Structural Requirements for Expression of Factor Va Activity*

Thrombin activated factor Va (factor VIIa, residues 1–709 and 1546–2196) has an apparent dissociation constant (Kd,app) for factor Xa within prothrombinase of ∼0.5 nm. A protease (NN) purified from the venom of the snake Naja nigricollis nigricollis, cleaves human factor V at Asp697, Asp1509, and Asp1514 to produce a molecule (factor VNN) that is composed of a Mr 100,000 heavy chain (amino acid residues 1–696) and a Mr 80,000 light chain (amino acid residues 1509/1514–2196). Factor VNN, has a Kd,app for factor Xa of 4 nm and reduced clotting activity. Cleavage of factor VIIa by NN at Asp697 results in a cofactor that loses ∼60–80% of its clotting activity. An enzyme from Russell's viper venom (RVV) cleaves human factor V at Arg1018 and Arg1545 to produce a Mr 150,000 heavy chain and Mr 74,000 light chain (factor VRVV, residues 1–1018 and 1546–2196). The RVV species has affinity for factor Xa and clotting activity similar to the thrombin-activated factor Va. Cleavage of factor VNN at Arg1545 by α-thrombin (factor VNN/IIa) or RVV (factor VNN/RVV) leads to enhanced affinity of the cofactor for factor Xa (Kd,app ∼ 0.5 nm). A synthetic peptide containing the last 13 residues from the heavy chain of factor Va (amino acid sequence 697–709, D13R) was found to be a competitive inhibitor of prothrombinase with respect to prothrombin. The peptide was also found to specifically interact with thrombin-agarose. These data demonstrate that 1) cleavage at Arg1545 and formation of the light chain of factor VIIa is essential for high affinity binding and function of factor Xa within prothrombinase and 2) a binding site for prothrombin is contributed by amino acid residues 697–709 of the heavy chain of the cofactor.

The prothrombinase complex responsible for the generation of ␣-thrombin in the hemostatic process is composed of factor Va and factor Xa associated on a phospholipid membrane in the presence of Ca 2ϩ (1,2). Although factor Xa alone can convert prothrombin to ␣-thrombin, the prothrombinase complex has a catalytic efficiency five orders of magnitude greater than factor Xa acting alone (3). Plasma factor V circulates as a large single chain protein of M r 330,000 (4 -6). The cDNA sequences for human, murine, porcine, and bovine factor V have been re-ported previously (7)(8)(9)(10)(11). The factor V molecule is composed of triplicated "A" domains, duplicated "C" domains, and a "B" region. Human factor V is cleaved by ␣-thrombin at Arg 709 , Arg 1018 , and Arg 1545 and generates the active cofactor factor Va, which is composed of a heavy chain (A1-A2 domains, M r 105,000, amino acid residues 1-709) non-covalently associated with the light chain (A3-C1-C2 domains, M r 74,000, amino acid residues 1546 -2196). The interaction between the two chains is promoted by divalent cations (12,13).
Activation of factor V by ␣-thrombin is required for the interaction of the cofactor with factor Xa and prothrombin. Factor Va and factor Xa interact stoichiometrically in the absence of phospholipids with a K d of 0.8 M, and the interaction is dependent upon the presence of Ca 2ϩ (14). The K d for the factor Va-prothrombin interaction is 1 M and appears to be independent of the presence of Ca 2ϩ (15). In the presence of a membrane surface and Ca 2ϩ , the K d of the factor Va-factor Xa association is ϳ1 nM (3,16). Both chains of the cofactor are required for the interaction with factor Xa (17)(18)(19)(20). The factor Va-prothrombin interaction is promoted by the heavy chain of the molecule (15,18). The factor Va-membrane interaction, governed by a K d value of ϳ3 nM, occurs at diffusionally limited rates, involves both hydrophobic and Ca 2ϩ -dependent electrostatic interactions, and results in penetration of a portion of the light chain into the membrane bilayer (21)(22)(23)(24)(25). Two sites on the light chain of the cofactor appear to be responsible for the interactions of factor Va with the membrane surface. The factor Va-membrane complex is a "true receptor" for factor Xa on the platelet surface (26,27).
Although proteolytic cleavage of factor Va by APC 1 at Arg 506 / Arg 679 results in a 10-fold decrease in the affinity of the molecule for factor Xa, the same cleavages result in the elimination of its interaction with prothrombin (15,18,28,29). It has been shown that prothrombin and thrombin have two separate electropositive exosites (anion binding exosite I, ABE-I, and anion binding exosite II, ABE-II) that are responsible for the majority of the functions of the molecules (30 -39). Whereas ABE-I has been involved in the binding to thrombomodulin (40), fibrinogen (41), PAR1 (42), the COOH-terminal hirudin peptides (43), and heparin cofactor II (44) among others, ABE-II was found to be involved in the interaction with protease nexin (45) and antithrombin III (44). Data from separate laboratories have demonstrated that both exosites bind factors V and VIII (33,34,35). Interestingly, proexosite I of prothrombin, which is present at a low affinity state on the molecule, and its affinity for its ligands increases by ϳ100-fold following activation and formation of thrombin (33,39), was found to be directly involved in the productive interaction with factor Va within prothrombinase (32,33).
Thrombin, factor Xa, and the RVV-factor V activator can produce factor Va by discrete proteolytic cleavages (for details see reviews (1,2)). It has also been reported that a protease from the venom of Naja naja oxiana produces a factor Va-like molecule lacking a major portion of the carboxyl terminal domain of the heavy chain, with diminished cofactor activity (46). The reasons for this diminished cofactor activity, which is mostly observed at low factor Xa concentrations, have yet to be elucidated. The present study was undertaken to understand the role of the last 13 amino acids from the COOH-terminal portion of the heavy chain of the cofactor. Thus, our present study adds to the knowledge regarding the structure/function relationships required for an active factor Va molecule.

EXPERIMENTAL PROCEDURES
Materials and Reagents-Hepes, Sepharose CL-4B, 1-palmitoyl-2oleoylphosphatidylserine (PS) from bovine brain, and 1-palmitoyl-2oleoylphosphatidylcholine (PC) from hen egg were purchased from Sigma (St. Louis, MO). The thromboplastin reagent used in the clotting assays was purchased from Organon Teknika Corp. (Durham, NC). Dansylarginine-N-(3-ethyl-1,5-pentanediyl)amide (DAPA), glutamylglycinylarginyl chloromethyl ketone (EGR), prothrombin-Sepharose, thrombin coupled to agarose through the active site as described (30), human APC, human factor Xa, and factor Xa labeled in the active site with Oregon Green 488 ([OG 488 ]-EGR-hXa) as described (47,48), were from Hematologic Technologies Inc. (Essex Junction, VT). Human prothrombin, ␣-thrombin, and human factor V were purified as previously reported (49 -51). Human factor Va was prepared and purified as recently described using a heparin column (48). The enzyme from the snake venom N. nigricollis nigricollis (NN) was prepared as described (52). Bovine factor X was isolated as described (53, 54), activated to factor Xa using the activator from Russell's viper venom, and purified using benzamidine-Sepharose (Amersham Biosciences) as described (55). The synthetic peptide representing amino acid residues 697-709 of factor Va heavy chain (DYQNRLAAALGIR, D13R) was purchased from Peninsula Laboratories Europe Ltd. (Mereyside, UK). The synthetic pentadecapeptide representing factor V residues 337-351 from factor Va heavy chain previously shown to have no effect on prothrombinase activity (PVIPANMDKKYRSQH, P15H) as well as the peptide shown to contain a binding site for factor Xa (EYFIAAEEV, AP4Ј) were synthesized, in the Biotechnology Core of the Cleveland Clinic Foundation (Cleveland, OH), purified by high-performance liquid chromatography (HPLC), and characterized by mass spectrometry as described (48). Phospholipid vesicles composed of 75% PC and 25% PS were prepared as previously described (56). The concentrations of the phospholipid vesicles were determined by phosphorous assay (57).
Generation of Various Factor V Species-The various factor V species were generated by incubation of single chain human plasma factor V with the purified enzymes. Phospholipid vesicles slightly increase the rate of cleavage of factor V by NN, whereas PCPS vesicles diminish the rate of activation of factor V by ␣-thrombin (58). Thus, the experiments studying the effect of ␣-thrombin on factor V NN were conducted in the absence of PCPS vesicles. The nomenclature for the proteolyzed factor V products is as follows: factor V IIa : human factor V incubated with ␣-thrombin (1 unit/ml, ϳ10 nM) at 37°C for 5 min; factor V NN : human factor V (220 nM) incubated with NN venom enzyme (2.4 nM) at 37°C for 3 h; factor V IIa/NN : factor V NN incubated with ␣-thrombin (1 unit/ml) at 37°C for 10 min. In all cases, 20 nM hirudin was added following proteolytic treatment. To generate factor V RVV , 212 nM human factor V was incubated at 37°C with 4.2 nM RVV-V activator for 2 h. In some experiments following full activation to factor V RVV , 2.1 nM APC and 50 M PCPS vesicles were added and the reaction was allowed to proceed for 1 h. Aliquots were taken at the time points indicated in the figure legends and analyzed by SDS-PAGE.
Cleavage of Factor V IIa by NN Protease and Inactivation by APC-␣-Thrombin-cleaved human factor V IIa (212 nM) was incubated at 37°C with 50 M PCPS vesicles and 2.1 nM NN protease. After 3 h, 2.1 nM APC was added and the reaction was allowed to proceed for 1 h. At selected time intervals prior to and subsequent to addition of APC, aliquots of the mixture were withdrawn and assayed for cofactor activity in a clotting assay using factor V-deficient plasma. At the same intervals, aliquots were prepared for SDS-polyacrylamide gel electrophoresis analyses.
Determination of Factor Va Cofactor Activity-Cofactor activity was measured at selected time intervals by clotting assay using factor V-deficient plasma and standardized to the percentage of control (50). In a typical assay 50 l of factor V-deficient plasma was mixed with an equal volume of the sample to be analyzed. To start the assay 100 l of the PT reagent (Simplastin Excel) was added while rocking the tube at 37°C. The assay end point was determined by visualization of the fibrin clot. Control experiments using a solution of RVV-V activator or NN alone (in the absence of factor V) in the clotting assay, at the same concentrations that are used to cleave factor V, gave no interference with the clotting process (with or without normal plasma). Thus, trace amounts of snake venom enzymes when introduced in plasma during the clotting assay (or in the prothrombinase assay) did not interfere with the results obtained. The assay was initiated with the addition of the factor Va followed by the addition of thromboplastin.
Measurements of Rates of Thrombin Formation using DAPA-Functionally defined apparent dissociation constants (K d,app ) for factor Xa binding to factor Va-PCPS were inferred from an assay measuring ␣-thrombin generation at various factor Xa concentrations (59,60). These studies were performed using a PerkinElmer Life Sciences MPF-44A fluorescence spectrophotometer with a ex 280 nm (slit at 8 nm), em 550 nm (slit at 16 nm), and a 500-nm long pass filter in the emission beam. In a typical experiment a mixture (ϳ80 ml) composed of prothrombin (1.4 M), PCPS vesicles (20 M), and DAPA (3 M) was incubated in the dark for 20 min in 20 mM HEPES, 0.15 M NaCl, 5 mM CaCl 2 , pH 7.4, buffer. An aliquot of the mixture (1800 l) was added to a cuvette containing the indicated factor V species (0.5 nM), and the base line fluorescence was monitored for 15-20 s at ambient temperature. Factor Xa (100 pM to 10 nM) was then added to initiate the reaction. ␣-Thrombin generation was continuously monitored by the change in the fluorescence intensity of DAPA present in the reaction mixture. The use of DAPA in the reaction mixture prevents feedback reactions catalyzed by the newly formed thrombin.
Determination of the Apparent Binding Constants for Factor Xa-To determine the apparent dissociation constant (K d,app ) of factor Xa for the various factor Va species, a titration with factor Xa measuring prothrombin activation (1.4 M) was performed using concentrations ranging from 100 pM to 10 nM.
The K d,app for the factor Va-factor Xa interaction is given by Equation 1, where [Va] F and [Xa] F are the concentrations of free factor Va and factor Xa; the final concentration of factor Va within the mixture was 0.5 nM. The initial rate of thrombin formation for each factor Xa concentration was plotted as a function of the factor Xa concentration (using the software DeltaGraph PRO, 3.5, Delta Point Inc., Monterey, CA) using the quadratic equation that is the solution for Equation 1 (16,61).

2⅐[Va]
(Eq. 2) Assuming that n ϭ moles of factor Xa bound/mole of factor Va at saturation; throughout this study n ϭ 1; the stoichiometry of the factor Va-factor Xa interaction was fixed at 1. obs ϭ V 0 ϩ [Xa bound ⅐Vϱ], where obs is the rate at Xa T ϭ 0 and Vϱ is the rate at infinite factor Xa concentrations (i.e. when the fixed concentration of factor Va is saturated with factor Xa). Prior to each experiment, the maximum amount of complex formed under the above conditions was verified by replacing [ Inhibition of Thrombin Formation-The formation of thrombin was also analyzed using the fluorescent thrombin inhibitor DAPA as described (48) using a PerkinElmer Life Sciences LS-50B luminescence spectrometer (PerkinElmer Life Sciences LLC, Norwalk, CT) with ex ϭ 280 nm, em ϭ 550 nm, and a 500-nm long pass filter in the emission beam (Schott KV-500). The buffer used in all cases was composed of 20 mM Hepes, 0.15 M NaCl, 5 mM CaCl 2 , pH 7.4 (HBS(Ca 2ϩ ), "assay buffer"). The final concentration of factor Va in the mixture was 4 nM with factor Xa at 10 nM, prothrombin at 350 nM, PCPS vesicles at 10 M, in the presence of 2 mM CaCl 2 . Factor Va was incubated with the peptides (at concentrations ranging from 0.5 nM to 100 M) for 10 min and added to the assay. The initial rate of thrombin formation (nM IIa⅐min Ϫ1 ) was calculated as described (48). The concentration of peptide given in each figure is the final concentration of the peptide in the assay mixture. All data were initially analyzed and stored using the software FL WinLab (PerkinElmer Life Sciences) and further analyzed and plotted with the software Prizm (GraphPad, San Diego, CA).
Fluorescence Anisotropy Measurements-Fluorescence anisotropy of [OG 488 ]-EGR-hXa was measured using a PerkinElmer Life Sciences LS-50B luminescence spectrometer in L-format as recently described (48). Anisotropy measurements were performed in a quartz cuvette under constant stirring (low) with ex ϭ 490 nm, em ϭ 520 nm with a long pass filter (Schott KV-520) in the emission beam. At each addition, anisotropy was measured for 20 s and eight successive readings were averaged. In all cases, the total addition of peptide did not exceed 10% of the volume of the reaction. All data were initially analyzed and stored using the software FL WinLab (PerkinElmer Life Sciences) and further analyzed and plotted with the software Prizm (GraphPad). Some of the data were also plotted using DeltaGraph (DeltaPoint, Monterey, CA).
Direct Binding of the Peptides to Thrombin-Thrombin immobilized to agarose through the active site was used (30). Peptide solutions of D13R and P15H (48) were dissolved in water to a given concentration. In control experiments it was determined that the maximum amount of peptide retained by the thrombin-agarose column was approximately 400 g. The peptides were thus diluted in 20 mM Hepes, 0.1 M NaCl, pH 7.4, in a manner so that ϳ400 g was contained in each of the starting solutions. The two peptides used in this study contained aromatic amino acid residues; thus, their presence in the flowthrough or the eluate of the column was ascertained by the optical density. Position identification of bound material was verified by LC/MS as detailed below in the analytical facility of Dr. David Anderson at Cleveland State University.
Mass Spectrometry Instruments and Conditions-Eluate was separated from the salt components using an HPLC system (HP 1100, HPLC gradient system, Agilent Technologies, Palo Alto, CA) with a C18 column (1 mm ϫ 15 cm, GraceVydac, Hesperia, CA) and buffers A (0.3% acetic acid in water) and B (0.3% acetic acid in acetonitrile), and the eluted material was monitored using a Micromass Quatro II ESI-Triple quadrupole mass spectrometer (Waters, Milford, MA). The data were collected using a Compaq Professional Workstation (AP200, Hewlett-Packard, Palo Alto, CA) and analyzed by MassLynx version 3.

(Waters).
Gel Electrophoresis-SDS-PAGE analyses were performed using 5-15% and 4 -12% gradient gels according to the method of Laemmli (62). Proteins were visualized after staining with Coomassie Brilliant Blue in 50% methanol, 10% acetic acid followed by destaining by diffusion using a solution of 50% methanol and 10% acetic acid.
Amino Acid Sequence Analyses-Human factor V (630 nM) was incubated with PCPS vesicles (200 M) and the purified NN protease for 3 h at 37°C. The reaction was stopped by the addition of 2% SDS, 2% ␤-mercaptoethanol, and the mixture was heated for 5 min at 90°C. Approximately 40 g of total factor V digest was analyzed on an 8 -18% linear gradient SDS-PAGE gel. Similar experiments were performed with the RVV-cleaved cofactor. Following electrophoretic separation, proteins were transferred to a PVDF membrane, and the NH 2 -terminal sequences were determined using automatic Edman degradation on an Applied Biosystems 475A protein sequencing system as described (60) in the laboratory of Dr. Alex Kurosky (University of Texas, Medical Branch at Galveston).

RESULTS
Limited Proteolysis of Factor V by N. nigricollis nigricollis (NN) Protease-Factor V prior to the activation by ␣-thrombin typically displays 13-to 15-fold lower clotting activity than the active cofactor factor Va obtained following activation by ␣-thrombin (50). The apparent activity of single chain factor V is a consequence of a limited amount of cleavage, which produces the active cofactor during the initial phase of the assay (63,64). In experiments using purified factor Xa, prothrombin, and phospholipid in the presence of the fluorescent thrombin inhibitor DAPA, which prevents feedback activation of single chain factor V, measurements of factor V activity give results that are 400-fold lower than for thrombin-activated factor V (3). Thus, the single-stage clotting assay overestimates factor V cofactor activity, because the clotting end-point occurs when ϳ1% of the total circulating procofactor is activated and relies upon the adventitious activation of the procofactor by the initial thrombin produced during the assay. For isolated human factor V the single stage assay yields a value of 54 units/mg (Fig. 1, lane 1), which increases 15-fold to 740 units/mg upon activation with ␣-thrombin (factor V IIa , Fig. 1, lane 12). Treatment of factor V with the NN protease results in the progressive development of a species (factor V NN , Fig. 1, lanes 2-11) with a specific activity lower than factor V (21 units/mg, Fig. 1 Factor V was incubated with the NN protease as described under "Experimental Procedures." At selected time intervals aliquots of the mixture were withdrawn and assayed for cofactor activity in a clotting assay using factor V-deficient plasma (the activity of each sample is shown at the bottom of the figure in units/mg). Fragments were visualized following Coomassie Blue staining on a 4 -12% (linear gradient) SDS-PAGE. Lane 1, factor V, control; lanes 2-11, factor V at 1, 3, 5, 10, 20, 30, 60, 90, 120, and 180 min following the addition of the enzyme; lane 12, factor Va control activated with ␣-thrombin. Molecular weight markers are shown in lane 13. heavy chain of M r ϳ 100,000 and a light chain doublet of M r ϳ 80,000, compared with a heavy chain of M r 105,000 and light chain doublet of M r 74,000/72,000 for thrombin-activated factor V (factor V IIa , lane 12). These data suggest that the impaired cofactor activity of factor V NN is due either to the shorter heavy chain or the longer light chain or both. NH 2 -terminal amino acid sequence analyses of the fragments resulting from incubation of factor V with NN is shown in Fig. 2. The data show preferential cleavage by NN between the X and Asp residues of selected X-Asp-Z bonds (where X is any amino acid and Z is typically tyrosine), consistently producing fragments with a D-Y at the NH 2 terminus (Fig. 2B). Cleavage by NN results in new NH 2 termini at Asp 697 , Asp 1509 , and Asp 1514 . These products correspond to the fragments with M r 100,000 (residues 1-696, fragment a, Fig. 2A), M r 80,000 (residues 1509-/1514 -2196, fragments b/c, Fig. 2A), and a fragment of M r 65,000 (fragment d). The latter fragment exhibited an NH 2 -terminal sequence that matches a region of factor V starting at amino acid residue 697 (Fig. 2B). The arrowhead in Fig. 2B, column d, identifies the Arg 709 thrombin cleavage site illustrating that the heavy chain of factor V NN is 13 residues shorter than for factor V IIa . The enzyme also cleaves at another site on the B region of the molecule, giving rise to fragment d ( Fig. 2A) and another fragment (arrowhead below fragment d, Fig. 2A), which remains to be explicitly identified. A high background complicated the sequencing of this fragment; however, in one experiment an NH 2 -terminal sequence starting with D-F-S was found for the fragment depicted by the arrowhead in Fig. 2A. In the B domain of factor V there are several repeats of 9 amino acids and two of these repeats start with the sequence D-F-S (residues 1260 and 1305) (8). Thus, a cleavage site in the B region of factor V is suggested at either Asp 1260 or Asp 1305 . The high levels of carbohydrate present in this region may be responsible for the difficulties in obtaining an extended sequence (8). Overall, NN cleavage of factor V produces a factor "Va-like" molecule, factor V NN , composed of a heavy chain (residues 1-696) and a light chain (residues 1509-/1514 -2196). The reduced activity of factor V NN in the clotting assay (Ͻ3% activity of factor V IIa ) thus must be the consequence of either a shorter heavy chain or a longer light chain or both.
The Importance of the COOH Terminus of the Heavy Chain-To examine the factor V proteolytic product with a full-length light chain and truncated heavy chain, factor V IIa was further subjected to proteolysis by NN. Using the plasma clotting assay, the cofactor activity of factor V IIa (210 nM, 750 units/mg) was followed during treatment with NN protease (2.1 nM), with parallel analysis by SDS-PAGE. After 10-min incubation with NN, factor V IIa lost ϳ40% of its initial clotting activity (Fig. 3A); after 1 h, ϳ20% clotting activity remains. No further loss in clotting activity was observed (up to 2 h, Fig.  3A). The cleavage of factor V IIa at Asp 697 of the heavy chain is correlated with the loss of 80% of the clotting activity of factor V IIa (factor V NN/IIa , Fig. 3A, 180 min). The remaining cofactor activity of factor V IIa/NN can be eliminated by cleavage of the heavy chain by APC and phospholipids, at Arg 506 , Arg 306 , and Arg 679 (Fig. 3A, lanes 13-18, open circles), yielding a fragmentation pattern similar to that observed for APC inactivation of factor V IIa (65). The data demonstrate that removal of the last 13 residues (amino acid region 697-709) from the COOH terminus of the heavy chain of factor V IIa results in a molecule with greatly diminished activity when measured in a clotting assay.
Activation of Factor V by Russell's Viper Venom Factor V Activator and the Importance of the Light Chain-In contrast to the NN protease, RVV-V activator is an efficient activator of factor V. Although several laboratories have used the RVV-V activator to activate factor V (12,66,67) no explicit description of the activation/cleavage sites has been reported. Popular consensus is that the RVV-V protease activates factor V to produce a species with only a light chain, equivalent (by electrophoretic migration) and with similar clotting activity to factor V IIa (12,66,67). We have reported that APC-inactivated membranebound factor V (cleaved at Arg 306 , Arg 506 , Arg 679 , and Lys 994 ) is further cleaved by the RVV-V activator at Arg 1018 and Arg 1545 (65). However, additional RVV-V cleavages within the region 1-1018 could not be ruled out, because those experiments were complicated by further APC fragmentation of the products (65). Fig. 4 shows that incubation of factor V with catalytic amounts of the RVV-V protease results in a species, factor V RVV , which FIG. 2. Identification of the proteolytic cleavages of the factor V molecule by the enzyme contained in the snake venom of N. nigricollis nigricollis. Human factor V was incubated with PCPS vesicles and the purified NN protease as described under "Experimental Procedures." After electrophoresis, the proteins were transferred to a PVDF membrane. A, one of the PVDF membranes that was used for the NH 2 -terminal sequence analysis stained with Coomassie Blue and represents the fragments derived from human factor V following cleavage by NN. The letters a-d depict fragments of human factor V. The arrowhead underneath fragment d denotes a fragment that is derived from the B region of factor V and does not stain with Coomassie Blue. The NH 2 -terminal sequence of this fragment could not be determined positively. B, NH 2 -terminal sequence of the fragments a-d shown in A. The numbers in parentheses indicate picomoles of amino acid at the given cycle. Fragment b was found to contain two amino acid sequences. According to the amino acid sequence derived from the cDNA of the human cofactor (8) these sequences are identified to represent cleavage by NN at Asp 1509 and Asp 1514 . The arrowhead at the right of the sequence shown for fragment d depicts Arg 709 , which is a specific cleavage site by ␣-thrombin for activation of human factor V (8).
possesses cofactor activity in the plasma clotting assay similar to the ␣-thrombin-activated cofactor, factor V IIa (approaching 700 units/mg, Fig. 4A). The activation of factor V by the RVV-V activator is a slower process than ␣-thrombin activation at similar catalyst and substrate concentrations. In addition to the light chain (M r 74,000), factor V RVV displays a M r 150,000 heavy chain (Fig. 4B). NH 2 -terminal amino acid sequence analyses of the fragments obtained following cleavage of factor V by the RVV-V activator (not shown) confirm our previous findings with the APC-treated factor V (65) consistent with the conclusion that RVV-V activator cleaves the procofactor at Arg 1018 and Arg 1545 .
The subunit compositions for each factor V molecule and their clotting activities are illustrated in Fig. 5, and their characteristics are reported in Table I. Lane 1 shows thrombinactivated factor V, followed by factor V RVV , which possesses a normal light chain and a heavy chain that is longer than that of factor V IIa . Treatment of factor V with NN produces factor V NN (Fig. 5) with a shorter heavy chain and longer light chain than factor V IIa . Factor V NN possesses minimum clotting activity. Factor V NN can be further cleaved by ␣-thrombin at Arg 1545 to reduce the light chain to normal size (factor V NN/IIa ). Coincidentally, the clotting activity of the molecule increases to ϳ50% of that of factor V IIa . Cleavage of factor V RVV with ␣-thrombin produces a species indistinguishable from factor V IIa (factor V RVV/IIa ), whereas factor V NN/RVV appears with a normal light chain and truncated heavy chain (identical to factor V NN/IIa ).
Functional Activity of the Factor V Species Derived from NN and Thrombin-Cleavage of factor V by NN produces a molecule exhibiting impaired cofactor activity either because of impaired factor Xa receptor capability, diminished catalytic effector properties, or both. To compare the receptor capacities of the factor V IIa , factor V NN , factor V RVV , factor V NN/IIa , and factor V NN/RVV , thrombin generation was measured in the DAPA assay as a function of added factor Xa (Fig. 6). The use of DAPA in the reaction also inhibited feedback cleavage by thrombin of the factor V NN and factor V RVV species. The curves in Fig. 6 represent best fits to Equation 2 (see "Experimental Procedures"), and each yields a K d,app describing on a functional basis the affinity between factor Xa and the factor V species (16, 61). These dissociation constants and corresponding activities in the clotting assay are listed in Table I

FIG. 4. Activation of human factor V by RVV-V activator and inactivation by APC.
A, human factor V was incubated at 37°C with RVV-V activator for 2 h. After full activation to factor V RVV , APC and PCPS vesicles were added, and the reaction was allowed to proceed for 1 h. Factor V clotting activity was measured as described under "Experimental Procedures." B, 4-12% linear gradient SDS-PAGE. Fragments were visualized by Coomassie Blue staining: Lane 1, factor V, control; lanes 2-10, factor V and 4.2 nM RVV at 1, 3, 5, 10, 15, 30, 60, 90, and 120 min; lanes 11-19, factor V RVV , PCPS and APC at 1, 3, 5, 7, 10, 15, 30, and 60 min. Vertical, arrows indicate the addition of APC to the reaction mixture. The position of the heavy chain of factor V RVV is indicated at left. species, and the results of these are also given in Table I. Compared with factor V IIa (Fig. 6A, filled circles), factor V NN (Fig. 6A, open circles) exhibits both reduced factor Xa affinity and diminished clotting activity at limiting concentrations of factor Xa (Table I). The K d,app for factor V NN with human factor Xa was ϳ4 nM (with a 1:1 stoichiometry). The effector efficiency of V NN was not affected at saturating concentrations of factor Xa. Upon treatment with ␣-thrombin or RVV, factor V NN improves its affinity for factor Xa and its ability to serve optimally as receptor for factor Xa within prothrombinase (Fig. 6A, open  squares); however, only partial recovery of the clotting activity was observed (Table I). Similar results were observed when factor V RVV was treated with NN (K d,app for factor V RVV/NN of ϳ0.5 nM). Because both factor V IIa and factor V IIa/NN possess a fully formed light chain, resulting from cleavage at Arg 1545 (Fig. 6B, lanes 1 and 3), it appears that cleavage at Arg 1545 is enough to restore optimal binding of the cofactor for factor Xa. However, the clotting activities of factors V NN/IIa and V NN/RVV still remained impaired as compared with factor V IIa (Table I and Fig. 5). These data suggest that cleavage of the heavy chain at Asp 697 results in a cofactor with reduced capabilities within prothrombinase that are related to functions of the molecule other than the binding to factor Xa; i.e. the COOH-terminal residues of the heavy chain of the cofactor do not appear to be involved in the factor Va-factor Xa interaction or the factor Va-membrane interaction. The latter was demonstrated by Bakker et al. (46) using a factor Va molecule lacking the Asp 683 -Arg 709 domain of the heavy chain. Collectively the data demonstrate that at saturating concentrations of factor Xa (Ն10 nM, Fig. 6 and Table I) cleavage at Arg 1545 is not required for full expression of factor Xa cofactor activity, however, this cleavage is necessary for full expression of factor Va cofactor activity at limiting amounts of factor Xa, as those available in the clotting assay. Thus, although cleavage at Arg 1545 is not a prerequisite for expression of maximum factor Va cofactor activity, it is required for optimal factor Xa binding. Overall, the data also indicate that elimination of 13 amino acids from the COOH-terminal portion of the heavy chain of factor V IIa results in partial loss in cofactor activity with no diminution in the factor Xa binding capability.

Inhibition of Prothrombinase Activity by a Synthetic Peptide Containing the 13 Amino Acids from the COOH-terminal Portion of the Heavy Chain of Factor Va (D13R)-
To ascertain the importance of the region 697-709 for cofactor activity, we obtained a 13-amino acid peptide encompassing this region (D13R). This peptide was tested for inhibition on factor Va cofactor activity in a prothrombinase assay using purified reagents. In the prothrombinase assay purified human factor Va was preincubated with the synthetic peptide and added to a cuvette containing PCPS, prothrombin, DAPA, and factor Xa. D13R showed complete inhibition of factor Va cofactor activity (Fig. 7A, filled squares) with an IC 50 of ϳ12 M. Complete inhibition occurred at ϳ100 M peptide. The residual 6 -7% prothrombinase activity observed at this peptide concentration Single chain, human factor V was subjected to proteolysis by thrombin, N. nigricollis nigricollis, and/or RVV snake venom as indicated under "Experimental Procedures." Activity was determined by clotting assay and is expressed in units/mg at the bottom. Species are as follows: Lane 1, factor V IIa (single chain human factor V was incubated with ␣-thrombin); lane 2, factor V RVV ; lane 3, factor V NN ; lane 4, factor V NN/IIa (factor V NN incubated with ␣ -thrombin); lane 5, factor V RVV/IIa (factor V RVV incubated with ␣-thrombin); lane 6, factor V NN/RVV (factor V NN incubated with RVV-V activator). The subunit composition of each species is given in M r ϫ 10 Ϫ3 : HC ϭ heavy chain; LC ϭ light chain. b Activity was determined in a clotting assay using factor V-depleted plasma as described under "Experimental Procedures" (50). The numbers show the clotting activities of the factor Va species depicted in Fig. 5; the maximum activity (100% ϭ 749 units/mg) was arbitrarily assigned to the ␣-thrombin-activated species (Fig. 5, lane 1). c The K d of each factor Va species was calculated using the quadratic equation (Equation 2) described under "Experimental Procedures" with a stoichiometry fixed at 1 (n ϭ 1). d The velocity is given in nM Ila/min/0.5 nM prothrombinase assuming that at saturating concentrations of factor Xa all factor Va (0.5 nM) will be saturated with factor Xa, and, thus, the maximum enzyme concentration of active enzyme is 0.5 nM prothrombinase. e The species V NN/Ila or V IIa/NN and V NN/RVV or V RVV/NN have similar properties and subunit composition. Thus, the data relating to the properties and subunit composition of only one species of the pairs are reported.
is due to the activity of factor Xa alone. Under similar experimental condition a control pentadecapeptide from the middle portion of factor Va (P15H (48)) did not show any inhibition of prothrombinase function (Fig. 7A, filled triangles). The pentadecapeptide has 7 amino acids in common with D13R (in a different order), contains a tyrosine and an aspartic acid, derives from the factor Va sequence, and overall possess a similar charge as D13R. P15H is thus an adequate control for D13R.
Overall data demonstrate that D13R inhibits prothrombinase by impairing the interaction of the cofactor with one of the components of prothrombinase. It is noteworthy that D13R was previously found to be non-inhibitory in a clotting assay using factor V-deficient plasma (68). These discrepancies may result from the way the assays are conducted, and from the definition of factor Va activity per se in each assay. In the prothrombinase assay, the peptide is preincubated with factor Va and the end-point of the assay is measured by the rate of thrombin formation as measured by its complexation with DAPA. In the clotting assay, the end-point of the assay is the formation of a fibrin clot. We must thus conclude that the nature of an assay is important for the determination of the inhibitory potential of a peptide.
To verify if the peptide inhibits prothrombinase activity by impairing the direct interaction of the cofactor with factor Xa, we have employed an assay using a fluorescent derivative of the enzyme as recently described (48) 7B, filled squares). Under similar experimental conditions P15H didn't have any effect on the factor Va-[OG 488 ]-EGR-hXa interaction, whereas a previously described nonapeptide from the heavy chain of factor Va (AP4Ј) showed complete inhibition of the binary interaction with an IC 50 of ϳ8 M (Fig. 7B, filled  triangles) (48). These data are consistent with the data of Fig.  6, which indicated that deletion of 13 amino acids from the COOH-terminal part of the heavy chain of factor Va does not interfere with the binding of factor Xa to the cofactor (Fig. 6, factor Va NN/IIa ). Altogether these data suggest that D13R inhibits prothrombinase, because it interferes with other interactions in prothrombinase.
The mechanism of D13R inhibition of prothrombinase was addressed by investigating the effect of the peptide on the kinetic parameters of prothrombinase (K m and V max ) in the presence of varying concentrations of inhibitor (2-5 M, Fig.  7C). Under the conditions employed and in the presence of increasing concentrations of peptide, the saturation curves were sigmoidal (Fig. 7C). The curves tended toward a similar asymptote, which is the V max of prothrombinase (525 Ϯ 17 nM IIa/min), whereas the apparent K 0.5 of the enzymatic reaction increased with increasing concentration of inhibitor (Fig. 7C, 2 M (filled triangles), 3 M (filled circles), 4 M (filled diamonds), and 5 M (filled inverted triangles)). These data represent an unusual competitive inhibition mechanism were only free substrate (prothrombin) can produce thrombin in the presence of prothrombinase (69). According to this model D13R binds prothrombin in competition with the binding of prothrombin to prothrombinase (factor Va-factor Xa) (69). This unusual competitive mechanism was previously described for the inhibition of prothrombin activation by the sulfated version of Hir54 -65 peptide, which inhibited prothrombinase with a K i of ϳ3 M (33). The inhibition constant (K i ) of prothrombinase by D13R calculated from the IC 50 (Fig. 7A) using the K m obtained from Fig. 7C (0.4 M Ϯ 0.03, filled squares) was found to be 6.4 M. These values are in complete agreement and strongly suggest that the hirudin-like peptide D13R is an unusual competitive inhibitor of prothrombinase, because it inhibits prothrombin activation by interfering with the interaction of the substrate (prothrombin) with the enzyme (prothrombinase) rather than interfering with the active site of the enzyme like a classical competitive-type inhibitor (69).
Direct Binding of D13R to Thrombin and Prothrombin-All FIG. 6. Requirement for cleavage at Arg 1545 in factor V for high affinity binding of factor Xa to factor Va and optimum function within prothrombinase. A, each of the factor Va species were assayed for activity at a fixed concentration of 0.5 nM and various factor Xa concentrations in a purified prothrombinase assay as described under "Experimental Procedures." The graphs represent: factor V IIa , filled circles; factor V NN , open circles; factor V NN/IIa , open squares. The reaction was initiated by the addition of varying amounts of factor Xa (100 pM to 10 nM). ␣-Thrombin generation was continuously monitored by the change in fluorescence intensity of DAPA. The slope of the graph of prothrombin generation was calculated separately and plotted as a function of the concentration of factor Xa used. The solid lines represent hyperbolas that were obtained by fitting the data to Equation 2 shown under "Experimental Procedures" using non-linear least squares analysis with n fixed as a locked parameter at 1 (i.e. the factor Va:factor Xa stoichiometry was fixed to 1). The initial estimates were a K d of 1 nM and a fixed stoichiometry of 1 (n ϭ 1). B, the identities of the cofactor molecules assayed in A are shown. The symbols at the bottom of the gel represent the cofactor species used in A. The star identifies an intermediate fragment deriving from cleavage of factor V by the NN protease.  FIG. 7. Function of a peptide containing amino acid residues 697-709  (D13R). A, prothrombinase function. Increasing concentrations of D13R were incubated with factor Va. Factor Va cofactor activity was assayed in the presence of increasing concentrations of D13R as described under "Experimental Procedures" in a prothrombinase assay using purified reagents (filled squares). The control peptide P15H previously shown to have no effect on factor Va cofactor activity is shown by the filled triangles (48). The data with D13R represent the average value found in three independent experiments, whereas the results with P15H represent the average value found in two independent experiments. B, prothrombinase assembly. Reaction mixtures containing a [OG 488 ]-EGR-hXa at 10 nM and 10 M PCPS were titrated with factor Va to saturation (25 nM (48)). The preformed complexes were titrated with increasing concentrations of AP4Ј (filled triangles), P15H (filled inverse triangles), and D13R (filled squares). Delta r was obtained by subtracting the value of the anisotropy of [OG 488 ]-EGR-hXa alone (average of eight readings) in the absence of factor Va from the anisotropy observed in the presence of a given concentration of factor Va (average of eight readings) as described (48). the data presented thus far suggest that a binding site for prothrombin is located within D13R, which represents the last 13 amino acids from the heavy chain of factor Va. To ascertain that the peptide interacts directly with prothrombin and/or thrombin, we have used a prothrombin-Sepharose column and a thrombin-agarose column. Attempts to study the binding of D13R to prothrombin-Sepharose failed, because D13R was consistently found in the flowthrough of the column. Two facts may account for this result: 1) the peptide has low affinity for prothrombin-Sepharose and our experimental conditions did not allow for the binary interaction to occur and/or 2) ABE-I of prothrombin, which has been reported to interact with hirudinlike peptides and to interfere with prothrombin activation (33), may have been blocked during the coupling procedure. However, although it has been previously demonstrated that factor Va interacts with prothrombin and thrombin via ABE-I, it has been also shown that the affinity of ABE-I for its ligands is increased following activation and thrombin formation (33,39). We have thus used a thrombin-agarose column where thrombin was coupled to the agarose beads via the active site as described (30). Because D13R and the control peptide P15H both contain tyrosine residues, we can measure their optical density, thus we can identify the presence of the peptides in the collected fractions. The thrombin-agarose column was loaded with the peptides and washed extensively. Subsequently, elution was performed with 2 M NaCl. The results of these experiments are shown in Fig. 8. The control peptide, P15H (filled squares), which has 7 amino acids identical to D13R, was detected in the flowthrough of the thrombin-agarose column. In contrast, D13R bound to the thrombin-agarose column and eluted with the high salt buffer (Fig. 8, arrow, filled circles, and  filled triangles). The identity of D13R in the elution fractions from both runs was verified by LC/MS (not shown). The data demonstrate that D13R contains a binding site for thrombin. A direct comparison, of the amino acid sequence of D13R with the amino acid sequence of several hirudin-like peptides that interact with thrombin, suggests that the hirudin-like motif DYQ (30 -33, 70 -72) contained within D13R may be responsible for its interaction with the thrombin-agarose column.
Taken together the data indicate that prothrombinase assembly can occur following cleavages at Asp 697 , Asp 1509 , and Asp 1514 , whereas optimum function requires cleavage at Arg 1545 and generation of the light chain of factor Va. This cleavage appears to be required for optimum binding of factor Xa, which in turn is a prerequisite for subsequent expression of efficient catalytic effector function by the cofactor at low physiological concentrations of factor Xa. Thus, although deletion of residues 697-709 in the heavy chain of factor Va species did not affect factor Xa binding, the clotting function of factor Va was significantly affected by truncation at the COOH terminus. Thus, the COOH-terminal portion of the heavy chain of factor Va appears to contribute to the prothrombin binding site of prothrombinase and is required for expression of optimum clotting activity. Fig. 9 is a schematic representation of the factor V products derived following NN (Fig. 9A), RVV-V activator (Fig. 9B), and ␣-thrombin (Fig. 9C) treatment. NN cleaves factor V at Asp 697 and Asp 1509/1514 to produce a molecule that has less clotting activity than unactivated factor V (factor V NN , Fig. 9A). The present work demonstrates that formation of the light chain of factor Va (following cleavage at Arg 1545 of factor V) is a common feature of the cleaved products, which exhibit optimum association for factor Xa in the assembly of prothrombinase (Fig. 9). Thus, factor V IIa/NN , factor V RVV/NN , and factor V IIa display similar affinities for factor Xa (K d of 0.5, 0.57, and 0.3, respectively, Table I). Each of these three forms of factor V has a fully formed light chain (1546 -2196) and differs only in the COOH terminus of the heavy chain. With an extended light chain, as in factor V NN (where the NH 2 terminus has the additional residues 1509-/1514 -1545), a weaker factor V NN -factor Xa interaction was observed (K d of 4.1 nM). Thrombin cleavage at Arg 1545 restored the affinity of factor V NN for factor Xa to levels found for factor V IIa (K d of 0.5 nM, Fig. 9D). The current study indicates that formation of the light chain (1546 -2196) is required to capture factor Xa on the lipid surface and that the addition of as few as 30 additional residues in the light chain reduces the cofactor and clotting activity by an order of magnitude or more. This correlates with observations made during tissue factor-initiated coagulation, in which the cleavage of factor V at Arg 1545 coincides with explosive activation of prothrombin (73,74). It is noteworthy that a similar situation is observed for factor VIII where cleavage at Arg 1689 by ␣-thrombin in the light chain of the cofactor is required for activation of the cofactor and optimum activity within the intrinsic tenase complex (75).

DISCUSSION
Although the species factors V IIa/NN , V RVV/NN , and V IIa ex- FIG. 8. Direct interaction of the hirudin-like peptide from the COOH terminus of factor Va heavy chain with thrombin-agarose. Chromatography of control and acidic peptides was performed on a thrombin-agarose column as described (30). Each run represents 400 g of peptide. This amount of peptide was determined to be saturable for the thrombin-agarose column used. Elution was performed with 2 M NaCl and was started at the point indicated by the arrow. The presence of the peptides in the corresponding fractions was monitored by absorbance at 280 nm. Identification of D13R in the elution fractions was performed by LC/MS as described under "Experimental Procedures." Results show P15H (filled squares), D13R (first run, filled circles), and D13R (second run, filled triangles).
hibited similar affinities for factor Xa, factors V IIa/NN and V RVV/NN had lower and more variable activity (20 -40%) in clotting assays using factor V-deficient plasma (Fig. 9D). However, for prothrombin activation at high (i.e. saturating) concentrations of factor Xa, the catalytic efficiency of the factor V IIa/NN -factor Xa complex was as high as that observed for the factor V IIa -factor Xa complex. Furthermore, we have found that the missing amino acid sequence from the heavy chain of factor V NN represents a binding site for prothrombin and that the clotting activity of factor V IIa/NN is considerably impaired by removal of residues 607-709 from the COOH-terminal region of the heavy chain. Thus, this region of the molecule is crucial for cofactor activity under low factor Xa concentrations (physiological conditions).
Bakker et al. (46), using a protease purified from the venom of the snake N. naja oxiana, prepared a species of factor Va missing an even larger portion (residues 683-709) of the heavy chain. These investigators concluded that the region 683-709 is FIG. 9. Schematic representation of various factor V species. The domain structure of the human procofactor (composed of 2196 amino acids) is as follows (8): the heavy chain region (residues 1-709) contains two A domains (A1-A2), which are associated through a region (amino acids 304 -316) containing mostly basic amino acid residues. The COOH-terminal portion of the heavy chain (657-709) contains a cluster of acidic amino acids, three potential sulfation sites, and Ser 692 , which is phosphorylated by a platelet-derived casein kinase II-like kinase (76,77). The light chain of factor Va (amino acids 1546 -2196) includes one A and two C domains (A3-C1-C2). Both chains are connected in the factor V molecule through the highly glycosylated B region. The binding site for factor Xa (AP4Ј) recently described is illustrated as part of the A2 domain (48), whereas peptide D13R identified in this study is shown as part of the COOH-terminal portion of the heavy chain of the cofactor. A, cleavage of factor V by NN results in a heavy chain containing amino acid residues 1-696 and a light-chain-spanning amino acid region 1509/1514 -2196. It is noteworthy that all NN cleavage sites described herein are found near the location of potential sulfation sites (84). The prothrombin binding site located on D13R is released from the heavy chain of factor Va and is part of the B region. B, our sequencing data demonstrate that the RVV-factor V activator activates factor V following two cleavages only: at Arg 1018 and Arg 1545 to produce a heavy chain (amino acid region 1-1018) and a light chain (amino acid region 1546 -2196). It is important to note that the consequence of cleavage at Arg 1545 and Arg 1018 of factor V by RVV-V activator is the production of two fragments of M r 150,000 (amino acid residues 1-1018 for one and 1019 -1545 for the other, the latter being highly glycosylated). Distinction between these two fragments on a SDS-PAGE gel following staining with Coomassie Blue or silver nitrate is sometimes impaired because: 1) the two fragments comigrate and 2) the highly glycosylated fragment (amino acid residues 1019 -1545) does not always stain with either Coomassie Blue or silver nitrate. C, ␣-thrombin produces an active cofactor, factor Va, following cleavage of factor V at Arg 709 , Arg 1018 , and Arg 1545 . These cleavages release a heavy chain of M r 105,000 (amino acid region 1-709) and a light chain of M r 74,000 (amino acid region 1546 -2196). The two subunits of the molecule are associated through Ca 2ϩ ions. D, ␣-thrombin and RVV-V activator cleave factor V NN at Arg 1545 . This cleavage releases a light chain (amino acid region 1546 -2196) and restores binding to factor Xa. However, the cofactor with the shorter heavy chain still has impaired clotting activity because of impaired interaction with prothrombin. required for the binding of factor Xa and/or prothrombin to factor Va. Thus, whereas amino acid region 697-709 contains a binding site for prothrombin and appears to be required for prothrombin interaction and expression of factor Va clotting activity under conditions of limiting factor Xa concentrations (physiological conditions), it is also possible that the additional stretch from amino acid residues 683-696 may be also required for factor Xa and/or prothrombin binding. Thus, this highly acidic COOH-terminal segment of the heavy chain, which has also been found to be phosphorylated by a platelet-derived casein kinase II (76,77), appears to play a critical regulatory role for factor Va cofactor function under limiting concentrations of factor Xa.
We have demonstrated an interaction between D13R, a hirudin-like peptide, and thrombin-agarose. We have also shown that D13R is a competitive inhibitor of prothrombinase function. We can thus conclude that D13R contains at least a portion of the binding site of factor Va for prothrombin. Several studies have suggested a contribution from the heavy chain of factor Va to the interaction of prothrombin with prothrombinase. Our findings are in complete agreement with the data from Dharmawardana and Bock (30) and Anderson et al. (33) demonstrating a productive interaction between factor Va and ABE-I of thrombin and prothrombin and suggesting that this interaction involves hirudin-like motifs from the COOH terminus of the factor Va heavy chain (33). The only hirudin-like motif of D13R is the sequence Asp-Tyr-Gln (70). Thus, our findings together with those on record provide for a new binding site for thrombin and prothrombin on the COOH terminus of the heavy chain of factor Va to which the amino acid residues Asp 697 -Gln 699 contribute.
Factor V requires proteolytic processing for expression of full cofactor activity (3, 4 -6, 12, 66, 67). The activation of factor V by ␣-thrombin, factor Xa (78, 79), RVV-factor V activator (80,81), plasmin (68,82), cathepsin G (83), human neutrophil elastase (83), and other enzymes has been the subject of investigation the past decade and has produced some controversy. Early data using bovine factor V suggested that generation of the heavy chain of the cofactor following cleavage at Arg 713 by ␣-thrombin (the equivalent of Arg 709 in the human molecule) is sufficient for expression of maximum cofactor activity in an assay measuring ␣-thrombin generation and employing purified reagents and saturating concentrations of factor Xa (4). Using a similar assay and human plasma factor V, it was reported that cleavage of human plasma factor V at Arg 1018 by factor Xa and generation of a heavy chain portion of the cofactor (amino acid residues 1-1018) and of a M r 220,000 fragment (amino acid residues 1019 -2196, factor V Xa ) was sufficient to promote maximum cofactor activity (79). No increase in cofactor activity was found when factor V Xa was treated with ␣-thrombin (factor V Xa/IIa ), and no differences in activities were observed between V Xa and factor V IIa . In contrast, using a clotting assay (5) or an assay that measures ␣-thrombin formation and employs limiting factor Xa concentrations (67), it was reported that cleavage of factor V at Arg 1545 and generation of the light chain alone was sufficient for optimum factor Xa activity within prothrombinase. The data presented here suggest that at saturating concentrations of factor Xa, cleavage at Arg 709 is sufficient for optimum factor Va cofactor expression, whereas under conditions of limiting factor Xa, cleavages at both Arg 1545 and Arg 709 are required to promote maximum cofactor activity. Hence, results obtained in a clotting assay or using factor Xa concentrations below the K d for the factor Xa-factor Va bimolecular interaction within prothrombinase cannot and should not be compared with the data obtained when using saturating amounts of factor Xa to study the effect of various factor Va species on prothrombinase assembly and function. Thus, differences between different groups concerning the "active species of factor Va" are most likely related to the assay conditions used rather than to specific proteolytic cleavages of the factor V molecule.