Human platelets release platelet-derived growth factor (PDGF) from α-granules during platelet activation. We have previously shown that platelets have PDGF α-receptors, a transmembrane tyrosine kinase that takes part in negative feedback regulation during platelet activation. Here we have described a study of PDGF-induced tyrosine phosphorylation of platelet substrates and phosphoinositide 3-kinase (PI-3K) activity in collagen-stimulated platelets. By immunoblotting with phosphotyrosine antibodies of collagen-activated platelets we found that PDGF increased the phosphorylation of several platelet substrates, e.g. pp140, pp120 and pp85. PDGF inhibited collagen-induced platelet activation in the presence of inhibitors of autocrine stimulation, thus blocking the pure collagen-induced signal transduction. PDGF enhanced the collagen-induced formation of PtdIns(3,4) P 2 and PtdIns(3,4,5) P 3 as measured by HPLC. Wortmannin and LY294002, two unrelated inhibitors of PI-3K, were used to investigate the role of PI-3K in PDGF-induced platelet signalling. Incubation of platelets with wortmannin and LY294002 blocked the formation of three phosphorylated inositides as well as the inhibitory effect of PDGF on collagen-induced platelet activation. We conclude that the inhibitory effect of PDGF on platelet activation is PI-3K dependent. This is the first demonstration of a negative regulatory function of 3-phosphorylated inositides in platelets.

Previous studies have indicated different energy requirements for some platelet responses; these differences could, however, be due to inadequate methodology and differences in platelet preparation. The present study describes the effect of decreasing ATP availability on seven platelet responses measured in gel-filtered human platelets. The cells, prelabelled with 5-hydroxy[ 3 H]tryptamine, [ 3 H]- or [ 14 C]adenine, [ 32 P]P i or [ 3 H]arachidonate, were incubated with antimycin A and 2-deoxy- d -glucose. Platelet responses induced by thrombin and collagen (secretion only), level of metabolic ATP and the adenylate energy charge (AEC) were determined at various times during incubation. Platelet aggregation was rapidly inhibited after a lag of 5–15 min and with 50% inhibition at AEC = 0.55–0.60. Secretion of 5-hydroxy[ 14 C]tryptamine and ATP + ADP from dense granules and of fibrinogen and β-thromboglobin from α-granules were inhibited in parallel, without a lag and with 50% inhibition at AEC = 0.65–0.70. The inhibition of secretion of platelet factor 4 from the α-granules followed another pattern with 50% inhibition at AEC = 0.70–0.80. Breakdown of [ 3 H]-phosphatidylinositol, formation of [ 3 H]- and [ 32 P]-phosphatidate, liberation of [ 3 H]arachidonate and secretion of acid hydrolases were inhibited in parallel and inhibition was present at the start of incubation with 50% inhibition at AEC = 0.80–0.87. These results suggest that the responses have different energy requirements, increasing in the order: aggregation < dense granule and α-granule secretion < acid hydrolase secretion, phosphatidylinositol breakdown, phosphatidate formation and arachidonate liberation. The powerful inhibition of phosphatidylinositol breakdown by metabolic inhibitors suggests that energy-requiring steps are involved in the activation of phospholipase C.

1. Shape change, aggregation and secretion of dense-granule constituents in platelets differ in their dependence on cellular energy metabolism. The possibility that such a difference also exists between secretion of dense-granule constituents and acid hydrolases was investigated. 2. Human platelets were incubated with [ 14 C]adenine in plasma, and then washed and resuspended in salt solutions. The effects of incubating the cells with antimycin A and 2-deoxyglucose on the concentrations of [ 14 C]ATP, ADP, AMP, IMP and inosine plus hypoxanthine and on thrombin-induced secretion of ATP plus ADP and acid hydrolases were studied. The metabolic inhibitors only affected 14 C-labelled nucleotides, whereas thrombin only liberated unlabelled ATP and ADP. 3. The extent of secretion decreased progressively with time during incubation with the metabolic inhibitors. At any time the secretion of acid hydrolases, β- N -acetylglucosaminidase, β-glucuronidase and β-galactosidase was inhibited to a greater extent than secretion of ATP plus ADP (dense-granule secretion). 4. Incubation with the metabolic inhibitors shifted the log (dose)–response relationship to higher thrombin concentrations, and with a greater shift for acid hydrolase secretion than for dense-granule secretion. 5. Antimycin, when present alone, caused a marked decrease in the rate of acid hydrolase secretion, but had no effect on dense-granule secretion. 6. These results further support the view that acid hydrolase secretion and dense-granule secretion are separate processes with different requirements for ATP energy. Acid hydrolase secretion, but not dense-granule secretion, appears to depend on a simultaneous rapid generation of ATP, which can be accomplished by oxidative, but not by glycolytic, ATP production.

1. Platelets containing adenine nucleotides labelled with 3 H and 14 C in vitro were aggregated biphasically with ADP and adrenaline. Amounts of ATP and ADP as well as the radioactivity of ATP, ADP, AMP, IMP, hypoxanthine and adenine were determined in platelets and plasma at different stages of aggregation. 2. ATP and ADP were released during the second aggregation phase and had a low specific radioactivity compared with the ATP and ADP retained by the cells. The specific radioactivity of intracellular nucleotides increased during release. The parameters observed with ADP and adrenaline as release inducers were the same as for collagen and thrombin. 3. Release induced by all four inducers was accompanied by conversion of cellular [ 3 H]ATP into extracellular [ 3 H]-hypoxanthine. By variation of temperature, inducer concentration, time after blood withdrawal and use of acetylsalicylic acid, the aggregation pattern caused by adrenaline and ADP could be made mono- or bi-phasic. Release or second-phase aggregation was intimately connected with the ATP–hypoxanthine conversion, whereas first phase aggregation was not. 4. The [ 3 H]ATP–hypoxanthine conversion started immediately after ADP addition. With adrenaline it usually started with the appearance of the second aggregation phase. The conversion was present during first phase of ADP-induced aggregation only if a second phase were to follow. 5. When secondary aggregation took place while radioactive adenine was being taken up by the platelets, increased formation of labelled hypoxanthine still occurred, but there was either no change or an increase in the concentration of labelled ATP. 6. Biphasically aggregated platelets converted [ 3 H]adenine more rapidly into [ 3 H]-ATP and -hypoxanthine than non-aggregated platelets. Addition of [ 3 H]adenine at different stages of biphasic aggregation showed that more [ 3 H]hypoxanthine was formed during than after the release step. 7. We conclude that ADP and adrenaline, like thrombin and collagen, cause extrusion of non-metabolic granula-located platelet adenine nucleotides. During release metabolic ATP breaks down to hypoxanthine, and this process might reflect an ATP-requiring part of the release reaction.

1. The effects of ATP, PP i and EDTA on the skeletal-muscle pyruvate kinase reaction at various concentrations of magnesium (where ‘magnesium’ refers to total Mg 2+ , both free and in the form of complexes) were investigated. The reaction rate was determined as the amount of pyruvate formed in a recorded time of incubation. 2. At 44m m -magnesium the K m values for ADP and phosphoenolpyruvate were unaltered by the presence of ATP up to 6·8m m in systems buffered with either tris–hydrochloric acid or glycylglycine–sodium hydroxide, but the K m values were different in these systems. The K m for one substrate was independent of the concentration of the second substrate. 3. At 10m m -magnesium in the tris–hydrochloric acid system ATP inhibited the reaction competitively with respect to ADP and phosphoenolpyruvate. In the glycylglycine–sodium hydroxide system the inhibition appeared to be non-competitive. At 10m m -magnesium the K m values were lower than at 44m m -magnesium and dependent on the system used. 4. In the tris–hydrochloric acid system the reaction rate rose with increasing magnesium concentration up to a maximum at a concentration 10–20 times that of ADP. Further increase inhibited the reaction and at 44m m -magnesium the rate was 25–50% of its maximum. This inhibition paralleled that produced by increasing trimethylammonium chloride concentrations and was not due to a specific effect of the Mg 2+ ion. 5. In the presence of 6·8m m -ATP no reaction occurred below 4–6m m -magnesium, and further increase apparently abolished the inhibition as the reaction rate increased and became equal to those obtained in the absence of ATP at 10–25m m -magnesium. Further increase in magnesium concentration gave reaction rates that were slightly higher in the presence of ATP than in its absence. The maximal rate in the presence of ATP was distinctly lower than in its absence. When 6·8m m -PP i or 6·8m m -EDTA was present the variations in reaction rate with rising magnesium concentration were similar to that obtained in the presence of ATP below 6–8m m -magnesium but further increase in the magnesium concentration resulted in an increase in the rate up to a maximum comparable with that of the control. The effect of pure chelation was thus a displacement of the reaction maximum to higher magnesium concentrations without changing the maximal rate. When correction had been made for this effect, ATP gave inhibition at 44m m -magnesium that was competitive with respect to ADP ( K i 2·1×10 −2 m ). This degree of inhibition is far less than was reported earlier and its importance for the mechanism of the pyruvate kinase reaction is discussed.