Molecular basis of factor VIII inhibition by human antibodies. Antibodies that bind to the factor VIII light chain prevent the interaction of factor VIII with phospholipid.

Most antibodies to factor VIII have recently been shown to react with discrete regions of the factor VIII light chain (within the C2 domain) and/or the factor VIII heavy chain (within the amino-terminal segment of the A2 domain). The mechanism by which these antibodies, usually designated "factor VIII inhibitors," interfere with factor VIII function has been examined by determining their effect on factor VIII binding to a phospholipid. Factor VIII-phosphatidylserine binding was prevented by all seven factor VIII inhibitors that had strong factor VIII light chain reactivity and reduced by two inhibitors with weak anti-light chain reactivity. None of four inhibitors with heavy chain reactivity prevented factor VIII-phosphatidylserine interaction, though a partial reduction (less than 50%) was noted for the intact IgG preparations. However, when Fab' fragments were substituted, no detectable reduction in factor VIII-phosphatidylserine binding was noted for the anti-heavy chain inhibitors and complete inhibition was retained by the anti-light chain inhibitors. These data suggest that a subset of factor VIII inhibitors, those that bind to light chain determinants, inactivate factor VIII by preventing its effective interaction with phospholipid.

Inhibition of human factor VIIIa by anti-A2 subunit antibodies.

Human inhibitory alloantibodies and autoantibodies to Factor VIII (FVIII) are usually directed toward the A2 and/or C2 domains of the FVIII molecule. Anti-C2 antibodies block the binding of FVIII to phospholipid, but the mechanism of action of anti-A2 antibodies is not known. We investigated the properties of a patient autoantibody, RC, and a monoclonal antibody, 413, that bind to the region which contains the epitopes of all anti-A2 alloantibodies or autoantibodies studied to date. mAb 413 and RC were noncompetitive inhibitors of a model intrinsic Factor X activation complex (intrinsic FXase) consisting of Factor IXa, activated FVIII (FVIIIa), and synthetic phospholipid vesicles, since they decreased the Vmax of intrinsic FXase by > 95% at saturating concentrations without altering the Km. This indicates that RC and mAb 413 either block the binding of FVIIIa to FIXa or phospholipid or interfere with the catalytic function of fully assembled intrinsic FXase, but they do not inhibit the binding of the substrate Factor X. mAb 413 did not inhibit the increase in fluorescence anisotropy that results from the binding of Factor VIIIa to fluorescein-5-maleimidyl-D-phenylalanyl-prolyl-arginyl-FIXa (Fl-M-FPR-FIXa) on phospholipid vesicles in the absence of Factor X, indicating it does not inhibit assembly of intrinsic FXase. Addition of Factor X to Fl-M-FPR-FIXa, FVIIIa, and phospholipid vesicles produced a further increase in fluorescence anisotropy and a decrease in fluorescence intensity. This effect was blocked completely by mAb 413. We conclude that anti-A2 antibodies inhibit FVIIIa function by blocking the conversion of intrinsic FXase/FX complex to the transition state, rather than by interfering with formation of the ground state Michaelis complex.

Transgenic pigs produce functional human factor VIII in milk.

Deficiency or abnormality of coagulation factor VIII (FVIII) causes a bleeding disorder called hemophilia A. Treatment involves FVIII concentrates prepared from pooled human plasma or recombinant FVIII (rFVIII) prepared from mammalian cell culture. The cost of highly purified FVIII or rFVIII is a major factor in hemophilia therapy and restricts prophylaxis. We have sought to generate a new source of rFVIII by targeting expression of the human FVIII cDNA to the mammary gland of transgenic pigs using the regulatory sequences of the mouse whey acidic protein gene. The identity of processed heterodimeric rFVIII was confirmed using specific antibodies, by thrombin digestion and activity assays. The secretion of as much as 2.7 micrograms/ml of rFVIII in milk was over tenfold higher than in normal plasma. Up to 0.62 U/ml of rFVIII was detected in an assay in which rFVIII restored normal clotting activity to FVIII-deficient human plasma.

A soluble recombinant factor VIII fragment containing the A2 domain binds to some human anti-factor VIII antibodies that are not detected by immunoblotting.

Human factor VIII (fVIII) inhibitors are pathologic antibodies that inactivate fVIII. A cDNA clone was modified to encode fVIII amino acid residues 373-740 for expression in a baculovirus vector in insect cells. The encoded protein fragment H2 was produced as a soluble, secreted protein, and it was used to test inhibitor plasmas for the presence of antibodies that were not detected by immunoblotting. Seven of 13 inhibitors that bound only to the fVIII light chain by immunoblotting also bound to fragment H2 in an immunoprecipitation assay. Thus multi-chain inhibitor reactivity of inhibitors is more frequent than previously reported. One of these inhibitors was shown to share the epitope for other inhibitors that bind to H2 within amino acid residues 373-541 in immunoblotting assays. The sensitive immunoprecipitation assay described allows determination of relative H2 binding capacity of the total IgG and epitope localization of inhibitors that cannot be similarly characterized by immunoblotting.

A factor VIII neutralizing monoclonal antibody and a human inhibitor alloantibody recognizing epitopes in the C2 domain inhibit factor VIII binding to von Willebrand factor and to phosphatidylserine.

A neutralizing monoclonal antibody, NMC-VIII/5, recognizing the 72 kDa thrombin-proteolytic fragment of factor VIII light chain was obtained. Binding of the antibody to immobilized factor VIII (FVIII) was completely blocked by a light chain-specific human alloantibody, TK, which inhibits FVIII activity. Immunoblotting analysis with a panel of recombinant protein fragments of the C2 domain deleted from the amino-terminal or the carboxy-terminal ends demonstrated binding of NMC-VIII/5 to an epitope located between amino acid residues 2170 and 2327. On the other hand, the epitope of the inhibitor alloantibody, TK, was localized to 64 amino acid residues from 2248 to 2312 using the same recombinant fragments. NMC-VIII/5 and TK inhibited FVIII binding to immobilized von Willebrand factor (vWF). The IC50 of NMC-VIII/5 for the inhibition of binding to vWF was 0.23 micrograms/ml for IgG and 0.2 micrograms/ml for F(ab)'2. This concentration was 100-fold lower than that of a monoclonal antibody NMC-VIII/10 which recognizes the amino acid residues 1675 to 1684 within the amino-terminal portion of the light chain. The IC50 of TK was 11 micrograms/ml by IgG and 6.3 micrograms/ml by F(ab)'2. Furthermore, NMC-VIII/5 and TK also inhibited FVIII binding to immobilized phosphatidylserine. The IC50 for inhibition of phospholipid binding of NMC-VIII/5 and TK (anti-FVIII inhibitor titer of 300 Bethesda units/mg of IgG) was 10 micrograms/ml.

Dominant C2 domain epitope specificity of inhibitor antibodies elicited by a heat pasteurized product, factor VIII CPS-P, in previously treated hemophilia A patients without inhibitors.

From June, 1990, to November, 1991, in The Netherlands and Belgium, 16 previously treated severe hemophilia A patients (PTP) developed inhibitors after exposure to factor VIII CPS-P, a new heat pasteurized product. A previously untreated patient (PUP) also developed an inhibitor to CPS-P. In inhibitor neutralization assays with recombinant fVIII C2 and A2 domain polypeptides, plasmas from 14 PTPs were > or = 79% neutralized by C2 and < 10% by A2, but the PUP plasma was partially neutralized by C2 (48%) and A2 (28%). Immunoprecipitation assays of the PTP and PUP plasmas with the fVIII heavy chain and with recombinant C2 and A3-C1 polypeptides confirmed that the C2 dominant immune response to CPS-P was found only in the PTPs. Competition of the binding of 2 inhibitors to 125I-CPS-P by unlabeled CPS-P and another plasma fVIII was similar, demonstrating that the antibody response was not directed to epitopes only present in CPS-P. We propose that the immunogenicity of the CPS-P C2 domain was altered by heat pasteurization.

Inhibitors in German hemophilia A patients treated with a double virus inactivated factor VIII concentrate bind to the C2 domain of FVIII light chain.

To reduce the risk of transmission of hepatitis A virus, an Octapharma produced factor VIII (fVIII) concentrate treated with solvent detergent (FVIII-SD) was further pasteurized after purification. This product, Octavi SDPlus (FVIII-SDP), was marketed in Europe in 1993 to 1995. Inhibitors appeared from September to October, 1995, in 12 of 109 previously treated German hemophilia A patients. A study of similarly treated Belgian patients, who also developed inhibitors, had shown antibodies to the fVIII light chain (domains A3-C1-C2) only. In the present study, the epitope specificity of 8 German inhibitor plasmas was also found to be restricted to the light chain. In radioimmunoprecipitation assays to localize the light chain epitope(s), antibody binding to heavy chain (domains A1-A2-B) was 11-148 fold lower than to the C2 domain, and binding to recombinant A3-C1 was barely detectable. These results were supported by >95% neutralization of a high responder inhibitor titer by the C2 domain.

An alloantibody recognizing the FVIII A1 domain in a patient with CRM reduced haemophilia A due to deletion of a large portion of the A1 domain DNA sequence.

We report the development of a FVIII inhibitor in a patient with severe, cross reacting material reduced (CRM(R)) haemophilia A. The level of Factor VIII antigen (FVIII:Ag) measured by ELISA using anti-C2 monoclonal and alloantibodies was 1.9 U/dl. This baseline FVIII:Ag level was increased to 8.3 U/dl after administration of DDAVP. The anti-FVIII inhibitor titer was 2.9 Bethesda U/ml. DNA analysis showed a large deletion of the FVIII gene from exon 4 to 7, corresponding to amino acid residues 111-317 included within the A1 domain. The size of the gene deletion was approximately 28 kb. 5' and 3' breakpoints were identified by sequencing in intron 3 and intron 7, respectively. FVIII mRNA was detected in the patient's peripheral lymphocytes and the deletion spanning exon 4 to 7 was confirmed at the RNA level. Immunoprecipitation experiments using 125I labeled A1, A2 and light chain demonstrated that the inhibitor reacted only with the 54 kDa A1 domain. The inhibitor activity was more than 95% neutralized by A1 domain polypeptide. Our findings suggest a close relationship between the inhibitor epitope and the specific gene deletion with regard to the pathogenesis of the inhibitor in this patient.

Anti-factor VIII inhibitor alloantibodies recognizing the A2 domain in the human factor VIII heavy chain poorly bind to porcine factor VIII.

Anti-factor VIII (FVIII) inhibitor alloantibodies from 11 patients with hemophilia A, along with five anti-FVIII neutralizing monoclonal antibodies, were examined for differences in their reactivities with the A2 and C2 domains of human and porcine FVIII. None of the patients had been previously treated with porcine FVIII. Six inhibitors which specifically recognized the human FVIII C2 domain bound to both the 76-kDa porcine FVIII light chain and its 69-kDa proteolyzed fragments, showing cross-reactivity against porcine FVIII between 33 and 100%. Two A2-specific inhibitors did not react with porcine FVIII. The cross-reactivity was low (0-0.5%). The inhibitors recognizing both C2 and A2 reacted with the 76- and 69-kDa bands of porcine FVIII light chain, with cross-reactivity of between 11 and 33%. Monoclonal antibodies recognizing A1 (C-5) and A2 (JR8) did not react with the porcine FVIII. No anti-porcine FVIII neutralizing activity was detected in these antibodies. Monoclonal antibodies to the amino-terminal portion of A3 (NMC-VIII/10 and C-2) poorly reacted with the 76-kDa band, the cross-reactivities being 0 and 0.5%, respectively. NMC-VIII/5 recognizing C2 which competes with the C2-specific inhibitor, reacted with both the 76- and 69-kDa fragments showing cross-reactivity of 13%. These findings suggest that porcine A2 is antigenically different from human A2. The A2-specific inhibitor is a useful indicator for therapy with porcine FVIII.

In hemophilia A and autoantibody inhibitor patients: the factor VIII A2 domain and light chain are most immunogenic.

Factor VIII (fVIII) is a protein cofactor essential for blood coagulation, and it binds in the factor Xase complex to factors IXa, X, and phospholipid. In about 30% of severe hemophilia A patients, treatment with fVIII leads to production of anti-fVIII antibodies. Anti-fVIII autoantibodies also rarely appear in normal individuals. Those antibodies that inactivate fVIII (inhibitors) prevent optimal fVIII therapy. Inhibitor epitopes were previously localized to the fVIII A2, A3, and C2 domains and to an acidic amino acid region between A1 and A2. Such anti-fVIII antibodies interfere with fVIII binding to components of the factor Xase complex and prevent blood coagulation. When total anti-fVIII titers were determined for each fVIII domain in 43 inhibitor plasmas by immunoprecipitation (IP) and inhibitor neutralization assays, the anti-light chain (LCh) antibody titer was highest, anti-A2 was intermediate, and anti-A1 and anti-B were low. The relative immunogenicity of the fVIII domains in hemophilic and autoantibody inhibitor patients was similar.

Human anti-factor VIII antibodies: epitope localization and inhibitory function.

Recombinant polypeptides derived from coagulation factor VIII (fVIII) have been used to determine the epitopes and characteristics of human pathologic anti-fVIII antibodies. The results of immunoprecipitation assays indicate that 70% of patient plasmas contain antibodies to the A2 as well as the C2 domains of the fVIII protein. The same polypeptides were used for inhibitor neutralization assays to demonstrate that about 60% of plasmas contain 2 or 3 different antibodies which collectively make up the inhibitor titer. The results of neutralization assays indicate that there is a third important inhibitor epitope within the light chain outside C2. Anti-A2 antibodies prevent normal function of the factor Xase complex of the intrinsic pathway of blood coagulation. Anti-C2 antibodies prevent the binding of fVIII to phospholipid and to von Willebrand factor, both of which are important for normal fVIII function.

Epitope specificity and inactivation mechanisms of factor VIII inhibitor antibodies.

The domain specificity of anti-factor VIII (FVIII) inhibitor antibodies was determined in assays using FVIII domains generated by thrombin cleavage or expressed as recombinant polypeptides to neutralise the inhibitor. The results revealed the existence of three major types of inhibitors, and various combinations of these antibodies were found in haemophilic and autoantibody patients. Anti-A2 domain inhibitors prevent normal function of the FVIII/factor IXa (FIXa)/phospholipid complex in an unknown manner. Binding of FVIII to phospholipid and to von Willebrand factor is blocked by anti-C2 domain antibodies, and the binding of FVIII to FIXa is prevented by anti-A3 domain antibodies. A rare type of inhibitor prevents release of activated FVIII from von Willebrand factor (vWf), and another probably interferes with FVIII binding to factor X (FX) because it shares the epitope of a monoclonal antibody with this property.

Properties of anti-factor VIII inhibitor antibodies in hemophilia A patients.

Blood coagulation factor VIII functions in the intrinsic pathway of blood coagulation as a cofactor by enhancing the assembly of its complex with factors IX and X on the surface of activated platelets. This requires factor VIII interaction with these two proteins, von Willebrand factor (vWF), and phospholipids on the platelet surface. Once factor VIII and factor IX are activated by proteolytic cleavage, the complex is able to activate factor X to factor Xa by proteolysis. In hemophilia A patients with severe factor VIII deficiency, about 30% respond to factor VIII infusion therapy immunologically to produce antibodies that inactivate the infused factor VIII and others that are noninhibitory. An assay that measures only the inhibitor antibodies demonstrated that the factor VIII A2, A3, and C2 domains are the most immunogenic, and domains A1 and B are poorly immunogenic or not immunogenic. The specific antibody responses to A2, A3, and C2 vary considerably among individuals, and epitopes for inhibitor antibodies have been determined for all three. The anti-C2 inhibitors prevent factor VIII binding to phospholipids and vWF, and anti-A3 inhibitors prevent binding to factor IX (IXa). An inhibitor binding site for factor X has been localized to the A1 domain acidic region, leading to inhibition of factor VIII/factor X binding by antibodies. This inhibitor mechanism is rare. Because a second binding site for factor IX was localized to the factor VIII A2 domain, it is likely but not proven that prevention of factor IX binding to factor VIII is the inhibitor mechanism for this epitope.

A mechanism for inhibition of factor VIII binding to phospholipid by von Willebrand factor.

von Willebrand factor (vWf) acts as a carrier for blood coagulation factor VIII (fVIII) in the circulation. The amino-terminal 272 residues of mature vWf contain a high affinity fVIII binding site. Upon thrombin activation, fVIII is released from vWf, thereby allowing its binding to phospholipid which is required for its procoagulant activity. Although phospholipid and vWf compete for fVIII binding, it was previously suggested that their binding sites are not closely juxtaposed within the fVIII protein because only amino-terminal vWf proteolytic fragments larger than SPIII-T4 (1-272) were able to block the binding of fVIII to phospholipid. We have demonstrated, however, that SPIII-T4 is able to inhibit fVIII binding to phosphatidylserine (PS) in a dose-dependent fashion, but only at concentrations higher than those used in previous experiments. Our demonstration that the Kd values for vWf and SPIII-T4 for fVIII are 0.52 nM and 48 nM, respectively, explain this discrepancy. Inhibition (> 95%) of SPIII-T4 binding to fVIII by a purified recombinant fVIII C2 domain polypeptide demonstrated that SPIII-T4 binds directly to C2, as we had previously shown for vWf. The similarity of the C2 binding sites for vWf and SPIII-T4 was further confirmed by the identical inhibitory effects of synthetic peptides and monoclonal antibodies (mAbs) on vWf-fVIII or SPIII-T4 fVIII binding. In both cases, binding was inhibited by synthetic peptide 2303-2332, containing a PS binding site, and by mAb NMC-VIII/5 Fab' (epitope within C2 residues 2170-2327). We propose that vWf, via residues 1-272, and PS compete for fVIII binding because they recognize overlapping sites within fVIII C2 domain residues 2303-2332.

A recombinant factor VIII A2 domain polypeptide quantitatively neutralizes human inhibitor antibodies that bind to A2.

Human antibodies that inactivate coagulation factor VIII (fVIII), known as inhibitors, have been shown by immunoblotting or immunoprecipitation assays to bind predominantly to epitopes within the A2 and/or C2 domains of the fVIII protein. Because these assays simply measure antibody binding, a soluble recombinant polypeptide containing the fVIII A2 domain was used to develop a quantitative inhibitor neutralization assay for antibodies that bound only to A2 by immunoblotting assay. Antibodies from six of eight inhibitor plasmas were fully neutralized by A2 (> or = 90%), whereas two were only partially neutralized. These results established the fVIII inhibitor properties of anti-A2 antibodies. In immunoprecipitation assays, five of the eight inhibitors also had significant levels of anti-light-chain antibody. In one case, this light-chain antibody was shown to have inhibitor activity. Because it did not bind to the C2 domain, this antibody appears to define a new inhibitor epitope within the fVIII light chain. Another inhibitor, which was partially neutralized by A2, was not neutralized by the light chain, even though it contained anti-light-chain antibodies by immunoprecipitation assay. Our results show additional complexities of the immune response to fVIII.

The acidic region of the factor VIII light chain and the C2 domain together form the high affinity binding site for von willebrand factor.

A binding site for von Willebrand factor (vWf) was previously localized to the carboxyl terminus of the C2 domain of the light chain (LCh) of factor VIII (fVIII). The acidic region of the LCh, residues 1649-1689, also controls fVIII.vWf binding by an unknown mechanism. Although anti-acidic region monoclonal antibodies prevent formation of the fVIII.vWf complex, the direct involvement of the acidic region in this binding has not been demonstrated. By limited proteolysis of LCh with Staphylococcus aureus V8 protease, we prepared 14- and 63-kDa LCh fragments, which begin with fVIII residues 1672 and 1795, respectively. Using surface plasmon resonance to measure binding interactions, we demonstrated that the 14-kDa fragment binds to vWf, but its affinity for vWf (Kd 72 nM) was 19-fold lower than that of LCh. This was not due to an altered conformation of the acidic region within the 14-kDa fragment, since its affinity for an anti-acidic region monoclonal antibody was similar to that of LCh. All LCh derivatives lacking the acidic region (thrombin-cleaved LCh, recombinant C2, and 63-kDa fragment) had also greatly reduced affinities for vWf (Kd 564-660 nM) compared with LCh (Kd 3.8 nM). In addition, the similar affinities of these derivatives for vWf indicated that apart from its acidic region, the LCh contains no vWf binding site other than the one within C2. The reduced affinities of the LCh derivatives lacking the acidic region for monoclonal antibody NMC-VIII/5 (epitope, C2 residues 2170-2327) indicated that removal of the acidic region leads to a conformational change within C2. This change is likely to affect the conformation of the vWf binding site in C2, which overlaps the epitope of NMC-VIII/5; therefore, the acidic region also appears to be required to maintain the optimal conformation of this vWf binding site. Our results demonstrate that the acidic region and the C2 domain are both directly involved in forming a high affinity binding site for vWf.

Human inhibitor antibodies specific for the factor VIII A2 domain disrupt the interaction between the subunit and factor IXa.

Factor VIIIa, a heterotrimer of the A1, A2, and A3-C1-C2 subunits, increases the catalytic efficiency for factor IXa-catalyzed activation of factor X. A significant fraction of naturally occurring, anti-factor VIII inhibitor antibodies reacts with the A2 domain. Utilizing the capacity for isolated A2 subunit to stimulate factor IXa activity, we show that a panel of these inhibitors block this activity. Inhibition of activity parallels the antibody potency as measured in the Bethesda assay. These antibodies also block the A2-dependent increases in fluorescence anisotropy of fluorescein-Phe-Phe-Arg factor IXa. Similar to the IgG fractions, a peptide representing the sequence of the inhibitor epitope (A2 residues 484-509) blocked the A2-dependent stimulation of factor IXa. These results indicate that antibodies possessing this specificity directly inhibit the interaction of A2 subunit with factor IXa, thus abrogating the contribution of this subunit to cofactor activity. Furthermore, these results also suggest that factor VIII residues 484-509 contribute to a factor IXa-interactive site.