Abstract We previously postulated that thrombin might initiate platelet aggregation and the release reaction by acting at the platelet surface. We have now demonstrated surface binding of thrombin using highly purified bovine thrombin labeled with 125I. At least two classes of binding sites have been demonstrated. When studies are done at relatively high thrombin concentrations, platelets bind 50,000 thrombin molecules per platelet at saturation with an apparent dissociation constant of 2.9 units per ml (30 nm thrombin). Binding studies carried out at lower thrombin concentrations disclose that approximately 500 thrombin molecules bind per platelet with an apparent dissociation constant of 0.02 unit per ml (0.21 nm thrombin). The other properties of these two types of thrombin-binding sites were identical. Diisopropyl fluorophosphate-treated thrombin (DIP-thrombin) binds to platelets with equal affinity and competes for the same sites as native thrombin, even though DIP-thrombin does not cause platelet aggregation or the release reaction. Experiments using high resolution electron microscopic autoradiography established that 125I-thrombin binds to the cell surface. The mean position of 400 grains was 490 ± 90 A (S.E.M.) external to the platelet plasma membrane, suggesting that bound thrombin is located in the glycoprotein coat of the platelet surface. Similar localization was found for 125I-lentil phytohemagglutinin, which is known to bind to a platelet surface oligosaccharide. Addition of excess unlabeled thrombin rapidly displaced g95% of 125I-thrombin previously bound to platelets, further suggesting that thrombin does not enter the cell. Although DIP-thrombin competitively inhibits binding of native thrombin, DIP-thrombin increases the extent of 14Cserotonin release when platelets are exposed to suboptimal concentrations of native thrombin. These findings suggest that thrombin induces the release reaction by a complicated mechanism involving both binding and proteolytic activities.
Tollefsen et al. (1974) studied this question.