A Novel Molecular Mechanism In The Interplay Of Platelet GPIb-VWF-Fibrin In Thrombus Formation

It is established that the interplay between von Willebrand factor (VWF) and fibrinogen is relevant for the platelet-platelet interaction in thrombus formation. Although the concentration of fibrinogen is much higher than that of VWF in blood, normally, they do not interact. In fact, VWF in plasma has a poor binding activity for fibrin(ogen)-coated surface, which exposes fibrin-specific sequences when it is immobilized. This observation suggests that 1) the site recognized by fibrin is inaccessible in plasma VWF, and 2) a conformational change in the VWF protein is a pre-requisite to bind fibrin. The latter may be supported by a previous study indicating that VWF has to bind to its platelet receptor glycoprotein (GP)Ib first in order to bind fibrin. Another line of evidence suggests that the binding of VWF to the polymerizing fibrin facilitates the interaction of platelet GPIb with the fibrin-bound VWF during the growth of the thrombus, a mechanism apparently defective in Bernard-Soulier syndrome (BSS) and von Willebrand disease (VWD). Consistent with this, fibrin monomer, like ristocetin, can induce the binding of VWF to GPIb. The C1C2 domains have been described as the fibrin-binding domain in VWF, and it is unknown whether the structures of these domains change to bind to fibrin. In contrast, conformational changes within the A1A2 domains of VWF induced by ristocetin result in the binding to GPIb and the protease ADAMTS-13. Therefore, we investigated a potential contact site for fibrin among the A1A2 domains of VWF. We characterized the interaction of each plasma (p)VWF, purified pVWF, recombinant A1A2A3 domains protein, single A1, and single A2 domains with fibrinogen, fibrin, or D-dimer. We also examined the effect of ristocetin in the VWF-fibrin binding. Both pVWF and purified pVWF bound to immobilized fibrin(ogen) in a ristocetin dependent manner. An A1A2A3 mutant expressing increased GPIb-binding activity bound to fibrin(ogen) in the absence of ristocetin. Both A1 and A2 domains, but not the A3 domain had binding activity for immobilized fibrin(ogen). Both soluble fibrin monomer and D-dimer, unlike fibrinogen in solution, had a significant binding activity for the immobilized A2 domain. Moreover, the A2 domain was capable of delaying fibrin polymerization in vitro. The use of conformation-specific monoclonal antibodies demonstrated that, like collagen, fibrin(ogen) induces a change in the structure of the A1 domain, and platelets interacted effectively with the fibrin(ogen)-bound A1 domain under high shear stress. Interestingly, in contrast to A3 or A1domain, flowing platelets did not adhere to the fibrin(ogen)-coated surface incubated with the A2 domain. In conclusion, we have demonstrated a previously unknown fibrin(ogen)-binding activity for the A1 and A2 domains in VWF. These observations reveal new mechanisms that may help to understand the pathophysiology of the bleeding in VWD, in which a number of natural occurring mutations that cause the disease are found within the A1 and A2 domains, and the poor fibrin polymerization observed in BSS. Furthermore, these outcomes may have implications for a better understanding of the high risk for thrombosis when the levels of both VWF and fibrinogen are increased.