Coagulation factor VIII regulates von Willebrand factor homeostasis invivo

Abstract

Background: Coagulation factor VIII (FVIII) and von Willebrand factor (VWF) circulate as a noncovalent complex, but each has its distinct functions. Binding of FVIII to VWF results in a prolongation of FVIII's half-life in circulation and modulates FVIII's immunogenicity during hemophilia therapy. However, the biological effect of FVIII and VWF interaction on VWF homeostasis is not fully understood.

Objectives: To determine the effect of FVIII in VWF proteolysis and homeostasis in vivo.

Methods: Mouse models, recombinant FVIII infusion, and patients with hemophilia A on a high dose FVIII for immune tolerance induction therapy or emicizumab for bleeding symptoms were included to address this question.

Results: An intravenous infusion of a recombinant B-domain less FVIII (BDD-FVIII) (40 and 160 μg/kg) into wild-type mice significantly reduced plasma VWF multimer sizes and its antigen levels; an infusion of a high but not low dose of BDD-FVIII into Adamts13^+/- and Adamts13^-/- mice also significantly reduced the size of VWF multimers. However, plasma levels of VWF antigen remained unchanged following administration of any dose BDD-FVIII into Adamts13^-/- mice, suggesting partial ADAMTS-13 dependency in FVIII-augmented VWF degradation. Moreover, persistent expression of BDD-FVIII at ∼50 to 250 U/dL via AAV8 vector in hemophilia A mice also resulted in a significant reduction of plasma VWF multimer sizes and antigen levels. Finally, the sizes of plasma VWF multimers were significantly reduced in patients with hemophilia A who received a dose of recombinant or plasma-derived FVIII for immune tolerance induction therapy.

Conclusion: Our results demonstrate the pivotal role of FVIII as a cofactor regulating VWF proteolysis and homeostasis under various (patho)physiological conditions.

Keywords: ADAMTS-13; coagulation factor VIII; proteolysis; regulation; thrombosis; von Willebrand factor.

Fig. 1.. Characterization of recombinant FVIII protein and mutant mice

. A. SDS-PAGE with Coomassie blue staining for a protein marker (Lane 1) and a purified recombinant B domain-deleted FVIII (DBBFVIII) (Lane 2); a, b, and c denote the intact, heavy chain, and light chain, respectively. B. Plasma ADAMTS13 activity in wild-type (WT), Adamts13^+/− (A13^+/−), and Adamts13^−/− (A13^−/−) mice determined by an ELISA-based assay using a pooled human normal plasma as a calibrator; C. Plasma FVIII activities in fVIII^−/− mice 13 weeks following administration of nothing or AAV8-BDD-FVIII. Data in panels B and C were presented as the mean ± standard deviation (SD). Kruskal-Willis one way analysis and Mann-Whitney analysis were performed to determine the difference among three groups in panel B and between two groups in panel C. Here, four stars (_****) indicates a p value <0.001.

Fig. 2.. Effect of recombinant BBD-FVIII on plasma VWF multimer distribution and antigen levels in wild type mice

. A and B/C show the multimer distribution and the ratios of high (H) to low (L) molecular weight of VWF multimers, respectively, in wild-type (WT) mice prior to (pre) and 24 hours following an infusion of recombinant BDD-FVIII (40 μg/kg or 3.0 nM). D and E/F are the multimer distribution and the H/L ratios of VWF multimers, respectively, in WT mice pre and 24 hours of post infusion of BDD-FVIII (160 μg/kg or 12.1 nM). B/E. Paired individual values of H/L ratios of VWF multimers pre and post BDD-FVIII infusion. C/F. Individual values and mean ± SEM of H/L ratios of VWF multimers pre and post BDD-FVIII infusion. G/H and I/J show the plasma VWF antigen levels pre and 24 hours of post BDD-FVIII infusion at 40 μg/kg (3.0 nM) and 160 μg/kg (12.1 nM), respectively. G/I. Paired individual values of VWF antigen pre and post BDD-FVIII infusion. H/J. Individual values and mean ± SEM of VWF antigen pre and post BDD-FVIII infusion. A paired t-test was performed for determine the statistical significance of the difference of two different time points. Here *, **, and *** indicate a p value less than 0.05, 0.01, and 0.005, respectively.

Fig. 3.. Effect of recombinant BDD-FVIII on plasma VWF multimer distribution and its antigen levels in Adamts13^+/− mice.

A and B/C show the multimer distribution and the ratios of high (H) to low (L) molecular weight of VWF multimers, respectively, in Adamts13^+/− (A13^+/−) mice pre and 24 hours post Infusion of BDD-FVIII at 40 μg/kg (or 3.0 nM). D and E/F are the multimer distribution and the ratios of H/L molecular weight of VWF multimers, respectively, in Adamts13^+/− (A13^+/−) mice pre and 24 hours post infusion of BDD-FVIII at 160 μg/kg (or 12.1 nM). G/H and I/J are the plasma VWF antigen levels pre and 24 hours of post infusion of BDD-FVIII at 40 μg/kg (or 3.0 nM) and 160 μg/kg (or 12.1 nM), respectively. A paired t-test was performed for determine the statistical significance of the difference of two different time points. Here, ns and **** indicate p value greater than 0.05 and less than 0.0001, respectively.

Fig. 4.. Effect of recombinant BDD-FVIII on plasma VWF multimer distribution and antigen in Adamts13^−/− mice.

A and B/C show the plasma multimer distribution and the ratios of high (H) to low (L) molecular weight of VWF multimers, respectively, in Adamts13^−/− (A13^−/−) mice pre and 24 hours of post infusion of BDD-FVIII at 40 μg/kg (or 3.0 nM). D and E/F are the plasma multimer distribution and the ratios of H/L molecular weight of VWF multimers, respectively, in Adamts13^−/− (A13^−/−) mice pre and 24 hours post infusion of BDD-FVIII at 160 μg/kg (or 12.1 nM). G/H and I/J show the plasma VWF antigen levels in A13^−/− mice pre and 24 hours of post infusion of BDD-FVIII at 40 μg/kg (or 3.0 nM) and 160 μg/kg (or 12.1 nM), respectively. A paired t-test was performed for determine the statistical significance between two different time points. Here, ns and *** indicate p value greater than 0.05 and less than 0.005, respectively.

Fig. 5.. Plasma VWF multimer distribution and antigen levels in fVIII^−/− mice following AAV8-mediated expression of recombinant BDD-FVIII.

A. An outlined protocol for administration of AAV8-BDD-FVIII vector and blood collection. B and C. Plasma VWF multimer distribution in the untreated fVIII^−/− mice (control) and mice received AAV8-BDD-FVIII (1×10¹¹ vg/mouse), respectively. D. The individual values and mean ± SEM of the ratios of high (H) to low (L) molecular weight of VWF multimers in fVIII^−/− mice without or with expression of BDD-FVIII. E. The individual and mean ± SEM of plasma VWF antigen levels. Mann-Whitney test was performed to determine the statistical significance. Here, * and ** indicate the p value less than 0.05 and 0.01, respectively.

Fig. 6.. Plasma VWF multimer distribution and its antigen levels in haemophilia A patients with immune tolerance induction or emicizumab therapy.

A. A representative image showing plasma VWF multimer distribution in 4 haemophilia A patients (Pt1-Pt4) who were treated with (+) immune tolerance induction (ITI) or emicizumab (Emic.) or without (−). Plasma from 4 other haemophilia A patients without recent FVIII exposure (HA-82, HA-84, HA-85, and HA-86) was used for additional controls. B, C and D show the ratios of high (H) to low (L) molecular weights of plasma VWF multimers, VWF antigen levels, and FVIII antigen, respectively, in haemophilia A patients who did not have recent FVIII exposure (HA), those who treated with ITI, and those with Emici. The data are presented as the median and interquartile range (IQR). Kruskal-Wallis test was performed to determine the statistical significance. Here, ns, *, and ** indicate the p value greater than 0.05, less than 0.05, and 0.01, respectively.