of quadruplicate measurements are expressed as the molar concentration of thrombin generated or as the percentage of increase above samples added to wells with no receptor but otherwise treated in an identical fashion.
Thrombin generation was also quantified in the following manner. Prothrombin was incubated in wells coated with ␣ IIb ␤ 3 in the presence of mAb 7E3 (20 g/ml) or the same concentration of mouse IgG. After 3 h of incubation, Spectrozyme TH was added to each well to a final concentration of 25 mM, and then prothrombin activation was initiated by addition of 2 g/ml Factor Xa. The changes of absorbance were monitored using Molecular Devices microplate reader.
Data Analyses-The means Ϯ S.D. were calculated using the programs Excel and Sigmaplot. For curve fitting the program Sigmaplot was used.

RESULTS AND DISCUSSION
Direct Binding of Prothrombin to ␣ IIb ␤ 3 -In view of the evidence linking thrombin generation and ␣ IIb ␤ 3 on platelets (1,8), we sought to determine whether prothrombin could bind directly to the receptor. For these analyses, prothrombin was radiolabeled to a specific activity of 5 ϫ 10 6 cpm/g. When characterized by SDS-PAGE, the radiolabeled prothrombin migrated as a single major band with the same mobility as nonlabeled prothrombin (Fig. 1B). The direct interaction of 125 Iprothrombin with purified ␣ IIb ␤ 3 from platelets was then evaluated by immobilizing the receptor on microtiter plate wells, a system utilized in numerous studies of this integrin (16,17,21,22). As shown in Fig. 1A, a saturable interaction was observed. At the highest concentration of 125 I-prothrombin added (200 g/ml), a 40-fold excess of nonlabeled prothrombin inhibited binding by Ͼ85%. By comparison, 125 I-prothrombin binding to immobilized BSA was minimal and similar to that observed with ␣ IIb ␤ 3 -coated plates in the presence of excess nonlabeled prothrombin. The stoichiometry of the maximal number of prothrombin molecules specifically bound to ␣ IIb ␤ 3 , defined by substrating the binding to the BSA-coated plates, to the number of immobilized ␣ IIb ␤ 3 receptors, quantitated with 125 I-CRC64, an ␣ IIb ␤ 3 -specific mAb (14), was 1.22 (molar ratio), suggesting that each ␣ IIb ␤ 3 is capable of binding a single prothrombin molecule.
Two separate approaches verified that authentic prothrombin bound to ␣ IIb ␤ 3 . First, the radioactivity bound and eluted from the immobilized receptor had the same mobility as the added prothrombin on SDS-PAGE (Fig. 1B). Moreover, the eluted material reacted with a mAb to prothrombin (Fig. 1B). Second, potential artifacts due to radioiodination were excluded by detecting the binding of prothrombin to ␣ IIb ␤ 3 in an enzyme-linked immunosorbent assay format with a prothrombin mAb. The absorbance developed with the prothrombin mAb was 1.352 with ␣ IIb ␤ 3 -coated wells compared with 0.124 when the prothrombin was added to BSA-coated wells.
Typical of the binding of adhesive ligands to ␣ IIb ␤ 3 , the interaction of prothrombin with the receptor was cationdependent; Ca 2ϩ and Mn 2ϩ supported binding, and EDTA blocked the interaction ( Fig. 2A). Three different mAbs to ␣ IIb ␤ 3 , 7E3, CRC64, and 10E5, which are known to block the binding of adhesive ligands to the receptor (13,14,23), inhibited prothrombin binding to ␣ IIb ␤ 3 (Fig. 2B). Fibrinogen, a major physiological ligand of ␣ IIb ␤ 3 , also inhibited prothrombin binding, but an unrelated protein, transferrin, did not (Fig.  2B). Known peptide ligands of ␣ IIb ␤ 3 , GRGDSP, a potent cyclic RGD peptide (11) and the fibrinogen ␥-chain peptide H12 (not shown) were effective inhibitors of prothrombin binding, whereas a control peptide had no effect (Fig. 2B). Both prothrombin and plasminogen contain kringle domains, but plasminogen did not inhibit prothrombin binding (Fig. 2C). In addition, thrombin did not produce significant inhibition of prothrombin binding. Taken together, this inhibitory profile is similar to that of fibrinogen and other adhesive protein ligands that bind to ␣ IIb ␤ 3 via an "RGD recognition specificity" (22).
Prothrombin Binding to Platelets-The capacity of ␣ IIb ␤ 3 on intact platelets to interact with prothrombin was evaluated. As shown in Fig. 3A, 125 I-prothrombin exhibited dose-dependent binding to washed human platelets; maximal binding was observed at the physiological concentration of 100 g/ml. This binding was cation-dependent; it was supported by Ca 2ϩ and inhibited by EDTA (Fig. 3A). The involvement of ␣ IIb ␤ 3 in the interaction was demonstrated by the capacity of mAb 7E3 to inhibit binding. The extent of inhibition by the mAb was ϳ50% (Fig. 3B). Stimulation of platelets with 400 nM phorbol myristate acetate (Fig. 3B) or 10 M ADP (not shown) did not en-hance prothrombin binding to the platelets. Thrombin stimulation of the platelets did substantially increase 125 Iprothrombin binding to platelets, but the increase was insensitive to ␣ IIb ␤ 3 mAbs and is likely to reflect prothrombin interaction with other platelet constituents, such as newly exposed phospholipids (24,25). Verifying the functional state of the platelet preparations used in these analyses, 125 I-fibrinogen did not bind to the resting platelet preparations but did bind to the platelets stimulated with the various agonists. Moreover, as shown in Fig. 3B, fibrinogen did not inhibit 125 Iprothrombin binding to the nonstimulated platelets, whereas it inhibited 125 I-prothrombin binding to the stimulated cells by ϳ50%.
The direct binding of prothrombin to ␣ IIb ␤ 3 on intact platelets was further verified by the demonstration that this ligandinduced expression of LIBS. LIBS are not expressed by ␣ IIb ␤ 3 when the receptor is unoccupied but are evoked when it is occupied by a ligand with an RGD recognition specificity (19,26). As shown by fluorescence-activated cell sorter analyses (Fig. 3C), prothrombin induced expression of the epitope recognized by the prototypic LIBS mAb, LIBS I. The mean fluorescence intensity of the platelets increased from 17.5 (arbitrary units) in the absence of prothrombin to 40.5 in the presence of prothrombin (100 g/ml). A saturating concentration (50 M) of the cyclic RGD peptide induced a fluorescence signal of 79.6.
Recognition of prothrombin by ␣ IIb ␤ 3 on platelets also was demonstrable in adhesion assays. Platelets, radiolabeled with 51 Cr, adhered specifically to immobilized prothrombin (4.3-fold increase compare with immobilized BSA). mAbs to ␣ IIb ␤ 3 and GRGDSP inhibited adhesion to the immobilized ligand (Fig.  3C). These data indicate that platelets adhere to prothrombin in an ␣ IIb ␤ 3 -dependent manner.
Effect of Prothrombin Binding to ␣ IIb ␤ 3 on Thrombin Generation-A functional consequence of prothrombin binding to ␣ IIb ␤ 3 was demonstrated with respect to its activation to thrombin. Initial rates of thrombin generation were measured in ␣ IIb ␤ 3 -coated plates in the presence of 7E3 or nonimmune mouse IgG. The results of the typical experiment are shown in Fig. 4. mAb 7E3, which inhibits prothrombin binding to ␣ IIb ␤ 3 (Fig. 2B), reduced the rate of thrombin generation by 25-30%. As summarized in Table I, blockade of prothrombin binding to ␣ IIb ␤ 3 by 7E3 reduced thrombin generation induced by Factor Xa or by Factor Xa-Factor Va to a similar extent. Accelerated prothrombin activation by ␣ IIb ␤ 3 was also noted when the chromogenic substrate was added directly to the ␣ IIb ␤ 3 -coated wells (not shown).
In these kinetic analyses, only a small fraction (ϳ0.2%; Fig.  1A) of the total prothrombin is directly bound to the receptor when the activators are added. To achieve the observed 25% receptor-mediated acceleration requires a several hundred-fold enhancement in the rate of activation of bound as compared FIG. 3. Interaction of prothrombin with ␣ IIb ␤ 3 on platelets. A, binding of prothrombin to platelets. Isolated platelets were suspended in Tyrode's buffer containing 0.1% BSA at concentration 2 ϫ 10 8 /ml in the presence of 1 mM Ca 2ϩ (q) or 10 mM EDTA (E). 125 I-Prothrombin was added at different concentrations as indicated. After 75 min of incubation at room temperature, platelet-bound ligand will be separated by centrifugation through 20% sucrose, and the cell-bound radio-activity was measured in a ␥-counter. Values are the means Ϯ S.D. of three representative experiments. B, effect of mAb 7E3 and fibrinogen on prothrombin binding to resting or PMA-stimulated platelets. Prior to adding 125 I-prothrombin at 30 g/ml, platelets were preincubated without any agents, with 7E3 (20 g/ml), or with fibrinogen (1 M) in the presence of 1 mM CaCl 2 (open bars). Platelets were stimulated with 400 nM PMA for 15 min and then preincubated with or without fibrinogen (1 M), and prothrombin at a final concentration of 30 g/ml was added to each sample (solid bars). Data are the means Ϯ S.D. of four representative experiments. C, induction of LIBS by prothrombin binding to platelets. Isolated platelets were incubated without ligands (lines 1 and 2), with prothrombin (100 g/ml, line 3) or with cyclic RGD peptide (50 M, line 4) for 60 min at room temperature. LIBS I (lines 2, 3, and 4) or control mouse IgG (line 1) was added at 50 g/ml for 30 min. After washing, the platelets were incubated with FITC-labeled goat antimouse IgG, and the platelets were analyzed by flow cytometry. D, platelet adhesion to immobilized prothrombin. Platelets were labeled by 51 Cr, gel-filtered, and suspended in Tyrode's buffer containing 0.1% BSA and 1 mM Ca 2ϩ at 5 ϫ 10 7 /ml. The platelet suspension (100 l) was added to microtiter wells coated with prothrombin (10 g/ml) or with BSA and incubated for 120 min in the presence of control IgG, mAb 7E3, and mAb CRC64 (20 g/ml each) or GRGDSP or control peptide (100 M each). Specific adhesion to prothrombin was calculated by subtracting the number of platelets adherent to BSA (in the absence or the presence of inhibitors) from the number of platelets adherent to prothrombin. with free prothrombin (assuming that the rate-determining step in the reaction is the direct activation of prothrombin as opposed to the release of thrombin from the receptor, the cleavage of the synthetic substrate, etc). Moreover, with intact platelets, local generation of thrombin would activate the cells to express many additional binding sites to further accelerate prothrombin activation. Thus, the binding prothrombin to ␣ IIb ␤ 3 has a potential for explosive amplification of thrombin generation.
Why does prothrombin but not thrombin bind to ␣ IIb ␤ 3 ? As a speculation, the RGD sequence at residues 187-189 within the thrombin region of prothrombin may be exposed in the zymogen and mediate binding to ␣ IIb ␤ 3 but may not be surfaceoriented in thrombin (27). This RGD sequence is proximal to the active site residue of thrombin, Ser 195 . Based on the crystal structure of ␣-thrombin (28,29), the side chain of Arg 187 is exposed at the hydrated surface of the molecule, but the carboxylate side chain of Asp 189 lies at the bottom of the S1 specificity pocket (27). Thus, the RGD sequence is unlikely to be available for the interaction with receptors such as the ␤ 3 integrins. However, this segment resides in a different orientation in crystal structure of the prothrombin activation intermediate, prethrombin 2 (30). Therefore, the RGD sequence may be available for interactions with integrins in prothrombin as well as certain intermediate products formed during its conversion to thrombin. In agreement with this hypothesis are data that some cleavage fragments of prothrombin can support RGD-dependent cell adhesion (31), whereas thrombin must be denatured to exhibit such activity (32). Mutational analyses of the RGD sequence in prothrombin offers an approach to test this hypothesis. However, because the RGD sequence resides proximal to the active site of thrombin, the enzymatic potential of such prothrombin mutants may be compromised.
Several corollaries arise from this hypothesis. First, depending upon the exposure of the RGD sequence, certain prothrombin activation intermediates may bind well to ␣ IIb ␤ 3 and others may not. Such differential interactions with ␣ IIb ␤ 3 may influence the pathways of prothrombin activation (33). Second, because several integrins recognize RGD sequences (34), prothrombin binding may not be restricted to ␣ IIb ␤ 3 . Our preliminary studies indicate that prothrombin also can bind to ␣ V ␤ 3 . This particular interaction could have a significant impact on prothrombin activation on the surface of the endothelium. Third, a pathway in which a zymogen, in this case prothrombin, binds to an integrin and is released to the local cell surface environment upon activation, in this case thrombin, to potentially react with other receptors may be a common mechanism for control of protease activity. There is an another example of an integrin that can bind a zymogen and amplify its activation. This scenario applies to Factor X, which binds to ␣ M ␤ 2 (Mac-1, CD11b/CD18) on monocytic/myeloid cells and supports the cell-associated formation of factor Xa (35,36).
In the study of Reverter et al. (1), the acceleration of prothrombin activation by ␣ IIb ␤ 3 on platelets was measured in fibrinogen-depleted plasma, i.e. in the absence of an abundant ligand for ␣ IIb ␤ 3 . Because fibrinogen only binds with high affinity to activated ␣ IIb ␤ 3 , we anticipated that fibrinogen would be an effective competitor of prothrombin binding to activated but not to nonstimulated platelets, and the data supported this prediction. These observations raise the possibility that the circulating platelet may be pre-armed with a small amount of prothrombin bound to ␣ IIb ␤ 3 . This bound prothrombin might serve as a primary target for initial activation by the Factor Xa-Va complex, which assembles on the platelet surface. As thrombin is formed, it is released from ␣ IIb ␤ 3 and can stimulate the platelet to amplify the cell's prothrombin binding capacity by exposing new phospholipid binding sites. Because the ␣ IIb ␤ 3 becomes activated as a result of platelet stimulation, whether unoccupied or occupied by residual prothrombin, it would bind fibrinogen or other available ligands to support the adhesive functions of the platelet. Thus, prothrombin binding to ␣ IIb ␤ 3 may play an important physiological role in the early events of thrombus formation. Consistent with this postulated functional role, the RGD sequence is evolutionarily conserved (37).

TABLE I
Effect of ␣ IIb ␤ 3 on thrombin generation ␣ IIb ␤ 3 -coated wells were preincubated with prothrombin (60 g/ml) in the presence mAb 7E3 or control mouse IgG (20 g/ml, each) for 3 h at 37°C in the presence of 1 mM CaCl 2 . Thrombin generation was initiated by adding Factor Xa (2 g/ml) or Factor Xa ϩ Factor Va (2 g/ml each). After 12 min for Factor Xa-induced activation or 6 min for Factors Xa/Va-induced activation, samples were removed from the incubation mixture and added to a prewarmed chromogenic thrombin substrate, Spectrozyme TH (0.5 mM). The means of V max determined using the program Softmax were used to calculate the thrombin concentration using thrombin of known specific activity for calibration. a Thrombin generated in the presence of control Ig was assigned a value of 100%.