Proexosite-1 on Prothrombin is a Factor Va-dependent Recognition Site for the. Prothrombinase Complex

JBC Papers in Press. Published on May 14, 2003 as Manuscript M302707200 Proexosite-1 on Prothrombin is a Factor Va-dependent Recognition Site for the Prothrombinase Complex Lin Chen, Likui Yang and Alireza R. Rezaie* From the Edward A. Doisy Department of Biochemistry and Molecular Biology, Saint Louis University School of Medicine, Saint Louis, Missouri 63104 Running Title: Proexosite-1 on prothrombin The research discussed herein was supported by grants awarded by the National Heart, Lung, and Blood Institute of the National Institutes of Health (HL 68571 to ARR). *Address of Corresponding Author: Alireza R. Rezaie, Ph.D. Department of Biochemistry and Molecular Biology St. Louis University School of Medicine 1402 S. Grand Blvd. St. Louis, MO 63104 Phone: (314) 577-8130 Fax: (314) 577-8156 E-mail: rezaiear@slu.edu 1 Copyright 2003 by The American Society for Biochemistry and Molecular Biology, Inc.

SUMMARY While the contribution of basic residues of exosite-1 to the catalytic function of thrombin has been extensively studied, their role in the specificity of prothrombin recognition by factor Xa in the prothrombinase complex (factor Xa, factor Va, PC/PS vesicles and Ca 2+ ) has not been examined. In this study, we prepared several mutants of prethrombin-1 (prothrombin lacking Gla and Kringle-1 domains) in which basic residues of this site (Arg 35, Lys 36, Arg 67, Lys 70, Arg 73, Arg 75, and Arg 77 in chymotrypsinogen numbering) were individually substituted with a Glu. Following expression in mammalian cells and purification to homogeneity, these mutants were characterized with respect to their ability to function as zymogens for both factor Xa and the prothrombinase complex. Factor Xa by itself exhibited similar catalytic activity toward both the wild type and mutant substrates; however, its activity in the prothrombinase complex toward most of mutants was severely impaired. Further kinetic studies in the presence of Tyr 63 -sulfated hirudin 54-65 peptide suggested that while the peptide inhibits the prothrombinase activation of the wild type zymogen with a K D of 0.5-0.7 µm, it is ineffective in inhibiting the activation of mutant zymogens (K D =2-30 µm). These results suggest that basic residues of proexosite-1 on prothrombin are specific recognition sites for factor Xa in the prothrombinase complex. 2

INTRODUCTION Prothrombin is a vitamin K-dependent serine protease zymogen that is proteolytically converted to thrombin by the prothrombinase complex (factor Xa, cofactor Va, negatively charged phospholipid vesicles and Ca 2+ ) in the final step of the blood clotting cascade (1-6). Factor Xa specifically catalyzes the cleavage of two peptide bonds after the two basic residues, Arg 273 and Arg 322, to convert prothrombin to thrombin (1). Although factor Xa by itself can catalyze the cleavage of both peptide bonds on the substrate, its catalytic efficiency is improved by greater than five orders of magnitude when it is assembled into the prothrombinase complex (1,3). Results of several kinetic studies have indicated that such a dramatic improvement in the rate of prothrombin activation by the prothrombinase complex is derived from an 100-fold decrease in the apparent K m and a greater than 1000-fold enhancement in the k cat of the activation reaction (3,7). The improvement in the apparent K m of the activation reaction is mediated through the Ca 2+ -dependent assembly of both prothrombin and factor Xa on negatively charged membrane surfaces via their N-terminal Gla domains (3,7). However, the improvement in the k cat of the activation reaction is believed to arise from the factor Vamediated protein-protein interaction between factor Xa and prothrombin in the prothrombinase complex (7,8). The mechanism of the factor Va-mediated protein-protein interaction that improves the catalytic efficiency of factor Xa in the prothrombinase complex is under intensive investigation. Based on recent kinetic data, it has been hypothesized that factor Va binding to factor Xa allosterically exposes a secondary binding site exosite remote from the catalytic pocket on the protease that is a specific recognition site for interaction with the prethrombin-2 (prothrombin lacking the Gla and both Kringles-1 and 2 domains) portion of the substrate (9). In support of this hypothesis, it has been demonstrated that factor Va enhances the k cat of both prothrombin and prethrombin-2 activation by factor Xa to a similar extent (9,10). The putative cofactor mediated interaction sites, on either the protease or the substrate, have not been identified. However, an active-site inhibited 3

thrombin and a C-terminal fragment derived from the cleavage of thrombin by chymotrypsin at Trp 148 has been shown to competitively inhibit the activation of prethrombin-2 by factor Xa in the prothrombinase complex (9). Based on such results, it has been hypothesized that the factor Vamediated factor Xa interactive site on prothrombin is located on the protease domain which does not include either the fibrinogen recognition (exosite-1) or the heparin-binding site (exosite-2) (9). Other studies have recently demonstrated that the heavy chain of factor Va contains a binding site for exosite-1 of thrombin and that this site is also present in a low affinity precursor state proexosite-1 on prothrombin (11,12). Thus, an alternative hypothesis for the mechanism of the cofactor function of factor Va is that the cofactor in the prothrombinase complex may provide a binding site for direct interaction with the proexosite-1 of prothrombin (13,14). In support of this hypothesis, it has been demonstrated that the factor Va-mediated acceleration of prothrombin activation by the prothrombinase complex can be specifically inhibited by an exosite-1 specific peptide ligand derived from the C-terminal domain of the leech inhibitor, hirudin (14). The exosite-1 of thrombin plays a pivotal role in the catalytic function of thrombin. This site has several basic residues which can interact with nearly all natural substrates, inhibitors and cofactors of thrombin including fibrinogen, factors V and VIII, PAR-1, heparin cofactor II, and thrombomodulin (TM) 1 (15-19). Previous mutagenesis of basic residues of exosite-1 has been demonstrated to dramatically impair the reactivity of thrombin with all of these macromolecules (20-23). Despite an extensive characterization of basic residues of exosite-1 in thrombin, the contribution of these residues to the specificity of prothrombin recognition by factor Xa in the prothrombinase complex has not been studied. In this study, we substituted all basic residues of this site including: Arg 35, Lys 36, Arg 67, Lys 70, Arg 73, Arg 75, and Arg 77 (chymotrypsinogen numbering (24)) with Glu in individual constructs in prethrombin-1 (prothrombin lacking both the Gla and Kringle-1 domains) and expressed the mutant proteins in mammalian cells as described (25). Following purification to homogeneity, the properties of mutant proteins were analyzed with respect to their ability to function 4

as zymogens for factor Xa in both the absence and presence of factor Va on phospholipid vesicles. It was discovered that while factor Xa activates the wild type and prethrombin-1 mutant zymogens with a comparable rate in the absence of factor Va, the protease exhibits a dramatic catalytic defect toward mutant substrates in the presence of the cofactor. Further kinetic studies in the presence of the proexosite specific peptide, Tyr 63 -sulfated hirudin 54-65 (Hir 54-65 (SO - 3), suggested that the hirudin peptide inhibits the prothrombinase activation of prethrombin-1 with a K D of 0.7 µm. However, the competitive inhibitory effect of the hirudin peptide in the prothrombinase activation of mutants was impaired at varying degrees which correlated well with the extent of impairments observed in the activation of mutant zymogens. These results suggest that basic residues of proexosite-1 are specific recognition sites for factor Xa in the prothrombinase complex. Interestingly, we also discovered that the epidermal growth factor-like domains 4-6 of TM (TM4-6) can inhibit the prothrombinase activation of prethrombin-1 with a K D of 0.5 µm. The possible physiological significance of this finding is discussed. 5

EXPERIMENTAL PROCEDURES Construction and Expression of Mutant Proteins- The expression of wild type prethrombin-1 (prothrombin lacking both Gla and Kringle-1 domains) and prethrombin-2 (prothrombin lacking Gla, Kringle-1, and Kringle-2 domains) by the pnut-pl2 expression/purification vector system in baby hamster kidney (BHK) cells has been described previously (10,26). Prethrombin-1 mutants in the chymotrypsinogen numbering system (24): Arg 35 Glu and Ala (R35E, R35A), Lys 36 Glu (K36E), Arg 67 Glu (R67E), Lys 70 Glu (K70E), Arg 73 Glu (R73E), Arg 75 Glu (R75E) and Arg 77 Glu (R77E) were prepared by PCR mutagenesis methods as described (26). Following confirmation of accuracy of mutations by DNA sequencing, the mutant constructs were expressed in BHK cells using the same expression/purification vector system described above. All derivatives were purified to homogeneity by immunoaffinity chromatography using the Ca 2+ -dependent monoclonal antibody, HPC4, as described (26). TM4-6 was expressed in HEK293 cells and purified to homogeneity as described (27). Human plasma proteins, antithrombin and factors Va and Xa were purchased from Haematologic Technologies Inc. (Essex Junction, VT). Phospholipid vesicles containing 80% phosphatidylcholine and 20% phosphatidylserine (PC/PS) were prepared as described (28). The chromogenic substrate S2238 was purchased from Kabi Pharmacia/Chromogenix (Franklin, OH). The chromogenic substrate N-p-Tosyl-Gly-Pro-Arg-p-Nitroanilide (GPR-pNA) and Tyr 63 -sulfated hirudin 54-65 (Hir 54-65 (SO - 3) were purchased from Sigma (St. Louis, MO). Prethrombin-1 Activation- The initial rate of prethrombin-1 activation by factor Xa was studied in both the absence and presence of factor Va on PC/PS vesicles. In the absence factor Va, the time course of activation of each prethrombin-1 derivative (2 µm) by factor Xa (5 nm) was measured at room temperature in 0.1 M NaCl, 0.02 M Tris-HCl (ph=7.5) containing 0.1 mg/ml bovine serum albumin (BAS), 0.1% polyethylene glycol (PEG) 8000 and 2.5 mm CaCl 2 (TBS/Ca 2+ ). At different time intervals, small aliquots of activation reactions were transferred to wells of a 96-well assay plate 6

containing 20 mm EDTA and the rate of thrombin generation was determined from the cleavage of S2238 (100 µm) at 405 nm by a V max Kinetic Microplate Reader (Molecular Devices, Menlo Park, CA). The concentration of thrombin generated was determined from standard curves prepared from the cleavage rate of S2238 (100 µm) by known concentrations of wild type and mutant thrombins. In the presence of factor Va, the concentration-dependence of prethrombin-1 (0.3-20 µm) activation by human factor Xa (0.05-1 nm) in complex with a saturating concentration of human factor Va (30 nm) was measured on PC/PS vesicles (35 µm) in TBS/Ca 2+. After 1-30 min of incubation at room temperature, the reactions were terminated by addition of EDTA and the initial rate of thrombin generation was measured from the cleavage rate of S2238 as described above. The apparent K m and k cat values for prethrombin-1 activation were calculated from the Michaelis-Menten equation. It was ensured that the factor Va concentration (30 nm) was in excess in all activation reactions. Thus, the factor Va (0.1-10 nm) concentration dependence of activation reactions by factor Xa (0.2 nm) were carried out on PC/PS vesicles (35 µm) in TBS/Ca 2+ using 1 µm wild type or mutant prethrombin-1. The (K d(app) ) values were calculated from hyperbolic dependence of activation rates on the concentrations of factor Va as described (29). In all reactions, it was ensured that less than 10% of prethrombin-1 was activated at all concentrations of the substrates. Prethrombin-1 Activation in the presence of Hir 54-65 (SO - 3)- The inhibitory effect of the hirudin peptide on the kinetics of prethrombin-1 and prethrombin-2 activation by both factor Xa and prothrombinase was studied. Thus, the rate of activation of each prethrombin-1 derivative (1 µm) by factor Xa alone (5 nm) or factor Xa (0.05-1 nm) in complex with factor Va (30 nm) on PC/PS vesicles was monitored in the presence of increasing concentrations of the hirudin peptide in the same TBS/Ca 2+ buffer system. The concentration of thrombin generated in each reaction was calculated from standard curves as described above except that GPR-pNA was used as the chromogenic substrate since the cleavage rate of this substrate has been reported not to be affected by the occupancy of exosite-1 by the hirudin peptide (18). To simplify comparisons of the hirudin peptide 7

dependence of the activation reactions, the data for all activation reactions were normalized to maximal thrombin generation in the absence of the peptide. The dissociation constants (K D ) for the interaction of the hirudin peptide with prethrombin-1 was calculated from equations 1 and 2 as described (14). V obs is the observed initial rate of prethrombin-1 (Pre-1) activation; V lim is the limiting rate at a saturating hirudin peptide (Hir) concentration; V o is the initial rate of activation in the absence of the hirudin peptide; K D is the dissociation constant for the hirudin peptide binding to prethrombin-1; and [Pre-1.Hir] represents the prethrombin-1-hirudin peptide complex concentration. The quadratic binding equation assumes that the concentration of prethrombin-1 in complex with either factor Xa or factor Va is small enough to be neglected under experimental conditions where [Pre-1] o is in excess of the initial concentrations of both the enzyme and the cofactor (14). Cleavage of Chromogenic Substrates- Approximately one mg of each prethrombin-1 derivative was activated to completion and thrombin mutants were purified by cation exchange chromatography on a Mono S column using a linear gradient of 0.1-0.6 M NaCl as described (25). The concentration of thrombin mutants were determined by their absorbance at 280 nm, assuming a molecular weight of 36.6 kda and an extinction coefficient (E 1% 1cm ) of 17.1 and by stoichiometric titrations with a known concentration of antithrombin as described (25). The steady-state kinetics of hydrolysis of S2238 (1.5-100 µm) by thrombin derivatives (0.5 nm) were studied in TBS/Ca 2+ as described (25). The rate of hydrolysis was measured at 405 nm at room temperature in a V max Kinetic Microplate Reader as described above. The K m and k cat values for the chromogenic substrate hydrolysis were calculated from the Michaelis-Menten equation and the catalytic efficiencies were expressed as the ratios of k cat /K m. 8

RESULTS Expression and Purification of Recombinant Proteins- Wild type and mutant prethrombin-1 derivatives were expressed in BHK cells using the pnut-pl2 expression/purification vector system as described previously (26). All recombinant proteins were purified to homogeneity by an immunoaffinity chromatography using the Ca 2+ -dependent monoclonal antibody HPC4 as described (26). Except for elevated K m(app) values, factor Xa in the prothrombinase complex activates both prethrombin-1 and prethrombin-2 with similar and normal V max values (10,30). Thus, these truncated substrates are ideal reagents for probing the extent that protein-protein interactions in the prothrombinase complex contribute to the high specificity of the catalytic reaction. Amidolytic Activity- Since the zymogenic properties of prethrombin-1 mutants by the prothrombinase complex are studied from the initial rate of thrombin generation in an amidolytic activity assay, it was essential to determine the kinetic parameters for the cleavage of the chromogenic substrate S2238 by mutant thrombins. Thus, following activation and purification on a Mono S column, the concentration of the active site of mutants were determined by stoichiometric titration with antithrombin as described (25). The concentration of active enzymes correlated well with the concentration of substrates determined based on their absorbance at 280 nm (within 90-100%). Kinetic parameters for the hydrolysis of S2238 are presented in Table I. With the exception of the K70E mutant, which exhibited a dramatically impaired K m value, all other mutants cleaved S2238 with kinetic constants which were similar to those observed for wild type thrombin. These results strongly suggest that with the exception of K70E, the mutations did not adversely affect the folding, charge stabilizing system or the reactivity of the catalytic pockets of mutant enzymes. The interaction of p-aminobenzamidine with the K70E mutant was also impaired 3-fold (K i = 38 µm and 107 µm for wild type and mutant thrombin, respectively). Thus, the conformation of the P3-P1 binding pocket of the K70E mutant must have been altered. 9

Prethrombin-1 Activation by Factor Xa and the Prothrombinase Complex- The initial rate of activation of prethrombin-1 derivatives by factor Xa was studied both in the absence and presence of factor Va on PC/PS vesicles. As shown in Fig. 1A, factor Xa exhibited similar activity toward both the wild type and mutant zymogens in the absence of factor Va. However, the activation of all mutants by factor Xa in the prothrombinase complex was markedly impaired (Fig. 1B). The concentration dependence of zymogen activation indicated that the prothrombinase complex activated wild type prethrombin-1 with apparent K m and k cat values of 9.1 ± 0.7 µm and 1529 ± 56 mol/min/mol, respectively. In the case of mutants, these values could only be determined for the R75E mutant (17.1 ± 2.8 µm and 913 ± 87 mol/min/mol) since the rate of thrombin generation was dramatically impaired and remained linear for up to 20 µm substrate (the highest concentration available) for all other mutants (Fig. 1B). Since the saturation of the initial rate of activation was not feasible with mutant substrates for the kinetic analysis, the overall extent of impairment with each proexosite-1 mutant residue was estimated from the initial rate of activation of a limiting concentration of each mutant substrate (at least 10-fold below K m values) in the presence of increasing concentrations of factor Va on PC/PS vesicles. A saturable dependence of thrombin generation on factor Va concentrations was observed yielding both K d(app) values for the factor Xa-factor Va interaction and the maximum rate of thrombin generation with all derivatives. The results presented in Table II suggested that while the K d(app) for the enzyme-cofactor interaction was independent of the substrate, the second-order rate of thrombin generation was impaired at varying degrees with all mutant substrates. The most impairment (75-150- fold) was observed with R67E and K70E mutants. The activation of all other mutants was impaired 4-18-fold (Table II). These results clearly suggest that none of the mutant residues under study interact with factor Xa in the absence of factor Va, however, they are critical recognition sites for the cofactor and/or the protease in the prothrombinase complex. 10

Inhibition of Activation by Hir 54-65 (SO - 3)- The Tyr 63 sulfated hirudin peptide is proven to be a useful probe to analyze the interaction of different ligands with exosite-1 of thrombin (15,31). It was recently demonstrated that the hirudin peptide can also interact with the proexosite-1 of the substrate prothrombin (14). Thus, to determine whether the loss of specific interactions of proexosite-1 residues with prothrombinase accounts for the defective catalytic reactions with different prethrombin-1 derivatives, the initial rates of activation of mutant substrates were measured in the presence of increasing concentrations of the hirudin peptide. The results presented in Fig. 2A and Table II indicated that the hirudin peptide inhibits the initial rate of wild type prethrombin-1 activation with a K D of 0.8 µm. However, the ability of the peptide to inhibit activation of prethrombin-1 mutants was impaired at varying degrees. Interestingly, the degree of impairment in K D values correlated well with the degree of impairment in the observed activation rates of mutant substrates. Thus, the hirudin peptide exhibited no inhibitory effect toward the prothrombinase activation of either R67E or K70E mutants which were also ineffective substrates for the activation complex. Parallel with impairment in the activation rates, the K D values were elevated 3-40 fold with all other mutants (Table II). It should be noted that the highest concentration of the hirudin peptide in the prothrombinase assays was 40 µ M (shown only up to 30 µm in Fig. 2), thus a larger uncertainty was associated with the K D values for R73E and K36E and that the activation of both mutants (with the exception of R67E and K70E) was impaired the most. It should be noted that the inhibitory property of the peptide was specific for factor Xa in the presence of factor Va since in the absence of the cofactor no inhibition of thrombin generation was observed (Fig. 2B). It is also known that the fragment-2 domain of prethrombin-1 can enhance the rate of substrate activation by factor Xa in the presence of factor Va (32). This raises the possibility that the hirudin peptide directly or indirectly interferes with the effect of fragment-2 in the activation reaction. However, as shown in Fig. 2B, the inhibitory effect was independent of the fragment-2 since a similar K D of 0.6 ± 0.1 µm was obtained when prethrombin-2 rather than prethrombin-1 was used in the prothrombinase inhibition reaction. 11

The binding of epidermal growth factor-like domains 4-6 of thrombomodulin (TM4-6) to exosite-1 of thrombin changes the macromolecular substrate specificity of thrombin from a fibrinogen clotting procoagulant enzyme to a protein C activating anticoagulant one (17,33,34). It is known that TM4-6 binds to exosite-1 of thrombin with a K D of 5 nm (35,36). TM4-6 is a highly specific ligand for the exosite-1 of thrombin and unlike the hirudin peptide, it is not known if TM4-6 can also interact with proexosite-1 of the substrate. This question was investigated by examining the ability of the TM fragment to inhibit the activation of prethrombin-1 by the prothrombinase complex. As shown in Fig. 3, TM4-6 effectively inhibited the rate of substrate activation by factor Xa specifically in the presence of factor Va with a K D of 0.5 ± 0.1 µm that is slightly better than the corresponding value observed with the hirudin peptide, but 100-fold weaker than its interaction with exosite-1 of thrombin. Taken together, these results confirm the proposal that the interaction of proexosite-1 with the cofactor/enzyme of the prothrombinase complex is required for normal prothrombin activation and that the loss of the zymogenic properties of mutants can be unequivocally attributed to the loss of this interaction. 12

DISCUSSION Owing to its pivotal role in numerous physiological processes, the structure and function of thrombin is being extensively studied by many laboratories. It has been well established that in addition to its catalytic pocket with a trypsin-like S1 specificity, two basic regions remote from the catalytic pocket of thrombin (exosites-1 and 2) play essential roles in determining the macromolecular substrate and inhibitor specificity of thrombin (15,37). Structural and mutagenesis data have indicated that the occupancy of both exosites by various ligands can regulate the proteolytic activity of thrombin both in the procoagulant and anticoagulant pathways. Thus, the interaction of fibrinogen and PAR-1 with exosite-1 of thrombin is required for the protease to function in the procoagulant pathway (15,18,20). Similarly, the interaction of both cofactors V and VIII with this site is a prerequisite for their activation by thrombin and subsequent amplification of the blood coagulation cascade (22,23,38). On the other hand, the binding of TM4-6 to exosite-1 changes the specificity of thrombin from a procoagulant to an anticoagulant enzyme by enabling the protease to rapidly activate the precursor protein of the anticoagulant pathway, protein C (17,33,34). Despite a wealth of knowledge about the role of exosite-1 in the catalytic function of thrombin, much less is known about the role of this site in the zymogenic properties of the precursor prothrombin in the prothrombinase complex. The current study was undertaken to address this question by a mutagenesis approach. Factor Xa activated all prethrombin-1 mutants at comparable rates in the absence of factor Va suggesting that proexosite-1 of prothrombin does not interact with the protease in the absence of the cofactor in a detectable manner. On the other hand, the catalytic efficiency of factor Xa toward most mutants was severely impaired in the presence of factor Va. The greatest degree of impairment ( 75-150-fold) was observed during the prothrombinase activation of R67E and K70E mutant zymogens. With other mutants, the impairments ranged from 4-fold for R75E, 7-fold for R35E, 9-fold for R73E and 15-20-fold for K36E and R77E. These results suggest that all basic residues of proexosite-1 are important recognition sites on the substrate for interaction with factor Va/factor Xa 13

in the prothrombinase complex. It is not known if the degree of impairment in the activation rates directly reflects the order of importance of the mutant residues in their interaction with the activation complex. This is because the proper folding of all mutant zymogens cannot be ascertained. In particular, the x-ray crystal structure of thrombin suggests that both Arg 67 and Lys 70 make intramolecular salt bridge/hydrogen bond contacts with Glu 80 (37). The observations that the K m for the hydrolysis of S2238 with K70E thrombin was impaired and the K i for interaction with p- aminobenzamidine was also elevated suggest that the conformation of the P3-P1 binding pocket of the mutant enzyme has been altered. These results, together with the previous observation that the active site pocket and exosite-1 of thrombin are allosterically linked (39,40) strongly suggest that the mutagenesis of Lys 70 disrupts the integrity of exosite-1 in the mutant thrombin. If the same ionic interactions also exist between Lys 70 and Glu 80 in the precursor protein, it is possible that mutagenesis of Lys 70 has also disrupted the integrity of proexosite-1 in the mutant zymogen. However, R67E thrombin exhibited normal amidolytic activity, thus mutagenesis of this residue has likely no adverse effect on the active site pocket and possibly also not on the integrity of exosite-1 in the mutant thrombin. Accordingly, the mutagenesis of Arg 67 may have no significant effect on the structure of proexosite-1 in the mutant zymogen. All other mutants are expected to fold properly since they are not known to be involved in intramolecular salt bridges in thrombin and presumably also not in prothrombin. This is consistent with the observation that their amidolytic activity has not been adversely affected. Thus, the degree of impairments in the substrate properties of these mutants likely reflects their order of importance for their recognition specificity of the prothrombinase complex. This proposal is also consistent with the observation that the competitive effect of the hirudin peptide was impaired with the proexosite-1 mutants of the substrate and that the degree of impairment correlated well with losses in the zymogenic activities of mutants. Taken together, these results firmly establish that basic residues of proexosite-1 on prothrombin are factor Va-dependent recognition sites for the prothrombinase complex. 14

The mechanism by which factor Va improves the catalytic efficiency of factor Xa in the prothrombinase complex is largely unknown. An attractive hypothesis is that the binding of factor Va to factor Xa allosterically exposes a cryptic substrate recognition exosite remote from the catalytic pocket of the protease. In support of this hypothesis, it has been demonstrated that the active-site inhibited thrombin and a specific chymotrypsin fragment of thrombin can competitively inhibit the activation of prothrombin by factor Xa in the prothrombinase complex (9). Based on the sequence of the inhibitory chymotrypsin fragment, the putative site is believed to include neither the fibrinogen nor the heparin binding exosites (9). Thus, the underlying cause of the defective zymogenic properties of proexosite-1 mutants cannot be attributed to the loss or weakening of an interactive site on mutants for the proposed factor Va-mediated exosite on factor Xa. On the other hand, in a different study, an important role for a direct interaction between factor Va and the proexosite-1 of the substrate has been proposed (14). This latter hypothesis is based on the observation that the exosite-1-specific peptide ligand derived from the C-terminal domain of hirudin could competitively inhibit the activation of prothrombin and prethrombin-1 by factor Xa in the presence but not in the absence of factor Va (14). In agreement with the latter hypothesis, results of this study suggest that a direct interaction between basic residues of proexosite-1 and an acidic region of factor Va may account for the loss of the zymogenic activities of mutant proteins. Several lines of evidence suggest that the complementary proexosite-1 interactive site on factor Va may be located on the C-terminal end of the factor Va heavy chain, possibly involving residues within 659-698 (41-44). This region of factor Va has a high homology for the hirudin peptide and contains several functionally important sulfated Tyr residues which are critical for both the activation of procofactor V by thrombin and the cofactor activity of factor Va in the prothrombinase complex (41-43). It should be emphasized that our results do not exclude the possibility that other functionally important interactive sites exist on prothrombin that can directly or indirectly interact with factor Xa in the prothrombinase complex. 15

Finally, the observation that TM4-6 can also effectively inhibit the activation of prethrombin- 1 by the prothrombinase complex suggests that TM can also bind to proexosite-1 of prothrombin. It is not known whether or not this property of TM can play a significant role in the modulation of the prothrombinase activity under physiological conditions. In particular the K D for this interaction ( 500 nm) is relatively high and the binding of prothrombin to membrane surfaces has been reported to counteract the interaction of the hirudin peptide with proexosite-1 of the substrate (14). However, the negative effect of the membrane interaction is reported to be dependent on the composition of the phospholipid vesicles (14). Noting the plasma concentration of prothrombin ( 1.5 µm) and the high concentration of TM in small vessels ( 500 nm) (45,46), a significant portion of prothrombin is expected to exist in dissociable complex with TM in microcirculation. Based on our results, following conversion of prothrombin to thrombin, the affinity of exosite-1 for TM is enhanced 100-fold. Thus it is conceivable that the TM interaction with proexosite-1 can regulate the prothrombinase activity under certain physiological conditions. More importantly, when the TM-associated prothrombin is activated to thrombin, the protease can not dissociate from the cofactor, insuring that both procoagulant and anticoagulant thrombins are concurrently generated. This may be highly critical for the regulation of the blood coagulation cascade. 16

ACKNOWLEDGEMENT We would like to thank Audrey Rezaie for proofreading of the manuscript. 17

REFERENCES 1. Mann, K. G., Jenny, R. J., and Krishnaswamy, S. (1988) Ann. Rev. Biochem. 57, 915-956 2. Furie, B. and Furie, B. C. (1988) Cell 53, 505-518 3. Rosing, J., Tans, G., Govers-Riemslag, J. W. P., Zwaal, R. F. A. Z., and Hemker, H. C. (1980) J. Biol. Chem. 255, 274-283 4. Nesheim, M. E., Kettner, C., Shaw, E., and Mann, K. G. (1981) J. Biol. Chem. 256, 6537-6540 5. Pei, G., Powers, D. D., and Lentz, B. R. (1993) J. Biol. Chem. 268, 3226-3233 6. Tracy, P. B., Rohrbach, M. S., and Mann, K. G. (1983) J. Biol. Chem. 258, 7264-7267 7. Nesheim, M. E., Taswell, J. B., and Mann, K. G. (1979) J. Biol. Chem. 254, 10952-10962 8. Boskovic, D. S., Giles, A. R., and Nesheim, M. E. (1990) J. Biol. Chem. 265, 10497-10505 9. Betz, A. and Krishnaswamy, S. (1998) J. Biol. Chem. 273, 10709-10718 10. Rezaie, A. R. and Yang, L. (2001) Biochem. Biophys. Acta 1528, 167-176 11. Dharmawardana, K. R. and Bock, P. E. (1998) Biochemistry 37, 13143-13152 12. Dharmawardana, K. R., Olson, S. T., and Bock, P. E. (1999) J. Biol. Chem. 274, 18635-18643 13. Anderson, P. J., Nesset, A., Dharmawardana, K. R., and Bock, P. E. (2000) J. Biol. Chem. 275, 16428-16434 14. Anderson, P. J., Nesset, A., Dharmawardana, K. R., and Bock, P. E. (2000) J. Biol. Chem. 275, 16435-16442 15. Fenton, J. W.,II (1995) Thromb Haemostas. 74, 493-498 16. Rydel, T. J., Ravichandran, K. G., Tulinsky, A., Bode, W., Huber, R., Roitsch, C., and Fenton, J. W.,II (1990) Science 249, 277-280 17. Esmon, C. T. (1993) Thromb. Haemost. 70, 1-5 18. Liu, L., Vu, T. H., Esmon, C. T., and Coughlin, S. R. (1991) J. Biol. Chem. 266, 16977-16980 19. Ye, J., Liu, L., Esmon, C. T., and Johnson, A. E. (1992) J. Biol. Chem. 267, 11023-11028 20. Wu, Q., Sheehan, J. P., Tsiang, M., Lentz, S. R., Birktoft, J. J., and Sadler, J. E. (1991) Proc. 18

Natl. Acad. Sci. (USA) 88, 6775-6779 21. Tsiang, M., Jain, A. K., Dunn, K. E., Rojas, M. E., Leung, L. L. K., and Gibbs, C. S. (1995) J. Biol. Chem. 270, 16854-16863 22. Myles, T., Yun, T. H., Hall, S. W., and Leung, L. L. K. (2001) J. Biol. Chem. 276, 25143-25149 23. Myles, T., Yun, T. H., and Leung, L. L. K. (2002) Blood 100, 2820-2826 24. Bode, W., Mayr, I., Baumann, U., Huber, R., Stone, S. R., and Hofsteenge, J. (1989) EMBO J. 8, 3467-3475 25. Rezaie, A. R. (1998) Protein Sci. 7, 349-357 26. Rezaie, A. R. (1997) Biochemistry 36, 7437-7446 27. Rezaie, A. R. and Esmon, C. T. (1992) J. Biol. Chem. 267, 26104-26109 28. Smirnov, M. D. and Esmon, C. T. (1994) J. Biol. Chem. 269, 816-819 29. Rezaie, A. R. and He, X. (2000) Biochemistry 39, 1817-1825 30. Krishnaswamy, S. and Walker, R. K. (1997) Biochemistry 36, 3319-3330 31. Liu, L., Ye, J., Johnson, A. E., and Esmon, C. T. (1991) J. Biol. Chem. 266, 23632-23636 32. Esmon, C. T. and Jackson, C. M. (1974) J. Biol. Chem. 249, 7791-7797 33. Dang, Q. D., Vindigni, A., and Di Cera, E. (1995) Proc. Natl. Acad. Sci. (USA) 92, 5977-5981 34. Castellino, F. J. (1995) Trends Cardiovasc. Med. 5, 55-62 35. Kurosawa, S., Galvin, J. B., Esmon, N. L., and Esmon, C. T. (1987) J. Biol. Chem. 262, 2206-2212 36. Lentz, S. R., Chen, Y., and Sadler, J. E. (1993) J. Biol. Chem. 268, 15312-15317 37. Bode, W., Turk, D., and Karshikov, A. (1992) Protein Sci. 1, 426-471 38. Esmon, C. T. and Lollar, P. (1996) J. Biol. Chem. 271, 13882-13887 39. Verhamme, I. M., Olson, S. T., Tollefsen, D. M., and Bock, P. E. (2002) J. Biol. Chem. 277, 6788-6798 40. Fredenburgh, J. C., Stafford, A. R., and Weitz, J. (1997) J. Biol. Chem. 272, 25493-25499 19

41. Jenny, R. J., Pittman, D. D., Toole, J. J., Kriz, R. W., Aldape, R. A., Hewick, R. M., Kaufman, R. J., and Mann, K. G. (1987) Proc. Natl. Acad. Sci. (USA) 84, 4846-4850 42. Hortin, G. L. (1990) Blood 76, 946-952 43. Pittman, D. D., Tomkinson, K. N., Michnick, D., Selighsohn, U., and Kaufman, R. J. (1994) Biochemistry 33, 6952-6959 44. Bakker, H. M., Tans, G., Thomassen, M. C. L. G. D., Yukelson, L. Y., Ebberink, R., Hemker, H. C., and Rosing, J. (1994) J. Biol. Chem. 269, 20662-20667 45. Busch, C., Cancilla, P., DeBault, L., Goldsmith, J., and Owen, W. (1982) Lab. Invest. 47, 498-504 46. Esmon, C. T. (1993) Ann. Rev. Cell Biol. 9, 1-26 20

FOOTENOTES Abbreviations Prethrombin-1, prothrombin mutant in which the Gla and Kringle-1 domains have been deleted by recombinant DNA methods; TM, Thrombomodulin; TM4-6, TM fragment containing the epidermal growth factor-like domains 4, 5, and 6; PEG, polyethylene glycol; BSA, bovine serum albumin. 21

FIGURE LEGENDS Figure 1. Activation of prethrombin-1 derivatives by factor Xa and the prothrombinase complex. A. In the absence factor Va, factor Xa (5 nm) and PC/PS vesicles (35 µm) was incubated with prethrombin-1 derivatives: wild type ( ), R35E ( ), K36E ( ), R67E ( ), K70E ( ), R73E ( ), R75E ( ), and R77E ( ) (2 µm each) at room temperature in TBS/Ca 2+. At different time intervals, small aliquots of the activation reactions were transferred to wells of a 96-well assay plate containing 20 mm EDTA and the rate of thrombin generation was measured from the cleavage rate of S2238 as described under Experimental Procedures. B. Factor Xa (0.05-1 nm) in complex with factor Va (30 nm) and PC/PS vesicles (35 µm) was incubated with different concentrations of prethrombin-1 derivatives: wild type ( ), R35E ( ), K36E ( ), R67E ( ), K70E ( ), R73E ( ), R75E ( ), and R77E ( ) at room temperature in TBS/Ca 2+. After 1-30 min of incubation at room temperature, the reactions were terminated by addition of EDTA and the initial rate of thrombin generation was measured as described above. Solid lines in both wild type and R75E thrombins are nonlinear regression fits of kinetic data to the Michaelis-Menten equation and all others are fits to a linear equation. Figure 2. Loss of factor Va-dependent inhibitory effect of Hir 54-65 (SO - 3) on factor Xa activation of prethrombin-1 derivatives. A. The inhibitory effect of the hirudin peptide was monitored by incubating each prethrombin-1 derivative (1 µm) with factor Xa (0.05-1 nm) in complex with a saturating concentration of factor Va (30 nm) on PC/PS vesicles (35 µm) in the presence of increasing concentrations of the hirudin peptide shown on x-axis. The initial rate of thrombin generation was measured by an amidolytic activity assay using GPR-pNA and the data were normalized to % of activity at each concentration of the inhibitor (100% in the absence of the inhibitor) as described under Experimental Procedures. The symbols are: wild type prethrombin-1 ( ), R35E ( ), K36E ( ), R67E ( ), K70E ( ), R73E ( ), R75E ( ), and R77E ( ). Solid lines are best fit of data according to equations 1 and 2. The K D values determined from the slope of these 22

curves are presented in Table II. B. The same as A except that prethrombin-1 ( ) or prethrombin-2 ( ) in the presence of factor Va and prethrombin-1 ( ) in the absence of factor Va used in the activation reactions. Figure 3. The TM4-6 concentration-dependence of prethrombin-1 inhibition by factor Xa in the absence and presence of factor Va. The inhibitory effect of TM4-6 was monitored by incubating prethrombin-1 (1 µm) with 5 nm factor Xa in the absence ( ) or 0.1 nm factor Xa in the presence ( ) of a saturating concentration of factor Va (30 nm) on PC/PS vesicles (35 µm) in the presence of increasing concentrations of TM4-6 shown on x-axis. The initial rate of thrombin generation was measured by an amidolytic activity assay using GPR-pNA and the data were analyzed as described under the legend of Fig. 2. The nonlinear regression of data according to equations 1 and 2 yielded a K D of 0.5 ± 0.1 µm for the TM4-6 inhibition of the activation reaction. 23

Table I- Kinetic constants for the cleavage of the chromogenic substrate S2238 by thrombin derivatives. K m (µm) k cat (s -1 ) K m /k cat (µm s -1 ) WT 6.4 ± 0.4 99.7 ± 7.9 16 ± 2 R35E 4.9 ± 0.4 108.1 ± 10.2 22 ± 4 R35A 3.9 ± 0.8 97.7 ± 7.9 25 ± 7 K36E 4.9 ± 0.2 93.3 ± 4.8 19 ± 2 R67E 4.9 ± 0.4 113.0 ± 10.9 23 ± 4 K70E 205 ± 15 92.5 ± 2.2 0.45 ± 0.04 R73E 5.1 ± 1.1 96.7 ± 4.9 19 ± 5 R75E 6.3 ± 0.4 123.8 ± 11.6 20 ± 3 R77E 5.8 ± 0.5 116.4 ± 11.9 20 ± 4 The kinetic constants were calculated from the cleavage rate of increasing concentrations of S2238 (1-100 µm) by each thrombin derivative (0.5 nm) in TBS/Ca 2+. Kinetic values are the average of three measurements ± standard errors. 24

Table II- Kinetic constants for interaction of factor Xa with factor Va (K d(app) ), maximum rate of thrombin generation (V), and dissociation constants (K D ) for binding of Hir 54-65 (SO - 3) to prethromnin- 1 derivatives. K d(app) (nm) V (nm/min/nm) K D (µm) WT 1.5 ± 0.2 104.2 ± 4.5 0.8 ± 0.1 R35E 1.1 ± 0.1 14.8 ± 0.2 2.3 ± 0.2 K36E 0.9 ± 0.1 7.0 ± 0.3 19.3 ± 8.4 R67E 0.7 ± 0.2 1.4± 0.1 ND K70E 0.8 ± 0.2 0.7 ± 0.1 ND R73E 1.5 ± 0.1 11.0 ± 0.4 33.3 ± 14.9 R75E 1.2 ± 0.1 28.8 ± 1.0 5.1 ± 0.8 R77E 0.8 ± 0.1 5.8 ± 0.2 11.6 ± 3.8 The apparent K d values for the factor Xa-factor Va interaction and the maximal rate of thrombin generation were determined from the saturable cofactor-dependent increase in the initial rate of prethrombin-1 (1 µm) activation on PC/PS vesicles in TBS/Ca 2+ as described under Experimental Procedures. The K D values for binding of Hir 54-65 (SO - 3) to prethrombin-1 derivatives were determined from nonlinear regression analysis of inhibition kinetic data (shown in Fig. 2) according to equations 1 and 2 as described in the text. ND, not determinable. All values are the average of at least two measurements ± standard errors. 25

Figure 1 26

Figure 2 27

Figure 3 28

Proexosite-1 on prothrombin is a factor Va-dependent recognition site for the prothrombinase complex Lin Chen, Likui Yang and Alireza R. Rezaie J. Biol. Chem. published online May 14, 2003 Access the most updated version of this article at doi: 10.1074/jbc.M302707200 Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts