VWF self-association: more bands for the buck

In this issue of Blood, Dayananda and coworkers demonstrate that fluid-phase VWF can bind homotypically to platelet VWF, strengthening platelet-platelet association and enhancing platelet activation.¹ 

The hemostatic function of von Willebrand factor (VWF) depends on its degree of multimerization, which results from posttranslational assembly of the nascent VWF chains in the endoplasmic reticulum and Golgi apparatus. Intracellular assembly is promoted by the VWF propeptide, a 741-amino acid sequence at the N-terminus of the nascent VWF polypeptide that acts as a disulfide isomerase to promote the formation of unique disulfide bonds between VWF dimers to form an array of VWF multimers that are stored in endothelial Weibel-Palade bodies or platelet α-granules. These newly synthesized multimers can be extremely large (ultra-large VWF [ULVWF]) and hyperadhesive. Upon their secretion into the blood and under shear stress, ULVWF multimers are proteolytically cleaved by the plasma metalloprotease ADAMTS13 to produce a population of VWF multimers with lower reactivity for platelets that efficiently support primary hemostasis. Over the past several years, it has become increasingly appreciated that another important mechanism exists for regulating the hemostatic activity of VWF: its ability to self-associate. This phenomenon was first reported by Savage et al²  who showed that fluid-phase VWF multimers could homotypically associate with VWF multimers that were immobilized onto a collagen surface; the self-associated VWF multimers supported platelet adhesion under shear stress. Fluid-phase VWF lacking either the A1 or the A3 domain still bound immobilized wild-type VWF; later studies showed that VWF self-association involved multiple domains³  and—in the case of plasma VWF binding to ULVWF attached to an endothelial surface—required free thiols in plasma VWF.⁴  The self-association has been reported to be reversible,^2,3  although both studies examined this issue indirectly. VWF self-association appears to be facilitated by unfolding of the molecule, as it is increased by both shear stress^5,6  and ristocetin.⁷  One apparent consequence of VWF self-association is an increased local density of platelet-binding sites, resulting in enhanced platelet adhesion and surface coverage.^2,5 

The structure of self-assembled VWF multimers was characterized by immunofluorescence microscopy by Barg et al⁵  who showed that purified VWF multimers assembled into a macroscopic network of cross-linked fibers on a collagen surface under flow. This network of VWF fibers bound platelets under shear stress and was degraded by ADAMTS13 and other plasma proteases. Similarly, ristocetin, which mimics the effect of shear stress on VWF molecules, also promoted self-association of VWF multimers into a network of fibers in solution.⁷ 

In this issue, Dayananda et al demonstrate that soluble VWF can associate with VWF bound to GPIbα on platelets.¹  VWF multimers lacking the A1 domain (ΔA1-VWF) homotypically associated with normal plasma VWF molecules on GPIbα on the platelet surface in a shear-dependent manner. Assembly of multiple VWF molecules on the platelet surface increases the magnitude of tensile force exerted on the platelet through GPIbα, which triggers platelet activation. These novel findings suggest that dynamic self-association of VWF under shear stress not only promotes initial platelet adhesion at the site of vascular injury, but may also contribute to platelet aggregation and pathological thrombus growth, particularly in response to inflammation or under abnormal flow conditions. They alsoraise the possibility that augmented or acceler-ated VWF self-association could contribute to conditions such as thrombotic thrombocytopenic purpura (TTP). It is intriguing to note, for example, that the full-blown syndrome of microvascular thrombosis in TTP continues unabated in the face of extreme thrombocytopenia. Could it be that enhanced VWF self-association on small blood vessels plays as important a role in forming microthrombi in the later stages of the illness as platelet binding? If so, TTP treatments that target the GPIbα—VWF interaction may not be as effective as anticipated.

Might the thick VWF strands produced by VWF self-association⁴  by themselves be able to account for the schistocytosis that is such a prominent feature of TTP, much as in disseminated intravascular coagulation they are attributed to fibrin strands?

These are only 2 of the many questions to be answered regarding the intriguing phenomenon of VWF self-association. Also awaiting elucidation are the precise sequences that mediate the homotypic interaction, whether changes in the concentration, multimer composition, or reactivity of circulating VWF change the rate of self-association, and whether there are other molecules in the blood that modulate the phenomenon. Stay tuned.