Removal of B-domain sequences from factor V rather than specific proteolysis underlies the mechanism by which cofactor function is realized.

Factor V, the precursor of factor Va, circulates in plasma with little or no procoagulant activity. Activity is generated following limited proteolysis indicating that the conversion of factor V to factor Va results in appropriate structural changes, which impart cofactor function. We have produced recombinant partial B-domain-truncated derivatives of factor V (FV(des811-1491) and FV(des811-1491) with Arg(709) and Arg(1545) mutated to Gln) to investigate whether discrete proteolysis within the B-domain followed by a conformational transition is responsible for activation. Direct binding fluorescence measurements as well as steady-state kinetic assays were employed to assess the ability of these factor V derivatives to assemble and function in prothrombinase. In contrast to human factor V, single-chain B-domain-truncated factor V bound to FXa membranes with an affinity that was identical to factor Va. Additionally, it was found that, once this modified derivative was assembled in prothrombinase, it functioned in an equivalent manner to factor Va. Taken together these data support the hypothesis that proteolysis within the B-domain of factor V, although necessary, is incidental to the mechanism by which cofactor function is realized. Instead, our results are more consistent with the interpretation that proteolytic activation of factor V simply eliminates steric and/or conformational constraints contributed by the B-domain that otherwise interfere with discrete binding interactions that govern the eventual function of factor Va.

The generation of thrombin at the appropriate time and place is crucial for balancing normal hemostasis and thrombosis. Thrombin is formed subsequent to proteolysis of two peptide bonds in prothrombin by the macromolecular enzyme complex prothrombinase, which assembles through reversible interactions between the serine protease factor Xa (FXa), 1 and the cofactor protein factor Va (FVa), on a suitable membrane surface in the presence of Ca 2ϩ ions (1). Factor Va is essential for the rapid generation of thrombin, and its removal from this enzyme complex profoundly reduces the rate of prothrombin activation (2).
The precursor of FVa is factor V (FV), a large (M r ϭ 330,000) heavily glycosylated, single-chain, multidomain (A1-A2-B-A3-C1-C2) protein, which is synthesized in the liver and is homologous to factor VIII (2,3). Factor V is considered a procofactor and is not known to function within the prothrombinase complex (4,5). This is consistent with the observation that FV binds very weakly, if at all, to FXa and prothrombin (6 -8) and indicates that proteolytic conversion of FV to FVa leads to appropriate structural changes that impart cofactor function.
Thrombin is established as the most robust activator of FV (2). Proteolysis occurs at Arg 709 , Arg 1018 , and Arg 1545 generating FVa IIa , a heterodimer composed of an N-terminal 105-kDa heavy chain associated via Ca 2ϩ ions to the C-terminal 74/71-kDa light chain (6, 9 -11). The large, heavily glycosylated Bdomain, spanning amino acids 710 -1545, is not necessary for cofactor activity and is released during activation (9).
The mechanism by which FV is maintained in an inactive state is not well understood. It is also not clear how macromolecular binding sites on the heavy and/or light chains are expressed following thrombin-mediated proteolysis. At least two mechanisms could underlie these processes. One possibility, drawn from the known mechanism of serine protease zymogen, protease-activated receptor (PAR), and fibrinogen activation, implies that discrete proteolysis followed by the release of a new N terminus and a conformational change accounts for activation (12)(13)(14). Support for this mechanism comes from studies using either recombinant FV or proteolyzed derivatives of FV, which demonstrate that cleavage at Arg 1545 is sufficient for activation and that release of the B-domain from the heavy chain is not necessary for expression of cofactor activity (6,(15)(16)(17)(18)(19). A second possibility, employed by proteins such as aspartic and cysteine proteases, implies that the prosegment or activation peptide sterically blocks functional binding sites and activation results from the removal of these sequences (12). Elimination of these steric sequences can be achieved by proteolysis, changes in pH, covalent modification, or noncovalent interactions with regulatory factors (20). Kane et al. (15,21) have described a recombinant partial B-domainless form of FV (FV des811-1491 ) that provides at least some support for this mechanism. This FV derivative is secreted as a single-chain protein and appears to have constitutive but partial activity of 30 -38% compared with FVa. However, the findings with this derivative also suggest an important role for discrete proteolysis, because full activity was only obtained following cleavage by thrombin or RVV-V (factor V activator from Russell's viper venom).
Numerous reports have attempted to correlate bond cleavage in FV with the generation of cofactor activity (9, 10, 15-19, 22, 23). Most of these studies have largely relied on functional measurements with thrombin generated in situ. This complicates interpreting the requirements for cofactor activation, because it is difficult to eliminate feedback proteolysis of FV. Additionally, it has been recognized that assay conditions can influence the conclusion as to which fragments of FV(a) are considered active further complicating interpretations (24). These potential shortcomings suggest limitations to the current understanding of the relationship between specific proteolysis in FV and the development of cofactor activity. In the current study we have expressed recombinant B-domain-truncated derivatives of FV and have used a series of physical and kinetic measurements to assess the role of B-domain sequences in contributing to the mechanism by which FV is maintained as a procofactor.
Construction of Recombinant FV del811-1491 (rFV-DT)-The FV cDNA in the pMT2 expression plasmid was purchased from ATCC (Manassas, VA). The cDNA was digested with SalI and subcloned into the pED expression vector, which was a generous gift from Dr. Monique Davis (Genetics Institute) (36). A partial B-domainless derivative of FV, FV del811-1491 , was constructed using the technique of splicing by overlap extension (37). This derivative has FV cDNA sequences corresponding to amino acids 811-1491 from the B-domain deleted. Specific oligonucleotides used were as follows: primer A, 5Ј-GAAGAGGTGGGAATAC-TT-3Ј corresponds to the cDNA sequence coding for amino acid residues 319 -325; primer B, 5Ј-CTCAATGTAATCTGTATCCTCTATAGGGT-C-3Ј in which the first 15 bases correspond to cDNA sequence coding for residues 1496 -1492 and the last 15 bases correspond to cDNA sequence coding for residues 810 -806; primer C, 5Ј-GACCCTATAGAGGATAC-AGATTACATTGAG-3Ј in which the first 15 bases correspond to cDNA sequence coding for residues 806 -810 and the last 15 bases correspond to cDNA sequence coding for residues 1492-1496; and primer D, 5Ј-TCTGTCCATGATAAGAAATGG-3Ј corresponds to the FV cDNA sequence coding for residues 1877-1871. The resulting 2632-bp fragment was digested with Bsu36I and SnaBI, gel purified, and sub-cloned into pED-FV digested with the same enzymes. To ensure the absence of polymerase-induced errors, the entire modified FV cDNA was sequenced. This rFV derivative has been previously described by Kane et al. (21), and the protein product has Asp 810 fused to Thr 1492 and was thus renamed to rFV-DT based on the nomenclature used in the FVIII field (38).
A rFV-DT mutant with Arg 709 and Arg 1545 replaced by Gln (rFV-DT QQ ) was generated with the QuikChange site-directed mutagenesis kit (Stratagene) using two complementary oligonucleotides containing the desired mutation, the sense strand for the R709Q mutation being 5Ј-CTGCAGCATTAGGAATTCAGTCATTCCGAAACTCATC-3Ј and for the R1545Q mutation 5Ј-GCAGCATGGTACCTCCAGAGCAACAA-TGGAAACAG-3Ј.
Expression and Purification of rFV-DT-Factor V expression plasmids were introduced into baby hamster kidney (BHK) cells using LipofectAMINE 2000 (Invitrogen) and pSV2neo as the selectable marker plasmid. Approximately 8 days later, 48 G418-resistant colonies from each construct were selected for expansion in 24-well plates. Following an initial screen for FV clotting activity, high producers were placed in T25 flasks and screened for protein production by enzymelinked immunosorbent assay and clotting activity. Protein expression levels of the clones varied from 0.5 to 4.0 g/10 6 cells/24 h. Selected clones were expanded into triple flasks (Nalge-Nunc, Naperville, IL) and cultured in Dulbecco's modified Eagle's medium/F-12 media supplemented with ITS, 2.5 mM CaCl 2 , and 1.0 mg/ml Albumax (Invitrogen). Conditioned media was collected for 4 -6 days, centrifuged, and stored at Ϫ20°C in the presence of protease inhibitors (10 mM benzamidine, 10 M APMSF).
Three liters of conditioned media was thawed at 37°C followed by the addition of protease inhibitors (10 M APMSF, 1 M FPR-CH 2 Cl, 0.5 ml/liter aprotinin). The media was loaded onto an anti-human FV-Sepharose column equilibrated in 20 mM Tris, 0.15 M NaCl, 1 mM benzamidine, pH 7.4. The column was washed with 20 mM Tris, 0.35 M NaCl, 1 mM benzamidine, pH 7.4, and then eluted with 20 mM Tris, 1.65 M NaCl, 1 mM benzamidine, pH 7.4. Fractions containing FV activity were pooled, dialyzed versus 20 mM Hepes, 150 mM NaCl, and 5 mM Ca 2ϩ , pH 7.4, and concentrated by ultrafiltration (Millipore). The final yield was ϳ2 mg of rFV-DT/liter of conditioned media, and the purified protein was stored at Ϫ80°C. Protein purity was assessed by SDS-PAGE using pre-cast 4 -12% gradient gels (Invitrogen) under reducing conditions followed by staining with Coomassie Brilliant Blue R-250.
Purification of rFVa-Purified rFV-DT (3 M, 10 mg) was incubated with 10 nM ␣-thrombin at 37°C for 15 min followed by the addition of 10 M FPR-CH 2 Cl. Following activation of rFV-DT, rFVa was purified by ion exchange chromatography as described (31).
Analytical Ultracentrifugation-Molecular weights were determined in a Beckman Optima XL-I analytical ultracentrifuge using interference optics. Sedimentation velocity was measured at 20°C with a protein concentration of 0.5 mg/ml in 20 mM Hepes, 150 mM NaCl, 5 mM CaCl 2 , pH 7.4, at 40,000 rpm with an AN60Ti rotor. Sedimentation coefficients and molecular weights were determined by g(s*) analysis (39). Extinction coefficients were determined by differential refractometry by the procedure described by Babul and Stellwagen (40).
FV-specific PT-based Clotting Assay-Factor V/Va (200 nM) derivatives were prepared in 20 mM Hepes, 0.15 M NaCl, 2 mM CaCl 2 , 0.1% polyethylene glycol 8000, pH 7.5 (assay buffer). For experiments in which pretreatment with thrombin was intended, samples (200 nM) were incubated at 37°C for 10 min with 2 nM thrombin, followed by the addition of 3 nM hirudin. Samples were then diluted to less than 1 nM in assay buffer with 0.1% albumin, and specific clotting activity using FV-deficient plasma was performed as described (18).
Fluorescence Intensity Measurements-Samples (2.5 ml) in assay buffer were maintained at 25°C in 1-ϫ 1-cm 2 stirred quartz cuvettes, and steady-state fluorescence intensity was measured using ex ϭ 480 and em ϭ 520 nm with a long pass filter (KV500, Schott, Duryea, PA) in the emission beam. Measurements, including controls, were performed essentially as described (29,41).

Kinetics of Macromolecular Substrate
Cleavage-Initial velocities of prethrombin-2 or prothrombin cleavage by prothrombinase were determined discontinuously at 25°C as described (42,43). Progress curves of prethrombin-2 activation were performed using the following reaction conditions: PCPS (50 M), DAPA (3 M), and prethrombin-2 (1.4 M) were incubated with the various cofactors (3.0 nM) in assay buffer, and the reaction was initiated with FXa (3.0 nM). At various time points, aliquots of the reaction mixture were quenched, thrombin generation was determined by using the chromogenic substrate S2238, and samples were also prepared for Western blotting analysis (see below). Dissociation constants (K d ) for FXa binding to membrane-bound cofactor were determined using the following reaction conditions: PCPS (50 M), DAPA (3 M), and prethrombin-2 (1.4 M) were incubated with various concentrations (0.3-18 nM) of cofactor in assay buffer, and the reaction was initiated with FXa (5 nM). The kinetic parameters of prothrombinase-catalyzed prothrombin or prethrombin-2 activation (K m and V max ) were determined as follows: using prothrombin as a substrate, PCPS (50 M), and the various cofactors (20 nM) were incubated with increasing concentrations of prothrombin (0 -5 M), and the reaction was initiated with FXa (0.1 nM); using prethrombin-2 as a substrate, assay mixtures contained PCPS (50 M), cofactor (20 nM), and various concentrations of prethrombin-2 (0 -20 M), and the reaction was initiated with FXa (5 nM).
Western Blot and Densitometric Analysis-Recombinant FV-DT QQ (20 ng) was subjected to SDS-PAGE (4 -12% gradient gel) and then electroblotted to nitrocellulose at 30 V for 1 h. The membrane was probed with a monoclonal antibody directed against the FVa light chain as previously described (18). The secondary antibody used was a goat anti-mouse IgG coupled to horseradish peroxidase (Rockland, Gilbertsville, PA). Visualization of bands was performed by photographic detection of enhanced chemiluminescence using film (Eastman Kodak Co., Rochester, NY). The film was imaged with transmitted light using a Kodak DC290 digital camera, and the bands were analyzed with the image analysis program Image J, Version 1.30v (National Institutes of Health). The exposure times and aperture settings for the imaging step were selected to yield a linear densitometric response.
Data Analysis-Data were analyzed according to the referenced equations by nonlinear least squares regression analysis using the Marquardt algorithm (44). The qualities of the fits were assessed by the criteria described (45). Fitted parameters are reported Ϯ 95% confidence limits. Dissociation constants (K d ) and stoichiometries (n) for the interaction between FXa and membrane-bound FVa were obtained from the dependence of the fluorescence intensity or initial rate of prethrombin-2 activation on the concentrations of cofactor (46). Initial velocity measurements of macromolecular substrate (prothrombin or prethrombin-2) cleavage by prothrombinase were analyzed by fitting the data to the Henri-Michaelis-Menten equation (47), to yield fitted values for K m and V max .

RESULTS
Expression and Purification of rFV-DT and rFV-DT QQ -Recombinant FV-DT and rFV-DT QQ were expressed in BHK cells and purified to homogeneity by immunoaffinity chromatography with typical yields of ϳ2 mg/liter. SDS-PAGE analyses ( Fig. 1) indicated that rFV-DT and rFV-DT QQ migrated with an apparent molecular weight of 200,000. Recombinant FV-DT was completely processed to rFVa following incubation with thrombin, whereas rFV-DT QQ was not significantly cleaved. Plasma-derived FV (PD-FV) and PD-FVa used in this study as controls are included on the gel. Recombinant FVa (rFVa) used as an additional control was purified following treatment of rFV-DT with thrombin and migrated on the gel in a manner similar to PD-FVa.
Physical Characterization of Recombinant Proteins-Sedimentation velocity studies were conducted to assess hydrodynamic molecular weights, extinction coefficients as well as homogeneity (Table I). The data obtained for PD-FVa and rFVa agreed well with published reports for bovine FVa (48 -50). The determined molecular weight of rFV-DT was consistent with SDS-PAGE (Fig. 1) and with the results of mass spectrometry (M r ϭ 208,000) using a surface-enhanced laser desorption/ ionization instrument (SELDI, Ciphergen, Malvern, PA). Nterminal sequence analysis revealed that rFV-DT was correctly processed by BHK cells as the N terminus was found to be: AQLRQFYVAAQG.
Assessment of Cofactor Activity by Clotting Assay-Initial experiments focused on the clotting activity of purified rFV-DT and rFV-DT QQ prior to and following treatment with thrombin (Fig. 2). Treatment of PD-FV with thrombin resulted in a 10-to 15-fold increase in cofactor activity, with levels comparable to those observed with purified PD-FVa and rFVa. In contrast, rFV-DT and rFV-DT QQ had activities comparable to FVa and pretreatment of these FV derivatives with thrombin resulted in only a minor change in their clotting activities. These results suggest that rFV-DT and rFV-DT QQ behave as fully active  cofactors and not procofactors. However, an alternative explanation is that these proteins are more rapidly activated by thrombin or other proteases produced during the clotting assay compared with PD-FV.
Binding of FVa and rFV-DT to Membrane-bound FXa-The ability of the FV derivatives to assemble into the FXa-FVa membrane ternary complex was assessed by equilibrium binding fluorescence measurements. Using a fixed concentration of OG 488 -FXa, subsequent titration with increasing concentrations of PD-FVa or rFVa (Fig. 3A), yielded a saturable increase in fluorescence intensity with dissociation constants (K d ) and stoichiometries (n) comparable to previously published values using similar methodologies (Table II) (29). In contrast, no significant increase in fluorescence intensity was observed when using single-chain human PD-FV (Fig. 3A) even if the titration was extended to 175 nM. Additional experiments also established that the procofactor did not compete for binding to pre-formed PD-FVa-OG 488 -FXa-PCPS (data not shown). The data yielded a lower limit estimate K d Ͼ500 nM for the interaction between FV and FXa on a membrane surface. Thus the conversion of FV to FVa leads to a large change (Ͼ500-fold) in its affinity for FXa on the membrane surface.
In contrast to PD-FV, single-chain rFV-DT and rFV-DT QQ bound to FXa on the membrane surface with an affinity identical to FVa (Fig. 3B and Table II). SDS-PAGE analysis of each of the FV derivatives at the beginning and end of the experiment (taken from cuvette) indicated that they were not cleaved during the experiment (Fig. 3, insets). These data establish that rFV-DT and rFV-DT QQ bind with high affinity to FXa membranes and indicate that proteolysis of these proteins is not a requirement for high affinity binding. Overall the data are consistent with the conclusion that the exposure of a FXa binding site on FVa results from removal of steric and/or conformational constraints contributed by portions of the B-domain present in FV but not rFV-DT.
Assessment of Cofactor Activity-A concern with functional measurements is rapid feedback proteolysis of rFV-DT or rFV-DT QQ by FXa and thrombin during assessment of initial rate. To accommodate this point, reactant concentrations were chosen to minimize proteolysis, and a reversible inhibitor of thrombin (DAPA) was included in the assay. Progress curves of the conversion of prethrombin-2 to thrombin were generated using equimolar concentrations of cofactor (3 nM) and FXa (3 nM). Under these conditions, PD-FV had very little activity, and the progress curve was characterized by an obvious lag indicating an increase in the rate of thrombin formation with time ( Fig. 4). This observation has been attributed to the proteolytic activation of the procofactor by thrombin or FXa during the assay (4,5). In contrast, thrombin generation increased linearly over time, with no obvious lag with rFV-DT and rFV-DT QQ . The initial rates of thrombin generation for PD-FVa, rFVa, and rFV-DT QQ were essentially equivalent, whereas the rate obtained with rFV-DT was ϳ1.4-fold greater.
A more extensive assessment of functional activity was conducted using increasing concentrations of cofactor at a single, fixed concentration of FXa and PCPS (Fig. 5). Each of the data sets was analyzed to extract the inferred equilibrium dissociation constants for membrane bound cofactor to FXa (Table II). These data agreed with the dissociation constants and stoichiometries obtained from direct binding measurements in the absence of substrate and indicate that rFV-DT and rFV-DT QQ bind to FXa-PCPS with an affinity identical to FVa.
Cofactor function of each of the derivatives was further assessed by initial velocity studies using prethrombin-2 as a substrate (Fig. 6). Using saturating concentrations of membranes and cofactor the resulting steady-state kinetic constants (Table III) indicated that once assembled in prothrombinase rFV-DT and rFV-DT QQ function equivalently to PD-FVa and rFVa. Similar results were obtained when prothrombin was   Fig. 3). For the kinetic measurements, initial, steady-state rates were determined in assay buffer containing various concentrations of cofactor species, 50 M PCPS, 1.4 M prethrombin-2, 3.0 M DAPA, and 5 nM FXa as described under "Experimental Procedures" (see Fig. 5 (Table III).

Assessment of Feedback Proteolysis during Functional
Measurements-Significant proteolysis during the preincubation phase prior to FXa addition or cleavage by thrombin or FXa during the initial phases of the reaction could account for the observed activities seen with rFV-DT and rFV-DT QQ . Preliminary experiments indicated that rFV-DT and rFV-DT QQ were proteolyzed during the initial rate measurement. The extent of proteolysis was dependent upon reaction time and the concentrations of cofactor and FXa. Under all conditions examined, rFV-DT was proteolyzed to a greater degree than rFV-DT QQ . Because proteolysis could not be completely eliminated, the relationship between rFV-DT QQ cleavage and expression of cofactor activity was directly assessed. Cleavage was monitored by Western blotting, and the disappearance of rFV-DT QQ was analyzed densitometrically. Analysis of the status of rFV-DT QQ prior to the addition of FXa, but following the 5-min preincubation phase revealed that the protein was not significantly proteolyzed (Fig. 7A). These results were confirmed using an antibody directed against the FVa heavy chain (data not shown). Following addition of FXa, a significant amount of proteolysis was seen over the 2-min time course. The principal product observed was a ϳ87/85-kDa fragment that is derived from the C-terminal region of the protein. Based on the size of the fragments, the doublet likely represents multiple sites of cleavage within the 155-amino acid B-domain or a single cleavage product with differential carbohydrate content at position Asn 2181 (51).
Densitometric analysis was performed to assess the time-dependent disappearance of rFV-DT QQ and determine whether this correlated with cofactor activity. A plot of the rate of thrombin generation using either rFV-DT QQ or rFVa (left axis; first derivative of data from Fig. 4), and the density of the 216-kDa fragment (right axis) versus time is shown in Fig. 7B. As expected, the rate of thrombin generation at any given time during the initial velocity measurement remained essentially unchanged for either rFV-DT QQ or rFVa, indicating that these proteins have a constant level of activity upon introduction into the assay. In contrast, the density of the 216-kDa fragment decreased linearly over time, indicating that cleavage of rFV-DT QQ does not correlate with the expression of cofactor activity, because all of the activity is observed prior to the disappearance of this peptide. Similarly, a plot of the total density of the ϳ87/85-kDa fragment over time revealed that the appearance of this peptide did not correlate with the expression of activity (data not shown). Taken together these data indicate that rFV-DT and rFV-DT QQ must have intrinsic cofactor activity that is comparable to FVa even in the absence of proteolysis.

DISCUSSION
The majority of circulating proteins involved in the hemostatic process are synthesized as inactive precursors. They express binding sites important to their function following proteolysis and liberation of prosegments or activation peptides. The paradigm for this type of activation mechanism is the zymogen to protease transition in the chymotrypsin-like serine protease family where bond cleavage at a highly conserved site unmasks a new N terminus, which acts as an intramolecular ligand facilitating a conformation change that leads to activation (12). This type of mechanism is also employed in the activation of PAR and in the clotting of fibrinogen (13,14). One alternative to this activation strategy is that structural deter- The lines were drawn following analysis of all data sets to a rectangular hyperbola, and the fitted kinetic constants can be found in Table III. The data are representative of two similar experiments. minants contributing to protein function are rendered sterically inaccessible by prosegments or activation peptides. Activation is achieved following removal of these sequences via proteolysis or by some other means (12).
The current work provides strong evidence that FV activation simply results from the elimination of steric and/or conformational constraints contributed by the B-domain that otherwise interfere with binding interactions with FXa and possibly prothrombin. These findings set the mechanism of FV activation apart from the paradigms established for serine protease zymogens. One hallmark of that mechanism is the inability to generate constitutively active proteins when prosegments or activation peptides are truncated (28,52). Our results indicate that this is not the case with B-domain-truncated rFV-DT, which exhibits functional activity comparable to FVa. Surprisingly, the mechanism responsible for the expression of FVa cofactor activity also appears to be distinct from that of FVIII. For example, procofactor activity is maintained in FVIII derivatives, which lack major portions of the B-domain (53)(54)(55)(56)(57)(58). Additionally, biochemical and mutational analysis indicate that cleavage of FVIII at Arg 372 between the A1 and A2 domains is critical for the generation of cofactor activity (58,59). These observations imply that the B-domains of FV and FVIII appear to function quite differently, at least with respect to maintaining the procofactor state.
Evidence to support our conclusions derives from experiments that establish that, in the absence of intentional proteolysis, rFV-DT and rFV-DT QQ assemble and function in prothrombinase in an equivalent manner to FVa. Direct binding fluorescence measurements indicate that single-chain rFV-DT and rFV-DT QQ bind to FXa membranes with an affinity that is identical to FVa. However, we could not detect any significant binding of FV to FXa membranes, extending previous observations using chromatography-based measurements (6). These results markedly contrast the recent findings by Steen and Dahlbä ck (17) who report that rFV bound to FXa membranes with an affinity of 10 nM. Thus our data indicate that removal of significant amounts of the B-domain in the absence of proteolysis results in the exposure of the FXa binding site on the heavy and/or light chains. Exposure of this site could simply result from removal of the large heavily glycosylated B-domain that is physically blocking it. Alternatively, the B-domain could function allosterically by making intramolecular contacts with heavy and/or light chain sequences inducing changes that mask the FXa binding site. A third possibility is that the generation of rFV-DT has forced or constrained heavy-and/or light-chain sequences such that their modified positions match a conformation identical to FVa. At present our data cannot distinguish between these possibilities. With respect to our findings that human FV does not detectably bind active siteblocked FXa, it should be noted that FXa membranes are known to activate the procofactor, indicating the two proteins must interact (5,60). These observations collectively suggest that FV and FVa bind to FXa membranes differently with active site interactions playing a dominant role in FV but not FVa recognition.
Although the conclusions from the direct binding measurements were unambiguous, determining whether or not these FV derivatives have functional activity presented a challenge. However, combining initial velocity measurements with Western blotting and densitometric analyses, we were able to establish that rFV-DT QQ and by extension rFV-DT, must have con-  stitutive cofactor activity. Although proteolysis of these FV derivatives was clearly evident during the initial rate period, it did not correlate with the expression of functional cofactor activity. We also consistently found that rFV-DT in our purified component assays appeared to have a 1.3-to 1.5-fold increase in cofactor activity compared with FVa and rFV-DT QQ , a result that was not found in the clotting assay. The molecular basis for this observation is not entirely clear but may relate to rapid and extensive proteolysis of this FV derivative during initial velocity studies. Overall the findings of our functional measurements highlight the complexity of ascribing levels of activity to various FV derivatives using assay systems in which active proteases are present.
Recombinant FV-DT was originally described by Kane et al. (21) and was shown to have partial cofactor activity of 30 -38% relative to FVa when transiently expressed in COS-7 cells. The molecular basis for this partial constitutive activity, however, was not investigated. Full activity was achieved following proteolysis at Arg 1545 by RVV-V or thrombin, indicating that discrete proteolysis resulting in liberation of the light chain is required for maximal activity (15). Our characterization of rFV-DT yielded different results that lead to the new and surprising conclusion that discrete proteolysis does not underlie the mechanism by which cofactor function is realized. A reasonable explanation for the opposing observations is that various post-translational modifications in COS-7 versus the BHK-derived protein are different, particularly the carbohydrate content. Recombinant FV-DT has 19 potential N-linked glycosylation sites, 7 of which are contained within the short 155-amino acid B-domain. It is possible that the density and composition of the carbohydrate between COS-7 and BHKderived rFV-DT could influence their levels of constitutive activity. For example, Bruin et al. (61) have noted that removal of sialic acid from FV resulted in a 2-fold increase in clotting activity. Whether these differences exist and contribute to changes in functional activity remains to be established; however, it is tempting to speculate that carbohydrate in addition to specific sequences within the B-domain play a role in maintaining FV as a procofactor. The generation of FV derivatives with variable amounts of B-domain sequence may help determine whether the length of this domain or its sequence, carbohydrate content, or its charge distribution contributes to the ability of this region to maintain FV in a procofactor state. The development of improved methods for producing recombinant FV derivatives in high yields presented in this report should facilitate these studies and permit the assessment of cofactor function by novel approaches requiring large amounts of protein.
In summary, the results of the present study establish that the FV-B-domain plays a predominant role in maintaining the procofactor state. Although it is well established that proteolysis within the B-domain of FV correlates with activation, our findings suggest that it is incidental and that this correlation cannot account for the mechanism by which cofactor function is realized. The removal of B-domain constraints, either through proteolysis or through recombinant truncation, results in the same outcome, a fully active cofactor that assembles and functions within prothrombinase. These findings set the mechanism of FV activation apart from that of serine protease zymogen activation and challenge existing ideas explaining the relative importance of specific proteolysis to the mechanism of FV activation.