Characterization of a factor VIII immunogenic site using factor VIII synthetic peptide 1687-1695 and rabbit anti-peptide antibodies

Characteristics, mechanisms of action, and epitope mapping of anti-factor VIII antibodies

The development of anti-factor VIII (FVIII) antibodies (Abs), also called inhibitors, is currently one of the most serious complications arising during the treatment of hemophilia A patients. Improved prevention and eradication of these Abs remain a challenge both for clinicians and scientists. Numerous studies in the literature have reported on their epitope specificity, on their mechanism of FVIII inactivation, as well as on the methods used for their detection. In this review, we summarize the current knowledge on the nature (isotypes, kinetic properties), epitope properties, and mechanisms of action of anti-FVIII Abs. Furthermore, we present methods for detection and epitope characterization of anti-FVIII Abs with emphasis on the Luminex technique susceptible to facilitate the monitoring of changes in the epitope specificity of these Abs.

Basic aspects of inhibitors to factors VIII and IX and the influence of non-genetic risk factors

The appearance of polyclonal antibodies inhibiting the function of exogenous factors VIII (FVIII) and IX (FIX) continues to be a major challenge in the treatment of patients with congenital haemophilia. Why these inhibitors develop in 10-20% of patients with haemophilia A, and in 1-5% of patients with haemophilia B, remains largely unexplained. The antibodies, however, are characterized by several features that may have implications for the immune process by which they occur. The FVIII antibodies are mainly directed towards the A2, A3 and C2 domains, thereby interfering with the function of the factor Xase complex, the binding of FVIII to von Willebrand factor, and the binding of FVIII to phospholipid membranes. The FIX epitopes are localized to the NH(2)-terminal gamma-carboxyglutamic acid region and the serine protease domain. Genetic risk factors are known to be of importance in the development of inhibitors, whereas the impact of non-genetic factors is less clear. However, based on studies of related subjects, it is obvious that non-genetic factors are of importance as well. Putative factors currently debated include age at the start of treatment, treatment in association with immune challenges, the type of product, and the mode of administration. Most of the findings reported to date, however, derive from small cohorts that have not been sufficiently well characterized with respect to genetic risk profile. Therefore, additional studies are required to quantify the impact of non-genetic factors on the pathophysiologic process of inhibitor development.

Environmental and genetic factors influencing inhibitor development

Inhibitor formation occurs at a frequency of 20% to 30% in severe hemophilia A, and 3% in hemophilia B. Today, it represents the major complication in patient care and renders classical substitution therapy ineffective. Genetic factors, such as factor VIII (FVIII) gene mutations and immune response genes, particularly the major histocompatibility complex, have been shown to constitute decisive risk factors for the development of inhibitors. In severe hemophilia A and B, those mutations that result in the absence or severe truncation of the FVIII/factor IX (FIX) proteins are associated with the highest risk for inhibitor formation, indicating that a major driving force in inhibitor development is the presentation of a novel antigen to the patient's immune system. An alternative pathomechanism may underlie inhibitor development in patients with mild hemophilia A. Missense mutations, especially those in the C1/C2 domains, may alter the immunogenicity of the FVIII protein, eliciting an inhibitor response against the mutated epitope. In some patients with hemophilia B, especially those with large deletions to the FIX gene, a severe allergic reaction occurs simultaneously with inhibitor onset. Despite the obviously strong genetic predisposition, discordant inhibitor status in monozygotic hemophilia A twins demonstrates that environmental factors also play a role in the development of inhibitors.

Inhibitor development in correlation to factor VIII genotypes

Alloantibodies (inhibitors) against factor VIII (FVIII) develop in 20-30% of patients with severe haemophilia A and render classical FVIII substitution therapy ineffective. Several studies have shown that genetic factors, the type of FVIII gene mutation and immune response genes (e.g. the Major Histocompatibility Complexes), influence the risk of inhibitor formation. In particular, the type of FVIII gene mutation has proven to be a decisive risk factor. Patients with severe molecular gene defects (e.g. large deletions, nonsense mutations, intron-22 inversion) and no endogenous FVIII synthesis have a 7-10 times higher inhibitor prevalence than patients with milder molecular gene defects (e.g. missense mutations, small deletions, splice site mutations). To date, at least 10 distinct classes of mutations have been shown which have differing risks of associated inhibitor formation. A challenging observation in inhibitor patients is the heterogeneity of the antibody epitopes with respect to their number and their specificity. At least five epitopes in the FVIII molecule have been identified that constitute the targets for antibodies in most inhibitor patients. These epitopes are located in the ar3 region and the A2, A3, C1, C2 domains which correspond to the functional binding sites of the ligands of the FVIII protein. At present, the determinants of the characteristics of these epitopes and the subsequent inhibitor titre are unknown. A relationship of the mutation site and the epitope localization has been shown for some individual patients with mild haemophilia A. However, in severely affected haemophilia A patients, the influence of patient genetics on inhibitor titre and epitope specificity has yet to be elucidated.

Epitope mapping of factor VIII inhibitor antibodies of Chinese origin

Epitopes recognized by factor VIII (FVIII) inhibitors of Chinese origin were analysed by immunoblotting with full-length recombinant FVIII (rFVIII), thrombin-activated FVIII (FVIIIa) and 16 FVIII fusion proteins synthesized by bacteria. Twenty-eight patients, 12 with haemophilia A and 16 with autoimmune diseases, were recruited. Antibodies from 22 patients showed reactivity with rFVIII, 20 with FVIIIa, and one reacted only with FVIII fusion proteins. Of these 22 cases, most were reactive with A2-a2 and A3-C1-C2 of FVIII(a). Of the nine cases that depicted binding to the fusion proteins, three were reactive with the A domains, three with only the B domain, and the other three with both the A and B (or C) domains. An epitope for a neutralizing antibody of a haemophilia A patient, designated TWN-112, was localized to residues 323-390, specified by FVIII fusion proteins. The same epitope also appeared on an FVIII-expression phage library screening. Immunoabsorption of antibodies from TWN-112 with the epitope reduced the neutralizing activity of the inhibitor by 33%. The incidence of a1 of FVIII is higher, and that of a3 is lower, than previously reported. Two novel epitopes, reported for the first time in this paper, were localized on the 8B2 (amino acid residues 1022-1204) and 8A2(V) (residues 673-740) fusion proteins. These two epitopes were able to reduce inhibitory antibody activity by 24% and 25% respectively. Changes of FVIII fragment specificity were also observed in one of six patients for whom multiple samples, collected at different times, were available. Our initial finding showed that the FVIII inhibitors in these Chinese patients shared epitopes with those of patients from very different genetic backgrounds, suggesting a common mechanism for the development of FVIII inhibitors.

Mapping of the minimal domain encoding a conformational epitope by lambda phage surface display: factor VIII inhibitor antibodies from haemophilia A patients

Haemophilia A patients who receive repeated transfusion of fVIII concentrates often develop inhibitor alloantibodies, resulting in reduced efficacy of the therapy. Determination of fVIII epitopes for the alloantibodies is essential for an understanding of their inhibitory effect on blood coagulation. Random fragments of fVIII displayed on lambda phage particles were selected using two patient plasmas immobilized onto the surface of a microtiter plate. A set of clones defined the minimal domain that consisted of 157 amino acid residues including cysteine at both boundaries. The minimal domain absorbed most of the binding activities of the plasmas to fVIII, suggesting that the domain contains a major determinant for the plasmas. Site-directed mutagenesis and chemical denaturation of the domain confirmed that a tertiary structure formed by the disulfide bridge was recognized by the antibodies. The epitope domain defined overlaps with fVIII binding sites to vWf and phospholipid, and may play an important role in blood coagulation. Thus, the bacteriophage lambda surface display may be useful for mapping the minimal folding domain of various protein antigens that contain a conformational epitope.

The Inhibitor Antibody Response Is More Complex in Hemophilia A Patients Than in Most Nonhemophiliacs With Factor VIII Autoantibodies

Approximately 25% of hemophilia A patients infused with factor VIII (fVIII) mount an immune response, which leads to its inactivation. Anti-fVIII autoantibodies are also seen rarely in individuals with normal fVIII. We have previously demonstrated that some anti-A2 and anti-C2 domain antibodies are fVIII inhibitors and that many patients have additional inhibitors with a fVIII light chain (LCh) epitope outside C2. Because the contribution of the different antibodies to the plasma inhibitor titer had been examined in a limited number of patients (14), we report in this study a more extensive analysis of 55 plasmas. The dominant inhibitors in 62% (13 of 21) of autoantibody plasmas were directed only against C2 or A2, but not both, whereas this pattern was found in only 15% (5 of 34) of hemophilic plasmas. In addition, anti-A2 inhibitors were present in 71% (24 of 34) of hemophilic plasmas, but only 33% (7 of 21) of autoantibody plasmas. These results demonstrated that the inhibitor response in hemophiliacs was more complex and the epitope specificity was somewhat different. A comparison of hemophiliacs treated only with plasma fVIII or recombinant fVIII showed no significant differences in the complexity of the inhibitor response, as ≥ 2 different inhibitor antibodies were present in 78% (18 of 23) of the former and 82% (9 of 11) of the latter. In contrast, the major inhibitors in 35% (8 of 23) of hemophiliacs treated with plasma fVIII were directed against C2 and another LCh epitope within residues 1649-2137, but not A2, while none (0 of 11) treated with recombinant fVIII had this pattern.

The Inhibitor Antibody Response Is More Complex in Hemophilia A Patients Than in Most Nonhemophiliacs With Factor VIII Autoantibodies

Approximately 25% of hemophilia A patients infused with factor VIII (fVIII) mount an immune response, which leads to its inactivation. Anti-fVIII autoantibodies are also seen rarely in individuals with normal fVIII. We have previously demonstrated that some anti-A2 and anti-C2 domain antibodies are fVIII inhibitors and that many patients have additional inhibitors with a fVIII light chain (LCh) epitope outside C2. Because the contribution of the different antibodies to the plasma inhibitor titer had been examined in a limited number of patients (14), we report in this study a more extensive analysis of 55 plasmas. The dominant inhibitors in 62% (13 of 21) of autoantibody plasmas were directed only against C2 or A2, but not both, whereas this pattern was found in only 15% (5 of 34) of hemophilic plasmas. In addition, anti-A2 inhibitors were present in 71% (24 of 34) of hemophilic plasmas, but only 33% (7 of 21) of autoantibody plasmas. These results demonstrated that the inhibitor response in hemophiliacs was more complex and the epitope specificity was somewhat different. A comparison of hemophiliacs treated only with plasma fVIII or recombinant fVIII showed no significant differences in the complexity of the inhibitor response, as ≥ 2 different inhibitor antibodies were present in 78% (18 of 23) of the former and 82% (9 of 11) of the latter. In contrast, the major inhibitors in 35% (8 of 23) of hemophiliacs treated with plasma fVIII were directed against C2 and another LCh epitope within residues 1649-2137, but not A2, while none (0 of 11) treated with recombinant fVIII had this pattern.

Hypothesis for the control of clotting factor VIII inhibitory antibodies by decreasing potency of helper T-cell-recognized epitopes in factor VIII

The study of the immunobiology of FVIII inhibitors may lead to new therapies for this potentially severe complication of haemophilia A and to new principles for the use of therapeutic proteins. In order to characterize the idiotype-anti-idiotype networks regulating FVIII inhibitors, we developed rabbit anti-idiotypic sera to 7 murine inhibitors and found at least 12 independent FVIII loci to which inhibitors could be raised. Rabbit antisera to the FVIII peptide, Ser1687-Thr1695, characterized one functional site to which about 46% of patients' inhibitor sera reacted. The multiplicity of inhibitor-recognized epitopes in FVIII makes it impractical, at the present time, to develop clinically useful specific anti-idiotypic therapies for FVIII inhibitors. Alternatively, one might induce genomic mutations in recombinant FVIII molecules to decrease immunogenicity of epitopes recognized by T helper cells. Methods to design such altered therapeutic proteins are presented, based on changing the longitudinal hydrophobic strip-of-helix which is in or near many T-cell-presented epitopes.