A Murine FVIIa Variant with Increased Tissue Factor-Independent Intrinsic Activity Corrects the Murine Hemophilia B Phenotype Following AAV Administration: Implications for Therapeutic AAV Vector Doses.

We have previously designed a gene transfer approach using a modified FVII transgene that is cleaved intracellularly, secreted in the active form (FVIIa) and results in phenotypic correction of hemophilia B mice, following adeno-associated virus (AAV)-mediated, liver-directed gene delivery. Research into FVIIa variants with increased catalytic efficiency will not only provide information on the mechanism of action of high-dose FVIIa [tissue factor (TF)-dependent or independent], but may also lower the vector doses required in gene therapy settings. However, the use of such variants in an animal model of hemophilia is lacking. To address this, based on published observations using human FVIIa, we have generated a murine FVIIa variant (L305V/A314E/K337A/I374Y, mFVIIa-VEAY) with substitutions located in the catalytic domain of murine FVIIa. Purified mFVIIa and mFVIIa-VEAY from conditioned medium were almost exclusively in the activated form. Mouse FVII zymogen was purified as a single chain, indicating the lack of autoactivation during purification. Using clotting-based assays for the extrinsic (PT) and intrinsic pathway (aPTT), we found that variant VEAY had almost identical extrinsic activity to unsubstituted mFVIIa but substantially increased intrinsic activity (between 3–7 fold) using FVIII, FIX or FVII deficient plasmas, suggesting that this variant has TF-independent clotting activity. From in vitro kinetic assays, binding to human or murine soluble TF was identical for both mFVIIa and mFVIIa-VEAY (2.0 and 3.2 nM for murine TF, respectively; 86.7 and 71.7 nM for human TF, respectively). Similarly, using the physiological substrate (human FX) at saturating TF conditions, FXa generation was identical for both mFVIIa and mFVIIa-VEAY. However, in the absence of TF, mFVIIa-VEAY exhibited substantial proteolytic activity towards FX (6.6 fold), when compared to mFVIIa (which results in a minimal rate of FXa generation). In rotational thromboelastometry with citrated hemophilia B whole mouse blood and mFVIIa-VEAY or mFVIIa added at a dose of 1μg/ml (corresponding to the clinically effective dose of 90μg/kg), mFVIIa-VEAY exhibited a reduction in the clot time (CT) relative to both mFVIIa-treated hemophilia B blood or wild-type blood [61 ± 3 sec vs. 79 ± 13 sec (mFVIIa-treated) and 106 ± 13 sec (wild-type)]. Lastly, we administered 1.2 × 10^12 vector genomes of AAV8-mFVIIa or AAV-mFVIIa-VEAY into the hepatic circulation of hemophilia B mice. At 2 weeks post transduction, AAV-mFVIIa-VEAY treated mice exhibited superphysiological correction of aPTT and PT, considerably below that of wild type animals (aPTT: 17.4 vs. 28.2 sec; PT: 20.8 vs. 39.7, respectively). In contrast, AAV-mFVIIa treated mice exhibited near normalization of phenotype, indicating that phenotypic correction of the hemophilia B phenotype may be achieved using lower AAV-mFVIIa-VEAY vector doses, compared to mFVIIa. The effect of mFVIIa-VEAY in TF-dependent mouse injury models as well as a low TF mouse background is currently under investigation. In conclusion, we describe a murine FVIIa variant with increased TF-independent activity in vitro; we demonstrate that the variant also shows increased activity in vivo. This variant mFVIIa has potential for lowering the AAV dose in vivo for use in gene transfer applications and may allow for the extension of this approach to a large animal model of hemophilia, since high doses of vector are required to treat hemophilic dogs using the wild-type molecule.