Proteinase-activated receptors: structural requirements for activity, receptor cross-reactivity, and receptor selectivity of receptor-activating peptides.

We have used three distinct bioassay systems (rat aorta (RA) relaxation; rat gastric longitudinal muscle (LM) contraction; human embryonic kidney 293 (HEK293) cell calcium signal) to evaluate the activity and receptor selectivity of analogues of the receptor-activating peptides derived either from the thrombin receptor (TRAPs, based on the human receptor sequence, SFLLRNPNDK...) or the proteinase-activated receptor 2 (PAR2APs, based on the rat receptor sequence SLIGRL...). Our main focus was on the activation of PAR2 by PAR2APs and the cross-activation of PAR2 by the TRAPs. In the RA and LM assay systems, PAR2APs that were either N-acetylated (N-acetyl-SLIGRL-NH2) or had a reverse N-terminal sequence (LSIGRL-NH2) were inactive, either as agonists or antagonists. An alanine substitution at position 3 of the PAR2AP (SLAGRL-NH2) led to a dramatic reduction of biological activity, as did substitution of threonine for serine at position 1 (TLIGRL-NH2). However, alanine substitution at PAR2AP position 4 caused only a modest reduction in activity, resulting in a peptide (SLIARL-NH2) with a potency equivalent to that of the human PAR2AP, SLIGKV-NH2. The order of potency of the PAR2APs in the RA, LM, and HEK assay systems was SLIGRL-NH2 > SLIARL-NH2 > SLIGKV-NH2 > TLIGRL-NH2 > SLAGRL-NH2. In HEK cells, none of the PAR2APs activated the thrombin receptor (PAR1). However, in the HEK cell assay, the TRAP, SFLLR-NH2, activated or desensitized both PAR1 and PAR2 receptors, whereas the xenopus TRAP, TFRIFD-NH2, activated or desensitized selectively PAR1 but not PAR2. By constructing human-xenopus hybrid peptides, we found that the TRAPs, TFLLR-NH2, and SFLLFD-NH2 selectively activated the thrombin receptor in HEK cells without activating or desensitizing PAR2. In contrast, the TRAPs SFLLRD-NH2 and AFLLR-NH2 activated or desensitized both PAR1 and PAR2. The order of potency for the TRAPs in all bioassay systems was SFLLR-NH2 approximately equal to SFLLRD-NH2 approximately equal to TFLLR-NH2 > SFLLFD-NH2 > TFRIFD-NH2. We conclude that the N-terminal domain of the PAR2AP as well as positon 3 plays important roles for PAR2 activation. In contrast, the first and fifth amino acids in the TRAP motif, SFLLR-NH2, do not play a unique role in activating the thrombin receptor, but if appropriately modified can abrogate the ability of this peptide to cross-desensitize or activate PAR2, so as to be selective for PAR1. The PAR1- and PAR2-selective peptides that we have synthesized will be of use for the evaluation of the roles of the PAR1 and PAR2 receptor systems in vivo.