Thomas C. Detwiler

Inhibition of platelet secretion by an antagonist of intracellular calcium.

The role of calcium in platelet secretion has been studied using an intracellular calcium antagonist, 8-(N,N-diethylamino)-octyl 3,4,5-trimethoxybenzoate-HCL (TMB-8). TMB-8 caused a dose-dependent inhibition of both thrombin- and calcium ionophore (A23187)-induced secretion. Addition of calcium (but not Mg 2+ ) with ionophore overcame the inhibition as did simultaneous addition of thrombin and ionophore, and we conclude that TMB-8 inhibits a calcium dependent step in both thrombin- and ionophore-induced secretion, possibly by blocking release of calcium from an internal “trigger” pool. These experiments provide the most direct evidence yet that calcium is required for platelet secretion.

Inhibition of calpain in intact platelets by the thiol protease inhibitor E-64d

E-64d, a membrane permeant derivative of E-64c, a thiol protease inhibitor (Tamai et al. (1986) J. Pharmacobio-Dyn. 9, 672-677), was tested for ability to inhibit calpain activity in intact platelets. Calpain activity was measured by proteolysis of actin-binding protein and talin, two known substrates of calpain. Incubation of platelets with E-64c (not permeant) or E-64d before lysis prevented proteolysis after lysis. When the platelets were incubated with E-64c or E-64d and then washed to remove the drugs before lysis, only E-64d inhibited proteolysis. When platelets were incubated with E-64c or E-64d and then activated with A23187 plus calcium, a treatment that activates intraplatelet calpain, only E-64d inhibited proteolysis. These results indicate that E-64d can enter the intact cell and inhibit calpain.

The action of calcium-dependent protease on platelet surface glycoproteins

The action of exogenous calcium-dependent protease (CDP) on tritium-labeled surface glycoproteins was analyzed by incubation of labeled, washed human platelets with CDP partially purified from human platelets. Labeled glycoproteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by fluorography. Incubation of the labeled platelets with the protease led to a loss (calcium-dependent) from the platelets of glycoproteins Ib and V and concomitant appearance in the supernatant solution of glycocalicin (a proteolytic fragment of glycoprotein Ib), glycoprotein V, and other, unidentified glycoproteins. These changes in surface label were accompanied by alterations in three parameters of platelet function. Compared to control platelets, the CDP-treated platelets were activated by thrombin more slowly and showed less saturable and nonsaturable binding of thrombin. The CDP-treated platelets, but not the controls, aggregated on addition of fibrinogen, indicating that treatment with CDP had exposed fibrinogen receptors. The alterations in surface glycoproteins and functional parameters were compared over a 1000-fold range of CDP treatment. The decreased binding of thrombin and the exposure of fibrinogen receptors were correlated with the release of surface glycoproteins to the supernatant solution, but the slow activation by thrombin was observed under conditions where no release of labeled glycoproteins was detected (i.e., brief incubations with low concentrations of CDP). Activation of the endogenous CDP with 2.5 mM calcium chloride plus the ionophore A23187 was accompanied by hydrolysis of actin-binding protein, a known substrate, and release to the supernatant solution of labeled glycocalicin and glycoprotein V plus a faster-migrating glycoprotein not released by exogenous protease. This effect was observed in the presence of leupeptin, which completely inhibited action of exogenous protease, suggesting that platelet calcium-dependent protease may modify the platelet surface in ways that can cause alterations of platelet function.

A second form of actin: platelet microfilaments depolymerized by ATP and divalent cations.

Abstract A 45,000-dalton protein has been isolated from platelet actomyosin by sucrose gradient centrifugation or Sepharose 6-B column chromatography in the presence of Ca-ATP. It has been characterized as actin on the basis of (i) comigration with actin during gel electrophoresis in sodium dodecyl sulfate or in urea at pH 9.5 and 2.4, (ii) its amino acid composition, which includes about 1 mole of 3-methylhistidine per 45,000 g, and (iii) its ability to form 7-nm-thick microfilaments that bind rabbit muscle heavy meromyosin to produce characteristic arrowhead structures. This protein can be distinguished from other actins by the conditions for polymerization-depolymerization; it depolymerizes under conditions used to polymerize other actins and it polymerizes in the absence of ATP and divalent cations. On the basis of polymerization properties we propose that conventional actin, also present in platelets, has a direct contractile function while the second form of actin is primarily structural.

Binding of thrombin to human platelet plasma membranes

Binding of 125I-labeled thrombin to isolated human platelet plasma membranes was studied. Two classes of sites, one with high and one with low affinity for thrombin, were demonstrated. The apparent dissociation constants for the high and low affinity sites were 3.2 and 600 nM, respectively, similar to values obtained with intact platelets. Maximum binding was within 10 s, the shortest time measured, and then decreased with time to a constant level of binding within 45 s. When the equilibrium was perturbed by dilution, the system re-equilibrated with less thrombin bound than in a control that was dilute before mixing thrombin and membranes. Neither the time-dependent decrease nor the dilution effect were observed with phenylmethylsulfonyl-125I-labelled thrombin, an irreversibly inhibited thrombin, suggesting that these phenomena may involve a thrombin-catalyzed modification of the membranes leading decreased binding.

Control of energy metabolism in platelets. A comparison of aerobic and anaerobic metabolism in washed rat platelets

Abstract 1. 1. To obtain information about control of energy metabolism in platelets, washed rat platelets were incubated under oxygen or nitrogen, either without glucose or with 4 mM glucose. 2. 2. Respiration inhibited glycolysis by 75% without glucose and by 50% with glucose. Glucose inhibited respiration by 50% and stimulated glycolysis from 2- to 4-fold. 3. 3. Insulin had no effect on glucose metabolism. 4. 4. In the absence of glucose, glycogen consumed was accounted for as lactate produced under aerobic as well as anaerobic conditions. With glucose, all glucose plus glycogen consumed was not accounted for as lactate, but the stoichiometry was the same with both aerobic and anaerobic incubations. These data are interpreted as indicating oxidation of some endogenous substrate other than glucose or glycogen. 5. 5. Without glucose, the ratio ATP/ADP declined during anaerobic incubation. The ratio ATP/ADP was higher with glucose than without. 6. 6. Under aerobic conditions, oxidative phosphorylation accounted for about 55% of total ATP production with glucose and almost 90% without glucose. 7. 7. Comparison of mass action ratios with apparent equilibrium constants indicated phosphofructokinase (EC 2.7.1.11) and pyruvate kinase (EC 2.7.1.40) as potential control enzymes. Changes in levels of metabolites with altered flux indicated that phosphofructokinase was a major glycolytic control under the conditions of these experiments. 8. 8. Citric acid cycle intermediates were measured, but no control points could be identified.

Thrombin-platelet interactions: An assessment of the roles of saturable and nonsaturable binding in platelet activation☆

Thrombin binds to platelets and induces platelet activation, but the relationship of binding to activation is not clear. To better define this relationship, we have analyzed parameters of binding and activation by alpha-thrombin and by three analogous proteases that activate platelets somewhat differently. The proteases were nitro-alpha-thrombin, a derivative with nitrated tyrosine, gamma-thrombin, a product of partial proteolysis of alpha-thrombin, and trypsin, a homologous protease. Nitro-alpha-thrombin and native alpha-thrombin activated platelets similarly, whereas gamma-thrombin and trypsin activated to a slightly lesser extent than alpha-thrombin and only after a distinctive delay. alpha-Thrombin and nitro-alpha-thrombin bound to platelets to about the same extent, but only alpha-thrombin showed evidence of saturable binding. Hirudin, a thrombin inhibitor, blocked both platelet activation and saturable binding by alpha-thrombin. With nitro-alpha-thrombin, hirudin blocked platelet activation, but it had no effect on binding. gamma-Thrombin and trypsin bound less than alpha-thrombin and with no evidence of saturable binding. There were identical relationships between the total amount bound and the extent of platelet activation for the four proteases (some show no saturable binding) but distinct differences in the relationships of total amount bound and the rate of activation; similar rates of activation required the binding of three to five times more gamma-thrombin or trypsin than alpha-thrombin. That is, without saturable binding, activation was slower. These data thus show a correlation between total amount bound and extent of activation but no correlation between amount saturably bound and the extent of platelet activation. Conversely, the rate of activation is more closely correlated with saturable binding than with total binding. We conclude that high-affinity saturable binding is not essential for thrombin-induced platelet activation but that it may accelerate the reaction.

Control of energy metabolism in platelets. The effects of thrombin and cyanide on glycolysis

1. n1. The effects of thrombin on platelet glycolysis were studied by measuring levels of glycolytic intermediates. Comparisons were made with the effects of inhibition of respiration by cyanide. n n2. n2. Glycogen phosphorylase and phosphofructokinase were identified as major regulatory enzymes with both cyanide and thrombin. n n3. n3. Activation of both enzymes occurred at the same time with cyanide. With thrombin, activation of phosphorylase was observed sooner (2.5 sec after addition of thrombin) than activation of phosphofructokinase (10 sec after addition of thrombin). n n4. n4. The results are interpreted as indicating that thrombin activates phosphofructokinase allosterically by changing the energy charge of the adenine nucleotide pool as a result of increased utilization of ATP, while activation of phosphorylase is by conversion of phosphorylase b to a, probably by a Ca2+-activated phosphorylase kinase.

Regulation of platelet phosphorylase.

A sensitive fluorimetric enzyme assay was developed for study of activation of glycogen phosphorylase (EC 2.4.1.1) in intact platelets and in platelet extracts. Activity was calculated as AMP independent (activity in the absence of AMP), total (activity in the presence of 1 mM AMP), and AMP dependent (difference between AMP independent and total). The following observations were made with intact rat platelets. (1) Stimulation of platelets with thrombin caused a 7-fold increase in total activity, with increases in both AMP-dependent and AMP-independent activities. Maximum activation was obtained within 10 s after addition of thrombin. (2) The divalent cation ionophore A23187 caused a similar, though less pronounced, activation of phosphorylase. (3) Acceleration of glycogenolysis by inhibition of respiration with cyanide caused similar changes in phosphorylase activity but with the maximum effect observed only after 45 s. (4) Dibutyryl cyclic AMP had two effects; it partially activated phosphorylase and blocked further activation by thrombin, but not A23187. Similar effects were observed with human platelets, but low resting levels of phosphorylase activity could not be maintained so that changes were not as large as with rat platelets. Experiments with extracts of rat platelets gave the following results. (1) Phosphorylase activity in many extracts of non-stimulated platelets could be increased by incubation with Mg2+-ATP and Ca2+; ethyleneglycol-bis-(beta-aminoethylether)-N,N-tetraacetic acid (EGTA) partially inhibited. (2) In some extracts there was essentially no activation by incubation with Mg2+-ATP and Ca2+, but addition of cyclic AMP GAVE PARTIAL ACTIVATIon while addition of rabbit muscle phosphorylase kinase gave full activation. (3) Incubation of extracts of thrombin-stimulated platelets caused conversion of AMP-dependent to AMP-indeptndent activity. It is concluded that platelet phosphorylase exists in an inactive and two active forms. Conversion of the inactive to the active forms and of the AMP-dependent to the AMP-independent form is catalyzed by a kinase(s) that requires Ca2+ for full activity and is activated through a cyclic AMP-mediated process. The major change following physiological stimulation is an increase in both active forms, with little change in their ratio.

Thiol-disulfide exchange by thrombospondin: Evidence for a thiol and a disulfide bond protected by calcium

Thrombospondin (Tsp), a protein secreted by activated platelets, forms disulfide-linked complexes with thrombin [K. J. Danishefsky, R. J. Alexander and T. C. Detwiler (1984) Biochemistry 23, 4984]. Thiols and disulfide bonds of Tsp were analyzed, and a search was made for other Tsp covalent complexes. Platelets in 1 mM EDTA were activated with ionophore A23187, and the secreted proteins were analyzed by gel electrophoresis in sodium dodecyl sulfate. One millimolar dithioerythritol (DTE) decreased the electrophoretic mobility of Tsp, indicating reduction of an intrachain disulfide bond; Ca2+ prevented this effect. Electrophoresis of single-chain Tsp prepared with 50 mM DTE in either EDTA or Ca2+ also revealed a Ca2+-stabilized intrachain disulfide bond. Ca2+ prevented the retention of Tsp on an activated thiol-Sepharose column, indicating protection of a thiol by Ca2+. Incubation at 37 degrees C for 60 min resulted in complexes with apparent mass much greater than 500 kDa. Formation of complexes was prevented by N-ethylmaleimide, by a temperature less than 25 degrees C, and by Ca2+ or Mg2+. From pH 6 to 9, complexes formed better at lower pH. Two-dimensional (nonreduced/reduced) electrophoresis revealed Tsp but no other constituents of the complexes. With 10 nM thrombin, complexes formed faster and included thrombin; Ca2+ only partially inhibited. The complex was very susceptible to dissociation by low concentrations (2.5 mM) of DTE. It is concluded that Tsp has a reactive thiol and an intrachain disulfide bond that are protected by Ca2+. When these groups are unprotected, there is intermolecular thiol-disulfide exchange.