Proteolysis at Arg740 Facilitates Subsequent Bond Cleavages during Thrombin-catalyzed Activation of Factor VIII*

Thrombin activates factor VIII by proteolysis at three P1 residues: Arg372, Arg740, and Arg1689. Cleavage at Arg372 and Arg1689 are essential for procofactor activation; however cleavage at Arg740 has not been rigorously studied. To evaluate the role for cleavage at Arg740, we prepared and stably expressed two recombinant B-domainless factor VIII mutants, R740H and R740Q to slow and eliminate, respectively, cleavage at this site. Specific activity values for the variants were ∼50 and 20%, respectively, that of wild-type factor VIII. Activation of factor VIII R740H by thrombin showed an ∼40-fold reduction in the rate of A2 subunit generation, which reflected an ∼20-fold reduction in cleavage rate at Arg372. Similarly, a ∼40-fold rate reduction in cleavage at Arg1689 and consequent generation of the A3-C1-C2 subunit were observed. Rate values for A2 and A3-C1-C2 subunit generation were reduced by >700-fold and ∼140-fold, respectively, in the R740Q variant. These results suggest that initial cleavage at Arg740 affects cleavage at both Arg372 and Arg1689 sites. Results obtained evaluating proteolysis of the factor VIII mutants by factor Xa revealed more modest rate reductions (<10-fold) in generating A2 and A3-C1-C2 subunits from either variant, suggesting that factor Xa-catalyzed activation of factor VIII was significantly less dependent upon prior cleavage at residue 740 than thrombin. Overall, these results support a model whereby cleavage of factor VIII by thrombin is an ordered pathway with cleavage at Arg740 facilitating cleavages at Arg372 and Arg1689, which result in procofactor activation.

Factor VIII is a plasma protein in which defects or deficiencies cause hemophilia A, the most frequently occurring severe inherited bleeding disorder. The activated form of factor VIII, factor VIIIa, functions as a cofactor for factor IXa, increasing its catalytic efficiency by several orders of magnitude in the phospholipid-and Ca 2ϩ -dependent conversion of factor X to factor Xa (1). Factor VIII is synthesized as an ϳ300-kDa single chain polypeptide (corresponding to 2332 amino acids) and is designated as six domains based on internal homologies: NH 2 -A1-A2-B-A3-C1-C2-COOH (2,3). Moreover, preceding the A3 domain and following the A1 and A2 domains are short segments containing high concentrations of acidic residues designated a1 (residues 337-372), a2 (residues 711-740), and a3 (1649 -1689). Factor VIII is processed by cleavage at the B-A3 junction to generate a divalent metal ion-dependent heterodimeric protein composed of a heavy chain (A1-a1-A2-a2-B domains) and a light chain (a3-A3-C1-C2 domains) (2)(3)(4)(5).
The inactive factor VIII procofactor is converted to factor VIIIa through limited proteolysis catalyzed by thrombin or factor Xa (6,7). Thrombin is believed to act as the physiological activator of factor VIII, as association of factor VIII with von Willebrand factor impairs the capacity for the membrane-dependent factor Xa to efficiently activate the procofactor (6 -8). Activation of factor VIII occurs through proteolysis by either protease via cleavage of three P1 residues at Arg 740 (A2-B domain junction), Arg 372 (A1-A2 domain junction), and Arg 1689 (a3-A3 junction) (6). After factor VIII activation, there is a weak electrostatic interaction between the A1 and A2 domains of factor VIIIa (9,10) and spontaneous inactivation of the cofactor occurs through A2 subunit dissociation from the A1/A3-C1-C2 dimer (11).
Cleavage at Arg 372 exposes a cryptic functional factor IXainteractive site in the A2 domain (11), while cleavage at Arg 1689 liberates factor VIII from von Willebrand factor (12) and contributes to factor VIIIa specific activity (13,14), thus making both sites essential for procofactor activation. Although it is known that cleavage at Arg 740 represents a fast step relative to cleavage at other sites in the activation of factor VIII (15), the role for this event is unknown. In the present study, cleavage at Arg 740 is examined using recombinant factor VIII variants possessing single point mutations of R740Q and R740H. Our results indicating altered rates of generation of factor VIIIa subunits dependent upon the residue at position 740 suggest an ordered mechanism for thrombin activation of factor VIII, whereby initial cleavage at Arg 740 is required for efficient cleavage at both Arg 372 and Arg 1689 . Alternatively, factor Xa-catalyzed cleavages at Arg 372 and Arg 1689 show less dependence upon initial cleavage at Arg 740 , suggesting that activation of factor VIII by this proteinase proceeds by a different mechanism.

EXPERIMENTAL PROCEDURES
Reagents-The monoclonal antibody 1A4 recognizing the A3 domain was a generous gift from Bayer. The anti-A2 domain factor VIII monoclonal antibody R8B12 (9) was obtained from Green Mountain Antibodies. The monoclonal antibody ESH8 (16) recognizing the C2 domain was purchased from American Diagnostica. The reagents human ␣-thrombin, factor IXa, fac-* This work was supported by National Institutes of Health Grants HL38199 and HL76213. An account of this work was presented at the 48th meeting of the American Society of Hematology, Orlando, FL on December 9, 2006. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. tor X, and factor Xa were purchased from Enzyme Research Laboratories. Recombinant nonsulfated hirudin and horseradish peroxidase-labeled streptavidin were from Calbiochem. Chromogenic factor Xa substrate Pefa-5523 (Pefachrome FXa) was purchased from Centerchem. Factor VIII-deficient plasma was prepared as previously described (17). Phospholipid vesicles containing 20% phosphatidylserine, 40% phosphatidylcholine, and 40% phosphatidylethanolamine were prepared using N-octyl glucoside as previously described (18). The Bluescript factor VIII vector (pBS factor VIII) and B-domainless factor VIII expression construct RENeo factor VIII were kindly provided by Pete Lollar and John Healey (Emory University).
Construction, Expression, and Purification of Recombinant Factor VIII-B-domainless factor VIII cDNA was restricted from the factor VIII expression construct FVIIIHSQ-MSAB-NotI-RENeo using the endonucleases XhoI and NotI and cloned into the pBluescript II K/S-vector. The B-domainless factor VIII cDNA was further restricted using endonucleases SacII/ApaI and subcloned into the pBluescript II K/S-vector. The R740H and R740Q mutations were introduced into the construct using the Stratagene QuikChange site-directed mutagenesis kit as previously described (19). The presence of only the desired mutation was confirmed using dideoxy sequencing. The mutated factor VIII cDNA was then ligated back into the factor VIII expression construct, and subjected to a second round of dideoxy sequencing to confirm that only the desired mutation was present. FuGENE6 (Roche Applied Science) was used to transfect the factor VIII expression vector into BHK cells. The selection, subcloning, and cloning of stable transfectants were performed by standards methods, and the cloned cells were cultured in roller bottles for protein expression (20). The conditioned media was collected daily, and the expressed proteins were purified by SP-Sepharose (Amersham Biosciences) column chromatography as previously described (20). A one-stage clotting assay was used to detect active fractions. The concentrations of purified factor VIII proteins were determined by the method of Bradford (21). Factor VIII specific activity values were determined using one-stage clotting and Coomassie dye binding assays (21,22). Resultant factor VIII was more than 90% pure as judged by SDS-PAGE with the main contaminant as albumin. Factor VIII samples were quick-frozen and stored at Ϫ80°C.
Factor Xa Generation Assay-The rate of conversion of factor X to factor Xa was monitored in a purified system (23). For the factor VIII activation time course following thrombin addition, factor VIII (1 nM) was reacted with thrombin (0.05 nM) in the presence of phospholipid vesicles (10 M) at 22°C. Samples were removed at indicated time intervals, thrombin activity was inhibited by the addition of hirudin (0.1 units/ml), and the reactions were immediately assayed following addition of factor IXa (20 nM) and factor X (300 nM). For the experiments assessing the contribution of factor Xa to factor VIII activation, factor VIII (1 nM) and factor IXa (20 nM) were reacted in the absence or presence of thrombin (20 nM) for 1 min. Hirudin (10 units/ml) was added, and reactions were initiated after 1 min by addition of factor X (300 nM). The reactions were terminated with EDTA (50 mM) at the indicated times. Rates of factor Xa generation were determined by the addition of the chromogenic substrate Pefa-5523 (0.46 mM final concentration). Reactions were read at 405 nm for 5 min using a V max microtiter plate reader (Molecular Devices). To assess the K i (apparent), various concentrations of the R740Q factor VIII (0 -60 nM) were added to a reaction containing wild-type factor VIII (5 nM) and thrombin (0.05 nM) in the presence of phospholipids (10 M) for 1 min. Thrombin was inactivated by addition of hirudin (0.1 units/ml), and each sample was reacted with factor IXa (20 nM) and factor X (300 nM). Aliquots were removed at appropriate times to assess initial rates of product formation, added to tubes containing EDTA (50 mM final concentration), and processed as described above.
Cleavage of Factor VIII by Thrombin or Factor Xa-Factor VIII (100 nM) was reacted with 2.5 nM thrombin in a buffer containing 20 mM HEPES (pH 7.2), 0.1 M NaCl, 5 mM CaCl 2 , and 0.01% Tween 20. Alternatively, factor VIII (100 nM) was reacted with factor Xa (2.5 nM) in the presence of phospholipid vesicles (10 M) in the above buffer. Reactions were run at 22°C and samples were taken at indicted time points during the time course. The reactions were terminated by addition of SDS-PAGE sample buffer and boiling for 3 min.
Electrophoresis and Western Blotting-Samples were run by SDS-PAGE on 8% polyacrylamide gels. Electrophoresis was carried out using a Bio-Rad mini gel apparatus at 175 V for 1 h. Proteins were transferred to polyvinylidene fluoride membrane for Western blotting. Blots were probed using the anti-factor VIII monoclonal antibodies R8B12 or 1A4, followed by incubation in a goat anti-mouse alkaline phosphatase-linked secondary antibody (Sigma). The signal was detected using the ECF (enhanced chemifluorescence) system (Amersham Biosciences), and the blots were scanned at 570 nm using Storm 860 (Molecular Devices). Densitometric scans were quantitated from linear density regions of the blots using Image Quant software (Molecular Devices).
Data Analysis-All experiments were performed at least three separate times and the average values are shown. Western blots were analyzed by densitometry and nonlinear least squares regression and the initial time points were fitted to the single-exponential Equation 1, where A o is 100 nM; the total A2 or A3-C1-C2 generated, k is the rate constant in minutes Ϫ1 , and t is the time in minutes. Rates presented were determined by normalizing the rate constant by the concentration of thrombin and factor Xa.
Kinetic parameters were determined by Xa generation analysis. K m and V max were calculated by fitting the data using nonlinear least-squares regression analysis to the Michaelis-Menten Equation 2, where v o is the initial velocity in (nM/min), and [S] is the concentration of wild-type factor VIII in nM. The K i for R740Q factor VIII was calculated by fitting the data to the competitive inhibition Equation 3 by nonlinear least squares regression analysis, where [S] is the concentration of wild-type factor VIII in nM, [I] is the concentration of R740Q factor VIII in nM, and K i is the inhibitor dissociation constant for R740Q factor VIII.
Mass Spectrometry-Wild-type and R740H factor VIII (250 nM) was digested with thrombin (12.5 nM) in non-denaturing conditions (50 mM ammonium bicarbonate buffer, pH 8.0) for 2 h at 22°C. Digestions were terminated by addition of 0.1% trifluoroacetic acid (final concentration). MALDI-TOF 2 mass spectra of the samples were collected on a Bruker Autoflex III smartbeam MALDI-TOF/TOF mass spectrometer at the Proteomics Center at the University of Rochester Medical Center. The MS/MS spectra were collected using a positive ion mode using a lift method to fragment the parent peak. The tandem mass spectra were identified using the MASCOT program to match experimental masses to the factor VIII sequence.

Characterization of Recombinant
Factor VIII Proteins-While earlier studies have identified mechanistic roles for cleavage of factor VIII at Arg 372 and Arg 1649 during its conversion from procofactor to cofactor, limited information exists on the role(s) served by cleavage at the relatively fast reacting Arg 740 separating the A2 and B domains in factor VIII heavy chain. To evaluate thrombin cleavage at Arg 740 , two recombinant factor VIII mutants, R740H and R740Q were prepared and stably expressed. These residues were selected based upon a prior study evaluating cleavage at the Arg 372 site (24). In that study, the thrombin-catalyzed cleavage rate at His 372 site was reduced ϳ80fold relative to Arg at this position, whereas no detectable cleavage was observed for the P1 Gln 372 mutation. Thus, these mutations were employed at the 740 position to, respectively, slow and eliminate proteolysis at this site. Specific activity values for the wild type, R740H, and R740Q factor VIII as measured using a one-stage clotting and Coomassie dye binding assays were 2.8 Ϯ 0.3 units/g, 1.5 Ϯ 0.2 units/g, and 0.6 Ϯ 0.1 units/g, respectively. These values suggested a near-normal phenotype (ϳ50% wild type) for the R740H variant, whereas the R740Q mutation approximated a mild hemophilic phenotype with a specific activity that was ϳ20% that of wildtype factor VIII.
Thrombin-catalyzed Cleavage of Recombinant Factor VIII and the Generation of A2 Subunit-To examine cleavage at residue 740 and its effect on other activating cleavages in factor VIII, reactions using catalytic levels of thrombin (2.5 nM) with factor VIII (100 nM) were performed and results visualized using SDS-PAGE and Western blotting. Wild-type or mutant B-domainless factor VIII expressed in BHK cells is secreted in two forms, a single chain (contiguous heavy chain and light chain) and the heterodimer. These forms along with the corresponding cleavage sites yielding factor VIIIa are schematically illustrated in Fig. 1A. Although the recombinant factor VIII is described as B-domainless, there is a short peptide remnant of the B-domain separating the A2 domain and the a3 segment FIGURE 1. Generation of A2 subunit following thrombin-catalyzed cleavage of recombinant factor VIII mutants. Panel A is a schematic of the secreted B-domainless factor VIII single chain (i) and heterodimer (ii) forms, and thrombin cleavage sites. Factor VIII (100 nM) was reacted with thrombin (2.5 nM) for the indicated times as described under "Experimental Procedures." Samples were run on 8% polyacrylamide gels followed by Western blotting using an anti-A2 antibody. Panel B shows the reaction time course of wild-type and R740H factor VIII, while panel C shows the time course of wild-type and R740Q factor VIII. Panel D shows quantitative densitometry of A2 subunit (nM) generation from blotting data from panels B and C, where OE indicates wild type; f, R740H; and ࡗ, R740Q. Initial time points were fit to the single exponential equation using non-linear least squares regression. The abbreviations WT, SC, A1-A2-BЈ, a3, and LC represent wild type, single chain, heavy chain with the small B-domain remnant, the N-terminal acidic region (residues 1649 -1689) that precedes the A3 domain in the light chain, and intact light chain, respectively. (Fig. 1A). Results from time courses of the thrombin-catalyzed cleavage reactions of wild type, R740H, and R740Q proteins probed by Western blotting using the anti-A2 domain-specific monoclonal antibody, R8B12 (25) are shown in Fig. 1, B and C, respectively. In examining cleavage of wild-type factor VIII, efficient cleavage of both single chain and heavy chain resulted in the rapid generation of the A2 subunit over the time course.
The cleavage profiles for the Arg 740 mutants were substantially altered from that of wild-type factor VIII. The R740H mutant showed reduced rates of cleavage at all thrombin-sensitive sites as judged by retention of significant levels of single chain factor VIII over an extended time course. Furthermore, factor VIII heavy chain (contiguous A1-A2) also persisted with little generation of A2 subunit as compared with the wild-type material. Two cleavage intermediates derived from single chain factor VIII were detected in this reaction that were less prominent in the reaction with wild-type factor VIII. One represented a contiguous A2-light chain possessing a molecular mass intermediate between single chain and factor VIII heavy chain, and derived from cleavage at Arg 372 , but not His 740 or Arg 1689 . The second was designated A2-a3 and derived from cleavage at Arg 372 and Arg 1689 , but not His 740 . The origin of this latter fragment was confirmed following blotting with antibody 10104 (data not shown), which is specific for the a3 segment. The overall effect of the R740H mutation was an ϳ44-fold reduction in the rate of generation of A2 subunit 3 as a result of impaired cleavage at Arg 372 and His 740 ( Fig. 1D and Table 1). Based upon summing band densities for the intermediates A2-LC and A2-a3, and A2 subunit, all of which represent products derived following cleavage at Arg 372 , the rate of cleavage at this site was reduced ϳ17-fold compared with wild-type factor VIII. These results suggest that slowing thrombin-catalyzed proteolysis by incorporating a non-optimal P1 His at residue 740 yields significant reductions in subsequent cleavage rates at Arg 372 and Arg 1689 (see also below), which are required to yield the active cofactor.
To further test the hypothesis that cleavage at residue 740 regulates rates at subsequent sites, the P1 Arg 740 residue was replaced with a non-cleavable Gln residue. Results from this analysis (Fig. 1C) showed further reduced rates of A2 subunit generation. Although weak band densities for the products were observed even at extended time points, the rate of A2 generated was determined to be reduced an additional ϳ17fold with the Gln mutation (Table 1). Furthermore, because this is a non-cleavable substrate, the A2 subunit generated likely terminates with the B-domain remnant. These data are consistent with a model where the rate of cleavage at Arg 740 facilitates subsequent cleavage at Arg 372 suggesting an ordered reaction pathway.
Cleavage at Arg 1689 in Factor VIII Proteins by Thrombin-Previous results have suggested that the presence of heavy chain promotes thrombin proteolysis of light chain (14). The above observations monitored using the anti-A2 domain antibody support this contention. To directly assess effects of mutation at residue 740 on cleavage at the a3-A3 junction (Arg 1689 ), Western blot reactions similar to those above were run with the A3 domain-specific antibody, 1A4. Blotting data evaluating rates of generation of A3-C1-C2 subunit for the wild type and R740H and R740Q variants are shown (Fig. 2, A and B, respectively). These blots were quantitated by scanning densitometry, and results are presented in Fig. 2C. Because generation of A3-C1-C2 is derived from a single cleavage at Arg 1689 , rates for cleavage and subunit generation are equivalent. The cleavage rate for thrombin at Arg 1689 was reduced ϳ45-fold for R740H factor VIII and ϳ143-fold for that site in R740Q factor VIII compared with the wild-type protein (Table 1).
These values were similar to the observed cleavage rate reductions at Arg 372 for these variants, although rates for A2 subunit generation were retarded compared with A3-C1-C2 subunit generation due to the requirement for additional cleavage at the mutated P1 740 residue to generate A2 subunit. Therefore, under these conditions, generation of A2 subunit becomes rate-limiting in the activation pathway. Overall, these results indicate cleavage at residue 740 is required to facilitate subsequent cleavages at both Arg 372 and Arg 1689 .
Identification of the Liberated B Domain Peptide from R740H Factor VIIIa-Because cleavage at residue 740 liberates a small peptide remnant of the B-domain, validation that cleavage was occurring at the P1 His residue was determined by MALDI-TOF/TOF mass spectrometry. Thrombin digests of both wildtype and the R740H variant were prepared as described under "Experimental Procedures." Mass spectrometry yielded a peak at 1538 m/z in both wild-type and R740H factor VIII (data not shown). Using MASCOT, the peak was identified as the factor VIII A2-B junction region possessing the sequence: SFSQN-PPVLKRHQ. These results confirm His 740 as an authentic thrombin cleavage site.
Thrombin Activation of Factor VIII Proteins-To investigate thrombin proteolysis of the factor VIII mutants and the generation of cofactor activity, a time course of thrombin activation of factor VIII was performed using a factor Xa generation assay as described under "Experimental Procedures" (Fig. 3). The wild-type and factor VIII variants (1 nM) were reacted with thrombin (0.05 nM) for indicated times, and proteins were assayed immediately in a purified system by Xa generation. Activity of the wild-type protein increased rapidly, peaking

Rates of subunit generation during factor VIII activation by thrombin and factor Xa
Subunit generation rates for A2 and A3-C1-C2 subunits by thrombin and factor Xa cleavage of wild-type and mutant factor VIII were estimated by non-linear regression analysis of the data shown in Figs. 1, 2 between 1 and 2 min after which this activity level quickly decayed to ϳ10% peak activity at 30 min. The loss of activity over the time course results from inactivation of factor VIIIa by A2 dissociation from the A1/A3-C1-C2 dimer (26). In comparison to wild-type factor VIII, the R740H variant showed a significantly altered activation profile. Upon addition of thrombin, cofactor activity increased slowly, reaching a maximal activity level that represented ϳ20% the level observed for wild type. Moreover, the characteristic peak of factor VIIIa activity was replaced by broad plateau. Although the factor VIIIa R740H appears more stable over the time course, the rate of procofactor activation is balanced by the rate of inactivation of the labile cofactor (27). Therefore, the low activity plateau suggests a reduced rate of procofactor activation by thrombin.
These data were supported by reacting R740Q factor VIII with thrombin in a similar factor Xa generation assay (Fig. 3). Evaluation of this variant showed no significant activation of the protein over an extended time course. These data are consistent with Western blot analysis suggesting that cleavage at Arg 740 is a prerequisite for efficient generation of factor VIIIa by thrombin.

Kinetic Parameters of Thrombin Activation for Wild-type and R740H
Factor VIII-The kinetics of thrombin proteolysis of factor VIII resulting in procofactor activation was measured indirectly using the factor Xa generation system as described under "Experimental Procedures." Variable amounts of wild-type factor VIII (0 -50 nM) were activated with a limited thrombin concentration (0.05 nM) for 1 min in the presence of phospholipid vesicles (10 M). Subsequent addition of a high concentration of factor IXa (20 nM) was used to drive factor VIIIa into formation of the intrinsic factor Xase complex. Factor Xa generation reactions were initiated by addition of factor X (300 nM). Under these conditions, the resultant rate of product factor Xa formation is directly proportional to the concentration of factor VIIIa in the reaction and can be used to assess the K m for thrombin-catalyzed conversion of procofactor to cofactor. Data from thrombin activation of factor VIII were fitted to the Michaelis-Menten equation using nonlinear least squares regression (data not shown). Using wild-type factor VIII, the V max value for factor Xa generated was 312 Ϯ 30 min Ϫ1 . The K m (apparent) for thrombin activation of factor VIII was determined using the V max for factor Xa generated and was found to be 8.4 Ϯ 0.1 nM.
To investigate thrombin binding to R740Q factor VIII, the variant was used as a competitor of wild-type factor VIII activation by thrombin (Fig. 4). Wild-type factor VIII (5 nM) was reacted with thrombin (0.05 nM) in the presence of phospholipid vesicles (10 M) for 1 min in the absence or presence of variable concentrations of factor VIII R740Q. After the addition of hirudin, the resultant factor VIIIa was reacted with factor IXa (20 nM) and factor X (300 nM), and rates of factor Xa generation were determined. Inhibition increased in a dose-dependent manner with a 10-fold excess of mutant yielding Ͼ70% FIGURE 2. Generation of A3-C1-C2 subunit following thrombin-catalyzed cleavage of recombinant factor VIII mutants. Panels A and B, recombinant factor VIII wild-type, R740H, and R740Q (100 nM) were reacted with thrombin (2.5 nM) for the indicated times and subjected to SDS-PAGE and blotting using the 1A4 antibody. In panel C, densitometry data from blots are plotted as described in the legend to Fig. 1. Panel D shows results obtained using recombinant factor VIII wild type (100 nM) reacted with a lower thrombin concentration (0.125 nM). The symbols OE indicate wild type; f, R740H; and ࡗ, R740Q. Initial time points were fit to the single exponential equation using non-linear least squares regression.
inhibition of thrombin activation of wild-type factor VIII. The apparent K i value for R740Q factor VIII was 9.6 Ϯ 0.2 nM, which is similar to the K m (apparent) of wild-type factor VIII. These results suggest the R740Q mutation does not appreciably impair binding of thrombin to substrate factor VIII and are consistent with the activation of procofactor being largely driven by exosite interactions (28,29).
Factor Xa Activation of Factor VIII Variants-The above results suggest the rate reductions observed in requisite cleavages to yield factor VIIIa by thrombin appear greater than what would be predicted from the more modest effects on specific activity values obtained from one-stage clotting assays. One explanation for these results is the activation of the factor VIII Arg 740 mutants results from the action of a proteinase other than thrombin. This hypothesis was tested by assessing the factor VIII variants in factor Xa generation assays run in the absence or presence of prior procofactor activation by throm-bin. While inclusion of factor VIIIa into a factor Xa generation assay results in a linear rate of product factor Xa formed, replacement with the procofactor and the absence of added thrombin yields a lag in the appearance of product. This observation has been long appreciated and results from the need for factor VIII to be activated by low levels of factor Xa generated in situ, which cleaves the same bonds in factor VIII as thrombin (6).
Wild-type and factor VIII variants (1 nM) were reacted in the absence or presence of thrombin (20 nM) and evaluated in factor Xa generation assays as described under "Experimental Procedures." Fig. 5 shows the extent of factor Xa generation as a function of time for both wild-type and mutant proteins. A linear rate of Xa generation was observed for wild-type factor VIII treated with thrombin, whereas in the absence of thrombin, the wild-type factor VIII produced a characteristic lag in the time for factor Xa formed. Both mutants showed reduced levels of factor Xa formed when pretreated with thrombin, suggesting limited levels of cofactor activation. However, similar and appreciable amounts of factor Xa were formed at the later time points independent of whether the factor VIII variants were treated with thrombin. These results suggested that factor Xa generated in these reactions was capable of activating the procofactors with potentially higher efficiency than observed with thrombin.
Cleavage in Wild-type and Mutant Factor VIII Forms by Factor Xa-Factor Xa-catalyzed cleavage of R740H and R740Q factor VIII forms were evaluated using similar methods employed as above for thrombin. Proteolysis of factor VIII (100 nM) by factor Xa (2.5 nM) was performed in the presence of phospholipids (10 M), and products were evaluated following SDS-PAGE and Western blotting using antibody R8B12 to assess generation of the A2 subunit (Fig. 6) or antibody 1A4 to assess generation of the A3-C1-C2 subunit (Fig. 7). While factor Xa  . Competition of thrombin activation of wild-type factor VIII by R740Q factor VIII. Various concentrations of recombinant factor VIII R740Q and wild-type factor VIII (5 nM) were reacted with thrombin (0.05 nM) for 1 min in the presence of phospholipids (10 M). Thrombin was inactivated by addition of hirudin (0.1 units/ml) and factor Xa generation was initiated by addition of factor X (300 nM) and factor IXa (20 nM) as described under "Experimental Procedures." The 100% value obtained in the absence of R740Q was ϳ100 nM/min. Data were corrected by the amount of factor Xa generated by R740Q factor VIII, which was Ͻ5% of total. Initial rates of factor Xa generation are plotted as a function of R740Q factor VIII concentration and fitted to competitive inhibition equation by nonlinear least squares regression. Experiments were performed at least three separate times, and average values are shown. cleavage of the wild-type factor VIII yielded a similar pattern of reaction products as with thrombin, the rate of factor VIIIa A2 subunit formed was ϳ2-fold slower than that observed for thrombin ( Fig. 6 and Table 1). Results evaluating proteolysis of R740H by factor Xa revealed relatively lower levels of the A2-LC and A2-a3 intermediates as generated by thrombin and faster rates of generation of A2 subunit, such that the rate of formation of this subunit from R740H factor VIII was ϳ20% that observed using wild-type factor VIII (Fig. 6, A and C and Table 1). Because the disparity in A2 formation rates was ϳ40-fold comparing these two substrates using thrombin, this result suggested that factor Xa cleaves the R740H substrate more efficiently than thrombin. Consistent with this observation, a Western blot time course for the R740Q mutant revealed that the generation of A2 subunit was reduced by ϳ10-fold compared with wild type (Fig. 6C and Table 1). This result contrasts with an apparent Ͼ700-fold reduction in A2 generation obtained with thrombin utilizing the R740Q substrate (Table  1). Because A2 subunit is essentially derived from the heavy chain of R740Q factor VIII rather than single chain, due to failure of factor Xa to cleave a P1 Gln, these results suggest that cleavage at the Arg 372 site is markedly increased (Ͼ25-fold) with factor Xa compared with thrombin when cleavage at residue 740 is abrogated.
Evaluation of reactions in factor VIII light chain yielded similar results. Quantification of blotting data (not shown) indicated that wild-type factor VIII cleavage by factor Xa at Arg 1689 produced the A3-C1-C2 subunit at a rate that was ϳ7-fold slower than thrombin ( Fig. 7 and Table 1). The R740H factor VIII mutant yielded a slower rate (ϳ5-fold) of A3-C1-C2 subunit generation over the time course, and this value was 9-fold faster compared with the fold reduction observed following reaction with thrombin (Fig. 2). Similarly, an ϳ8-fold reduction was observed in factor Xa-catalyzed cleavage at Arg 1689 (Fig. 7) using R740Q factor VIII as compared with an ϳ140-fold rate reduction observed using thrombin ( Fig. 2 and Table 1). Taken together with the above data evaluating cleavage at the Arg 372 site, these results indicate that factor Xa exhibits less dependence upon prior cleavage at Arg 740 to affect cleavages leading to procofactor activation. Overall, these observations may explain the higher than predicted specific activity values obtained for the two variants using the one-stage clotting assay and suggests a different mechanism of action for the two activating proteinases.

DISCUSSION
Thrombin-catalyzed activation of factor VIII occurs following proteolysis at three P1 Arg residues, Arg 372 , Arg 740 , and FIGURE 6. A2 subunit generated following factor Xa cleavage of factor VIII. Wild-type and mutant recombinant factor VIII (100 nM) were reacted with factor Xa (2.5 nM) in the presence of phospholipids (10 M) for the indicated times as described under "Experimental Procedures." Samples were run on 8% polyacrylamide gels followed by Western blotting using an anti-A2 antibody. Panel A shows the factor Xa time course for wild-type and R740H factor VIII, while panel B shows the time course of wild-type and R740Q factor VIII. Panel C shows quantitative densitometry of the A2 generation (nM) from blotting data from panels A and B. The symbols OE indicate wild type; f, R740H; and ࡗ R740Q. Initial time points were fit to the single exponential equation using non-linear least squares regression. FIGURE 7. A3-C1-C2 subunit generated following factor Xa cleavage of factor VIII. Recombinant factor VIII wild-type, R740H, and R740Q (100 nM) were reacted with factor Xa (2.5 nM) in the presence of phospholipids (10 M) for the indicated times as described under "Experimental Procedures." Samples were run on 8% polyacrylamide gels followed by Western blotting using an anti-A3 antibody. Data were derived from densitometric analysis of A3-C1-C2 generation (nM) from blotting data (data not shown). Symbols OE indicate wild type; f, R740H; and ࡗ, R740Q. Initial time points were fit to the single exponential equation using nonlinear least squares regression. The inset shows A3-C1-C2 subunit generation from wild-type factor VIII (100 nM) activated by factor Xa (0.5 nM). Arg 1689 . Cleavage at Arg 740 is rapid relative to the other two sites (15) and results in the liberation of the B domain or its fragments from the factor VIII heavy chain. However, little available information has indicated a functional role for this cleavage. We now show using two point mutations at residue 740, a His substitution to markedly slow cleavage and a Gln mutation that is refractory to cleavage, that proteolysis at Arg 740 represents a critical step in the thrombin-catalyzed pathway for procofactor activation. This event appears necessary for facilitating cleavages at both Arg 372 and Arg 1689 , suggesting an ordered reaction mechanism for the activation of factor VIII.
The importance of the Arg 740 site to the mechanism for thrombin action was not predicted from the specific activity values obtained for the point mutants based on the one-stage clotting assay. The specific activity of R740H was 50% that of wild-type factor VIII and would yield a normal phenotype, while a specific activity of ϳ20% for the R740Q mutant would yield a mild hemophilia A phenotype. A recent study showed that replacement of Arg 372 with His yielded a factor VIII variant possessing a severe hemophilia A phenotype (1% wild type) and showed an ϳ80-fold reduced rate for thrombin catalyzed cleavage at His 372 (24). While the His 740 mutation resulted in a similar fold reduction in the rate of thrombin cleavage at this site as well as Arg 372 , as judged by A2 subunit generation data, these impairments were not reflected in the specific activity values. However, this discrepancy may be explained by the capacity for an alternative proteinase such as factor Xa to activate the procofactor. Thrombin is a more efficient activator of factor VIII than factor Xa, with earlier results evaluating catalytic efficiency values of ϳ3-fold (14) and ϳ5-fold (30) greater for the former proteinase. Consistent with these data, we observed ϳ2-fold and ϳ7-fold rate increases in generating A2 and A3-C1-C2 subunits in comparing thrombin to factor Xa using wild-type factor VIII. However, results from Western blotting revealed that factor Xa-catalyzed cleavages of factor VIII at Arg 372 and Arg 1689 showed a marked reduction in the dependence for prior cleavage at Arg 740 , consistent with factor Xa employing an activation mechanism different from that of thrombin.
Whereas the capacity for factor Xa-catalyzed procofactor activation may help to explain the failure to observe point mutations at Arg 740 as yielding hemophilic phenotypes, an additional reason could reflect the structure of the circulating form of factor VIII. Factor VIII circulates as a series of heterodimers possessing a constant sized light chain, but a heterogeneous heavy chain due to the presence of some, or all, of the B domain. Fractionation of factor VIII forms obtained from partially purified, human plasma-derived concentrates revealed heterodimers possessing discrete sizes of heavy chain from ϳ90 kDa to ϳ210 kDa (4). That study showed that the ϳ90-kDa heavy chain form appeared the most abundant, based upon staining density, and followed the same kinetics for activation by thrombin as heterodimers containing substantial lengths of B domain. In addition, it was noted that cleavage of the B-domain containing heavy chains proceeded through an intermediate of identical size to the ϳ90 kDa heavy chain form prior to generation of the final reaction products, A1 and A2 subunits. Although no C-terminal sequence information was available, the mass of the 90 kDa heavy chain was consistent with the absence of B domain sequences. Thus, the intracellular processing of factor VIII to the 90 kDa heavy chain form would serve to liberate PЈ residues following Arg 740 obviating the requirement for thrombin action at this site.
In an earlier study, Pittman and Kaufman (31) employed mutation and transient transfection of factor VIII variants to assess proteolysis and thrombin activation of factor VIII. Their results showed an R740I variant had little, if any, effect on specific activity as judged using a chromogenic assay. Evaluation of proteolysis showed the generation of factor VIIIa subunits, although the reaction was not complete. Therefore the authors concluded that cleavage at Arg 740 by thrombin was not a prerequisite for cleavage at Arg 372 or Arg 1689 . The reaction conditions employed in that study are worth noting in that factor VIII (ϳ0.1 units/ml) was reacted with thrombin (10 units/ml) for 30 min at 37°C. These conditions represent an ϳ600-fold molar excess of proteinase to substrate that was reacted over an extended time. Thus, any impairment on reaction rate due to the R740I mutation would likely have gone unnoticed.
More recently, our laboratory characterized a double mutation D392A/D394A within the A2 domain of factor VIII that possessed a severe hemophilic phenotype with factor VIII activity ϳ1% of the wild-type protein (29). This mutant protein was defective in thrombin-catalyzed cleavage at Arg 740 , whereas cleavage at Arg 372 was not appreciably impaired. One explanation for the lack of influence of the 740 site on reaction at Arg 372 is that the double mutation disrupts the interaction of thrombin with factor VIII, essentially uncoupling the linked proteolytic steps. We speculated in that study that these residues, which contribute to a cluster of high negative charge density localized near the N-terminal of the A2 domain, may interact with anion binding exosite 2 of thrombin, inasmuch as the binding of thrombin to the isolated A2 subunit was blocked by heparin but not hirudin (29). We further speculated that the very low specific activity of the mutant derived from failure to remove the remnant B domain linker from the C-terminal end of the A2 domain. This conclusion is not likely correct since we now identify the peptide by MALDI-TOF MS following cleavage of the wild-type and R740H proteins. Given the slow rate of cleavage at that site for the latter variant coupled with the largely benign effects of cofactor activity, we now believe some other functional defect, possibly the direct result of the mutation at residues 392 and 394, is responsible for the low specific activity of that variant.
The activation of factor VIII by thrombin is exosite-dependent (28) and this mechanism represents a recurrent theme in the proteolytic modulation of coagulation protein substrates. The ordered cleavage of factor VIII heavy chain described in this report parallels the APC-catalyzed proteolysis leading to the inactivation of factor Va. In that exosite-driven mechanism (32), an initial cleavage of factor Va heavy chain at Arg 506 is required to expose and facilitate cleavage at Arg 306 (33). Mutation of Arg 506 to Gln yields a thrombogenic factor Va that shows resistance to cleavage at Arg 506 yielding slow cleavage at Arg 306 and consequent reduced inactivation rates by APC (34). Interestingly, the homologous sites attacked by APC in factor VIIIa are cleaved independently (22), likely the result of the P1 Arg 336 and Arg 562 located on separate subunits.
In summary, our results indicate thrombin catalyzes the ordered cleavage of factor VIII heavy chain with initial attack at Arg 740 facilitating subsequent bond attack at Arg 372 and Arg 1689 , which contribute functionally to the development of cofactor activity. This sequential presentation of scissile bonds in the factor VIII heavy chain may occur by a ratcheting mechanism analogous to that proposed for the cleavage of prothrombin by prothrombinase via the meizothrombin pathway (35). However, the strong requirement for ordered cleavage by thrombin is not shared by the factor Xa activation pathway, which provides an alternative mechanism potentially limiting the role for Arg 740 in the activation of factor VIII.