Polyphosphate modulates blood coagulation and fibrinolysis

Abstract

Inorganic polyphosphate is an abundant component of acidocalcisomes of bacteria and unicellular eukaryotes. Human platelet dense granules strongly resemble acidocalcisomes, and we recently showed that they contain substantial amounts of polyphosphate, which is secreted upon platelet activation. We now report that polyphosphate is a potent hemostatic regulator, accelerating blood clotting by activating the contact pathway and promoting the activation of factor V, which in turn results in abrogation of the function of the natural anticoagulant protein, tissue factor pathway inhibitor. Polyphosphate was also found to delay clot lysis by enhancing a natural antifibrinolytic agent, thrombin-activatable fibrinolysis inhibitor. Polyphosphate is unstable in blood or plasma, owing to the presence of phosphatases. We propose that polyphosphate released from platelets or microorganisms initially promotes clot formation and stability; subsequent degradation of polyphosphate by blood phosphatases fosters inhibition of clotting and activation of fibrinolysis during wound healing.

Figures

Fig. 1.

PolyP activates clotting and delays fibrinolysis. (A) Combined clotting/fibrinolysis assays were conducted in 96-well plates by adding 75 μM polyP75 3 min before adding Ca2+ and uPA (done in duplicate; representative of ten experiments). In the presence of polyP, plasma clotted faster and the clots lysed more slowly. (B) Clotting assays were conducted as in A except that no uPA was added and plasma was preincubated for 2 min with 0 or 75 μM polyP75 before the addition of Ca2+. The first two bars on the left show that addition of polyP to pooled normal plasma (PNP) dramatically shortened the clotting time, whereas the next two bars show that addition of polyP to factor XII-deficient plasma (XII) did not. The two bars on the right are the results of, first, preincubating 40 μl plasma (factor XII-deficient or normal) with 75 μM polyP75 for 2 min, and second, adding an additional 40-μl aliquot of normal or factor XII-deficient plasma, respectively, followed immediately by the addition of Ca2+ to initiate clotting. PolyP dramatically shortened the clotting time only when preincubated with normal plasma, not with factor XII-deficient plasma, even though the final plasma mixtures were identical.

Fig. 2.

PolyP abrogates the anticoagulant effect of TFPI. (A) Tissue factor-induced clotting times were measured in a coagulometer using normal plasma and the indicated concentrations of recombinant TFPI, in the presence or absence of 25 μM polyP75.(B) Concentration dependence of the abrogation of TFPI's anticoagulant effect by polyP. Clotting assays were performed with normal plasma as in A containing 0 or 1 nM recombinant TFPI and varying concentrations of polyP75. Data are mean clotting times ± SEM. (C) Factor Va abrogates TFPI function. Clotting assays were performed as in panel A using factor V-deficient plasma supplemented with 1 nM factor Va in the presence (filled circles) or absence (open triangles) of 25 μM polyP75.(D) The ability of polyP to abrogate TFPI function depends on chain length. Clotting assays were performed as in panel B with 0 or 2 nM TFPI, and with 25 μM polyP of the indicated chain lengths. Mean clotting times (±SEM) are depicted. (E) Phosphatase digestion of polyP eliminated its ability to abrogate TFPI function. PolyP65 (7.5 mM) was digested for 2 h at room temperature with 0 or 100 units/ml calf intestinal alkaline phosphatase, then diluted for use in clotting assays. Clotting assays were performed as in B with 0 or 1 nM TFPI, and mixed with a dilution of the phosphatase reaction mixtures to yield a final concentration of 4.3 μM polyP65. Data are mean clotting times ± SEM. (F) Urea-PAGE analysis of polyP65 after 2 h incubation with (+) or without (–) alkaline phosphatase; polyP was stained with toluidine blue.

Fig. 3.

PolyP accelerates the activation of factor V and the generation of thrombin. (A) Activation of prothrombin is accelerated by polyP plus factor V. Reactions contained 6 nM factor V, 0.33 nM factor Xa, and 400 nM prothrombin with or without 22.5 μM polyP75. The y axis represents cleavage of the thrombin substrate, S-2238. (B) Activation of prothrombin is not accelerated by polyP in the presence of factor Va. Reactions were identical to A except that 0.45 nM factor Va was used in place of factor V. (C) PolyP accelerates factor V activation by factor Xa (Xa) and thrombin (IIa). Factor V (100 nM) was reacted with either 1 nM factor Xa or 0.1 pM thrombin for the indicated times, aliquots were inactivated by heating in SDS sample buffer and resolved on SDS PAGE followed by Western blotting, and the factor Va heavy chain was detected by using a specific monoclonal antibody. (D) Platelet releasates abrogate the anticoagulant function of TFPI. Clotting assays were performed as in Fig. 2 A, containing 0 or 1 nM TFPI and the indicated dilution of platelet releasate (1 = undiluted; 0 = no releasate added) prepared from platelets resuspended at 6.0 × 109 per ml. Data are mean ± SEM (n = 3) from a single donor's platelets, but are representative of releasates obtained from five separate donors. (E) PolyP shortens the lag to thrombin generation. Plasma clotting was initiated by tissue factor with or without 75 μM polyP75, and thrombin generation was monitored in real time by using a fluorogenic substrate for thrombin and analyzed in a Thrombinoscope. Lag time is the time to achieve 10 nM thrombin, whereas ETP is endogenous thrombin potential (integrated thrombin generation over 60 min). Data are mean ± SEM (n = 4); asterisks indicate statistical significance (P < 0.0001 using Student's t test).

Fig. 4.

PolyP inhibits fibrinolysis in a TAFI-dependent fashion. Combined clotting/fibrinolysis assays were performed in 96-well plates as in Fig. 1 A, except that 10 nM thrombin was added to rapidly initiate clotting. (A) PolyP inhibition of fibrinolysis is concentration dependent. Assays were performed with pooled normal plasma in the presence of 0 nM (open circles), 750 nM (open triangles), 3.75 μM (filled triangles), 7.5 μM (open squares), 37.5 μM (filled squares), or 75 μM (filled circles) polyP75.(B) The antifibrinolytic effect of polyP depends on chain length. PolyP of different chain lengths was included at 75 μM: PolyP25 (open triangles), polyP45 (filled triangles), polyP65 (open squares), polyP75 (filled circles), or no polyP (open circles). (C and D) Attenuation of fibrinolysis by polyP is TAFI dependent. (C) Assays with normal plasma were performed in the absence (open circles) or presence of 75 μM polyP75 (filled circles). Addition of 6.25 μM CPI accelerated clot lysis equally well in the absence (open triangles) and the presence of polyP75 (filled triangles). (D) Assays with TAFI-deficient plasma were performed without CPI or polyP (open circles), with 75 μM polyP75 but no CPI (filled circles), with 6.25 μM CPI but no polyP (open triangles), or with both 75 μM polyP75 and 6.25 μM CPI (filled triangles). Results in all four panels are the mean of duplicate wells (representative of three to five experiments).

Fig. 5.

Instability of polyP in plasma and serum. PolyP75 (10 mM) was incubated in human serum (squares) or heparinized plasma (circles) at 37°C. At different time points, aliquots were taken and the remaining polyP was determined after guanine isothiocyanate extraction as described (3).