Platelet-derived microparticles express high affinity receptors for factor VIII.

Factor VIII is a cofactor in the tenase enzyme complex which assembles on the membrane of activated platelets. A critical step in tenase assembly is membrane binding of factor VIII. Platelet membrane factor VIII-binding sites were characterized by flow cytometry using either fluorescein maleimide-labeled recombinant factor VIII or a fluorescein-labeled monoclonal antibody against factor VIII. Following activation by thrombin, most platelets bound factor VIII within 90 s. In addition, over the course of several minutes, membranous vesicles (microparticles) were shed from the platelet plasma membrane and each microparticle bound as much factor VIII as a stimulated platelet. Over 30 min, stimulated platelets (but not microparticles) lost the capacity to bind factor VIII. Factor VIII bound saturably to microparticles from platelets stimulated with thrombin, thrombin plus collagen, or the complement proteins C5b-9. The binding of factor VIII was compared to factor V, a structurally homologous coagulation cofactor. Analysis of microparticle binding kinetics yielded similar on and off rates for factor VIII and factor Va and KD values of 2-10 nM. In the presence of 20 nM factor Va, the binding of factor VIII to microparticles was increased, and there was a comparable increase in platelet tenase activity. At higher factor Va concentrations, factor VIII binding and tenase activity were inhibited. Conversely, factor VIII had a similar dose-dependent effect on factor Va binding and platelet prothrombinase activity. Synthetic phospholipid vesicles containing phosphatidylserine competed with microparticles for binding of factor VIII and factor Va. These studies indicate that activated platelets express a transient increase in high affinity receptors for factor VIII, whereas platelet-derived microparticles express a sustained increase in receptors. The binding characteristics of platelet membrane receptors for factor VIII are similar to those for factor Va.     

Cryoprecipitate of intermediate purity produced in a closed thaw-siphon system from DDAVP stimulated blood donor plasma.

The effects of intravenous administration of DDAVP to blood donors and the use of DDAVP plasma for the production of cryoprecipitate in the closed thaw-siphon system were evaluated. DDAVP treatment produced on the average a 3.2-fold rise in plasma levels of factor VIII. Von Willebrand factor antigen increased to a lesser extent. Cryoprecipitate prepared from 220-280 ml aliquots of DDAVP stimulated donor plasma contained 472 +/- 210 units of factor VIII and 276 +/- 130 units of von Willebrand factor antigen. The average yield of factor VIII was 57% of that in the prefrozen plasma. The specific activity of factor VIII in cryoprecipitate was 0.77 +/- 0.44 U/mg protein, comparable to that for intermediate purity concentrates. Thus, by the use of DDAVP and the thaw-siphon technique it is possible to produce cryoprecipitate 4-7 times as potent as conventionally manufactured preparations.     

Proteolytic activity of alpha 2-macroglobulin-enzyme complexes toward human factor VIII/von Willebrand factor

The low level of enzymatic activity of certain alpha 2-macroglobulin-proteinase complexes could be important to the function of factor VIII/von Willebrand glycoprotein since it is especially sensitive to proteolytic cleavage. To test this possibility, complexes of alpha 2-macroglobulin with plasmin, trypsin, and thrombin were formed in at least a 2:1 molar ratio of alpha 2-macroglobulin:proteinase and tested for effects on the factor VIII procoagulant activity of the factor VIII/von Willebrand glycoprotein. Neither the alpha 2-macroglobulin-trypsin complex nor the alpha 2-macroglobulin-plasmin complex affected factor VIII procoagulant activity. The behavior of the alpha 2-macroglobulin-thrombin complex was different. When alpha 2-macroglobulin and thrombin were incubated in a mole ratio of 3:1 or less, factor VIII procoagulant activity was enhanced to about the same extent as with free thrombin. Even at a 24:1 mole ratio, the mixture could produce 45% of the increase in factor VIII activity obtained with free thrombin. The isolated alpha 2-macroglobulin-thrombin complex could also activate the factor VIII procoagulant function to about 45% of the level obtained with an identical amount of uncomplexed thrombin. Analysis of the alpha 2-macroglobulin-125I-labeled thrombin complexes by rechromatography or by polyacrylamide gel electrophoresis in sodium dodecyl sulfate indicated that this activation was not due to free thrombin. We conclude that the alpha 2-macroglobulin-thrombin complex retains sufficient proteolytic activity to activate the procoagulant function of factor VIII/von Willebrand glycoprotein despite the latter being a very large substrate, having an estimated molecular weight of 1-20 million.     

Mn2+ binding to factor VIII subunits and its effect on cofactor activity

Metal ions, such as Ca2+ and Mn2+, are necessary for the generation of cofactor activity following reconstitution of factor VIII from its isolated light chain (LC) and heavy chain (HC). Titration of EDTA-treated factor VIII with Mn2+ showed saturable binding with high affinity (K(d) = 5.7 +/- 2.1 microM) as detected using a factor Xa generation assay. No significant competition between Ca2+ and Mn2+ for factor VIII binding (K(i) = 4.6 mM) was observed as measured by equilibrium dialysis using 20 microM Ca2+ and 8 microM factor VIII in the presence of 0-1 mM Mn2+. The intersubunit affinity measured by fluorescence energy transfer of an acrylodan-labeled LC (fluorescence donor) and fluorescein-labeled HC (fluorescence acceptor) in the presence of 20 mM Mn2+ (K(d) = 53.0 +/- 17.1 nM) was not significantly different from the affinity value previously obtained in the absence of metal ion (K(d) = 53.8 +/- 14.2 nM). The sensitization of phosphorescence of Tb3+ bound to factor VIII subunits was utilized to detect Mn2+ binding to the subunits. Mn2+ inhibited the phosphorescence of Tb3+ bound to HC and LC, as well as the HC-derived A1 and A2 subunits with a relatively wide range of estimated inhibition constant values (K(i) values = 169-1147 microM), whereas Ca2+ showed no effect on Tb3+ phosphorescence. These results suggest that factor VIII cofactor activity can be generated by Mn2+ binding to site(s) on factor VIII that are different from the high-affinity Ca2+ binding site. However, like Ca2+, Mn2+ did not alter the affinity for HC and LC association. Thus, Mn2+appears to generate factor VIII cofactor activity by a similar mechanism as observed for Ca2+following its association at nonidentical sites on the protein.     

Construction and characterization of an active factor VIII variant lacking the central one-third of the molecule

The primary structure of factor VIII consists of 2332 amino acids that exhibit 3 distinct structural domains, including a triplicated region (A domains), a unique region of 909 amino acids (B domain), and a carboxy-terminal duplicated region (C domains), that are arranged in the order A1-A2-B-A3-C1-C2. The B domain (residues 741-1648) of factor VIII is lost when factor VIII is activated by thrombin, which proteolytically processes factor VIII to active subunits of Mr 50,000 (domain A1), 43,000 (domain A2), and 73,000 (domains A3-C1-C2). To determine if the B domain is required for factor VIII coagulant activity, a variant was constructed by using recombinant DNA techniques in which residues 797-1562 were eliminated. This shortened the B domain from 909 to 142 amino acids. This variant factor VIIIdes-797-1652 was expressed in mammalian cells and was found to be functional. The factor VIIIdes-797-1562 protein was purified and shown to be processed by thrombin in the same manner as full-length factor VIII. The factor VIIIdes-797-1562 variant also bound to von Willebrand factor (vWF) immobilized on Sepharose. These results indicate that most of the highly glycosylated B domain of factor VIII is not required for the expression of factor VIII coagulant activity and its interaction with vWF.     

The functional domains of coagulation factor VIII:C

A lack of factor VIII:C, manifested as a bleeding disorder due to the absence of clot formation, is known as hemophilia A, an X chromosome-linked inherited disease afflicting 1-2 males/10,000. To determine the minimum functional domain(s) essential for factor VIII:C activity, we have expressed the amino-terminal (92-kDa) and carboxyl-terminal (80-kDa) proteolytic cleavage products as individual, secreted polypeptides in monkey cells without the 909-residue central region. We have found that neither terminal domain alone is able to promote coagulation in factor VIII:C-deficient plasma. However, when the 92- and 80-kDa peptides are co-expressed, clotting activity is readily detected. Thus, these two chains alone constitute an active or activatable complex. The central domain is required neither for activity nor for the assembly of an active complex from two chains expressed in trans. These results suggest that a truncated derivative of factor VIII:C may be useful in coagulation therapy.     

Thrombin-activated and factor Xa-activated human factor VIII: differences in cofactor activity and decay rate.

The decay of human coagulation factor VIIIa has been studied by kinetic methods that ensure no interference through proteolytic feedback. The rate of decay of factor VIIIa activity was found to vary with the activator used to activate factor VIII. Thrombin-activated factor VIII-von Willebrand factor complex (fVIII-vWf) decayed at a rate of 0.31 min-1, whereas factor Xa-activated fVIII-vWf decayed at 0.11 min-1 under the same conditions. Factor VIII free of von Willebrand factor (factor VIII: C), although decaying at a generally slower rate after activation, still showed a dependence of decay rate on activator: thrombin-activated factor VIII:C decaying at a rate of 0.06 min-1, and factor Xa-activated factor VIII: C at 0.01 min-1. Readdition of von Willebrand factor (18 micrograms/ml) to factor VIII:C did not alter the observed activity or decay rate. The decay of the two species of factor VIIIa was studied, using the fVIIIa-vWf complex, in the presence of varying levels of factor IXa. Plots of reciprocal decay rates vs factor IXa concentration were linear, and nearly parallel for the two factor VIIIa species, with a mean slope of 0.56 min.nM-1. In addition to these studies, we have confirmed previous studies showing that the two forms of factor VIIIa differ in cofactor activity, but they do so in the same ratio as in their decay rates. We suggest that this difference and that observed in decay rate have a common cause, and incorporate this into a potential kinetic model of factor VIII activation and decay.     

Factor VIII activity and factor VIII-associated antigen in healthy newborns and in newborns with stressing perinatal factors

Factor VIII coagulation activity (VIII:C) and factor VIII associated antigen (VIII:AGN) were determined in healthy newborns and in children with charging perinatal factors ("risk children"). VIII:C values of healthy newborns may be compared with those of grown-ups with normal coagulation. Risk children have somewhat higher values than newborns, the difference, however, being statistically not significant. The concentration of VIII:AGN is clearly increased in both groups on the first day of life. Moreover, VIII:AGN is being eliminated more slowly in risk children. The increased VIII:AGN concentrations are considered as a sequel of stress conditions caused by birth, whereas the discrepancy between VIII:C and VIII:AGN is due to a thrombin effect.     

A critical evaluation of the available methods for the determination of factor VIII von Willebrand

Von Willebrand factor (vWf) is the major component of the circulating factor VIII complex. The von Willebrand molecule includes factor VIII related antigen (VIIIR: Ag) which represents the molecular substrate of the von Willebrand activity expressed as Ristocetin cofactor (VIIIR:RCoF) activity. Several methods have been developed for VIIIR: Ag evaluation, among the first being the rocket-immunoelectrophoresis method of LAURELL. Radial immunodiffusion (MANCINI's method) was also used. Subsequently, radioimmunological assays, either as radioimmunoassay (RIA) or immunoradiometric assay (IRMA), were developed with improvements in sensitivity, so that levels of VIIIR: Ag lower than 0.1% of normal can be detected. More recently, an enzyme-linked immunosorbent assay (ELISA), characterized by the use of enzyme-conjugated antibody was proposed. This method shows a sensitivity similar to immunoradiometric methods but without using any dangerous reagent. Finally, a nephelometric method was proposed for factor VIII antigen evaluation. For a qualitative evaluation of von Willebrand factor crossed-immunoelectrophoresis and multimeric analysis can be used. In the first case, the use of precipiting antibodies against von Willebrand factor may demonstrate a peak with different characteristics related to the biochemical property of von Willebrand. Multimeric analysis in SDS-agarose gel electrophoresis followed by staining with labelled antifactor VIII antibodies gives information about different polymeric forms of circulating VIII/vW factor. Von Willebrand factor activity, expressed as its ability to induce platelet aggregation in the presence of the antibiotic Ristocetin, can be carried out using normal formalin fixed platelets, either with aggregometer or visual methods (glass slide test or tubes test and microtritation plate). The corrected evaluation of factor VIII complex by all these techniques together with the clotting activity assay allows a satisfactory study of factor VIII properties.     

Inactivation of human factor VIII by activated protein C: evidence that the factor VIII light chain contains the activated protein C binding site

Factor VIII is represented as a series of heterodimers composed of an 83(81) kDa light chain noncovalently bound to a variable size (93 to 210 kDa) heavy chain. Activated protein C inactivates factor VIII causing several cleavages of the factor VIII heavy chain(s). When factor VIII subunits were dissociated and component heavy and light chains isolated, the heavy chains were no longer a substrate for proteolysis by activated protein C. However, when factor VIII heavy chains were recombined with light chain, the reconstituted factor VIII activity was inactivated by activated protein C. The rate of factor VIII inactivation catalyzed by activated protein C was reduced by the presence of free light chain. The extent of this inhibition was dependent upon the concentration of light chain. Control experiments indicated that this protective effect of free light chain was not the result of inhibition of the activated protein C - lipid interaction. Fluorescence analysis demonstrated binding between the factor VIII light chain, chemically modified with eosin maleimide, and activated protein C, modified at its active site by dansyl-Glu-Gly-Arg chloromethyl ketone. Similar to proteolysis of factor VIII by activated protein C, this binding was dependent upon a lipid surface. Based upon the degree of fluorescence quenching, a spatial distance of 26 A was calculated separating the two fluorophores. These results demonstrate direct binding of activated protein C to the factor VIII light chain and suggest that this binding is an obligate step for activated protein C-catalyzed inactivation of factor VIII.     

The effect of carbohydrate depletion on procoagulant activity and in vivo survival of highly purified human factor VIII

Human factor VIII procoagulant protein (factor VIII) was purified using a modification of our previously described method, in which Sephacryl S-400 elution, rather than QAE-cellulose chromatography, served as the final purification step. The protein had a specific activity of more than 2500 U/mg and consisted of a single polypeptide (Mr 100 000) when analyzed by SDS-polyacrylamide gel electrophoresis. Factor VIII was shown to be a glycoprotein by staining with periodic acid-Schiff's reagent following electrophoresis. Treatment of factor VIII with a mixture of exo- and endoglycosidases caused a reduction by about 50% in the intensity of periodic acid-Schiff staining, as determined by scanning densitometry, and an increase in electrophoretic mobility (equivalent to a new Mr 95 000). Removal of this portion of the total carbohydrate had no significant effect on factor VIII clotting activity or on thrombin potentiation of clotting activity. The in vivo survival curves of a native and sugar-depleted 125I-labeled factor VIII both showed similar patterns of initial rapid decay to 60 and 40% activity, respectively, followed by a one-half decay time of 4 h for both. These results suggest that the carbohydrate portion of human factor VIII does not contribute significantly to either clotting function in vitro or to biological turnover in vivo.     

The relationship of biological and immunological activities of factor VIII

Factor VIII is an essential blood clotting factor which consists of two protein moieties, each with distinct biological functions and antigenic determinants. The immunological markers were originally seen as indicators of the biological activities; however this view has been increasingly challenged. We have investigated the biological and immunological properties of Factor VIII to clarify these relationships. Plasma stored at room temperature for 21 days lost biological activity, but retained immunological activity: The procoagulant activity was reduced to 35% and the ristocetin cofactor activity to 75.4% of their original levels; but the reactivities of both procoagulant antigen and Factor VIII related antigen were maintained. A dissociation of activities was also demonstrated in serum, in which the procoagulant activity was 10% and the procoagulant antigen 72% of corresponding plasma values. These results indicate that the antigenic reactivities are not appropriate markers for Factor VIII biological activity.     

Proteolytic processing of human factor VIII. Correlation of specific cleavages by thrombin, factor Xa, and activated protein C with activation and inactivation of factor VIII coagulant activity

Human factor VIII was isolated from commercial factor VIII concentrates and found to consist of multiple polypeptides with molecular weights ranging from 80 000 to 210 000. Immunological and amino acid sequence data identified these polypeptides as subunits of factor VIII. N-Terminal amino acid sequence analysis determined that the Mr 210 000 and 80 000 proteins are derived from the N- and C-terminal portions of factor VIII, respectively; Mr 90 000-180 000 polypeptides are derived from the Mr 210 000 polypeptide by C-terminal cleavages. Treatment of purified factor VIII with thrombin resulted in proteolysis of Mr 80 000-210 000 proteins and the generation of polypeptides of Mr 73 000, 50 000, and 43 000. Maximum coagulant activity of thrombin-activated factor VIII was correlated with the generation of these polypeptides. The proteolysis as well as activation of factor VIII by thrombin was found to be markedly dependent on CaCl2 concentration. Proteolysis of factor VIII with activated protein C (APC) resulted in degradation of the Mr 90 000-210 000 proteins with the generation of an Mr 45 000 fragment. This cleavage correlated with inactivation of factor VIII by APC. The Mr 80 000 protein was not degraded by APC. Factor Xa cleaved the Mr 80 000-210 000 factor VIII proteins, resulting in the generation of fragments of Mr 73 000, 67 000, 50 000, 45 000, and 43 000. Factor Xa was found to initially activate and subsequently inactivate factor VIII.(ABSTRACT TRUNCATED AT 250 WORDS)     

Lower inhibitor development in hemophilia A mice following administration of recombinant factor VIII-O-phospho-L-serine complex

Factor VIII is a multidomain protein composed of A1, A2, B, A3, C1, and C2 domains. Deficiency or dysfunction of factor VIII causes hemophilia A, a bleeding disorder. Administration of exogenous recombinant factor VIII as a replacement leads to development of inhibitory antibodies against factor VIII in 15-30% of hemophilia A patients. Hence, less immunogenic preparations of factor VIII are highly desirable. Inhibitory antibodies against factor VIII are mainly directed against immunodominant epitopes in C2, A3, and A2 domains. Further, several universal epitopes for CD4+ T-cells have been identified within the C2 domain. The C2 domain is also known to interact specifically with phosphatidylserine-rich lipid vesicles. Here, we have investigated the hypothesis that complexation of O-phospho-l-serine, the head group of phosphatidylserine, with the C2 domain can reduce the overall immunogenicity of factor VIII. The biophysical (circular dichroism and fluorescence) and biochemical studies (ELISA and size exclusion chromatography) showed that O-phospho-l-serine binds to the phospholipid-binding region in the C2 domain, and this interaction causes subtle changes in the tertiary structure of the protein. O-Phospho-l-serine also prevented aggregation of the protein under thermal stress. The immunogenicity of the factor VIII-O-phospho-l-serine complex was evaluated in hemophilia A mice. The total and inhibitory antibody titers were lower for factor VIII-O-phospho-l-serine complex compared with factor VIII alone. Moreover, factor VIII administered as a complex with O-phospho-l-serine retained in vivo activity in hemophilia A mice. Our results suggest that factor VIII-O-phospho-l-serine complex may be beneficial to increase the physical stability and reduce immunogenicity of recombinant factor VIII preparations.     

Fractionation of individual, biologically active factor VIII multimers

We have designed an electrophoretic system for the fractionation of individual, biologically active multimers of factor VIII. Human factor VIII, purified by gel filtration on Sepharose CL-2B from plasma cryoprecipitate, was submitted to electrophoresis without SDS on 2.0% polyacrylamide gels in 0.04 M Tris/0.06 M Tes buffer, pH 7.5. Staining with Coomassie blue revealed a series of protein bands. Measurement of electrophoretic mobility showed constant size intervals between adjacent bands. Electrophoresis in a second dimension, in the presence of SDS, resulted in an identical order of mobilities, suggesting that the different migration rates of factor VIII proteins in the first electrophoretic system were size- and not charge-dependent. After electrophoresis in the absence of SDS both factor VIII coagulant and ristocetin cofactor activities as well as factor VIII-related antigen were recovered by elution from gel slices. The distribution of activity peaks resembled that of Coomassie-stained factor VIII proteins found in control gels. We thus demonstrate that an electrophoretic fractionation of factor VIII multimers is possible even at neutral pH where factor VIII activities are retained.     

Standardization of factor VIII: establishment and use of secondary standards

Two secondary standards for use in routine assays of Factor VIII in therapeutic concentrates and in patients, plasmas, respectively, have been established in a multicenter collaborative study. In order to assess the effect of the adoption of these preparations as common Secondary Standards a comparative assay has been performed: one sample of a Factor VIII concentrate of intermediate purity and one plasma sample have been tested in two laboratories for Factor VIII:C activity using as reference, among others, the common working standard. Analysis of the results shows that with the plasma sample the differences of the estimates obtained with any of the references in our two laboratories were not statistically significant (P greater than 0.3), while with the concentrate sample the differences were always statistically significant (P less than 0.005). The study shows that the adoption of common working standards (besides the uniformity in assay method, reagents and basic equipment) is not sufficient to eliminate interlaboratory variation in the measurement of Factor VIII:C.     

Amphipathic helices support function of blood coagulation factor IXa

Blood coagulation factor IXa gains proteolytic efficiency upon binding to a phospholipid membrane. We have found that an amphipathic, membrane-binding peptide from the C2 domain of factor VIII, fVIII(2303)(-23), enhances proteolytic efficiency of factor IXa in the absence of phospholipid membranes. This enhancement is the result of a reduction in the K(M) for the substrate, factor X, with little effect on the k(cat). Enhanced function requires interaction of the gamma-carboxyglutamic acid (Gla) domains of factor IXa and factor X since (i) a synthetic peptide comprising the Gla domain of factor IXa and antibodies directed to the Gla domain of factor IXa inhibit this acceleration, (ii) the acceleration is Ca(II) dependent, and (iii) conversion of Gla-domainless factor X is not affected by the presence of fVIII(2303)(-23). The effect of fVIII(2303)(-23) on factor IXa parallels the enhanced function produced by phosphatidylserine-containing bilayers, and fVIII(2303)(-23) does not further enhance function of factor IXa when phospholipid vesicles are present. The critical feature of fVIII(2303)(-23) is apparently its amphipathic helix-forming structure [Gilbert, G. E., and Baleja, J. D. (1995) Biochemistry 34, 3022-3031] because other alpha-helical peptides such as a homologous peptide from the C2 domain of factor V and melittin have similar effects. Diastereomeric analogues of fVIII(2303)(-23) and melittin, which have reduced helical content, do not support factor IXa activity. A truncated peptide of fVIII(2303)(-23) with three C-terminal residues deleted retains alpha-helical content but loses capacity to enhance factor X cleavage, suggesting that a minimum length of alpha-helix is required. Although these results probably do not illuminate the physiologic function of the factor VIII peptide corresponding to fVIII(2303)(-23), they demonstrate a novel, membrane-mimetic role of amphipathic helical peptides in supporting function of factor IXa.     

A monoclonal antibody to factor IX that inhibits the factor VIII:Ca potentiation of factor X activation

A murine monoclonal antibody (IgG1k, Kd approximately 10(-8) M) specific for an epitope located on the heavy chain of human factor IXa was used to study structure-function relationships of factor IX. The antibody inhibited factor IX clotting activity but did not impair activation of factor IX either by factor XIa/calcium or by factor VIIa/tissue factor/calcium. The antibody also did not impair the binding of factor IXa to antithrombin III. Moreover, the antibody did not prevent calcium and phospholipid (PL) from inhibiting the binding of factor IXa to antithrombin III. The antibody also failed to impair activation of factor VII by factor IXa/calcium/PL. Furthermore, the antibody did not interfere with the very slow activation of factor X by factor IXa/calcium/PL. In contrast, the antibody did interfere with factor X activation when reaction mixtures also contained factor VIII:Ca/von Willebrand factor. The marked acceleration of factor X activation observed in control mixtures was not observed in mixtures containing the antibody. Similar results were obtained in reaction mixtures containing the Fab portion of the antibody and factor VIII:Ca free of von Willebrand factor. In additional experiments, factor VIII:Ca/von Willebrand factor was found to inhibit the binding of the antibody to 125I-factor IXa as determined using an immunosorbent assay. Moreover, the antibody displaced factor VIII:Ca from the factor X activator complex (IXa/calcium/PL/VIII:Ca) as evidenced by an altered elution pattern on gel filtration chromatography. From these observations, we conclude that the antibody impairs the clotting activity of factor IXa through interference with its binding of factor VIII:Ca. This suggests a significant role for the heavy chain (residues of 181-415) of factor IXa in binding factor VIII:Ca.     

Metal ion-independent association of factor VIII subunits and the roles of calcium and copper ions for cofactor activity and inter-subunit affinity

Factor VIII circulates as a divalent metal ion-dependent heterodimer comprised of a light chain (LC) and a heavy chain (HC). Reassociation of factor VIII subunits was assessed using fluorescence energy transfer where LC and HC were labeled with acrylodan (Ac; fluorescence donor) and fluorescein-5-maleimide (Fl; fluorescence acceptor), respectively. The reduction of donor fluorescence due to the acceptor was used as an indicator of binding. Subunits associated with high affinity (K(d) = 53.8 nM) in the absence of metal ion and presence of EDTA. However, this product showed no cofactor activity, as measured by a factor Xa generation assay. In the presence of 25 mM Ca(2+), no increase in the intersubunit affinity was observed (K(d) = 48.7 nM) but specific activity of the cofactor was approximately 30% that of native factor VIII. At saturating levels of Fl-HC relative to Ac-LC, donor fluorescence decreased to 79.3 and 73.5% of its original value in the absence and presence of Ca(2+), respectively. Thrombin cleaved the heterodimers that were associated in the absence or presence of Ca(2+) with similar efficiency, indicating that the lack of activity was not the result of a defect in activation. Cu(2+) (0.5 microM) increased the intersubunit affinity by approximately 100 fold (K(d) = 0.52 nM) and the specific activity to approximately 60% of native factor VIII. The former effect was independent of Ca(2+), whereas the latter effect required Ca(2+). These results indicate that the intersubunit association in factor VIII is primarily metal-ion independent while divalent metal ions serve specific roles. Ca(2+) appears essential to promote the active conformation of factor VIII while Cu(2+) primarily enhances the intersubunit affinity.