Abstract 2213

Coagulation factor VIII (FVIII) binds von Willebrand factor (VWF) at the D′D3 domain with high affinity. FVIII almost always circulates as a FVIII-VWF complex in blood. Our recent study has demonstrated that FVIII accelerates proteolytic cleavage of VWF by ADAMTS13 under mechanically induced fluid shear stress (Cao et al, PNAS, 2008). In this study, we sought to determine: 1) the domain of FVIII required for the enhancing effect on VWF proteolysis by ADAMTS13 under these conditions; 2) the physiological relevance of this enhancing effect on plasma VWF homeostasis in a murine model. First, we employed the novel vortex-based shear assay developed in the laboratory to assess the rate enhancing effect of various recombinant FVIII variants on VWF proteolysis by ADAMTS13. In these experiments, purified plasma VWF (150 nM) was incubated for 10 min with recombinant ADAMTS13 (50 nM) in the presence of various concentrations of recombinant FVIII variants (0-20 nM) in a total volume of 20 μl under constant fluid shear generated with a bench top mini-vortexer. This maneuver appears to produce ∼75 dyn/cm2 of fluid shear stress. Proteolytic cleavage product (∼350 kDa, the dimer of two C-terminal fragments linked by a disulfide bond) was then determined by 5% SDS-polyacrylamide gel electrophoresis and Western blotting. We showed that addition of a recombinant light chain of FVIII (FVIII-LC) increased the formation of proteolytic cleavage product as a function of increasing FVIII-LC concentrations. This rate enhancing effect was similar to that of full-length recombinant FVIII and B-domainless FVIII (FVIII-SQ). However, addition of a heavy chain of FVIII (FVIII-HC) or a light chain variant lacking the acidic (a3) region (FVIII-Δa3LC) did not increase the formation of cleavage product under the same conditions. These results suggest that FVIII-light chain is sufficient for accelerating VWF proteolysis by ADAMTS13 under physiological conditions. The rate enhancing effect of FVIII-LC depends on its high affinity interaction with VWF. Second, we determined plasma VWF antigen and multimer distribution in fVIII-/- and fVIII+/+ (WT) mice with same genetic background (C57BL/6) prior to and after reconstitution with recombinant FVIII-SQ and variants via a hydrodynamic approach. Plasma VWF antigen was determined by a sandwich ELISA assay and plasma VWF multimers were determined by 1% agarose gel electrophoresis and Western blotting. We showed that plasma VWF antigen levels in the fVIII-/- mice (n=18) were increased by ∼2.0 fold as compared with those in the WT mice (n=14). No difference in the ratio of ultra large (UL)-VWF to dimer was observed between the fVIII-/- mice and the WT mice (p>0.05). These data suggest that lack of FVIII may impair plasma VWF homeostasis. To assess whether plasma FVIII affects VWF proteolysis in vivo, plasma VWF multimer distribution was determined by agarose (1%) gel electrophoresis and Western blotting in the fVIII-/- mice 48 hours after injection of a series of endotoxin-free expression plasmids encoding various FVIII variants/fragments (i.e. plasmids diluted in 2 ml normal saline and injected into tail vein within 5 seconds). We showed that the ratio of the UL-VWF multimers to the dimer in plasma of fVIII-/- mice receiving normal saline alone was 1.70 ± 0.59 (means ± SD) (n=10). However, the ratios in plasma of fVIII-/- mice receiving plasmids encoding canine FVIII-SQ, human FVIII-SQ, human FVIII-HC+LC, and FVIII-LC were 0.44 ± 0.37 (n=20), 0.88 ± 0.18 (n=9), 0.40 ± 0.20 (n=5), and 0.97 ± 0.29 (n=9), respectively. These ratios were dramatically reduced compared with that in fVIII-/- mice receiving normal saline alone (p values<0.05∼0.001). Our hydrodynamic injection approach resulted in plasma levels of FVIII-SQ, FVIII-LC+HC and FVIII-LC between 150% and 200% of the WT. In contrast, a hydrodynamic injection of plasmid encoding human FVIII-HC or FVIII-Δa3LC did not prevent the accumulation of plasma UL-VWF multimers in fVIII-/- mice. We therefore conclude that FVIII, through the high affinity binding interaction between its light chain and D′D3 domain of VWF, may play a critical role in maintaining VWF homeostasis under (patho) physiological conditions.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution