Inorganic polyphosphate (polyP) is widespread throughout biology, although its contributions to biological processes have been studied most extensively in microorganisms . In mammalian cells, a variety of disparate roles for polyP have been reported in recent years, including in cell proliferation , angiogenesis , bone mineralization , energy metabolism  and tumor metastasis . In 2004, Ruiz et al.  reported the remarkable finding that dense granules of human platelets contain abundant levels of polyP, and that polyP is efficiently secreted upon platelet activation. While microbial polyP is typically very heterodisperse and contains very long polymers (hundreds of phosphate units long) , platelet polyP is considerably shorter and much more uniform in size (about 60–100 phosphate units long) [7,9]. In 2006, we showed for the first time that polyP of the approximate size secreted by platelets is strongly procoagulant, leading to shortening of plasma clotting times and rendering clots more resistant to fibrinolysis .
Subsequent studies from our laboratory and others have provided a richer understanding of how polyP modulates blood clotting, revealing that it acts at multiple steps in the plasma clotting cascade, but always in a procoagulant manner . Long-chain polyP of the size found in bacteria (i.e. many hundreds of phosphate units long) is an extremely potent activator of the contact pathway of blood clotting  and may represent at least one of the long-sought (patho)physiologic activators of this rather enigmatic pathway . Medium-chain polyP (of the size secreted by activated platelets, and also a bit longer) is very active in accelerating the proteolytic activation of factor (F) V [9,10]. This has the consequence of accelerating the appearance of the thrombin burst during plasma clotting reactions, and also of totally abrogating the anticoagulant activity of tissue factor pathway inhibitor [9,10]. Medium-chain polyP also greatly accelerates the rate of back-activation of FXI by thrombin , and in fact polyP secreted by platelets may represent the endogenous cofactor that permits this reaction to occur at physiologically meaningful rates. Finally, medium-chain polyP enhances the thickness of fibrin fibrils and makes them more resistant to fibrinolysis [9,13,14]. Most of the above studies were conducted in vitro, but experiments with mice have shown that polyP can be both strongly prothrombotic and pro-inflammatory in vivo, the latter as a consequence of bradykinin release following activation of the contact pathway of clotting .
Although a limited number of studies of polyP’s effects on blood clotting have focused on detailed biochemical mechanisms [11,16], it is probably fair to say that we are still in a discovery phase in which the effects of polyP on specific proteins, or specific steps in blood clotting and inflammation, are still being identified. Thus, most recently, polyP has been shown to potentiate platelet activation by extracellular histones  and to elicite pro-inflammatory responses from endothelial cells . It seems very likely that polyP secreted from activated platelets [7,15] and mast cells  has other, undiscovered, roles in blood clotting, inflammation and innate immunity.
In this issue of the Journal of Thrombosis and Haemostasis, Montilla et al.  now report that polyP of the size secreted by human platelets bound tightly to von Willebrand factor (VWF) and, in fact, they found that polyP was present in purified VWF protein isolated from normal human platelets and plasma, indicating that this protein normally circulates with polyP bound to it. They further found that platelets of patients with type 1 von Willebrand disease had reduced levels of polyP compared with normal controls, further supporting the idea that polyP is normally associated with VWF in vivo. When Montilla et al.  treated normal plasma with an enzyme that specifically degrades polyP (exopolyphosphatase), they found that the ability of VWF to interact productively with platelet glycoprotein 1b was significantly reduced, which is direct evidence that the polyP bound to VWF enhances one of the important functions of this protein. On the other hand, they reported that binding of VWF to collagen, and multimerization of VWF, were unaffected by degrading polyP, indicating that polyP has specific, limited contributions to VWF function. The fact that they found that added polyP was able to increase the deficient ristocetin-cofactor activity of VWF from patients with type 1 von Willebrand disease raises the intriguing possibility of polyP being a treatment for at least some types of this bleeding disorder.
Heparin, like polyP, is a linear, highly anionic polymer. Interestingly, however, these two polyanions exhibit quite different effects on blood clotting. For example, polyP has no effect on the rate of inhibition of thrombin or FXa by antithrombin , while heparin strongly enhances the ability of antithrombin to inactivate these proteases. Another example is the autoactivation of ‘factor seven-activating protease’ (FSAP), which is strongly stimulated by polyp, but unlike heparin, polyP does promote inhibition of FSAP by antithrombin . In a similar vein, Montilla et al.  found that polyP promoted ristocetin-induced platelet agglutination (which is mediated by VWF). In contrast, heparin inhibits the activity of VWF .
This intriguing report by Montilla et al.  adds to a growing body of evidence that polyP secreted by activated platelets (and as very recently reported, mast cells ) exerts pleiotropic effects on blood clotting, inflammation and innate immunity. Doubtless there are more revelations to come for polyP, a highly multi-tasking molecule.