Noncovalent stabilization of the factor VIII A2 domain enhances efficacy in hemophilia A mouse vascular injury models

Abstract

An important negative regulator of factor VIIIa (FVIIIa) cofactor activity is A2 subunit dissociation. FVIII molecules with stabilized activity have been generated by elimination of charged residues at the A1-A2 and A2-A3 interfaces. These molecules exhibited reduced decay rates as part of the enzymatic factor Xa generation complex and retained their activities under thermal and chemical denaturing conditions. We describe here the potency and efficacy of 1 such stability variant, D519V/E665V, derived from B domain-deleted FVIII (BDD-FVIII). The major effect of A2 stabilization was on cofactor activity. D519V/E665V potency was increased twofold by the 2-stage chromogenic assay relative to BDD-FVIII. D519V/E665V demonstrated enhanced thrombin generation responses (fivefold by peak thrombin) relative to BDD-FVIII. In vivo consequences of enhanced cofactor activity of D519V/E665V included >fourfold increased maximal platelet-fibrin deposition after laser injury and twofold increased protection from bleeding in acute and prolonged vascular injury model in hemophilia A mice. These results demonstrate that noncovalent stabilization of the FVIII A2 subunit can prolong its cofactor activity, leading to differential enhancement in clot formation over protection from blood loss in hemophilia. The FVIII molecule described here is the first molecule with clear efficacy enhancement resulting from noncovalent stabilization of the A2 domain.

Figure 1

A representative sensogram (Biacore analysis) showing A2 domain dissociation from FVIIIa after activation of FVIII by thrombin. The top solid line represents D519V/E665V; the dashed line represents BDD-FVIII. The lower lines (dashed and solid) are the respective nonspecific antibody control channels. All data have been subtracted from a blank channel and are representative of 2 independent experiments.

Figure 2

TGA responses of FVIII variants. Representative thrombin profiles at 0.5 nM (A) and the dose-dependent peak thrombin responses (B) for BDD-FVIII and D519V/E665V. The estimated EC₅₀ for BDD and D519V/E665V are 3.3 and 0.7 nM, respectively, indicating a ∼fivefold difference in potency. The specific activities of BDD-FVIII and D519V/E665V tested were 6.02 and 11.25 U/µg, respectively.

Figure 3

Comparable profiles. PK profiles for D519V/E665V and BDD-FVIII in HemA mice dosed with 200 U/kg BDD-FVIII or D519V/E665V. MRT, mean residence time; t_1/2, terminal half-life; Vss, volume of distribution at steady state.

Figure 4

Enhanced thrombin generation with D519V/E665V results in increased platelet and fibrin response to vascular injury. Intravital microscopy of laser-injured vessels was performed in HemA mice and the kinetics of platelet accumulation (A) and fibrin deposition (B) in response to D519V/E665V (filled black histograms) and BDD-FVIII (hatched histograms). The quantitation of the maximum clot size based on platelet-associated (C) and fibrin-associated (D) fluorescence is also shown. Comparison of the median maximum fluorescence (horizontal bars) shows a statistically significant difference between clots formed by animals dosed with BDD-FVIII (open circles) and D519V/E665V (filled circles); P < .0001 and P < .008 for platelet and fibrin maximum fluorescence, respectively. The BDD-FVIII data were derived from analysis of 85 clots from 9 mice; the D519V/E665V data were derived from analyzing 40 clots from 4 mice.

Figure 5

Tail clip injury model showing comparable efficacy as BDD-FVIII with half the mass dose for D519V/E665V. HemA mice were dosed with buffer (excipient); 0.6, 2, or 6 µg/kg D519V/E665V or 4 µg/kg BDD-FVIII. The BDD-FVIII dose represents the ED₅₀ for this molecule in these studies. Statistical significance (*P < .05, 1-way ANOVA with Dunnett’s multiple comparison test) was achieved with 2 µg/kg D519V/E665V vs excipient, whereas 4 µg/kg BDD-FVIII was required to achieve the same level of statistical difference vs excipient. †0.6 U/kg = 2 µg/kg D519V/E665V = 4 µg/kg BDD-FVIII. ANOVA, analysis of variance.

Figure 6

Enhanced efficacy of D519V/E665V in a TVT injury model. HemA mice were treated with BDD-FVIII or D519V/E665V at the indicated dose 24 hours before TVT. HemA mouse 24-hour survival plotted against the dose. ED₅₀ values obtained for HemA mice dosed with BDD-FVIII or D519V/E665V are displayed, along with their respective chromogenic specific activities (right). Statistical significance (P = .05) was determined using analysis of covariance. Chromo SA, chromogenic specific activity.