A Chemical Genetic Analysis of Platelet Activation Identifies An Antithrombotic Allosteric Modulator That Acts through Helix 8 of Par1

Abstract 483

The essential challenge in pharmacologic control of platelet activation during thrombus formation is to achieve inhibition of platelet mediated thrombosis without impairing hemostasis. Inhibition of platelet GPCRs has focused primarily on inhibition of orthosteric binding sites, selecting for compounds that compete with endogenous ligands. However, small molecules that act at non-orthosteric sites can stabilize alternative GPCR conformations and modify GPCR function. Such allosteric modulators demonstrate saturable binding, selective effects on affinity and efficacy, and could provide improved physiologic control of platelet function. However, cell-based screens are required for the identification of such modulators. We screened a 16,000 compound library for inhibitors of platelet activation and identified a family of alkylated quinolines that modified Par1-mediated granule secretion. Activity of these compounds was determined by the length of their alkyl tails such that a two carbon addition to the tail converted a compound that inhibited Par1 activating peptide (SFLLRN)-induced P-selectin expression into a compound that augmented Par1-mediated activation. The most potent inhibitory compound, JF5 (IC₅₀ = 4 μM), blocked Par1 activating peptide-induced GTPase activity and GTP[γ-³⁵S] binding but not GTPγS-induced platelet activation in permeabilized platelets, thus indicating activity at Par1. The compound failed to inhibit platelet activation induced by a Par4 activating peptide, the thromboxane analog U46619, arginine vasopressin, PMA, or Ca²⁺ ionophore. Studies using cultured cells overexpressing specific GPCRs demonstrated that JF5 inhibited Par1-mediated activation, but failed to inhibit activation through TP, P2Y₁ or EP3 receptors. However, JF5 did demonstrate inhibition of the α_2A-adrenergic receptor. Radioligand binding studies confirmed non-competitive inhibition of the α_2A-adrenergic receptor by JF5. Comparative phenotyping using a battery of platelet GPCR agonists demonstrated that sensitivity to JF5 was limited to select GPCRs. Evaluation of the amino acid sequence of these receptors demonstrated that sensitive receptors possessed a constrained 8^th helix, as defined by a C-terminal palmitoylation site and interactions of the 8^th helix with TM7 and the i1 loop. The 8^th helix of GPCRs is oriented parallel to the intracellular face of the plasma membrane and is thought to transduce signals to cognate G_α subunits. To assess the hypothesis that the 8^th helix contributes to the inhibitory activity of JF5, the compound was evaluated for its ability to inhibit IP₃ generation in COS7 cells overexpressing wild-type Par1 or cells expressing a Par1 mutant in which residues 376–386 had been replaced with 3 alanine residues (ΔH8 Par1). JF5 completely inhibited SFLLRN-induced IP₃ generation in cells expressing wild-type Par1. Inhibition was reversed in the ΔH8 Par1 mutant, confirming the involvement of the 8^th helix in JF5 inhibitory activity. Despite its potent inhibition of Par1-mediated aggregation, secretion, and IP₃ generation, JF5 did not inhibit SFLLRN-induced shape change. To evaluate the premise that JF5 inhibited coupling to G_αq but not G_α12/13, we determined the effect of JF5 on SFLLRN-induced G_α12-dependent decreased barrier function by measuring changes in transepithelial resistance (TER) in MDCK cells overexpressing G_α12. JF5 demonstrated no inhibition of SFLLRN-mediated decrease in TER at concentrations up to 200 μM, whereas the orthosteric Par1 inhibitor, Sch79797, completely blocked SFLLRN-induced TER at 1 μM. These results confirmed that JF5 inhibits Par1 coupling to G_αq, but not G_α12. JF was also evaluated in a mouse model of thrombus formation following laser-induced injury of cremaster arterioles. JF5 inhibited thrombus formation with an IC₅₀ of 1 mg/kg and delayed thrombus formation by 2.3-fold. Evaluation of the 8^th helix of mouse Par4 demonstrated that, unlike the 8^th helix of human Par4, it contained a palmitoylation site and is predicted to have a constrained conformation based on interactions with TM7. These results indicate a role for the 8^th helix in conferring sensitivity to small molecules and show that this sensitivity can be exploited to control platelet activation during thrombus formation. JF5 will serve as a useful probe to evaluate the role of the 8^th helix in coupling of GPCRs to cognate G_α subunits.