Effects Of Biased PAR1 Ligands On Platelets and Endothelial Cells

Protease-activated receptor 1 (PAR1) is a widely expressed G protein coupled receptor (GPCR) that monitors the extracellular protease environment with high sensitivity. Activation of PAR1 by thrombin on platelets links the coagulation cascade to platelet activation and contributes to thrombus formation. Activation of endothelial cell PAR1 by thrombin results in loss of barrier function and apoptosis. In contrast, activation of PAR1 by activated protein C (APC) stimulates a genetic program that protects endothelial cells from barrier dysfunction and apoptosis induced by chemokines or cytotoxins. To identify new strategies for modulating PAR1 activity, we screened a 300,000 compound library in a platelet-based assay to identify compounds that inhibit PAR1-mediated activation. Four structurally unrelated compounds, termed parmodulins, inhibited platelet activation mediated through PAR1 without inhibiting platelet stimulation through other receptors including PAR4. We compared the activity of parmodulins to three known orthosteric inhibitors (i.e., compounds that act at the ligand binding site) of PAR1, including SCH79797, BMS-200261, and FR-171113. While orthosteric inhibitors blocked both platelet aggregation and platelet shape change induced by the PAR1 agonist SFLLRN, parmodulins inhibited PAR1-mediated aggregation, but spared shape change. PAR1 inhibition by orthosteric inhibitors was irreversible, while that of parmodulins was readily reversible upon washing. Orthosteric inhibitors had no activity against mouse platelets stimulated with the PAR4 agonist AYPGKF. In contrast, parmodulins inhibited activation of mouse platelets stimulated with AYPGKF even though they failed to inhibit activation of human platelets by AYPGKF. Since mouse PAR4 (mPAR4) and human PAR4 (hPAR4) differ in their cytoplasmic tail distal to the DPxxY motif, we evaluated the activity of all compounds in COS7 cells transfected with wild-type hPAR1 or with a hPAR1/hPAR4 chimera in which the cytoplasmic tail of hPAR1 was replaced with that of hPAR4. All compounds inhibited stimulation of [Ca2+]i flux mediated through wild-type hPAR1. However, only orthosteric inhibitors blocked the hPAR1/hPAR4 chimera, indicating that parmodulins act at the cytosolic tail of hPAR1. When tested in human umbilical vein endothelial cells (HUVECs), all orthosteric PAR1 inhibitors elicited apoptosis as detected by YEO-PRO-1 staining by approximately as much as TNF-a. In contrast, parmodulins failed to elicit apoptosis. In fact, incubation with parmodulins protected HUVECs from apoptosis induced by TNF-a, thrombin, or staurosporine. Parmodulin-mediated protection from apoptosis was completely reversed in HUVECs in which PAR1 had been knocked down using siRNA. When tested in vivo, infusion of 5 mg/kg parmodulin 2 resulted in a 72% reduction in platelet accumulation at sites of laser-induced injury in cremaster arterioles (P<0.05). In contrast, parmodulin 2 had no effect on bleeding times in a tail snip assay. Recent clinical trials of orthosteric PAR1 inhibitors as antiplatelet agents have shown that these compounds are associated with increased major bleeding. Our results demonstrate that this class of PAR1 inhibitors induces apoptosis in endothelium. In contrast, PAR1 ligands that target the cytosolic face of the receptor are cytoprotective in endothelium. This new class of biased PAR1 ligands block prothrombotic platelet activation, but simultaneously induce cytoprotective signaling in endothelial cells. The strategy of targeting the cytosolic face of receptors to achieve biased signaling could be broadly applicable to GPCRs.