Regulation of Factor VIII By Activated Protein C: Assessment In Vitro and In Vivo

Factor VIII (FVIII) plays an essential role in the activation of factor X (FX), acting as a factor IXa (FIXa) cofactor in the intrinsic Xase complex. Activated protein C (APC) is a serine protease that inactivates FV/Va. Resistance of FV/FVa to APC due to the Leiden mutation is the most common inherited thrombophilia. By cleaving FVIII/FVIIIa at Arg336 and Arg 562, APC can also inactivate FVIII; however, the role of this reaction, especially in vivo, is not well defined. Further, thrombophilic mutations at either of these sites have yet to be identified, calling into question the importance of this reaction. Understanding the role of APC in FVIII regulation is important as there are several novel therapeutics in development for hemophilia A (HA) that either bypass FVIII regulatory pathways or target inhibition of APC. Additionally, while data using a high-specific activity factor IX variant has been successful in a hemophilia B gene transfer trial, there is no such described hyperactive FVIII protein. In principle, manipulation of APC cleavage sites may result in a more procoagulant FVIII protein with enhanced hemostastic function that may serve as a novel therapeutic for HA in the setting of a recombinant protein treatment or gene therapy.

We sought to further understand the role of APC regulation of FVIII and to evaluate the in vivo hemostatic effect of an APC-resistant FVIII as a potential treatment strategy for HA. We conducted site-directed mutagenesis to introduce missense arginine to glutamine mutations at FVIII APC cleavage sites (R336Q and R562Q) on a B-domain deleted FVIII (FVIII-WT) backbone. The resulting recombinant, purified FVIII variant protein (FVIII-QQ) was then compared to FVIII-WT. Following incubation with APC, SDS-PAGE analysis of FVIII-QQ demonstrated no apparent cleavage by APC. In contrast, FVIII-WT demonstrated expected cleavage at R336 and R562, thereby confirming FVIII-QQ is resistant to APC cleavage. Surprisingly, one-stage clotting assay determination of FVIII-QQ specific activity was approximately 30% higher than FVIII-WT; this was unexpected given the absence of thrombomodulin or APC. We speculate this may be related to the ability of FXa to cleave FVIIIa at Arg336 and Arg562 inactivating FVIIIa that is likely disrupted in the FVIII-QQ variant (Eaton D et al. Biochemistry 1986 and Plantier JL et al. J Thromb Haemost 2010). To test the role of APC regulation of FVIII, we performed plasma-based in vitro thrombin generation assays in the presence of APC. Human FVIII deficient plasma was reconstituted to 1 nM (normal plasma FVIII concentration) with FVIII-QQ or FVIII-WT wherein increasing concentrations of APC were then added and thrombin generation was measured. Consistent with the known role of APC being an anticoagulant, the amount of thrombin generation decreased with increasing APC concentrations. However, the amount of APC necessary to inhibit 50% peak thrombin generation with FVIII-WT was 5.2 nM ± 0.3 nM versus FVIII-QQ was 11.3 nM ± 0.8 nM, which again demonstrated FVIII-QQ is resistant to APC inhibition. We suspect the observed decrease in thrombin generation with increasing APC concentrations is due to the role of APC in FV/FVa inhibition. Overall, our data support a role for APC inactivation of FVIIIa in this in vitro system. Interestingly, FVIII-QQ reliably demonstrated an approximate 2-fold increase in peak thrombin in the thrombin generation assay. This is being explored, but we again speculate this may be related to the effects of FXa on FVIII-WT at R336 and R562.

We next evaluated the in vivo hemostatic efficacy of FVIII-WT vs. FVIII-QQ in HA mice. HA mice were infused with FVIII-WT or FVIII-QQ protein prior to tail clip assay. Comparison of both proteins at a dose of 3 µg/kg demonstrated greater blood loss in HA mice treated with FVIII-WT vs. FVIII-QQ (p=0.006). Further, blood loss in HA mice treated with FVIII-WT was greater than wild type mice (p=0.007); in contrast, there was no difference between wild type and HA mice treated with 3 µg/kg of FVIII-QQ (p=0.26).

Our data demonstrate FVIII-QQ has a higher specific activity, requires 2-fold greater APC concentrations to inhibit thrombin generation in vitro and an approximate 2-fold greater in vivo hemostastic efficacy. Our results support that APC has physiologic significance in FVIII/FVIIIa inactivation and provide rationale for the study of APC in FVIII/FVIIIa regulation to develop potential novel HA therapeutics.