Abstract

Von Willebrand factor (VWF) plays an important role in primary hemostasis as it mediates platelet adhesion to the vessel wall and subsequent platelet aggregation at the site of vascular injury. Since the adhesive function of VWF depends on its multimer size, the loss of high molecular weight multimers (HMWM) results in hemorrhagic diathesis as seen in classical von Willebrand disease (VWD) type 2A, which represents the most common qualitative defect of VWD. The size distribution of VWF multimers in plasma is strongly influenced - if not regulated - by the specific VWF cleaving protease ADAMTS13 that cleaves VWF at the Y1605-M1606 bond in the A2 domain. Mutations in classical VWD type 2A cluster in the A2 domain and earlier studies suggest that some of the mutations make VWF more susceptible to ADAMTS13 dependent proteolysis (group 2) while others decrease the secretion of VWF HMWM (group 1). Our aim was to investigate the impact of VWF A2 domain mutations on ADAMTS13 dependent proteolysis of VWF. We used recombinant human ADAMTS13 (rhuADAMTS13) to digest recombinant full-length VWF and a fragment spanning the VWF A1-A2-A3 domains, harboring 13 different mutations that we found in patients with VWD type 2A. Proteolysis was monitored by VWF multimer analysis and by SDS-PAGE of the VWF A1-A2-A3 fragment. Cleavage of full-length VWF resulted in multimer patterns similar to that seen in plasma of patients with VWD type 2A, confirming the specifity of the reaction. Eleven VWF mutants (C1272S, G1505R, S1506L, M1528V, delR1569, R1597W, V1607D, G1609R, G1629E, G1631D, E1638K) showed less HMWM and more pronounced proteolytic fragments than wildtype (wt) VWF digested under the same conditions. Co-expression of the wt allele attenuated the proteolysis-permissive phenotype. The G1629E mutation resulted in highly increased proteolysis, suggesting an important role of this residue in the interaction between VWF and ADAMTS13. Surprisingly, G1505E and I1628T mutations failed to increase cleavage of the full-length VWF by rhuADAMTS13. However, when these mutations were introduced in the monomeric VWFA1-A2-A3 fragments they (like others) allowed cleavage of the Y1605-M1606 bond even under non-denaturing conditions, suggestive of an increased proteolytic susceptibility since wt VWF is only cleaved under denaturing conditions. The differences between the assays of full-length VWF and A1-A2-A3 domain fragment might be due to the lack of shear in our assay. This study provides direct evidence that in VWD type 2A, VWF with mutations in the A2 domain is subject to increased cleavage by ADAMTS13. This includes mutations previously designated as group 1 (G1505R, S1506L and V1607D) suggesting that increased susceptibility to ADAMTS13 is a more general property of VWF with A2 domain mutations. Therefore future therapies for patients with VWD type 2A might target VWF cleavage by ADAMTS13. RhuADAMTS13 and VWF constructs with mutations in the A2 domain are valuable tools to investigate VWF cleavage under varying conditions. Further work should address the question how shear influences ADAMTS13 dependent cleavage of VWF mutants.

Author notes

Corresponding author