The Hemostatic Impairment in Mild Hemophilia Α Possessing Arg593Cys Mutation Is Caused By the Attenuated Binding to Phospholipid

Approximately 10% of mild hemophilia A (HA) patients possessing an Arg593Cys mutation in the A2 domain of factor (F)VIII heavy chain enrolled in the HAMSTeRS database have been reported to develop the anti-FVIII inhibitor alloantibodies (Ab). The hemostatic characteristics and the mechanism(s) of the development of inhibitors in HA with this mutation remain unclear, however. We had a mild HA patient carrying this mutation with inter-assay variation in FVIII activity (FVIII:C) and FVIII antigen (FVIII:Ag). Patient’s FVIII:C (17.1 IU/dl) measured in aPTT-based clotting assay using the coagulometer (STart-4^®) by an electro-mechanical clot detection method was higher by 2-fold than that (8.1 IU/dl) using an optical clot detection system (MDA-II^®). In order to assess the precise FVIII activity, we applied the thrombin generation test. We evaluated the peak thrombin height (PeakTh) in the control plasmas consisting of FVIII-deficient plasma and serial dilutions of recombinant FVIII. PeakTh (319 ± 59 nM) in the patient plasma was corresponded to FVIII:C 15 IU/dl. The levels of pro-/anti-coagulant, and fibrinolytic factors were within normal range, except for FVIII. Of note, FVIII:Ag in the patient plasma was 4.2%, measured by a sandwich ELISA with the anti-FVIII polyclonal Abs recognizing the C2 domain, whilst the one by an ELISA with anti-A2 polyclonal Ab and anti-A1 monoclonal Ab (VIII-9222) was 14%, corresponded to FVIII:C level. FVIII:Ag in the normal plasma showed any little difference between both ELISAs. We focused on the impaired interaction between PL and FVIII in the patient plasma, since a PL-binding site on FVIII was located on the C2 domain. FVIII:C assessed by a one-stage clotting assay with various concentrations of PL (25, 12.5, 6.25, 1.5 μM) in the patient plasma exhibited the PL-dependent reduction and the % reduction to FVIII:C with PL(25μM) was 34%, 45%, 50%, respectively, whilst that in the control plasma was any little decreased. FXa generation in his plasma and a control plasma adjusted to 20 IU/dl of FVIII concentration was compared using various concentrations of PL. An apparent K_m for PL in the FXa generation showed an ~4-fold increase compared to that of control (29 ± 6.1 and 7.2 ± 1.5 μM, respectively), suggesting that the reactivity of FVIII with PL in the patient plasma appeared to be lower compared to that in normal control. To elucidate the mechanism(s) of the alteration in the reactivity to PL, a binding affinity of FVIII in the patient plasma to the immobilized anti-C2 mAbs ESH4 (epitope 2303-2332, overlapped with the PL-binding region) and ESH8 (epitope 2248-2285), anti-A1 VIII-9222 (epitope 337-372) was evaluated using a surface plasmon resonance (SPR-)based assay. The binding affinity to ESH4 in the patient plasma (K_d; 308 ± 44 nM) was apparently lower than that in the control plasma (K_d; 0.48 ± 0.16 nM), whilst those to the other mAbs were of little difference between the case and control plasmas (K_d; 1.38 ± 0.12 and 0.64 ± 0.33 nM for ESH8, respectively, and K_d; 0.54 ± 0.34 and 0.37 ± 0.26 nM for VIII-9222, respectively), supporting the impairment of C2 epitope (residues 2303-2332) on FVIII-Arg593Cys. Taken together, the Arg593Cys mutation in the A2 domain seemed to induce the conformational change at the remote site in the C2 domain resulting in the attenuated binding affinity to PL. This mechanism would account for the hemostatic impairment and also be associated with the development of the inhibitor in HA patients with this mutation.