FVIII A2 Domain Research Articles

Hybrid human-porcine factor VIII proteins partially escape the inhibitory effects of anti-factor VIII inhibitor alloantibodies having A2 or C2 domain specificity.

Porcine factor (pF)VIII has low cross-reactivity with anti-human (h)FVIII inhibitor alloantibodies. Clinical trials of pFVIII in congenital haemophilia A patients with inhibitor (PwHA-I) are in progress. Most polyclonal anti-hFVIII inhibitors recognize its A2 and/or C2 domain(s), and recombinant human-porcine hybrid (hp)FVIII proteins may escape neutralization by these inhibitors. To evaluate the ability of hpFVIII to limit the anti-FVIII activity of inhibitor alloantibodies. Three hybrid proteins were created by substituting the hFVIII A2, C2 domain or both with the corresponding domains of pFVIII [termed hp(A2), hp(C2) and hp(A2/C2), respectively]. The reactivity of these hybrids was assessed by one-stage clotting assays (OSA), thrombin generation assays (TGA) and rotational thromboelastometry (ROTEM) by adding them to FVIII-deficient samples. OSA demonstrated that the hybrid proteins avoided neutralization by anti-FVIII A2 or C2 monoclonal antibodies (mAb) and polyclonal inhibitor-antibodies (polyAb) from PwHA-I. In TGA, thrombin generation with hp(A2) and hp(A2/C2) was not attenuated in the presence of patient IgG recognizing anti-A2 domain. In contrast, that with hFVIII and hp(C2) was suppressed by this IgG to levels equivalent to those of FVIII-deficient plasma. With anti-A2/C2 polyAb, the activity of hp(A2/C2) was unaffected. ROTEM demonstrated that the addition of hp(A2) or hp(A2/C2) to anti-A2 polyAb shortened clot times/clot formation times, whilst hFVIII or hp(C2) were ineffective. Similarly with anti-A2/C2 polyAb, hp(A2/C2) restored coagulation potential to a greater extent than hp(A2) and hp(C2). Hybrid FVIII proteins containing porcine FVIII A2 and/or C2 domain(s) could support effective therapy in PwHA-I by avoiding neutralization.

Contribution of the Residues 488-490 on the Factor (F)VIII A2 Domain and the Residue 2239 on the FVIII C2 Domain for the Interaction between FVIII and Protein S

Factor (F)VIII functions as a cofactor in the Xase complex responsible for phospholipid (PL) surface-dependent conversion of FX to activated FX (FXa) by FIXa. Protein S (PS), a vitamin K-dependent anticoagulant protein, functions as a cofactor of activated protein C (APC) that inactivates FVIIIa and FVa. We have previously reported a new regulatory mechanism on coagulation that PS directly impaired the Xase complex by competing the FVIIIa-FIXa interaction (Takeyama, Br J Haematol. 2008), and identified at least two PS-interactive sites that were the 488-490 residues on FVIII A2 domain (Takeyama, Thromb Haemost. 2009) and the 2239 residue on FVIII C2 domain (Furukawa, ASH 2016). In the present study, to investigate the contribution of binding sites on FVIII A2 and C2 domains for the interaction between FVIII and PS, three recombinant B-domainless FVIII mutants substituted to alanine were expressed by BHK cell, such as the mutant of the 488-490 residues (S488A/R489A/R490A), the mutant of the 2239 residue (K2239A), and the mutant of both residues (S488A/R489A/R490A/K2239A), subsequently compared interaction with PS of the three mutants with wild-type (wt-) FVIII. First, we performed surface plasmon resonance-based assays to compare the direct binding of each FVIII mutant to PS. When PS was immobilized on the sensor chip and bound to the wt-FVIII and each FVIII mutant as an analyte, the K d value of wt-FVIII was 3.5 nM, while those of S488A/R489A/R490A, K2239A, and S488A/R489A/R490A/K2239A mutants were 43.7 nM, 6.7 nM, and 44.7 nM, respectively. These results demonstrated that the residues 488-490 and 2239 contributed to PS-interactive sites as described in the previous reports. Next, the effect of PS on the directly reduction of FVIIIa function was evaluated using the FXa generation assay with the three FVIII mutants and compared with wt-FVIII. After each FVIII mutant (1 nM) was reacted with PS (0-500 nM) in the presence of PL vesicles (20 µM) for 30 min, FVIII was activated by thrombin (30 nM), followed by the reaction with FIXa (40 nM) and FX (300 nM) for 1 min. Generated FXa in wt-FVIII was reduced in the presence of PS dose-dependently and was reduced to 7.7% at 125 nM of PS, compared to that in the absence of PS. However, generated FXa in K2239A, S488A/R489A/R490A, and S488A/R489A/R490A/K2239A were mildly reduced to 23.1 %, 23.6 %, and 29.2 % even at 500 nM of PS, respectively, compared to those in the absence of PS. Finally, to investigate the contribution of these FVIII and PS binding sites when PS plays as a cofactor of APC in FVIIIa inactivation, FXa generation assays were measured in each FVIII mutant in the presence of PS (150 nM) and APC (0-10 nM). In the presence of PS, generated FXa in wt-FVIII and each FVIII mutant were decreased in APC dose-dependently. At the 10 nM of APC, generated FXa in the wt-FVIII and K2239A were reduced to 48.1 % and 40.6 %, respectively, in comparison to the absence of APC, while generated FXa in S488A/R489A/R490A and S488A/R489A/R490A/K2239A mutants were mildly reduced to 65.8 % and 69.0 %, respectively. These findings suggested that the contribution of the 488-490 residues on the FVIII A2 domain was particularly great for the FVIII-PS interaction between the two PS binding sites identified on the FVIII A2 and C2 domains. In addition, the coexistence of the K2239A mutation on the FVIII C2 domain with S488A/R489A/R490A mutation further reduced the direct FVIII-PS interaction, suggesting that K2239A mutation further reduced the inhibitory effect of FVIII function by PS. On the other hand, mainly not the 2239 residue on the FVIII C2 domain but the 488-490 residues on the FVIII A2 domain were thought to affect the PS function as a cofactor in FVIIIa inactivation by APC. Overall, our data indicated that the 488-490 residues on the FVIII A2 domain were key residues for the FVIII-PS interaction including FVIIIa inactivation by APC, and the 2239 residue affected partially for only the FVIII-PS direct interaction.

Contact-dependent suppression of FVIII-specific memory B cells by BAR-Tregs

Abstract Anti-drug antibody formation poses tremendous obstacles for optimal treatment of hemophilia A (HA). In this study, we sought to utilize chimeric receptor-modified natural regulatory T cells (Tregs) to target FVIII-specific memory B cells, which are responsible for persistent anti-FVIII neutralizing antibodies (inhibitors) in HA patients. Thus, CD4+CD25hiCD304+ natural Tregs were FACS sorted from naïve C57BL/6 mice and retrovirally transduced to express a chimeric B-cell antibody receptor (BAR) containing the immunodominant A2 domain of FVIII. Plasmablast-depleted (CD138neg) splenocytes from FVIII/IFA immunized FVIII-knockout HA mice served as the source for FVIII-specific memory B cells, which were then specifically stimulated in vitro with FVIII and enumerated in B-cell ELISPOT assays. Adding A2-BAR Tregs (1 per 150 splenocytes), but not conventional T cells, to the CD138− splenocytes significantly suppressed the formation of anti-FVIII producing cells, compared to the non-relevant OVA-BAR Tregs control group. Transwell experiments confirmed that the suppression was contact-dependent. Interestingly, even in the presence of antibody to FVIII (so-called inhibitors), similarly prepared CD4+CD25hiCD127low A2-BAR human natural Tregs completely suppressed polyclonal anti-FVIII Elispots formation, indicating potential by-stander suppression by A2-BAR Tregs on all FVIII-specific B cells in the local milieu. In conclusion, we demonstrated in vitro that FVIII domain-expressing BAR Tregs could efficiently target and suppress FVIII-specific memory B cells. Supported by a Bayer Hemophilia Grant [AHZ] and NIH HL126727 [DWS].

Suppression of FVIII-Specific Memory B Cells by Chimeric BAR Receptor-Engineered Natural Regulatory T Cells.

Anti-drug antibody formation poses tremendous obstacles for optimal treatment of hemophilia A (HA). In this study, we sought to utilize chimeric receptor-modified natural regulatory T cells (Tregs) to target FVIII-specific memory B cells, which are responsible for persistent anti-FVIII neutralizing antibodies (inhibitors) in HA patients. Thus, CD4+CD25hiCD304+ natural Tregs were FACS sorted from naïve C57BL/6 mice and retrovirally transduced to express a chimeric B-cell antibody receptor (BAR) containing the immunodominant A2 domain of FVIII. Plasmablast-depleted (CD138neg) splenocytes from FVIII immunized FVIII-knockout HA mice served as the source for FVIII-specific memory B cells, which were specifically stimulated in vitro with FVIII and enumerated in a B-cell ELISPOT assays. Adding A2-BAR Tregs (1 per 150 splenocytes), but not conventional T cells, to the CD138– splenocytes significantly suppressed the formation of anti-FVIII antibody secreting cells (ASC), compared to the non-relevant OVA-BAR Tregs control group. The observation that A2-BAR Tregs can suppress the response to FVIII suggests that bystander suppression can occur in the local milieu in this system. Transwell experiments confirmed that the suppression was contact-dependent. Moreover, even in the presence of antibodies to FVIII (so-called inhibitors), similarly prepared CD4+CD25hiCD127low A2-BAR human natural Tregs completely suppressed polyclonal anti-FVIII ASC formation. In conclusion, we demonstrated in vitro that FVIII domain-expressing BAR Tregs could efficiently target and suppress FVIII-specific memory B cells.

Prevention and reversal of adverse humoral immune response by BAR-transduced regulatory T cells

Abstract Antigen-specific elimination of unwanted immune response, such as anti-FVIII neutralizing antibody formation in hemophilia A patients, is highly desirable but remains a challenge. Analogous to the chimeric antigen receptor (CAR) strategy successfully used in cancer immunotherapy, we have created a chimeric receptor comprising a protein antigen or its domain, linked with the transmembrane and signal transduction domains, CD28-CD3ζ. We termed it “BAR” for B-cell-targeting antibody receptor. CD4+CD25hiCD127low human Tregs were retrovirally transduced to express a BAR containing the immunodominant FVIII C2 (C2 BAR) or A2 domain (A2 BAR). Control human Tregs were transduced to express a BAR containing chicken ovalbumin (OVA BAR). FACS sorted, BAR transduced human CD4+ Tregs have been successfully expanded in vitro, and shown to maintain phenotypic Treg markers in terms of co-expression of Foxp3 and Helios. Pretreatment of naïve hemophilia A mice with a mixture of C2 BAR plus A2 BAR human Tregs completely prevented anti-FVIII antibody development in response to FVIII/IFA immunization for as long as 8 weeks. More importantly, BAR human Tregs also effectively suppressed anti-FVIII antibody formation in primed animals with detectible pre-existing anti-FVIII antibodies. Further mechanistic studies suggest that the FVIII-specific B cell response were directly suppressed while the T cell response remained intact. Taking together, we report here a successful proof of principle strategy utilizing FVIII-specific BAR Tregs to directly target FVIII-specific B cells, an approach which could be adapted to address other adverse immune responses as well.

Identification of a Factor X-Interactive Site on the Factor VIII A2 Domain

We have experienced a case of acquired hemophilia A with inhibitor recognizing only a factor (F) VIII A2 epitope, and reported the inhibitory mechanism for disappearing FVIII activity (Blood, 124, 4226, 2014). In summary, the patient’s inhibitor IgG bound to FVIII A2N (residue 372-562) fragment and inhibited Arg372 cleavage in FVIII by FXa, suggesting that FX(a) bound to FVIII A2 domain. ELISA-based assay showed that FVIII A2 fragment bound to FX (Kd; 338 nM). We hypothesized that FVIII A2 residues 400-429 might be FX binding site according to the 3-D model of FVIII molecule, and prepared synthetic peptides (400-409, 409-419, and 420-429). The 400-409 peptide inhibited the FVIII A2-FX interaction, suggesting that the 400-409 region contributed to FX-interactive site. In this current study, we further performed the localization of a FX-interactive site on the 400-409 region in the A2 domain. A purified FXa generation assay demonstrated the 400-409 peptide decreased the generation of FXa in a dose-dependent manner up to 38% of 100 μM (Ki; 23 ± 9 nM). In comparison, scrambled peptide of 400-409 decreased up to 10% of 100 μM. These data demonstrated that the 400-409 peptide inhibited the generated FXa, suggesting the 400-409 region contributed to regulate the coagulation function. Covalent cross-linking was observed between the biotinylated 400-409 peptide and FX following reaction with EDC (1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide) using SDS-PAGE. This cross-linking formation was blocked by the addition of unlabeled 400-409 peptide. N-terminal sequence analysis of the peptide-FX product demonstrated that two sequential residues (Lys408 and Ser409) could not be detected, supporting that two residues participate in cross-link formation. To confirm the significance of these residues in A2 domain for FX-binding, the mutant forms of the A2 domain, converted to alanine, were expressed in BHK system and purified. Compared with wild type FVIII (Kd; 10 ± 3 nM), the binding affinity of Ser409Ala FVIII mutant for FX was no significant difference (Kd; 14 ± 1 nM) on SPR-based assay. Lys408Ala or Lys408Ala/Ser409Ala double FVIII mutant, however, decreased the binding affinity by 3.6~4.3-fold (Kd; 36 ± 7 or 43 ± 2 nM, respectively), suggesting contribution of Lys408Ala to the binding interaction. For the functional evaluation of the association with FVIII mutants to FX, a FXa generation assay was repeated. Lys408Ala, Ser409Ala, or Lys408Ala/Ser409Ala FVIII mutant reacted with varying concentrations of FX decreased by 1.2~1.6-fold (Km; 53 ± 12, 69 ± 15, or 65 ± 15 nM, respectively) compared to wild type FVIII (Km; 43 ± 9 nM), supporting a contribution of these mutants to Km and overall catalytic efficiency. Vmax values were largely unaffected by the mutations with most values within approximately 30% of the wild-type value. On the other hand, Kcat/Km value of Lys408Ala, Ser409Ala, or Lys408Ala/Ser409Ala FVIII mutant were decreased by 0.5~0.7-fold (Kcat/Km; 1.0, 1.3, or 0.9 nM-1min-1, respectively) compared to wild type FVIII (Kcat/Km; 1.8 nM-1min-1), suggesting low catalytic efficacy of Lys408Ala and Ser409Ala. These results indicate that the 400-409 region in the FVIII A2 domain, and in particular Lys408 and Ser409, may contribute to a unique FX-interactive site. DisclosuresNogami:Chugai Pharmaceutical Co., Ltd.: Honoraria, Membership on an entity’s Board of Directors or advisory committees, Patents & Royalties, Research Funding; F. Hoffmann-La Roche Ltd.: Honoraria, Membership on an entity’s Board of Directors or advisory committees; Sysmex Corporation: Patents & Royalties, Research Funding. Shima:Sysmex Corporation: Patents & Royalties, Research Funding; Chugai Pharmaceutical Co., Ltd.: Consultancy, Honoraria, Membership on an entity’s Board of Directors or advisory committees, Patents & Royalties, Research Funding; F. Hoffmann-La Roche Ltd.: Consultancy, Honoraria, Membership on an entity’s Board of Directors or advisory committees.

Identification of a Protein S-Interactive Site on the Factor VIII C2 Domain

Factor (F)VIII functions as a cofactor in the tenase complex responsible for phospholipid (PL) surface-dependent conversion of FX to FXa by FIXa. On the other hand, protein S (PS) functions as a cofactor of activated protein C that inactivates FVIII(a) and FV(a). We have reported a new regulatory mechanism on coagulation that PS directly impaired the FXase complex by competing the FIXa-FVIIIa interaction (Takeyama, Br J Haematol. 2008;143:409), and identified the PS-interactive site on the FVIII A2 domain (Takeyama, Thromb Haemost. 2009;102:645). However, the contribution of FVIII light chain (LC) to PS-binding has not been determined. In this study, several approaches were employed to assess a PS-FVIII LC interaction. The binding of FVIII LC to active site-modified FIXa (EGR-FIXa) was inhibited by PS dose-dependently (Ki; 4.6 nM) on ELISA. Because FVIII C2 domain has a binding site for FIXa, we examined whether FVIII C2 domain bound to PS. PS bound to FVIII C2 domain (Kdapp; 283 nM) by the ELISA. A SPR-based assay also revealed that FVIII C2 domain bound to PS (Kd; 62 nM). We have reported that a FIXa-interactive site exits on FVIII C2 domain (residues 2228-2240) (Soeda, J Biol Chem. 2009;284:3379). Therefore, we assessed the interaction of the synthetic peptide spanning the residues (2228-2240) with PS. ELISA showed that the peptide bound to PS (Kdapp; 104 µM). SPR-based assay also revealed that the peptide bound to PS (Kd; 31.4 µM), although scrambled peptide failed to bind to PS. Covalent cross-linking was observed between the biotynated 2228-2240 peptide and PS following reaction with EDC (1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride) using SDS-PAGE. This cross-linking formation was blocked by the addition of the unlabeled peptide. Furthermore, N-terminal sequence analysis of the peptide-PS product showed that one basic residue (K2239) could not be detected, supporting that this residue participates in cross-link formation. These results indicate that the 2228-2240 residues in the FVIII C2 domain, in particular K2239, may contribute to a PS-interactive site. To inquire further into the interactive region of FVIII C2 domain to PS, we prepared a stably expressed recombinant B-domainless FVIII mutants (E2228A, D2233A, K2236A, and K2239A), which were polar amino acids in the 2228-2240 residues, subsequently compared interaction with PS of the mutants with B-domainless FVIII wild type. To evaluate the binding affinity to PS, we performed SPR-based assay with the wild type FVIII and the FVIII mutants. The assay showed K2239A bound to PS with ~1.9-fold higher Kd value (28.1 nM) than that of wild type (Kd; 14.7 nM), although E2228A, D2233A, or K2236A bound to PS (Kd; 12.7, 16.3, and 10.3 nM, respectively) with no significant difference compared with wild type. In addition, to evaluate the effect of mutants for FXa generation, we assessed FXa generation assay with wild type or K2239A in the presence or absence PS, because K2239 might contribute to a PS interaction of FVIII. After each FVIII (1 nM) was reacted with various concentrations of PS in the presence of PL (20 µM) for 30 minutes, FVIII was activated by thrombin (30 nM), followed by the reaction with FIXa (40 nM) and FX (300 nM) for 1 minute. Although generated FXa with both wild type and K2239A were reduced in the presence of PS dose-dependently, compared with its absence, the inhibition ratio of K2239A was less than that of wild type (18.2% and 31%, respectively in the presence of 500 nM PS). On the other hand, the Km value on FXa generation of K2239A for various concentrations of FIXa was ~1.8-fold higher than that of wild type (5.5±0.9 and 3.1±0.2 nM, respectively), suggesting that K2239 residue, involved in the FIXa-interactive site, might contribute to the inhibition of FVIII function by PS. In conclusion, FVIII C2 domain, in particular K2239, was possible to play an important role of the inhibitory mechanism to FVIII function by PS, due to the binding to PS. DisclosuresShima:Sysmex Corporation: Patents & Royalties, Research Funding; F. Hoffmann-La Roche Ltd.: Consultancy, Honoraria, Membership on an entity’s Board of Directors or advisory committees; Chugai Pharmaceutical Co., Ltd.: Consultancy, Honoraria, Membership on an entity’s Board of Directors or advisory committees, Patents & Royalties, Research Funding. Nogami:Sysmex Corporation: Patents & Royalties, Research Funding; F. Hoffmann-La Roche Ltd.: Honoraria, Membership on an entity’s Board of Directors or advisory committees; Chugai Pharmaceutical Co., Ltd.: Honoraria, Membership on an entity’s Board of Directors or advisory committees, Patents & Royalties, Research Funding.

Targeting FVIII-Specific B Cells Using BAR-Transduced Regulatory T Cells

Up to a third of hemophilia A (HA) patients receiving therapeutic FVIII develop neutralizing antibodies termed "inhibitors". Once inhibitors develop, clinical management of HA patients becomes extremely difficult. Thus, a rational solution would be to establish FVIII-specific immune tolerance to FVIII in high risk patients. To address this issue in a mouse model of human HA, we elected to use an antigen-specific regulatory T cell (Treg) approach. Analogous to the chimeric antigen receptor (CAR) strategy successfully used in cancer immunotherapy, we have created a chimeric receptor comprising a protein antigen or its domain, linked with the transmembrane and signal transduction domains, CD28-CD3ζ. We termed this receptor "BAR" for B-cell-targeting antibody receptor. Human Tregs (CD4+CD25hiCD127low) were retrovirally transduced with a BAR containing FVIII C2 domain (C2-BAR) or FVIII A2 domain (A2-BAR) and expanded successfully in vitro. These cells stained positively with anti-C2 and anti-A2 monoclonal antibodies, respectively, and maintained Treg phenotypic markers in terms of co-expression of Foxp3 and Helios. Control human Tregs were transduced with a BAR containing chicken ovalbumin (OVA-BAR). To test the hypothesis that BAR-transduced Tregs could directly and effectively suppress the activity of specific B cells, a xenogeneic model was employed. On day 0, FVIII-/- HA mice were injected intravenously with 106 transduced human Tregs. The mice were then immunized subcutaneously on day 1 with FVIII in incomplete Freund’s adjuvant, and anti-FVIII antibody development was followed. By two weeks after immunization, anti-FVIII antibodies could be detected in the control mice (n = 4). However, in the experimental group (n = 5) that received a mixture of equal number of C2-and A2-BAR Tregs, anti-FVIII antibody development was reproducibly completely blocked for at least 8 weeks. To examine the possible mechanism of BAR Treg suppression, purified B cells and T cells from "tolerized" (A2+C2-BAR) or "control" (OVA-BAR) recipients were mixed and tested for recall responses to FVIII in vitro. The results suggested that the FVIII-specific B cells were directly tolerized while the T-cell response remained intact. Taken together, we report here a successful approach utilizing FVIII-specific BAR-Tregs to directly target FVIII-specific B cells, an approach which could be adapted to address other adverse immune response as well. (Supported in part by a NIH grant HL127495) DisclosuresNo relevant conflicts of interest to declare.

A subset of high-titer anti–factor VIII A2 domain antibodies is responsive to treatment with factor VIII

AbstractThe primary B-cell epitopes of factor VIII (fVIII) are in the A2 and C2 domains. Within the C2 domain, antibody epitope and kinetics are more important than inhibitor titer in predicting pathogenicity in a murine bleeding model. To investigate this within the A2 domain, the pathogenicity of a diverse panel of antihuman fVIII A2 domain monoclonal antibodies (MAbs) was tested in the murine model. MAbs were injected into hemophilia A mice, followed by injection of human B domain-deleted fVIII. Blood loss after a 4-mm tail snip was measured. The following anti-A2 MAbs were tested: high-titer type 1 inhibitors 4A4, 2-76, and 1D4; 2-54, a high-titer type 2 inhibitor; B94, a type 2 inhibitor; and noninhibitory MAbs GMA-012, 4C7, and B25. All high-titer type 1 MAbs produced blood loss that was significantly greater than control mice, whereas all non-inhibitory MAbs produced blood loss that was similar to control. The type 2 MAbs were not pathogenic despite 2-54 having an inhibitor titer of 34 000 BU/mg immunoglobulin G. In addition, a patient with a high-titer type 2 anti-A2 inhibitor who is responsive to fVIII is reported. The discrepancy between inhibitor titer and bleeding phenotype combined with similar findings in the C2 domain stress the importance of inhibitor properties not detected in the standard Bethesda assay in predicting response to fVIII therapy.

Factor VIII A2 Domain Contains a Binding Site for Factor X

Acquired hemophilia A (AHA) is a rare hemorrhagic disease in which autoantibodies against coagulation factor (F) VIII impair the coagulation system. The inhibitors developed in AHA are polyclonal autoantibodies and the majority of FVIII inhibitors bind to the A2, A3, or C2 domains. Depending on the location of the epitope, different mechanisms of action for the anti-FVIII antibodies have been reported. Anti-A3 antibodies neutralize the procoagulant activity of FVIII by preventing its interaction with FIXa. Anti-C2 antibodies inhibit the binding of FVIII to phospholipid membrane and/or von Willebrand factor, whereas A2 and A3 inhibitors block the binding of FVIII to FIXa and FX, respectively, and obstruct the formation of the Xase complex. We have a case of AHA whose inhibitor recognizes only A2 domain and attempted several approaches to determine the mechanism of neutralizing FVIII. Thrombin and plasmin generation assay using patient’s plasma showed that the thrombin and plasmin generation in this AHA patient were decreased compared with that in congenital severe hemophilia A patient. Furthermore, FX generation (Coatest) in this AHA was also decreased compared with that in congenital severe hemophilia A patient (p<0.05). These results indicated that this inhibitor impaired the generation of Xase complex and might cause the severe bleeding disorder in this patient. The IgG subclass of inhibitor in our case was IgG1 and IgG4. Western blotting analysis using FVIIIa revealed that the inhibitor IgG recognized only A2 domain. Furthermore, western blotting analysis using FVIII A2 fragment, digested by activated protein C, showed that the inhibitor IgG bound to FVIII A2N (residue 372-562) fragment. It is known that FVIII A2 domain contains FIXa and thrombin binding sites. Western blotting analysis revealed that the inhibitor IgG inhibited Arg336 cleavage in FVIIIa by FIXa and Arg372 cleavage in FVIII by thrombin. However, the FXa-catalyzed cleavage at Arg372 in FVIII was inhibited by this inhibitor IgG. ELISA-based assay showed that the inhibitor IgG inhibited FX binding to FVIII A2. These results suggest that FX(a) binds to FVIII A2 domain. Therefore, to determine the direct binding of FX and FVIII A2 domain, ELISA-based assay was employed to assess this interaction. ELISA-based assay showed that FVIII A2 fragment bound FX in a dose-dependent manner with moderate affinity (Kd = 338 nM). FX inhibited FVIII A2 fragment binding to immobilized FX up to 70% with an inhibition constant (Ki = 254 nM) similar to the affinity constant. It is known that the residue 484-509 in the A2 domain interacts with FIXa. We hypothesized that FX binding site in the A2 domain might be in the opposite side of FIXa binding site in the A2 domain. According to the 3-D model of FVIII molecule, we prepared synthetic peptides corresponding to FVIII A2 residues 400-409, 409-419, and 420-429. To determine the specificity of these sequences for FX interaction, we examined the effects of these peptides on FVIII A2 binding to FX using ELISA-based assay. The 400-409 peptide inhibited the A2 and FX interaction up to 70%. In contrast, the 410-419 and the 420-429 peptides inhibited the interaction up to 30%. Covalent cross-linking was observed between the 400-409 peptide and FX following reaction with EDC using SDS-PAGE. These results indicate that FVIII A2 domain contains the binding site for FX(a), and the 400-409 region in the FVIII A2 domain contributes to a unique FX(a)-interactive site. DisclosuresNo relevant conflicts of interest to declare.

Factor VIII genotype characterization of haemophilia A affected patients with transient and permanent inhibitors: a comprehensive Argentine study of inhibitor risks

Inhibitor development against exogenous factor VIII is a severe impairment of replacement therapy affecting 18% of Argentine patients with severe haemophilia A (HA). To study the molecular predisposition for inhibitor development, we genotyped 260 HA patients with and without inhibitors, countrywide. The inhibitor-positive population (19 transients, 15 low responders, LR and 70 high responders, HR) of 104 severe-HA patients showed 59 Inv22 (intron 22 inversions), 18 small ins/del-frameshifts, 12 gross deletions, 12 nonsense, one splicing defect and two missense, p.Arg531Pro and p.Leu575Pro, both LR and thought to impair FVIII A2 domain secondary structure. In addition, a patient with mild HA and HR showed the missense p.Glu1704Lys associated with two neutral intronic substitutions potentially affecting the A3 domain. A case/control study (84/143) permitted estimation of F8 genotype-specific inhibitor risks [OR; prevalence (CI)] in severe-HA patients classifying a high-risk group including multi-exon deletions [3.66; 55% (19-100)], Inv22 [1.8; 24% (19-100)] and nonsense in FVIII-LCh [1.2; 21% (7-59)]; an average risk group including single-exon deletions, indel frameshifts and nonsense-HCh; and a low-risk group represented by missense defects [0.14; 3% (0.6-11)]. Analysis of inhibitor concordance/discordance in related patients indicated additional genetic factors other than F8 genotype for inhibitor formation. No significant inhibitor-predisposing factors related to FVIII product exposure were found in age- and F8 genotype-stratified populations of severe-HA patients. In conclusion, the Argentine HA patient series presents similar global and mutation-specific inhibitor risks than the HA database and other published series. This case-specific information will help in designing fitted therapies and follow-up protocols in Argentina.

Selection and characterisation of FVIII-specific single chain variable fragments

The development of inhibitory anti-FVIII antibodies is currently the most severe complication in the treatment of haemophilia A patients. Inhibitor eradication can be achieved by immune tolerance induction (ITI). Recent findings suggest a correlation between the FVIII-specific IgG subclass distribution and the duration or outcome of ITI. To quantify FVIII-specific IgG subclasses in patients' plasma FVIII-specific IgG standards are required. Here, the isolation of FVIII-specific single chain variable fragments (scFvs) from synthetic phage display libraries and the characterisation of their FVIII domain specificity are described. The isolated scFv 1G10, which binds to the FVIII A2 domain, was cloned into the context of the four human IgG (hIgG) subclasses and expressed in mammalian cells. Purified 1G10-hIgG1, -hIgG2, -hIgG3 and -hIgG4 are used as standards to determine the absolute amounts and relative contribution of the different FVIII-specific IgG subclasses in future studies. The results from these studies will eventually add to understanding the role of the FVIII-specific IgG subclass distribution as prognostic factor for the outcome of ITI.

Factor VIII Region 1811–1818 Is Involved in Factor IXa Binding and Factor VIIIa Stability

AbstractAbstract 3357Previously, we identified a role for the lysine residue couple 1967/1968 in the stability of activated factor VIII (FVIIIa). Using tandem mass tags (TMT 126/127) in combination with mass spectrometry, we identified lysine residues involved in the interaction between the A2 domain and the rest of heterodimer (A1/A3-C1-C2) of FVIIIa (Bloem et al., J Biol Chem 2012;287:5575–83). Upon FVIII activation and A2 domain dissociation, the highest increase in surface exposure occurred for the lysine couple 1967/1968, and functional studies confirmed the role thereof in FVIIIa stability. In addition to lysines 1967/1968 also other lysines displayed an increased surface exposure, including those in positions 1804, 1808, 1813 and 1818. The A3 domain region 1803–1818 has previously been implicated in interactions with ligands such as activated factor IX (FIXa). As such, one might expect increased surface exposure due to FVIII activation. On the other hand, A2 domain dissociation may have rearranged the A3 domain surface in this region. The relation between FIXa assembly and A2 domain retention was therefore explored in the present study.To unravel the role of region 1803–1818 in FVIIIa stability and FIXa binding, either region 1803–1810 or 1811–1818 was replaced by the corresponding regions of the homologous factor V. Additionally, as Asn1810 is N-linked glycosylated and this glycan is maintained in both chimeras, a variant that lacks this glycan (N1810C) was investigated.Factor × activation kinetics were used to investigate the apparent FIXa binding affinity of the FVIII variants. FXa generation was assessed on 15% phosphatidyl serine (PS) containing membranes. FIXa titration experiments showed that the affinity for the 1811–1818 variant is reduced (apparent Kd from 1.3 nM to 2.4 nM). Removal of the glycan and substitution of 1803–1810 has little or no effect on the apparent FIXa binding. FVIIIa-FIXa assembly on membranes containing 15% PS was studied using lipid- coated glass beads (lipospheres). Lipospheres were incubated with fluorescein-labeled FIXa and different FVIIIa concentrations. FIXa did only display liposphere binding in the presence of FVIIIa. Therefore, the mean fluorescent intensity on the lipospheres at increasing FVIIIa concentration could be used as a measure for FVIIIa-FIXa assembly on lipids. Results from these experiments showed that the 1811–1818 variant fails to promote FVIIIa-FIXa assembly, whereas the other variants behave like WT. To investigate FVIIIa stability, and thereby the role of the mutations on A2 domain dissociation, the activity of the variants was followed over time. Results showed that the 1811–1818 variant has a decreased half life of 1.5 min, compared to 6.9 min for WT. Also substitution of region 1803–1810 results in a lower half life, although to a lesser extent (2.8 min), whereas the glycan lacking variant behaves like WT (6.8 min). Incubation of FVIIIa variants with FIXa is known to stabilize FVIIIa and leads to an increased half life for all variants. However, the 1803–1810 variant is most efficiently stabilized by FIXa, shown by a 3-fold increase in half life, instead of 1.6-fold as seen for both WT and N1810C. The 1811–1818 variant is stabilized by 2.2-fold and therefore remains significantly less stable than WT.Together these results show that the 1811–1818 region is not only involved in FIXa binding, but additionally plays a major role in A2 domain retention. Region 1803–1810 also plays a role in FVIIIa stability, although to a lesser extent. Remarkably, the glycan at position Asn1810 does not influence neither FIXa binding nor FVIIIa stability, and apparently serves some other function. Disclosures:No relevant conflicts of interest to declare.

The Structural and Functional Diversity of the Humoral Immune Response to the A2 Domain of Human Factor VIII

Abstract 2247Hemophilia A is an X-linked recessive disorder that is caused by a deficiency or defect of factor VIII (fVIII) coagulant protein. The major complication of treatment is the development of anti-fVIII antibodies (inhibitors) in approximately 20–30% of patients with severe hemophilia A. The majority of these inhibitors are directed against the A2 or C2 domains (Prescott R et al. Blood 1997). This study examines the structural and functional diversity of the humoral immune response to the A2 domain of human fVIII.A panel of 24 murine anti-A2 monoclonal antibodies (MAbs) produced in our laboratory plus MAb413 (American Red Cross) and GMA012 (Green Mountain, Burlington, VA) were used in this study. Previous studies have shown that anti-C2 MAbs produced from murine anti-fVIII hybridomas had a similar spectrum of epitopes to those found in inhibitor patient plasmas (Meeks SL et al. Blood 2008). A competition sandwich ELISA with immobilized anti-A2 primary MAb, human fVIII, biotinylated anti-A2 secondary MAb and streptavidin–alkaline phosphatase conjugate for detection was used to determine overlapping epitopes. Each antibody was used as both a capture and detection antibody. Antibody pairs were classified as having non-overlapping or overlapping epitopes based on whether the binding of the secondary antibody was present or absent, respectively. Porcine/human hybrid fVIII proteins were employed in a direct ELISA to fine map the epitopes of the anti-A2 MAbs. The results of both the competition and human/porcine mapping ELISAs were compiled into a Venn diagram describing overlapping epitopes for all MAbs. Functional mapping of the MAbs included fVIII inhibitor titers by modified Bethesda assay, inhibition in a purified intrinsic Xase assay, and inhibition of thrombin cleavage of fVIII. Thrombin activation assays were run with varying concentrations of MAbs, and fVIII cleavage by thrombin was analyzed by SDS-PAGE.The competition ELISA results demonstrated 7 non-overlapping epitopes on the A2 domain of human fVIII (Figure 1). In addition, the human/porcine mapping ELISA revealed that the epitopes of the anti-A2 MAbs covered the majority of the A2 domain. The inhibitor titers of the anti-A2 MAbs ranged from non-inhibitory to 40,000 Bethesda units (BU)/mg IgG (Table). The inhibitory MAbs displayed both type I (greater than 95% inhibition at saturating MAb concentrations) and type II-(incomplete inhibition at saturating MAb concentrations) behavior. MAb413, a group D MAb, noncompetitively inhibits factor VIIIa cofactor activity without affecting thrombin cleavage. 2–54, a group G MAb, inhibits thrombin cleavage of both heavy and light chains. In contrast, 1D4, which overlaps groups B, E, and F, only inhibited light chain cleavage. [Display omitted] Overall these results indicate that the humoral immune response to the A2 domain of fVIII is complex in terms of both structural and functional epitopes. These anti-A2 MAbs were found to target 7 non-overlapping epitopes spanning the majority of the A2 domain. Elucidation of the structural and functional complexity of the anti-A2 repertoire will lead to a better understanding of the pathogenicity of A2 inhibitors.Table:A2 MAb CharacteristicsMAbInhibitor Titer (BU/mg)GroupStructural EpitopeB25100A444–5082G10500B468–484G323000C468–508MAb41321,000D484–5082–934E541–604B664000F604–7402–5433,000G508–541, 604–740 Disclosures:No relevant conflicts of interest to declare.

The Diversity of the Immune Response to the A2 Domain of Human Factor VIII In a Murine Hemophilia A Model

AbstractAbstract 2223Factor VIII (fVIII) inhibitory antibodies (fVIII inhibitors) are a significant source of morbidity in patients with hemophilia A. Approximately 30% of patients with severe hemophilia A will develop inhibitors. Most inhibitors are directed against either the A2 or C2 domains of fVIII. Anti-A2 antibodies have been reported to bind to at least 3 different regions of the A2 domain. Murine anti-human fVIII monoclonal antibody (MAb) MAb413 has an epitope that localizes to amino acids 484–508 and possibly blocks factor IXa binding to fVIII. R8B12 is a minimally inhibitory anti-A2 MAb that recognizes a discontinuous epitope that includes residues 497–510 and 584–593. CLB-Cag 9 is an anti-A2 MAb that recognizes amino acids 713–740 and decreases the rate of light chain cleavage by thrombin and factor Xa.The goal of this study was to investigate the diversity of the humoral immune response to the A2 domain of human fVIII in a murine hemophilia A model. A panel of 10 murine anti-A2 antibodies was obtained, which included 9 MAbs from anti-fVIII hybridomas produced in our laboratory and antibody MAb413. All 10 MAbs were used as primary antibodies in a competition ELISA using human fVIII as the antigen. The same panel was biotinylated and used as secondary antibodies. Antibody pairs were classified as having non-overlapping or overlapping epitopes based on whether the binding of the secondary antibody was present or absent, respectively. The competition ELISA yielded 3 distinct groups of structural epitopes (see Figure). The ability of the MAbs to inhibit porcine/human hybrid fVIII proteins in a one-stage coagulation assay was used to further delineate the position of the structural epitope. Group 1 contained 6 antibodies, including MAb413, with 4 distinct epitopes on the fVIII protein. Group 2 contained 2 MAbs, and the epitope for 2–54 localized to amino acids 508–541. Group 3 contained 2 MAbs, and the epitope for 1D4 localized to amino acids 605–740.Five of the 6 MAbs in Group 1 were Type I inhibitors with titers ranging from 330-40,000 Bethesda units (BU)/mg and one was non-inhibitory (see Table). In Group 2, both antibodies had high inhibitor titers (11,000 and 33,000 BU/mg) but were type II inhibitors with residual activities of ~20% at saturating concentration of antibody. Group 2 MAb 2–54 also was tested in a purified Xase assay and the residual fVIII activity was high (60-70%) at saturating concentration. Of the two MAbs in Group 3, one was non-inhibitory and the other was a Type I inhibitor with a titer of 7000 BU/mg. The concentration dependence of MAb413 (Group 1), 2–54 (Group 2), and 1D4 (Group 3) on fVIII cleavage and activity was tested by SDS-PAGE and the intrinsic Xase assay. With MAb413 a normal cleavage pattern was seen at concentrations that produced complete inhibition in the Xase assay. With 1D4, light chain, but not heavy chain, cleavage was inhibited in a concentration-dependent manner that corresponded to a decrease in Xase activity. This is consistent with CLB-Cag 9 which has a similar epitope at the C terminal end of the A2 domain. With 2–54, cleavage at all thrombin cleavage sites was partially inhibited, which corresponded to 30–40% inhibition in the Xase assay. This MAb represents a novel class of high titer, type II anti-A2 inhibitors that recognizes residues 508–541 and has similar in vitro characteristics to the Group BC anti-C2 MAbs. The elucidation of the structural and functional complexity of the anti-fVIII A2 repertoire should be useful in the characterization of the pathogenicity of A2 inhibitors. [Display omitted] Table:A2 MAb CharacteristicsMAbInhibitor Titer (BU/mg)Inhibitor KineticsGroupStructural Epitope4A440,000Type I1aMAb41321,000Type I1b484–5082-7638,000Type I1b2-10540,000Type I1b4F4330Type I1c2-93<1n/a1d2-5433,000Type II2508–5412-10111,000Type II21D47000Type I3605–7404C7<1n/a3 Disclosures:No relevant conflicts of interest to declare.

HLA-DRB1*1101-Restricted T Cells from Two Unrelated Mild Hemophilia a Inhibitor Subjects with Missense Genotype R593C Respond to the Same Minimal Epitope.

The development of inhibitory anti-FVIII antibodies in hemophilia A patients, referred to clinically as “inhibitors”, results from stimulation of T cells when the infused FVIII is recognized as foreign. T-cell stimulation follows the binding of specific epitopes within FVIII-derived peptides to HLA-DR molecules on the surface of antigen-presenting cells. We are investigating the HLA dependence of immune responses to FVIII and characterizing T-cell epitopes to better understand the process of T-cell stimulation leading to anti-FVIII antibody production. Inhibitors are less common in mild/moderately severe patients than in severe hemophilia A, but an elevated risk of inhibitor formation has been noted for patients with missense mutation R593C in the FVIII A2 domain. CD4+ T cells were isolated from two unrelated hemophilic subjects with this missense genotype, one Dutch and one American. Both had a DRB1*1101 haplotype and longstanding low- to moderate-titer inhibitors. Antigen-specific T cells from both subjects were identified by staining with fluorescent, peptide-loaded MHC Class II tetramers. The tetramer staining was carried out after stimulating the cells with synthetic peptides corresponding to the FVIII A2 domain sequence. Both hemophilic subjects, but not an HLA-matched healthy control, had a DRB1*1101-restricted response to peptide FVIII589-608, which contained the wild-type R593 sequence. Single-cell sorting and expansion with PHA facilitated the isolation of A2-specific T-cell clones from both subjects. A polyclonal population of A2-specific T cells was also isolated from the American subject's CD4+ T cells. All T-cell clones and the polyclonal T-cell line proliferated in response to the FVIII589-608 peptide and to FVIII. However, the cytokine profiles were somewhat different for peptide-stimulated cells from the two subjects. The single clone obtained from the Dutch subject produced modest levels (<750 pg/mL) of IFN-g, TNF-a, IL-4 and IL-10, while the cloned and polyclonal T cells from the American subject secreted similar levels of the latter three cytokines but had a much more pronounced Th1 polarization, secreting >10,000 pg/mL of IFN-g. Somewhat surprisingly, all of the cloned and polyclonal A2-specific cultured cells proliferated in response to the hemophilic peptide FVIII589-608,593C, despite the fact that this peptide elicited no response from primary T-cell cultures isolated from the American subject. FVIII589-608 bound to recombinant, monomeric DRB1*1101 protein with high affinity, while the affinity of the hemophilic peptide FVIII589-608,593C was threefold lower. FVIII residues 594-602 (FLPNPAGVQ) comprise a predicted high-affinity DRB1*1101-binding motif. Proliferation assays carried out with truncated peptides (FVIII592-608, FVIII593-608, and FVIII594-608) confirmed the responsiveness of the clones and polyclonal T-cell line to peptides containing this predicted motif, but proliferation decreased markedly when residue R593 was removed. The immunogenicity of this FVIII region with respect to other HLA types was evaluated by binding assays utilizing several recombinant DR molecules, in order to determine the relative promiscuity of this T-cell epitope. The effects of modifying amino acid residues within this epitope are also under investigation, as substitutions at “anchor” positions should decrease the immunogenicity of the corresponding peptides. Similar substitutions in recombinant FVIII may allow development of less immunogenic variants of human FVIII proteins.

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.

Identification of a thrombin‐interactive site within the FVIII A2 domain that is responsible for the cleavage at Arg372

FVIII is activated by cleavage at Arg(372), Arg(740), and Arg(1689) by thrombin. This study showed that an anti-A2 monoclonal antibody, with a specific epitope for residues 484-509, and anti-FVIII inhibitor alloantibodies with similar A2 epitopes, inhibited thrombin-catalyzed FVIII activation. Sodium dodecyl sulphate polyacrylamide gel electrophoresis analysis showed that cleavage at Arg(372) but not at Arg(740) occurred at approximately fourfold decreased rate in the presence of anti-A2 antibody. Peptide 484-509 also inhibited co-factor activation, consistent with inhibition of cleavage at Arg(372). Direct binding studies using active-site modified thrombin showed that a 484-509 peptide as well as the anti-A2 antibodies blocked the A2-thrombin binding. Furthermore, covalent cross-linking was observed between the 484-509 peptide and thrombin following reaction with 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide. Mutant A2 molecules in which the clustered basic residues in this sequence were converted to alanine were used to assess the binding reactions in a surface plasmon resonance-based assay. Mutants R484A, R489A, R490A, H497A and K499A possessed two to fivefold lower affinity than wild-type A2. These findings demonstrate that clustered basic residues within the 484-509 region of the A2 domain play a part of key role in thrombin-binding, which is responsible for thrombin-catalyzed FVIII activation by cleavage at Arg(372).

Epitope Identification and Generation of an Antigen-Specific T-Cell Clone in a Mild Hemophilia A Inhibitor Subject with Missense Genotype R593C.

The development of inhibitory anti-FVIII antibodies is a major clinical problem in hemophilia A. While less common in mild/moderately severe patients, the relative risk of inhibitor formation is elevated in patients with missense mutations in the FVIII A2 domain, especially those with an R593C genotype. In this study, T-cell responses to 14 FVIII A2 domain peptides with predicted DRB1*1101 MHC binding motifs were investigated using MHC class II tetramer reagents. CD4+ T cells were isolated from a hemophilic subject (DRB1*1101, 1302) with FVIII missense genotype R593C and stimulated with pooled peptides. This subject had developed a longstanding low titer inhibitor after receiving multiple FVIII infusions. Staining with fluorescent, peptide-loaded tetramers revealed that the hemophilic subject, but not an HLA-matched healthy control, had a DRB1*1101-restricted response to peptide A2589–608, which contained the wild-type R593 sequence. MHC class II tetramers with bound A2589–608 were used to sort antigen-specific T cells and then generate a T-cell clone recognizing this high avidity DRB1*1101-restricted epitope. FVIII residues 594–602 (FLPNPAGVQ) comprise a predicted high-affinity binding motif. Peptide binding assays confirmed that A2589–608 bound to recombinant, monomeric DRB1*1101 protein with high affinity. A peptide with the corresponding hemophilic sequence (A2589–608, 593C) bound with affinity that was threefold lower, but in a range that should be sufficient for presentation to T cells. The other 12 A2 peptides had a wide spectrum of binding affinities for the DRB1*1101 protein. Interestingly, prediction algorithms suggest that the peptides containing residue 593 occupy the MHC Class II peptide-binding groove in a register that places this residue just outside the p1 anchor position. Thus we anticipate that residue 593 modulates T-cell recognition of peptide A2589–608. This provides a potential explanation for inhibitor development in mild hemophilia A patients with a FVIII R593C genotype who are HLA-DRB1*1101. Immunogenicity of this FVIII region with respect to other HLA types is under investigation.

Clearance mechanisms of von Willebrand factor and factor VIII.

Clearance mechanisms of von Willebrand factor and factor VIII.