The X-linked bleeding disorder haemophilia A is due to a deficiency or functional defect of coagulation factor VIII. The bleeding tendency can be corrected by administration of factor VIII concentrates. A serious complication of factor VIII replacement therapy is the development of anti-factor VIII antibodies (inhibitors) that neutralize factor VIII activity. In recent years, the epitope-specifities and the inhibitory mechanisms of factor VIII inhibitors have gained increasing interest. The generation of factor VIII knock-out mice has opened the possibility of studying the immunobiology of inhibitor formation in murine models of haemophilia A. In spite of these recent developments however, the immunological mechanisms underlying the anti-factor VIII immune response have remained poorly understood so far. Most of our current knowledge is based on studies on inhibitor formation in the severe form of haemophilia. However, inhibitors also occur in patients with mild haemophilia A, in particular after a period of extensive factor VIII replacement therapy. These patients differ from severe haemophiliacs in that they have low levels of circulating factor VIII activity (5-25% of normal). The presence of endogenous factor VIII may have major impact on the immune response to exogenous factor VIII during replacement therapy. The studies presented in this thesis were performed to obtain a better understanding of the immunobiology of inhibitor development in mild haemophilia A. In the introduction (chapter 1), recent studies on the immunobiology of factor VIII inhibitors in haemophilia A patients are summarized and discussed. We have characterized the anti-factor VIII antibodies in patients with mild haemophilia A employing phage display technology. In chapter 2, anti-C2 antibodies were isolated and characterized from the repertoire of a mild haemophilia A patient. Our results provide evidence for the presence of two classes of anti-C2 antibodies that recognize distinct antigenic sites in factor VIII. The characteristics of the anti-C2 antibodies were further analysed in chapter 3, and compared to the epitopes of previously described murine monoclonal antibodies. The first class of anti-C2 antibodies bind to the epitope defined by monoclonal antibody ESH4. The second class of antibodies bind to the epitope defined by monoclonal antibody CLB-CAg 117. Antibodies belonging to this second class of antibodies were also isolated from a different patient with mild haemophilia A (chapter 4). The VH gene segment usage of the antibodies directed at the epitope defined by CLB-CAg 117 is less restricted compared to the first class of anti-C2 antibodies. Based on the long CDR3 region, we argue that this second class of antibodies originates from a pool of polyreactive human antibodies. In chapter 5, we describe the inhibitor development of a patient with mild haemophilia A caused by an Arg593 to Cys mutation. We have isolated and characterized anti-A2 antibodies using phage display and we have performed epitope-mapping studies of anti-factor VIII antibodies in plasma using immunoprecipitation analysis. The data presented in chapter 5 provide a possible explanation for anamnestic responses observed in patients with a history of inhibitor development. We propose that activation of a quiescent pool of memory B cells underlies the rise in inhibitor titer observed in haemophilia A patients with a history of inhibitor development. Chapter 6 describes the epitope specificities of anti-factor VIII antibodies in another patient from our cohort of mild haemophilia A patients with the Arg593 to Cys mutation. Results from this chapter and previous studies show that high responder patients with the Arg593 to Cys substitution develop inhibitory antibodies predominantly directed at the A2 domain of factor VIII. This suggests that inhibitor formation proceeds via a common mechanism in these patients. The role of HLA class II alleles in inhibitor formation was investigated by HLA genotyping of 42 patients with the Arg593 to Cys mutation. Our data suggest a weak association between inhibitor development and HLA class II alleles in mild haemophilia A patients with the Arg593 to Cys mutation. In Chapter 7, we present the characteristics of a mouse transgenic for human factor VIII with the Arg593 to Cys mutation (hufVIII-R593C mouse). The anti-factor VIII immune response was analysed in transgenic hufVIII-R593C mice crossed with factor VIII-deficient mice (exon 16 knock out, or E-16 KO mice). Serial intravenous injections of human factor VIII do not evoke an immune response in hufVIII-R593C/E-16 KO mice. The introduction of the human factor VIIIR593C transgene renders the mice tolerant to human factor VIII. However, when hufVIII-R593C/E-16 KO mice were subcutaneously injected with factor VIII in the presence of an adjuvant, loss of tolerance to factor VIII was observed. The results of chapter 7 demonstrate that hufVIII-R593C/E-16 KO mice provide a valuable model for studies directed at the mechanisms underlying inhibitor development in haemophilia A. Finally, chapter 8 provides a general overview that discusses the implications of our findings.
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.
[Show abstract][Hide abstract]ABSTRACT: Twenty six patients with mild or moderate haemophilia A and inhibitors are described. The inhibitor was detected at a median age of 33 years, after a median of 5.5 bleeding episodes. This usually following intensive replacement therapy. The median presenting inhibitor titre was antihuman 11.6 BU/ml, antiporcine 1.45 BU/ml. Plasma basal factor VIII level declined from a median of 0.08 IU/ml to 0.01 IU/ml following the inhibitor development. This caused spontaneous bleeding in 22 and a bleeding pattern similar to acquired haemophilia in 17. Bleeding was often severe and caused two deaths. The inhibitor disappeared spontaneously, or following immune tolerance induction, in 16 cases after a median of 9 months (range 0.5-46), with a return to the original baseline VIIIC level and bleeding pattern accompanied inhibitor loss. The inhibitor persisted in the remainder of the cases over a median period of 99 months (range 17-433 months) of follow-up. Inhibitors are an uncommon complication of mild haemophilia which frequently persist and may be associated with severe, life-threatening, haemorrhage. Forty-one percent of treated haemophilic family members had a history of factor VIII inhibitors, suggesting a familial predisposition to develop inhibitors in these kindreds. Sixteen patients from 11 families were genotyped. Seven different missense mutations affecting the light chain were detected and two in the A2 domain. Five patients from three families had a mutation causing a substitution of Trp2229 by Cys in the C2 domain which appears to predispose to inhibitor formation since 7 of the 18 affected individuals have a history of inhibitor development.
Full-text · Article · May 1998 · Thrombosis and Haemostasis
[Show abstract][Hide abstract]ABSTRACT: The A2 domain (residues 373-740) of human blood coagulation factor VIII (fVIII) contains a major epitope for inhibitory alloantibodies and autoantibodies. We took advantage of the differential reactivity of inhibitory antibodies with human and porcine fVIII and mapped a major determinant of the A2 epitope by using a series of active recombinant hybrid human/porcine fVIII molecules. Hybrids containing a substitution of porcine sequence at segment 410-508, 445-508, or 484-508 of the human A2 domain were not inhibited by a murine monoclonal antibody A2 inhibitory, mAb 413, whereas hybrids containing substitutions at 387-403, 387-444, and 387-468 were inhibited by mAb 413. This indicates that the segment bounded by Arg484 and Ile508 contains a major determinant of the A2 epitope. mAb 413 did not inhibit two more hybrids that contained porcine substitutions at residues 484-488 and 489-508, indicating that amino acid side chains on both sides of the Ser488-Arg489 bond within the Arg484-Ile508 segment contribute to the A2 epitope. The 484-508, 484-488, and 489-508 porcine substitution hybrids displayed decreased inhibition by A2 inhibitors from four patient plasmas, suggesting that there is little variation in the structure of the A2 epitope in the inhibitor population.
[Show abstract][Hide abstract]ABSTRACT: Inhibitory antibodies directed against factor VIII develop in a substantial number of patients with hemophilia A as a consequence of factor VIII replacement therapy. These antibodies usually recognize discrete epitopes within the A2 and/or the C2 domains of factor VIII. Here, we have characterized the antibodies present in the plasma of a patient affected by severe hemophilia A. The antibodies reacted readily with the metabolically labeled factor VIII light chain and fragments thereof when analyzed by immunoprecipitation. The inhibitory activity could be neutralized by the complete light chain, whereas only slight neutralization occurred with a fragment comprising the isolated C2 domain. Binding of the majority of antibodies to in vitro synthesized factor VIII fragments was dependent on the presence of amino acid residues Gln1778-Met1823, a region known to contain a factor IXa binding site. Functional characterization showed that purified IgG from the patient's serum inhibited binding of factor IXa to immobilized factor VIII light chain in a dose-dependent manner. These data indicate that human alloantibodies may inhibit factor VIII activity by interfering with factor IXa-factor VIIIa complex assembly.
[Show abstract][Hide abstract]ABSTRACT: Factor VIII (fVIII) functions as a cofactor of factor IXa in the intrinsic pathway of blood coagulation. Its absence or abnormality causes the bleeding disorder hemophilia A. About 23% of hemophiliacs who receive therapeutic fVIII infusions develop antibodies that inhibit its activity. We previously showed by inhibitor neutralization assays that the fVIII A2 and C2 domain polypeptides contain common inhibitor epitopes. Often hemophilic inhibitor plasmas were partially neutralized by C2 and more completely neutralized by fVIII light chain (A3-C1-C2), suggesting the presence of an additional major inhibitor epitope(s) within the A3-C1 domains. In immunoprecipitation assays, 17 of 18 inhibitor IgGs bound to recombinant 35S-A3-C1. Amino acids 1811-1818 of the A3 domain comprise a binding site for factors IX and IXa. Three inhibitor IgGs prevented binding of factor IXa to fVIII light chain, and the binding of each IgG to light chain was competed by A3 peptide 1804-1819. The generation of factor Xa by the fVIIIa/fIXa complex in a chromogenic assay was prevented by these inhibitors. Therefore, we propose that another important mechanism of fVIII inactivation by human inhibitors is the prevention of fVIIIa/fIXa association.
[Show abstract][Hide abstract]ABSTRACT: The finding that human factor VIII (fVIII) inhibitor antibodies with C2 domain epitopes interfere with the binding of fVIII to phosphatidylserine (PS) suggested that this is the mechanism by which they inactivate fVIII. We constructed a recombinant C2 domain polypeptide and demonstrated that it bound to all six human inhibitors with fVIII light chain specificity. Thus, some antibodies within the polyclonal anti-light chain population require only amino acids within C2 for binding. Recombinant C2 also partially or completely neutralized the inhibitor titer of these plasmas, demonstrating that anti-C2 antibodies inhibit fVIII activity. Immunoblotting of a series of C2 deletion polypeptides, expressed in Escherichia coli, with inhibitor plasmas showed that the epitopes for human inhibitors consist of a common core of amino acid residues 2248 through 2312 with differing extensions for individual inhibitors. The epitope of inhibitory monoclonal antibody (MoAb) ESH8 was localized to residues 2248 through 2285. Three human antibodies and anti-C2 MoAb NMC-VIII/5 bound to a synthetic peptide consisting of amino acids 2303 through 2332, a PS-binding site, but MoAb ESH8 did not. These antibodies also inhibited the binding of fVIII to synthetic phospholipid membranes of PS and phosphatidylcholine, confirming that the blocked epitopes contribute to membrane binding as well as binding to PS. In contrast, MoAb ESH8 did not inhibit binding. As the maximal function of activated fVIII in the intrinsic factor Xase complex requires its binding to a phospholipid membrane, we propose that fVIII inhibition by anti-C2 antibodies is related to the overlap of their epitopes with the PS-binding site. MoAb ESH8 did not inhibit fVIII binding to PS-containing membranes, suggesting the existence of a second mechanism of fVIII inhibition by anti-C2 antibodies.
[Show abstract][Hide abstract]ABSTRACT: The human blood coagulation factor VIII C2 domain (Ser2173-Tyr2332) contains an epitope recognized by most polyclonal inhibitory anti-factor VIII alloantibodies and autoantibodies. We took advantage of the differential reactivity of inhibitory antibodies with human and porcine factor VIII and mapped a major determinant of the C2 epitope by using a series of active recombinant hybrid human/porcine factor VIII molecules. A series of five C2-specific human antibodies and a murine anti-factor VIII monoclonal antibody, NMC-VIII/5, inhibited a hybrid containing a substitution of porcine sequence for Glu2181-Val2243 significantly less than human factor VIII. In contrast, four of the five patient antibodies and NMC-VIII/5 inhibited a hybrid containing a substitution of porcine sequence for Thr2253-Tyr2332 equally well as human factor VIII. Thus, a major factor VIII inhibitor epitope determinant is bounded by Glu2181-Val2243 at the NH2-terminal end of the C2 domain. Because C2 inhibitors block the binding of factor VIII to phospholipid and von Willebrand factor, for which binding sites have been localized to Thr2303-Tyr2332, these results imply that the segment bounded by Glu2181-Val2243 also is involved in these macromolecular interactions.