Farmacologiche Mario Negri, Consorzio Mario Negri Sud, Santa Maria lmbaro, ltaly. lntroduction nus near the first transmembrane domain (4, 3L, 57). In keep-ing with this model, numerous studies in a variety of cells have now shown that synthetic peptides beginning with the first 5 residues of the tethered ligand domain (SFLLR in the human receptor) can mimic the effects of thrombin, presumably by forcing the same conformational changes in the body of the receptor that are needed to promote guanine nucleotide exchange on G proteins located at the cytoplasmic face of the receptor. Implicit in this model is the conclusion that thrombin's role in receptor activation is limited to cleavage of the N-terminus and, perhaps, facilitating docking of the tethered ligand domain in the body of the receptor. This issue has not, however, been fully addressed. The identification of the thrombin receptor brought ques-tions about the existence of other receptors with the same mechanism of activation and about the ability of the cloned receptor to'account for all of the effects of thrombin on human platelets. Even though the thrombin receptor agonist peptides (or TRAPs) are much less potent on a molar basis than the peptide agonists for other G-protein-coupled receptors, their apparent selectivity for the thrombin receptor meant that for several years after the first description of the thrombin receptor, a peptide response in a particular cell was equated with the presence of the receptor. However, it is now clear that some of the peptides originally thought to be thrombin receptor specific will also activate PAR-2, includ-ing the prototypical SFLLRN (8, 61, 70). Furthermore, several studies have shown that the agonist peptides may not evoke all of the responses seen with thrombin, or even neces-sarily the same pattern of responses as those which are seen with thrombin (22, 45, 41, 48, 79, 85). This suggests that other thrombin response elements exist, a conclusion supported by the recent observation that deletion of the mouse homologue of the cloned thrombin receptor has no obvious effect on the response of mouse platelets to thrombin even though it does eliminate thrombin responses in other types of cells (21, 23). The nature of this additional receptor is unknown. It is also unknown whether an additional thrombin receptor is present on human platelets. Even before the knockout studies were completed, there was reason to believe that there might be more than one thrombin receptor in some animal species including rats where agonist peptides based on the autologous thrombin receptor sequence tail to cause platelet activation (22, 25). In contrast, human platelets

Platelets and endothelial cells exist in a protease-filled environment, particularly following tissue injury or inflam-mation. One such protease is thrombin, but others exist as well, including other coagulant enzymes, protein C, plasmin and urokinase. In addition, proteases such as cathepsin G, the granzymes and tryptase can be secreted locally by peripatetic leukocytes and mast cells. In some cases the mechanism by which extracellular proteases affect intracel-lular events are fairly well understood. In other cases, very