fVIII C2 Domain Research Articles

Factor VIII Proteins Having a Rationally Modified, Immunodominant T-Cell Epitope Demonstrate Normal Procoagulant Activity, Bind To VWF With High Affinity, and Are Markedly Less Stimulatory To FVIII-Specific Human T Cells

Neutralizing anti-factor VIII (FVIII) antibodies, referred to clinically as “inhibitors”, can develop as an alloimmune response in hemophilia A patients receiving FVIII infusions as replacement therapy. Immune Tolerance Induction, consisting of intensive FVIII treatment in an effort to tolerize patients to FVIII, is extraordinarily expensive and not always successful. New approaches to avoid inhibitor development, and to treat patients who develop this deleterious immune response, are needed. Initial stages of inhibitor development include FVIII uptake and processing into peptides by antigen-presenting cells such as dendritic cells (DCs), presentation of FVIII peptides on MHC Class II (HLA) receptors on the cell surface, and recognition of HLA-peptide complexes by one or more circulating T cells. Subsequent signaling through immunological synapses between antigen-presenting cells and T-cell receptors causes proliferation of effector T cells, which secrete cytokines promoting anti-FVIII antibody production.Our laboratory has been identifying T-cell epitopes in FVIII, which are amino acid sequences 11-15 residues long that bind to specific HLA-DR proteins and are in turn recognized by T cells. FVIII 2194-2205 comprises an immunodominant T-cell epitope that binds to HLA-DRB1*01:01. Peptide-MHC binding experiments established that the side chains of amino acid residues F2196, M2199, A2201 and S2204 fit into the HLA-DRB1*01:01 peptide-binding groove, allowing peptides containing this sequence to be presented on the surface of antigen-presenting cells. Fluorescent HLA-DRB1*01:01 tetramers loaded with FVIII2194-2213 were used to stain and isolate FVIII-specific T cells from 3 hemophilia A subjects who had an HLA-DRB1*01:01 allele. These cells were expanded in culture to generate T-cell clones and polyclonal lines that recognize this sequence and proliferate in response to it. Stimulation of clones with FVIII peptides containing systematic alanine substitutions demonstrated that MHC anchor residues F2196 and M2199 are important for the immunogenicity of this T-cell epitope. The clones were then stimulated with FVIII peptides and recombinant FVIII-C2 domain proteins in which F2196 was changed to a series of other residues. The effect of substitutions at M2199 was examined using the ProPred computer prediction program. Three substitutions that significantly reduced T-cell proliferation (F2196A, F2196L, F2196K) or were predicted to do so (M2199A, M2199W, M2199R) were introduced into recombinant, B-domain-deleted (BDD)-FVIII. The present study tests the hypothesis that less immunogenic FVIII proteins having normal FVIII procoagulant activity can be produced through rational modification of T-cell epitopes. The F2196K and M2199A muteins were expressed at levels similar to wild-type (WT)-BDD-FVIII in BHK-M cells and were purified from serum-free BHK-M cell supernatants. Purified CD14-positive monocytes from individuals with the HLA-DRB1*01:01 allele were differentiated into DCs and used for antigen-presentation assays in which 4 hemophilic T-cell clones and 2 polyclonal lines were added to the DCs and stimulated with FVIII2194-2213, WT-FVIII, WT-BDD-FVIII and the BDD-FVIII muteins. One clone proliferated weakly in response to both the WT and mutant proteins, while the remaining 3 clones and 2 lines showed markedly less proliferation in response to the BDD-FVIII muteins. BDD-FVIII-2196K and 2199A had specific activities similar to that of WT-BDD-FVIII (by chromogenic and clotting assays), and surface plasmon resonance confirmed that they retained high-affinity binding to von Willebrand factor. These results provide proof-of-principle for the design of less immunogenic FVIII proteins targeted to specific subsets of hemophilia A patients. Disclosures:Pratt:Bayer, Pfizer, CSL Behring: Research Funding; Puget Sound Blood Center Research Institute: sequence-modified FVIII variants Patents & Royalties.

Phenotypes of Allo- and Autoimmune Antibody Responses to FVIII Characterized by Surface Plasmon Resonance

Evidence of antibody isotype/subtype switching may provide prognostic value regarding the state of immune responses to therapeutic proteins, e.g. anti-factor VIII (FVIII) antibodies that develop in many hemophilia A patients, clinically termed “inhibitors”. A sensitive, high- information-content surface plasmon resonance (SPR) assay has been developed to quantify IgG subtype distributions and the domain specificity of anti-drug antibodies. Plasma samples from 22 subjects with an allo- or auto-immune reaction to FVIII were analyzed. Pre-analytical treatment protocols were developed to minimize non-specific binding and specific matrix interference due to von Willebrand factor-FVIII interactions. The dynamic range for IgG quantification was 0.2–5 µg/ml (∼1–33 nM), allowing characterization of inhibitor-positive samples. Subtype-specific monoclonal antibodies were used to quantify the IgG subtype distribution of FVIII-specific antibodies. Most samples obtained from multiply-infused inhibitor subjects contained IgG4 antibodies. Several distinct phenotypes were assigned based on the IgG subtype distribution: IgG1, IgG4, IgG1 & IgG4, and IgG1, IgG2 & IgG4. An IgG1-only response was found in mild/moderate HA subjects during early FVIII infusions, and analysis of serial samples followed antibody class switching as several subjects’ immune responses developed. Competition studies utilizing a recombinant FVIII-C2 domain indicated 40–80% of FVIII-specific antibodies in most samples were directed against this domain.

Design of less immunogenic therapeutic factor VIII proteins (P3169)

Abstract Hemophilia A patients lack functional factor VIII (FVIII), which is essential for normal blood clotting. Approximately 1 in 4 patients receiving replacement therapy by infusions of FVIII develop neutralizing antibodies that can cause serious bleeds. Tetramer-guided epitope mapping (TGEM) experiments identified an HLA-DRB1*01:01-restricted T-cell epitope in 2 subjects with missense substitution A2201P. The same epitope was identified using samples from a subject with severe hemophilia. T-cell clones were isolated by single-cell sorting and expansion of tetramer-positive CD4 T cells. Peptide-MHC binding assays using recombinant DR0101 and FVIII peptides with systematic arginine substitutions identified anchor residues F2196, M2199, A2201 and S2204. Recombinant FVIII-C2 domain with substitution F2196A abrogated proliferation of 4 clones. Six FVIII proteins with rational substitutions designed to interfere with epitope presentation (F2196A, F2196K, F2196L, M2199A, M2199W and M2199R) were produced in a BHK expression system. All had pro-coagulant activities, measured by both clotting and chromogenic assays, similar to that of wild-type FVIII. Their immunogenicity is now being evaluated. These experiments provide proof of principle for the design of less immunogenic, functional FVIII proteins targeted to specific subsets of patients.

Characterization and Solution Structure of the Factor VIII C2 Domain in a Ternary Complex with Classical and Non-classical Inhibitor Antibodies

The most significant complication for patients with severe cases of congenital or acquired hemophilia A is the development of inhibitor antibodies against coagulation factor VIII (fVIII). The C2 domain of fVIII is a significant antigenic target of anti-fVIII antibodies. Here, we have utilized small angle x-ray scattering (SAXS) and biochemical techniques to characterize interactions between two different classes of anti-C2 domain inhibitor antibodies and the isolated C2 domain. Multiple assays indicated that antibodies 3E6 and G99 bind independently to the fVIII C2 domain and can form a stable ternary complex. SAXS-derived numerical estimates of dimensional parameters for all studied complexes agree with the proportions of the constituent proteins. Ab initio modeling of the SAXS data results in a long kinked structure of the ternary complex, showing an angle centered at the C2 domain of ∼130°. Guided by biochemical data, rigid body modeling of subunits into the molecular envelope of the ternary complex suggests that antibody 3E6 recognizes a C2 domain epitope consisting of the Arg(2209)-Ser(2216) and Leu(2178)-Asp(2187) loops. In contrast, antibody G99 recognizes the C2 domain primarily through the Pro(2221)-Trp(2229) loop. These two epitopes are on opposing sides of the fVIII C2 domain, are consistent with the solvent accessibility in the context of the entire fVIII molecule, and provide further structural detail regarding the pathogenic immune response to fVIII.

Sequence-Modified Factor VIII Variants Having Reduced Immunogenicity

AbstractAbstract 39Neutralizing anti-factor VIII (FVIII) antibodies, or “inhibitors”, interfere with FVIII pro-coagulant activity, and persistent inhibitors can result in significant morbidity and mortality in hemophilia A (HA) patients and in individuals who develop autoantibodies to their endogenous FVIII. Inhibitor production follows stimulation of helper T cells by linear amino acid sequences in FVIII corresponding to HLA-restricted T-cell epitopes. An immunodominant HLA-DRB1*01:01-restricted T-cell epitope within a peptide corresponding to FVIII residues 2194–2213 was identified previously using blood samples from two mild HA subjects with hemophilic mutation A2201P. This same immunodominant epitope was found recently in inhibitor subjects with (a) a large F8 gene deletion and b) an F8 nonsense mutation in exon 12. The present study aims to identify amino acid substitutions in FVIII that will neutralize this T-cell epitope while preserving the pro-coagulant activity of the modified FVIII protein. MHC class II - peptide binding assays were carried out using truncated FVIII peptides to determine the shortest sequence with full binding affinity for a recombinant HLA-DR0101 protein, and subsequently using peptides having systematic arginine substitutions at each position to identify the specific residues that confer this binding affinity. The results indicated that FVIII2194–2205 is the minimal binding epitope and that residues F2196, M2199, A2201 and S2204 interact with the peptide-binding groove of HLA-DR0101. Next, four T-cell clones that all proliferate in response to this epitope but have different T-cell receptors were stimulated with 12 peptides having systematic alanine substitutions at each position of FVIII2194–2205. The F2196A substitution abrogated proliferation of all four clones. The M2199A-substituted peptide stimulated three of the clones more weakly than the wild-type peptide. Peptide binding and T-cell assays were next carried out with FVIII2194–2205 peptides in which the 19 common non-phenylalanine amino acids were substituted at position 2196. These results identified 12 different amino acid substitutions that decreased both MHC binding and T-cell proliferation more than 10-fold. The binding of FVIII2194–2205 and FVIII2194–2205, F2196A to 10 common HLA-DRB1 proteins was measured to determine the potential promiscuity of this epitope. Moderate or low affinity binding of FVIII2194–2205 (IC50 < 50 mM) to DR0401, DR0404, DR0901, DR1001, and DR1501 was observed. FVIII2194–2205, 2196A did not bind to any of the HLA-DRB1 proteins, suggesting that this substitution would not introduce a neo-epitope recognized by these other common MHC class II receptors. A recombinant FVIII-C2 domain protein with substitution F2196A was generated in E. coli and purified to homogeneity following a procedure that removes endotoxin. This FVIII-C2 mutein failed to stimulate the same four T-cell clones, all of which showed a strong, dose-dependent response to wild-type FVIII-C2. Recombinant B-domain-deleted FVIII (BDD-FVIII) proteins with substitutions F2196A, F2196L, F2196K, M2199A, M2199W and M2199R were expressed in BHK-M cell lines. Multiple cell lines were generated to express wild-type BDD-FVIII and each of these mutant proteins. Expression levels of the muteins were similar to that of wild-type BDD-FVIII except for the M2199W and F2196A variants, which had expression levels ∼30 and 10% that of wild-type BDD-FVIII, respectively. Specific activities of the muteins, measured using chromogenic and clotting assays, were similar to that of wild-type BDD-FVIII. Binding of these muteins to plasma-derived von Willebrand factor was evaluated by ELISAs, as a surrogate assay to indicate possible effects of specific mutations on FVIII half-life in the circulation. Their affinities for VWF ranged from ∼40–100% that of wild-type BDD-FVIII. Our results suggest that FVIII muteins with amino acid substitutions that abolish binding to DR0101 and retain reasonable FVIII functionality could be developed as less immunogenic therapeutic proteins, in order to avoid HLA-DRB1*01:01-restricted immune responses in HA patients with this common allele. The immunogenicity of this T-cell epitope and of the sequence-modified peptides and proteins in HA subjects with other HLA-DRB1 alleles is currently under investigation. Disclosures:Pratt:Puget Sound Blood Center: Employment, patent describing design of novel factor VIII proteins Other.

Backbone resonance assignments of the C2 domain of coagulation factor VIII

Factor VIII (FVIII, other clotting factors are named similarly) is a glycoprotein that circulates in the plasma bound to von Willebrand factor. During the blood coagulation cascade, activated FVIII (FVIIIa) binds to FIXa and activates FX in the presence of calcium ions and phospholipid membranes. The C1 and C2 domains mediate membrane binding that is essential for activation of the FVIIIa-FIXa complex. Here, (1)H, (13)C, and (15)N backbone chemical shift assignments are reported for the C2 domain of FVIII, including assignments for the residues in solvent-exposed loops. The NMR resonance assignments, along with further structural studies of membrane-bound FVIII, will advance understanding of blood-clotting protein interactions.

High-Resolution Mapping of B-Cell Epitopes on the Factor VIII C2 Domain Using Surface Plasmon Resonance

Abstract 203Neutralizing anti-factor VIII (FVIII) antibodies that develop in hemophilia A mice following exposure to human FVIII bind to several distinct regions (B-cell epitopes) on the FVIII C2 domain. This 19-kDa region of FVIII binds to thrombin and factor Xa during proteolytic activation of FVIII, and it also binds to phospholipids, von Willebrand factor and factor IXa; all of these interactions are essential for the cofactor effect of FVIII in the intrinsic tenase reaction, which is a critical control point in blood coagulation. The binding of antibodies to any of these surfaces on the C2 domain can block FVIII activity, causing serious bleeding that is difficult to control.In order to fine-map B-cell epitopes on FVIII, 60 recombinant N-terminal His-tagged FVIII-C2 proteins were generated in an E. coli expression system and purified to homogeneity. Each had a surface-exposed amino acid residue mutated to alanine. The binding kinetics of these FVIII-C2 muteins to 10 inhibitory anti-FVIII-C2 monoclonal antibodies (mAbs) was then evaluated by surface plasmon resonance (SPR). The muteins were injected over biosensors on which the mAbs were captured using a high affinity goat anti-mouse IgG that was covalently attached to the Biacore CM5 biosensor surface by amine coupling. The anti-FVIII-C2 mAbs were: ESH8, ESH4, 2–77, 2–117, 3D12, 3E6, I109, 1B5, 3G6 and I54. Single-cycle kinetic experiments were carried out for each FVIII-C2 mutein/mAb pair, and variants having a dissociation rate >4X that measured for wild-type FVIII-C2 were marked as candidates for the corresponding B-cell epitope. The dissociation rate, rather than the binding affinity, was found to be the best diagnostic measurement to identify side-chain substitutions that altered the antigen-antibody binding interaction. Initial SPR kinetics experiments were run in a “blinded” fashion, meaning the FVIII-C2 crystal structure (Pratt et al., Nature 402, 439–42, 1999) was not consulted either before or during analysis of the binding kinetics. Subsequent visualization of the crystal structure using a molecular graphics program allowed us to locate the sites of amino acid substitutions that perturbed the binding of particular mAbs to the FVIII-C2 surface. Amino acid substitutions that affected binding to multiple mAbs, or that occurred at sites that were non-contiguous with clusters of surface-exposed residues identified in this manner, were not considered part of a given epitope as they clearly affected the protein folding/stability. FVIII-C2 muteins used for this analysis bound to most of the mAbs with affinities similar to that of wild-type FVIII-C2, indicating that they were properly folded. For these proteins, observation of altered kinetics in binding to a particular mAb indicated a change had occurred at the antigen-antibody binding interface. Averaged kon and koff values were obtained from 10 SPR runs per mAb, measuring the binding kinetics to WT-FVIII-C2. These experiments yielded accurate kinetics and affinities of each mAb for the FVIII C2 domain.Clusters of surface-exposed residues with significantly altered binding kinetics identified specific sites comprising “functional” B-cell epitopes, defined as those residues contributing appreciable antigen-antibody binding affinity. All of these mAbs were previously shown, by competition ELISAs, to correspond to partially overlapping epitopes, designated A, AB, B, BC and C; binding of antibodies to these epitopes interfered with: (1) proteolytic activation of FVIII by thrombin or Factor Xa; (2) binding of FVIII to phospholipid surfaces and/or von Willebrand factor; (3) a chromogenic factor X activation assay (Meeks et al., Blood 110:4234–4242, 2007). Fine-mapping of these epitopes by SPR indicated that the mAbs assigned to classes A, B and C indeed bound to distinct regions of the FVIII-C2 surface. Identification of these epitopes also pinpoints functionally important regions on the FVIII C2 domain, and these results contribute further insight into the binding interactions leading to FVIII activation to FVIIIa and its participation in the intrinsic tenase complex at negatively charged phospholipid surfaces. Disclosures:Pratt:Bayer, CSL Behring, Pfizer: Research Funding; Puget Sound Blood Center: Patents & Royalties.

Determination of Anti-Factor VIIII Antibody Titers, Isotypes and Epitopes by SPR Analysis of Human Plasma

Abstract 202Development of neutralizing anti-factor VIII (FVIII) allo-antibodies (“inhibitors”) occurs in a significant proportion of congenital Hemophilia A (HA) patients receiving exogenous FVIII, and autoimmune responses to FVIII can also occur. Accurate early detection of these immune responses, including low-titer or non-neutralizing antibodies, could have prognostic value and thus assist clinicians in evaluating therapeutic options. Rapid, sensitive, high-information-content surface plasmon resonance (SPR) assays have been developed to quantify the titer, IgG subtype distribution and domain/epitope specificity of anti-FVIII antibodies by direct analysis of pre-treated plasma samples. Samples were injected directly onto the biosensor of a Biacore T-100 instrument, on which recombinant FVIII was captured via a covalently immobilized anti-FVIII monoclonal antibody (mAb), and the increase in resonance units (RU) was converted to antibody titers. The lower limit of detection was 100ng/mL (∼1nM ∼0.01 Bethesda units/mL), allowing characterization of anti-FVIII antibodies that could not be detected by the Bethesda assay. Subsequent serial injections of isotype-specific anti-human mAbs yielded quantitative measurements of the fractions of anti-FVIII antibodies with specific subtypes. Plasma samples from 200 HA subjects age 2 and above (with and without a measurable inhibitor) as well as samples from 3 autoimmune adults were analyzed. Distinct IgG subtype distributions were observed, with most inhibitor subjects exhibiting mixed IgG1 + IgG4, or IgG1 + IgG2 + IgG4 responses. Serial samples from 1 subject following initial FVIII infusions showed transient class switching with the appearance of anti-FVIII-IgG3.Epitope mapping was next carried out via competition studies utilizing (a) a panel of 11 monoclonal antibodies (mAbs) whose epitopes on the FVIII-C2 domain have been mapped at high resolution and (b) recombinant FVIII-C2 proteins. The mAb-competition assays allowed the (FVIII-C2) epitope specificity of anti-FVIII antibodies in plasma to be mapped at a similar resolution, identifying specific clusters of surface-exposed FVIII residues that comprise or overlap immunodominant epitope(s) recognized by human antibodies. Complementary solution-phase competition SPR assays utilized wild type FVIII-C2 protein and selected proteins from a panel of 60 FVIII-C2 point mutants (muteins) that were identified during epitope mapping of the mAbs by SPR. This second approach alleviated several technical issues associated with the earlier blocking studies that utilized mAbs to compete the polyclonal plasma IgG (e.g. dissociation of the blocking mAb and steric effects) and provided reproducible quantitative results. Because saturating concentrations of FVIII-C2 proteins were used, changes in the number of RUs measured before and after addition of WT-FVIII-C2 indicated the fraction of polyclonal anti-FVIII antibodies that recognized this domain. Notably, the IgG subtype distribution of all FVIII-C2 specific antibodies reflected the total IgG subtype distribution, indicating that affinity maturation of B-cell epitopes preceded antibody class switching. Interestingly, saturating amounts of either FVIII-C2 or the high-affinity anti-FVIII-C2 mAb BO2C11 completely competed binding of antibodies from 1 subject (with HA genotype A2201P) to FVIII captured on the biosensor surface, indicating that his antibody response was to a single B-cell epitope or else to closely overlapping epitopes. However, although rare subjects had a completely FVIII-C2 specific response or no binding to FVIII-C2, between 30–50% of the anti-FVIII antibodies in the vast majority of inhibitor-positive samples targeted the FVIII-C2 domain. This was true for inhibitor subjects with severe HA due to the intron 22 inversion, for subjects with other HA genotypes resulting in severe HA and for the 3 autoimmune subjects.Results of these competition SPR assays are directly translatable to our laboratory's efforts to design FVIII muteins that retain adequate pro-coagulant activity but can avoid neutralization by anti-FVIII antibodies in inhibitor patients' blood. FVIII muteins having amino acid substitutions that eliminate immunodominant B-cell epitopes are expected to have potential for development as less immunogenic/antigenic bypass therapies for patients who develop a high-titer inhibitor. Disclosures:Howard:Haplomics: Equity Ownership; Baxter, Bayer, Biogen: Research Funding; Haplomics: Patents & Royalties. Pratt:Bayer, CSL Behring, Pfizer: Research Funding; Puget Sound Blood Center: Patents & Royalties.

De-Immunization of Immunodominant HLA-DR Epitopes In the C2 Domain of Factor VIII

AbstractAbstract 2207Development of anti-factor VIII (FVIII) antibodies that neutralize FVIII activity is a major obstacle of replacement therapy in hemophilia A. To create a less immunogenic therapeutic protein, recombinant FVIII can be modified to reduce binding of FVIII sequences to HLA, based on predicted anchoring residues. This process is known as ‘de-immunization’. Previously, we found that immunogenic peptide epitopes in the C2 domain of FVIII could be modified to reduce MHC class II binding resulting in lower antigenicity in vitro in hemophilic E16 mice (H-2b). Here, we extended these studies to HLA-DR transgenic mice and identified immunodominant C2 peptides in E16, DR3 and DR4 mice as well as in DR3 mice crossed to E16 mice (DR3×E16). E16 and DR transgenic mice were immunized with the designated original high class II-binding peptides (ORG) or modified peptides (MOD) emulsified in CFA, and lymph node (LN) cells were tested in T-cell proliferation and ELISpot assays against the ORG and MOD peptides. We found that while p2191-2210 dominantly stimulates H-2b, p2271-89 is immunodominant in DR3 transgenic animals; DR3×E16 T cells recognized both epitopes as would be predicted. On the other hand, p2231-51 and p2310-30 were immunodominant in DR4 mice. We also found that modifications of dominant epitope peptides 2191, 2271, 2231, 2310 showed lower HLA class II binding and T-cell proliferation in corresponding mouse strains; stimulation of IL17-a and IFNγ cytokine production were reduced with modified 2191 in DR3xE16 mice. Direct immunization with rFVIII containing modified peptides is required to validate potential immunogenicity. These results will be utilized to design and produce a functional full-length modified rFVIII that can be used as a less immunogenic therapeutic protein for hemophilia A patients. (Supported by NIH grant R43 HL088834-01 to ADG.) Disclosures:Moise: EpiVax: Employment. Tassone: EpiVax: Employment. McClaine: EpiVax: Employment. Martin: EpiVax: Employment, Equity Ownership, Patents & Royalties. DeGroot: EpiVax: Employment, Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties. Scott: EpiVax: Consultancy.

Factor VIII Lacking the C2 Domain Retains Cofactor Activity in Vitro

Factor (F) VIII consists of a heavy chain (A1A2B domains) and light chain (A3C1C2 domains). The activated form of FVIII, FVIIIa, functions as a cofactor for FIXa in catalyzing the membrane-dependent activation of FX. Whereas the FVIII C2 domain is believed to anchor FVIIIa to the phospholipid surface, recent x-ray crystal structures of FVIII suggest that the C1 domain may also contribute to this function. We constructed a FVIII variant lacking the C2 domain (designated DeltaC2) to characterize the contributions of the C1 domain to function. Binding affinity of the DeltaC2 variant to phospholipid vesicles as measured by energy transfer was reduced approximately 14-fold. However, the activity of DeltaC2 as measured by FXa generation and one-stage clotting assays retained 76 and 36%, respectively, of the WT FVIII value. Modest reductions ( approximately 4-fold) were observed in the functional affinity of DeltaC2 FVIII for FIXa and rates of thrombin activation. On the other hand, deletion of C2 resulted in significant reductions in FVIIIa stability ( approximately 3.6-fold). Thrombin generation assays showed peak thrombin and endogenous thrombin potential were reduced as much as approximately 60-fold. These effects likely result from a combination of the intermolecular functional defects plus reduced protein stability. Together, these results indicate that FVIII domains other than C2, likely C1, make significant contributions to membrane-binding and membrane-dependent function.

Deimmunization of HLA-DR Epitopes in the C2 Domain of Human FVIII (144.29)

Abstract To reduce immunogenicity of therapeutic proteins, recombinant proteins can be modified to lower HLA binding of epitopes, a method (deimmunization) useful with monoclonal antibodies. Since the immune response to human clotting factor VIIII limits therapeutic efficacy, we reported that immunodominant peptides in the C2 domain of fVIII could be modified to reduce MHCII binding to display less antigenicity in vitro in a hemophilic fVIII KO mouse model (H-2b). We have extended these studies in humanized HLA-DR transgenic mice and identified responses to C2 peptides p2191-2210, p2231-51, p2254-69, p2271-89; and p2310-30 in DR2, DR3, DR4 mice, as well as in DR3 mice crossed to fVIII KO hemophilic mice. DR transgenic mice were immunized with the designated original peptides (ORG) or with rFVIII in CFA, and LN cells tested in T-cell proliferation assays against rFVIII and ORG peptides. While p2191 dominantly stimulates H-2b, other peptides are immunodominant in DR2, DR3 and DR4 mice. A different pattern was revealed (DR3 x fVIII KO) mice depending on whether they also express H-2b. Efforts to test the reactivity/antigencity and immunogenicity of modified (MOD) peptides predicted to have lower class II binding will be reported. These results will be utilized to design and produce a functional full-length modified rFVIII that can be used as a therapeutic protein in hemophilia A (Supported by NIH R43 HL088834-01).

A human FVIII inhibitor modulates FVIII surface electrostatics at a VWF-binding site distant from its epitope.

BO2C11 is a human monoclonal factor (F) VIII inhibitor. When bound to the C2 domain of FVIII, the Fab fragment of BO2C11 (Fab(BO2C11)) buries a surface of C2 that contains residues participating in a binding site for von Willebrand factor (VWF). BO2C11 has thus been proposed to neutralize FVIII by steric hindrance. The BO2C11 epitope on C2 overlaps with residues located at the periphery of the putative VWF binding site; hence, most of the residues that constitute the VWF binding site on C2 and a3 remain accessible for VWF interaction following BO2C11/FVIII complex formation. We thus investigated the contribution of alternative molecular mechanisms to FVIII inactivation by BO2C11. Continuum electrostatic calculations were applied to the crystal structure of C2, free or Fab(BO2C11)-complexed. In silico predictions were confirmed by site-directed mutagenesis and VWF-binding assays of the mutated FVIII. Binding of Fab(BO2C11) to C2 induced perturbations in the electrostatic potential of C2 and in the local electrostatic parameters of 18 charged residues in C2, which are distant from the BO2C11 epitope. Nine of the predicted electrostatic hotspots clustered on the VWF-binding site of C2. Mutation of some of the predicted electrostatic hotspots has been associated with hemophilia A and reduced VWF binding in vitro. Inhibitors may neutralize FVIII by alteration of protein surface electrostatics at a long distance from their epitope. Perturbation of the electrostatic environment of C2, either upon binding by anti-FVIII antibodies or consecutive to missense mutations in the F8 gene, may lead to hampered VWF binding and reduced FVIII residence time in circulation.

A Subset of High Titer Acquired Hemophilia Patients May Benefit from Treatment with High Dose Factor VIII.

Abstract 3476Poster Board III-413Approximately 30% of patients with severe hemophilia A will develop inhibitory antibodies (Abs) to factor VIII (fVIII inhibitors). The immune response to fVIII currently is the most significant complication in the management of patients with hemophilia A. In addition, autoimmune Abs to fVIII can develop in non-hemophiliacs, producing acquired hemophilia A, which frequently produces life- or limb-threatening bleeding. Patients with autoimmune hemophilia often have Abs with type II kinetics in which there is incomplete inactivation of fVIII at saturating concentrations of inhibitor. We have characterized the antibody response to the C2 domain of human fVIII in a murine hemophilia model and described 5 structural groups of Abs. Groups A, AB, and B are classical anti-C2 Abs that block fVIII and fVIIIa binding to phospholipid. Groups BC and C consist of non-classical anti-C2 Abs that inhibit the proteolytic activation of fVIII but do not block the binding of fVIII to phospholipid. Subsequently, we identified classical and non-classical anti-C2 Abs in human fVIII inhibitor plasmas. Most murine non-classical Abs have inhibitor titers greater than 10,000 Bethesda units/mg IgG. In a murine in vivo bleeding model, both type I classical C2 Abs, type II non-classical C2 Abs, and a type I anti-A2 Ab produced similar amounts of blood loss that were significantly greater than control mice injected with 180 U/kg of fVIII alone. Increasing the dose of fVIII to 360 U/kg overcame the bleeding diathesis produced by the type II MAbs, but not the type I Abs. These results were consistent with the in vitro Bethesda assay in which a type I anti-A2 Ab, 4A4, completely inhibited both 1 U/mL and 3 U/mL fVIII, while there was 40% residual activity at saturating concentrations of a type II anti-C2 Ab, 2-77, at either concentration of fVIII. To determine if similar in vitro characteristics exist in patients with acquired hemophilia, plasmas from 3 patients with high titer type II inhibitors were studied. All 3 plasmas primarily had C2 domain epitope specificity that included non-classical Abs. Plasma A7 additionally had detectable anti-A2 activity. Recovery of fVIII activity after a 2 h incubation at 37 °C at nominal added concentrations of 1 mL and 3 U/mL fVIII was compared (Table 1). At 3 U/mL added fVIII, recovery of activity in plasmas A4 and A5 was 1.1 U/mL and 0.51 U/mL, respectively, despite the presence of inhibitor titers of 18 and 11 Bethesda units (BU) per mL. The presence of anti-A2 Abs, which typically have type I kinetics, may have contributed to the overall lower recovery of activity in plasma A7. These results suggest that treatment with high-dose fVIII rather than bypassing agents may be warranted in patients with an inhibitor response dominated by non-classical anti-C2 Abs.Table 1Patient PlasmaInhibitor Titer (BU/mL)Recovered Activity at 1U/mL FVIII (U/mL)Recovered Activity at 3 U/mL FVIII (U/mL)A4180.311.1A5110.180.51A7620.070.12 Disclosures:No relevant conflicts of interest to declare.

Proteolytically Inactivatable Factor VIII is Less Immunogenic than Factor VIII in a Murine Hemophilia A Model.

Abstract 27Patients with hemophilia A have a congenital deficiency of the factor VIII (fVIII) protein due to a mutation in the fVIII gene that frequently leads to absence of detectable expression of fVIII. Accordingly, the therapeutic replacement fVIII protein potentially is recognized as non-self by the immune system. Thirty percent of patients with severe hemophilia A develop detectable inhibitory anti-fVIII antibodies (inhibitors). Additionally, greater than 90 percent of hemophilia A mice treated with human fVIII develop inhibitors using dosing schedule that mimics use in humans. Because fVIII is an immunologically foreign protein, it might be expected that a hemophilia A patient would make a fVIII inhibitor. However, intravenous injection of soluble proteins in either humans or rodents usually results in tolerance rather than a humoral immune response. One major difference between fVIII and other proteins is that it is released from its large carrier protein von Willebrand factor (VWF) and is potentially exposed to the immune system at sites of active hemostasis and inflammation. Heat-inactivated, denatured fVIII, which maintains all T-cell epitopes but lacks several B-cell epitopes, is less immunogenic than native fVIII, suggesting that fVIII-dependent thrombin generation along the intrinsic pathway of blood coagulation may provide co-stimulatory signals necessary for the immune response (Skupsky BS, Zhang A, Scott DW Blood 2008; 112:1220a). We constructed a B domain-deleted human fVIII mutant, designated fVIIIi, which contains alanine substitutions at two critical thrombin cleavage sites, Arg372 and Arg1689, and purified it to homogeneity. FVIIIi does not develop procoagulant activity and is not released from VWF in response to thrombin. Therefore fVIIIi is less likely than wild-type fVIII to be exposed to the immune system at sites of active hemostasis and inflammation. Additionally, VWF binds to the immunodominant fVIII C2 domain and potentially hides part of fVIII from the immune system. FVIIIi was antigenically intact judging from intact binding to a panel of11 mouse anti-fVIII monoclonal antibodies whose epitope specificity was represented by all five domains of BDD fVIII. The immunogenicity of wild-type fVIII and fVIIIi was compared in a murine hemophilia A model in which groups of 25 mice received 8 weekly injections of physiologic doses of fVIII. Plasma was collected weekly for total anti-fVIII antibody titers by ELISA and one week following the last injection for total anti-fVIII antibody titers, inhibitor titers by Bethesda assay and for epitope mapping. Mice treated with fVIIIi had significantly lower levels of inhibitory as well as total anti-fVIII antibodies than mice treated with wild-type fVIII. Domain mapping using single human domain hybrid human/porcine molecules as ELISA antigens revealed that hemophilia A mice broadly recognized all fVIII domains in response to either wild-type or fVIIIi, although fVIIIi produced less anti-light chain antibodies. Mice in both the wild-type fVIII and fVIIIi groups produced antibodies that recognized the phospholipid-binding site of the C2 domain, even though this site overlaps the VWF binding site on fVIII. There was no difference in the isotype spectrum of the antibodies made to fVIII or fVIIIi. This study indicates that inactivatable fVIII is less immunogenic than native fVIII and suggests that the immunogenicity of fVIII is related either to its interaction with VWF or to events triggered by activation of the coagulation mechanism. Disclosures:No relevant conflicts of interest to declare.

Automated Detection of Conformational Epitopes Using Phage Display Peptide Sequences

Precise determination of conformational epitopes of neutralizing antibodies represents a key step in the rational design of novel vaccines. A powerful experimental method to gain insights on the physical chemical nature of conformational epitopes is the selection of linear peptides that bind with high affinities to a monoclonal antibody of interest by phage display technology. However, the structural characterization of conformational epitopes from these mimotopes is not straightforward, and in the past the interpretation of peptide sequences from phage display experiments focused on linear sequence analysis to find a consensus sequence or common sequence motifs. We present a fully automated search method, EpiSearch that predicts the possible location of conformational epitopes on the surface of an antigen. The algorithm uses peptide sequences from phage display experiments as input, and ranks all surface exposed patches according to the frequency distribution of similar residues in the peptides and in the patch. We have tested the performance of the EpiSearch algorithm for six experimental data sets of phage display experiments, the human epidermal growth factor receptor-2 (HER-2/neu), the antibody mAb Bo2C11 targeting the C(2) domain of FVIII, antibodies mAb 17b and mAb b12 of the HIV envelope protein gp120, mAb 13b5 targeting HIV-1 capsid protein and 80R of the SARS coronavirus spike protein. In all these examples the conformational epitopes as determined by the X-ray crystal structures of the antibody-antigen complexes, were found within the highest scoring patches of EpiSearch, covering in most cases more than 50% residues of experimental observed conformational epitopes. Input options of the program include mapping of a single peptide or a set of peptides on the antigen structure, and the results of the calculation can be visualized on our interactive web server. Users can access the EpiSearch from our web server http://curie.utmb.edu/episearch.html.

Induction of ‘Tolerance’ in a T-Cell Clone from a Mild Hemophilia Patient

A major obstacle in the treatment of Hemophilia A is that patients can develop an inhibitory immune response to therapeutic doses of coagulation factor VIII (fVIII). Over the last decade, we have developed a B-cell delivered gene therapy approach to prevent the development of inhibitory antibodies (“inhibitors”) in fVIII knockout mice (see Lei and Scott, Blood 105: 4865, 2005). In our murine platform, activated primary spleen B cells or bone marrow cells are transduced with a retroviral vector encoding the fVIII A2 and/or C2 domain fused to an IgG heavy chain, and these cells are injected systemically into immunocompetent fVIII knockout animals. The recipients are rendered specifically tolerant to the encoded C2 and A2 domains, as evidenced by a >90% reduction of inhibitor titers, even in primed animals. To help evaluate the potential of this approach for translation, we are developing in vitro models for tolerance induction using human T-cell clones isolated from subjects with mild hemophilia A. The clones are isolated by single-cell sorting of CD4+ cells that are labeled by fluorescent HLA-DR tetramers complexed with peptides containing fVIII epitopes, followed by expansion with HLA-DR mismatched peripheral blood mononuclear cells (PBMC), phytohaemagglutinin, and interleukin-2. Our initial model utilizes a T-cell clone from an individual with mild hemophilia A due to fVIII missense genotype A2201P, which recognizes an HLA-DRA-DRB1*0101-restricted epitope within a synthetic peptide corresponding to fVIII residues 2194–2213. All of the antigen-specific T-cell clones isolated from this subject secreted interferon-gamma (IFN-γ) when stimulated by fVIII2194–2213 presented by irradiated HLA-DR-matched PBMCs or with plate-bound anti-CD3. Because of their robust response to a clinically relevant epitope in fVIII, one of these clones that expanded well in culture was chosen for initial testing of a modified gene therapy platform similar to that developed using the murine hemophilia A model. HLA-matched peripheral blood B cells were activated with antibodies to IgM or with CD40L-expressing fibroblasts and then transduced with a modified retroviral vector containing the human C2 domain sequence in-frame with the IgG sequence. These B cells were cultured with the hemophilic T-cell clone. After pre-treatment (“tolerance-induction step”), the cells were washed and then stimulated by plate-bound anti-CD3. The subsequent IFN-γ response (measured by ELIspots and ELISA) was dramatically reduced compared to the response of same T-cell clone cultured with mock-transduced B cells. The post-treatment reduction in IFN-γ secretion was equivalent to that induced after soluble anti-CD3 pre-treatment, a known method to induce T-cell anergy in vitro. Interestingly, IL-10 was produced during the tolerance induction (pre-treatment) phase, most likely from the activated B cells. Preliminary, parallel experiments with B cells transduced with a “gutless” adenovirus vector expressing C2-Ig did not result in a similar down-regulation of the T-cell response, suggesting that this non-integrating method of expressing antigens for tolerance is not effective, at least in this system. These results are the first to demonstrate in vitro modulation of cytokine responses using DR-restricted, fVIII-specific T cells from a hemophilia A subject. Further investigations using T-cell clones from hemophilic subjects with and without anti-fVIII antibodies will allow us to explore mechanisms of tolerance and may also suggest novel approaches to reduce inhibitor titers.

Nonclassical anti-C2 domain antibodies are present in patients with factor VIII inhibitors

AbstractThe antihuman factor VIII (fVIII) C2 domain immune response in hemophilia A mice consists of antibodies that can be divided into 5 groups of structural epitopes and 2 groups of functional epitopes. Groups A, AB, and B consist of classical C2 antibodies that inhibit the binding of fVIII to phospholipid and von Willebrand factor. Groups BC and C contain nonclassical C2 antibodies that block the activation of fVIII by thrombin or factor Xa. Group BC antibodies are the most common and display high specific inhibitory activity and type II kinetics. The C2 epitope groups recognized by 26 polyclonal human anti-fVIII inhibitor plasmas were identified by a novel competition enzyme-linked immunosorbent assay using group-specific murine monoclonal antibodies. Most of the anti-C2 inhibitor plasmas inhibited the binding of both classical and nonclassical antibodies. These results suggest that nonclassical anti-C2 antibodies contribute significantly to the pathogenicity of fVIII inhibitors.

The factor VIII C1 domain contributes to platelet binding

Activated factor VIII (FVIIIa) forms a procoagulant complex with factor IXa on negatively charged membranes, including activated platelet surfaces. Membrane attachment involves the FVIII C2 domain; involvement of the adjacent C1 domain has not been established. Binding of recombinant FVIII C1C2 and C2 proteins to platelets was detected by flow cytometry using (1) anti-C2 monoclonal antibody ESH8 followed by a phycoerythrin-labeled secondary antibody; (2) biotinylated C1C2 detected by phycoerythrin-labeled streptavidin, and (3) C1C2 and C2 site-specifically labeled with fluorescein. Highest binding and lowest background were obtained using fluorescein-conjugated proteins. More than 90% of activated platelets bound C1C2, compared with approximately 50% for equimolar C2. Estimates using fluorescent microbeads indicated approximately 7,000 C1C2-binding sites per platelet, approximately 1,400 for C2, and approximately 3,000 for fluorescein-labeled FVIIIa. Unlike C2 or FVIII(a), C1C2 bound to approximately 700 sites/platelet before activation. C1C2 binding to activated platelets appeared independent of von Willebrand factor and was competed effectively by FVIII(a), but only partially by excess C2. Fluorescein-labeled FVIIIa was competed much more effectively by C1C2 than C2 for binding to activated platelets. Two monoclonal antibodies that inhibit C2 binding to membranes competed platelet binding of C2 more effectively than C1C2. Thus, the C1 domain of FVIII contributes to platelet-binding affinity.

Antihuman factor VIII C2 domain antibodies in hemophilia A mice recognize a functionally complex continuous spectrum of epitopes dominated by inhibitors of factor VIII activation

The diversity of factor VIII (fVIII) C2 domain antibody epitopes was investigated by competition enzyme-linked immunosorbent assay (ELISA) using a panel of 56 antibodies. The overlap patterns produced 5 groups of monoclonal antibodies (MAbs), designated A, AB, B, BC, and C, and yielded a set of 18 distinct epitopes. Group-specific loss of antigenicity was associated with mutations at the Met2199/Phe2200 phospholipid binding beta-hairpin (group AB MAbs) and at Lys2227 (group BC MAbs), which allowed orientation of the epitope structure as a continuum that covers one face of the C2 beta-sandwich. MAbs from groups A, AB, and B inhibit the binding of fVIIIa to phospholipid membranes. Group BC was the most common group and displayed the highest specific fVIII inhibitor activities. MAbs in this group are type II inhibitors that inhibit the activation of fVIII by either thrombin or factor Xa and poorly inhibit the binding of fVIII to phospholipid membranes or von Willebrand factor (VWF). Group BC MAbs are epitopically and mechanistically distinct from the extensively studied group C MAb, ESH8. These results reveal the structural and functional complexity of the anti-C2 domain antibody response and indicate that interference with fVIII activation is a major attribute of the inhibitor landscape.

Non-Classical Anti-Factor VIII C2 Domain Inhibitors Are Present in the Plasmas of Most Factor VIII Inhibitor Patients.

Approximately 30% of patients with severe hemophilia A will develop inhibitory antibodies to factor VIII (fVIII inhibitors). The immune response to fVIII currently is the most significant complication in the management of patients with hemophilia A. In addition, autoimmune antibodies to fVIII can develop in non-hemophiliacs, producing acquired hemophilia A, which frequently produces life- or limb-threatening bleeding. These inhibitors primarily are directed against the A2 or C2 domains of fVIII. The human response to the C2 domain of fVIII classically has been thought to inhibit fVIII activity by blocking its binding to phospholipid. We recently characterized the antibody response to the C2 domain of human fVIII in a murine hemophilia model and described 5 structural groups of antibodies. Groups A, AB, and B are classical anti-C2 antibodies. Groups BC and C consist of non-classical anti-C2 antibodies that inhibit the proteolytic activation of fVIII but do not block the binding of fVIII to phospholipid. Most non-classical antibodies have inhibitor titers greater than 10,000 Bethesda units/mg IgG. To determine if non-classical antibodies are present in fVIII inhibitor patients, patient plasmas were tested in an ELISA for their ability to block the binding of representative antibodies from the different anti-human fVIII C2 antibody groups. Classical and non-classical monoclonal antibodies (MAbs) were biotinylated and serially diluted into either fVIII deficient plasma or patient inhibitor plasma and then added to microtiter wells coated with fVIII. The ability of patient plasma to block the binding of the murine MAbs to fVIII was determined. A total of 16 patient plasmas were assessed: 4 from patients with a C2 predominant response, 2 with a non-C2 predominant response, and 10 with unknown specificities. Three of the 4 patients with C2 predominant responses had non-classical anti-C2 antibodies, while the 2 with non-C2 predominant responses did not. In the unknown plasmas, 6 of 10 had evidence of non-classical antibodies. Figure 1 shows representative results of the effect of 3 patient plasmas on the binding of a biotinylated non-classical MAb to fVIII. Patient plasmas 1 and 2 blocked MAb binding while patient plasma 3 did not. This study indicates that the majority of patients with fVIII inhibitors have non-classical anti-C2 antibodies in their response to fVIII. [Display omitted]