Shape Change Induced in Human Platelets by Platelet-activating Factor CORRELATION WITH THE FORMATION OF PHOSPHATIDIC ACID AND PHOSPHORYLATION OF A 40,000-DALTON PROTEIN*

Washed human platelets that have been separated from plasma in the presence of prostacyclin are activated by the addition of platelet activating factor (PAF). Activation (shape change, serotonin release, and aggregation) correlates closely with the formation of phosphatidic acid and the phosphorylation of a 40,000-dalton protein. Platelet shape change, formation of phosphatidic acid, and protein phosphorylation precede aggregation and are induced at lower concentrations of PAF than those required to induce release of serotonin and platelet aggregation. Platelet shape change, formation of phosphatidic acid, and protein phosphorylation induced by PAF are not affected by trifluoperazine or indomethacin. This indicates that these responses are independent of the liberation of arachidonic acid from platelet phospholipids and the metabolism of arachi- donic acid via cyclooxygenase and lipoxygenase. These responses are, however, inhibited by prostacyclin. Platelet shape change is the first measurable physi-ologic response to platelet agonists and may be associ- ated with the stimulation of phospholipase C, inducing formation of 1,2-diacylglycerol and its phosphorylated product, phosphatidic acid. Transient formation of 1,2-diacylglycerol may also induce the specific activation of the protein kinase C that phosphorylates a 40,000-dalton protein. washed with 20 ml of prefiltered distilled water using a syringe pump (1.5 ml/min flow rate). Filters were dehydrated in a graded series of ethanol, critical point dried in a Polaron drier with liquid CO, as the transition fluid, and coated with platinum in a Polaron sputter coater (Model E5100). Platelets were examined in a JEOL JSM-35 scanning electron microscope at an accelerating voltage of 15 kV.

sequential actions of phosphatidylinositol-specific phospholipase C and 1,2-diacylglycerol-kinase (14). It has been suggested that phosphatidic acid could be an intracellular mediator of platelet activation as it is produced by a wide range of platelet stimuli and has the properties of a Ca2+ ionophore and a fusogen (1,14). Alternatively, the formation of phosphatidic acid in activated platelets may be a epiphenomenon of the generation by phospholipase C of 1,2-diacylglycerol. 1,Z-diacylglycerol has recently acquired significance in transmembrane signaling by virtue of its role as a specific activator of protein kinase C (15)(16)(17).
The present study indicates that PAF induces shape change in human platelets in parallel with the formation of phosphatidic acid and phosphorylation of a 40,000-dalton protein.
These changes depend on phospholipase C activation, precede platelet aggregation and release of serotonin, and are independent of the liberation and metabolism of arachidonic acid.

MATERIALS AND METHODS
Isolation of Human Platelets, Labeling with (32P)Orthophosphute, and Measurement of Phosphatidic Acid and Protein Phosphotylation-Materials were obtained as previously reported (1,8,18). Human platelets were isolated, labeled and stimulated as recently described (18). For measurement of protein phosphorylation, aliquots (0.1 ml) were quenched by adding 0.025 ml of 5 times concentrated Laemmli sample buffer (19) and incubat,ed a t 100 "C for 10 min. Aliquots containing 5-50 pg of protein were separated by electrophoresis through a 7.5% sodium dodecyl sulfate-polyacrylamide gel using the Laemmli buffer (19). Gels were then stained with Coomassie brilliant blue, dried, and their radioactivity determined by radioautography. Estimation of radioactivity of the 40,000-dalton protein was done by cutting out the specific areas of the gel which were placed in scintillation vials and heated for 2 h at 80 "C in 30% hydrogen peroxide. Then, scintillation fluid was added and radioactivity determined by liquid scintillation counting. For measurement of phosphatidic acid, aliquots (0.1 to 0.9 ml) were transferred into tubes containing 3.75 volumes of chloroform/methanol (1:2) for lipid extraction (18). Phosphatidic acid was separated and the release of ['Hlserotonin measured as previously reported (1, 8,18). All experiments are representative of a t least three that gave qualitatively very similar results. Results are within &lo% of the mean.
Scanning Electron Microscopy-Platelets were fixed in 0.1 M sodium phosphate buffer (pH 7.3) containing 3% glutaraldehyde. Samples were then filtered through a 0.2-pm nuclepore filter and washed with 20 ml of prefiltered distilled water using a syringe pump (1.5 ml/min flow rate). Filters were dehydrated in a graded series of ethanol, critical point dried in a Polaron drier with liquid CO, as the transition fluid, and coated with platinum in a Polaron sputter coater (Model E5100). Platelets were examined in a JEOL JSM-35 scanning electron microscope at an accelerating voltage of 15 kV.

Platelet activating Factor Induces Shape Change in Washed
Human Platelets-Platelet activating factor induces shape change in human platelets over a concentration range of 1 nM to 1 p M (Fig. 1). Scanning electron microscopy confirmed that the observed decrease in light transmission was due to platelet shape change. Fig. 2 shows the changes in platelet shape induced during a 30-s incubation of human platelets with 0.1 p~ PAF. Platelets change from smooth, discoid-shaped cells to spiny spheres with protrusion of pseudopods. The release of serotonin from platelets is observed at higher concentrations of PAF than those needed for induction of shape change, occurring only at PAF concentrations over 0.1 p~.
PAF-induced Shape Change in Human Platelets Is Associ-ated with the Formation of Phosphatidic Acid and Is Independent of the Metabolism of Arachidonic Acid-PAF-induced platelet shape change and the accompanying formation of phosphatidic acid show a similar sensitivity to inhibitors (Fig.  3). Neither response is inhibited by 50 PM trifluoperazine (a calmodulin-antagonist which inhibits phospholipases A2, Refs. 1 and 10) nor 10 P M indomethacin (an inhibitor of cyclooxygenase). Scanning electron microscopy examination also indicates that indomethacin does not prevent PAF-induced shape change (Fig. 2). On the other hand, prostacyclin (2-4 ng/ml) prevents both PAF-induced platelet shape change (Fig. 2) and the formation of phosphatidic acid (Fig. 3).
The PAF-induced formation of phosphatidic acid is also dependent on the concentration of PAF in the range 1 nM to 1 P M with a maximal effect at 0.1 PM (data not shown). At none of these PAF concentrations does trifluoperazine and indomethacin inhibit the PAF-induced formation of phosphatidic acid. However, the formation of phosphatidic acid is sensitive to prostacyclin at all PAF concentrations tested (data not shown).
I' AF Induces Simultaneous Formation of Phosphatidic Acid and Phosphorylation of a 40.000-Dalton Protein during Platelet Shape Change-Stimulation of human platelets by PAF induces shape change with simultaneous formation of phosphatidic acid and phosphorylation of a 40,000-dalton protein (Fig. 4). Both phosphorylated products reach maximal formation at about 20 s (Fig. 4).
Effect of Different Concentrations of Prostacyclin on PAFinduced Shape Change, Formation of Phosphatidic Acid, and Phosphorylation of a 40,000-Ilalton Protein-Prostacyclin (4 ng/ml) inhibits PAF-induced shape change and formation of phosphatidic acid (Fig. 3). We have further studied the effect of different concentrations of prostacyclin in an attempt to differentiate these various responses. Figs. 5 and 6 show one of three experiments that gave similar results and indicate that, within the limits of experimental error, the dependence of the various responses (shape change, formation of phosphatidic acid, and phosphorylation of a 40,000-dalton protein) on prostacyclin concentration is identical. This suggests that there is a common site of action for prostacyclin in eliciting each of these three responses.
Effect of Different Concentrations of Trifluoperazine on PAF-induced Shape Change, Formation of Phosphatidic Acid, and Phosphorylation of a 40,000-1)alton Protein-Concentra-tions of trifluoperazine ranging from 10 to 100 P M do not affect the degree of shape change, formation of phosphatidic acid, and phosphorylation of a 40,000-dalton protein (data not shown). Similarly, indomethacin (1-10 P M ) does not change the PAF-induced formation of phosphatidic acid and phosphorylation of the 40,000-dalton protein (data not shown).

DISCUSSION
Shape change, aggregation, and release of granule contents are physiological expressions of platelet activation. PAF stimulation of human platelets induces shape change which precedes the other two responses and can be studied separately.
Phosphatidic acid formation and phosphorylation of a 40,000-dalton protein occur parallel to the induction ofplatelet shape change by PAF. Formation of both these phosphorylated molecules and shape change are totally independent of the metabolism of arachidonic acid. Thus, trifluoperazine, which inhibits phospholipases of the A2 type (1, lo), does not affect PAF-induced shape change, formation of phosphatidic acid, or phosphorylation of a 40,000-dalton protein. Neither are these responses affected by indomethacin, which effectively inhibits cyclooxygenase and the consequent formation of active endoperoxides and thromboxanes. Platelet shape change and the related formation of phosphatidic acid induced by PAF are, however, inhibited by prostacyclin. Prostacyclin decreases the accumulation of phosphatidic acid in platelets by its effects on the phosphatidylinositol cycle (1,14) and this further emphasizes the interrelationship between appearance of phosphatidic acid and activation of platelets.
We have shown that there is a correlation between conditions favoring phosphatidic acid generation, protein phosphorylation, and platelet activation, suggesting that phosphatidic acid or its precursor, 1,2-diacylglycerol, may play a direct role in platelet responses. Phosphatidic acid may participate in Ca"+ gating (141, an event which is associated with activation of many cell types. Alternatively, 1,2-diacylglycerol may be related to transmembrane signaling since it is an activator of protein kinase C which, in turn, phosphorylates a 40,000dalton protein (15)(16)(17). Although both phosphorylated products well reflect platelet activation, it is not possible at the present time to determine whether one or both of these products are necessary for shape change.
We have recently studied the formation of phosphatidic acid during the shape change of human platelets induced by arachidonic acid (20) and thrombin or collagen (18). In all cases, the amount of phosphatidic acid produced is closely correlated with the degree of platelet activation. Furthermore, platelet activation was detected only when phosphatidic acid was formed. Accumulation of low and high levels of phosphatidic acid were associated with shape change and aggregation, respectively (20). In those studies, it was found that the induction of phosphatidic acid formation can follow two distinct pathways. 1) Arachidonic acid (20) and collagen (18) stimulate phospholipase C and phosphatidic acid formation via the initial production of endoperoxides or thromboxanes. 2) Thrombin, in contrast (18), stimulates phosphatidic acid formation directly by an endoperoxide-independent mechanism. The action of PAF is analogous to that of thrombin; i.e. it does not require endoperoxides or thromboxanes to trigger stimulation of phospholipase C, formation of phosphatidic acid, and platelet shape change.