Cathepsin G is a neutrophil-derived protease that has been shown to inhibit the effects of thrombin on some cells expressing thrombin receptors while acting as an agonist on others. The present studies examine whether cleavage of the thrombin receptor by cathepsin G can account for these diverse effects. When added to cells that normally respond to thrombin, cathepsin G prevented a subsequent cytosolic Ca2+ increase caused by thrombin, but had no effect on responses to the thrombin receptor agonist peptide, SFLLRN. These effects were observed in cells in which cathepsin G had little or no agonist effect (human umbilical vein endothelial cells and HEL cells), as well as in cells in which cathepsin G acted as an agonist (platelets and CHRF-288 cells). Binding studies using monoclonal antibodies with defined epitopes within the first 60 residues of the thrombin receptor N-terminus showed that incubation of platelets and endothelial cells with cathepsin G abolished the binding of all of the antibodies, while thrombin abolished only the binding of antibodies whose epitopes were N-terminal to the known thrombin cleavage site between Arg41 and Ser42. Analysis of peptide proteolytic fragments identified three potential cleavage sites for cathepsin G: Arg41-Ser42, Phe43-Leu44, and Phe55-Trp56. Cleavage at Phe55-Trp56 would account for both the observed loss of all of the antibody binding sites and the inhibition of receptor activation by thrombin. Two approaches were used to determine whether a solitary cleavage at Arg41-Ser42 could result in receptor activation. In the first, HEL cells were exposed to cathepsin G or thrombin in the presence of an antibody whose epitope includes Phe55. The antibody inhibited responses to thrombin, but augmented the response to cathepsin G. In the second, COS-1 cells were transfected with variant thrombin receptors in which Phe55 and Trp56 were mutated to alanine. Transfected wild-type receptors responded to thrombin, but not cathepsin G, while the variant receptors responded to both proteases. These results 1) suggest that the ability of cathepsin G to inhibit responses to thrombin, but not SFLLRN, is due to cleavage of the receptor at Phe55-Trp56, deleting the tethered ligand domain, and 2) show that cathepsin G can activate thrombin receptors, but only if the cleavage site at Phe55-Trp56 is mutated or otherwise protected.(ABSTRACT TRUNCATED AT 400 WORDS)