Synthesis and biological evaluation of novel ligands for the cannabinoid-like orphan G protein-coupled receptors GPR18 and GPR55

Abstract

GPR18 and GPR55 belong to the more than 100 orphan G protein-coupled receptors (GPCRs) whose natural signaling molecules remain unidentified. Both orphan GPCRs have some relations with the cannabinoid receptors CB₁ and CB₂. For example, Δ⁹-tetrahydrocannabinol (THC), the cannabinoid receptor agonist found in <em>Cannabis sativa</em>, can stimulate GPR18 and block the activation of GPR55. However, GPR18 and GPR55 have only low sequence similarity with CB receptors. mRNA encoding GPR18 is present mainly in cells and tissues of the immune system as well as in cancer. While GPR55 is expressed widely in the brain and in metabolically active tissues. In fact, it has been shown that GPR18 and GPR55 are involved in various pathophysiological processes, and both receptors may have potential as future drug targets. Physiological lipids have been postulated to activate the receptors, <em>N</em>-arachidonoylglycine (NAGly) for GPR18, and 1-lysophosphatidylinositol (1-LPI) for GPR55, but these have not been unambiguously confirmed so far. <br /><strong>Phenothiazine derivatives<br /></strong>Compounds of this series were derived from the 2-chlorophenothiazine scaffold. This compound came into focus after the discovery of antagonistic activities of perphenazine and related antipsychotics at GPR18 that were observed in a screening approach of a library of approved drugs. At first, the basic propylamino group, a typical feature of these drugs, was replaced with the purpose to develop GPR18 antagonists without inhibitory activity at dopamine or other neurotransmitter receptors. Interestingly, these compounds lost their effect at GPR18 but were active as inhibitors of GPR55. The most potent phenothiazine-derived GPR55 antagonist was 2-chloro-10-(3-phenylpropyl)-10<em>H</em>-phenothiazine (<strong>40</strong>, IC₅₀ = 5.29 μM). However, by exchange of the benzene ring in <strong>40</strong> for a primary amino group, the selectivity was shifted towards GPR18. Indeed, 2-chloro-10-(3-aminopropyl)-10<em>H</em>-phenothiazine (<strong>41</strong>, IC₅₀ = 4.89 μM) was the only derivative in this series capable of inhibiting the activation of GPR18 while being inactive at GPR55. Therefore, it could be assumed that an amino group was required for GPR18 antagonism. <br /><strong><em>N</em>-Acylamino acids<br /></strong>Several different <em>N</em>-acylamino acids have been found in mammalian brains as well as the periphery, and a few have been shown to possess biological effects. Via acylation and amide coupling methods, various conjugates of amino acids and natural fatty acids were prepared. Among the four amino acids employed in the present study, tryptophan as a constituent of the prepared <em>N</em>-acylamino acids appeared to be an important moiety for inhibiting GPR18 activation. Furthermore, arachidonic acid also appeared to be a well-tolerated residue by GPR18 with <em>N</em>-arachidonoyl-L-tryptophan methyl ester (<strong>97</strong>, IC₅₀ = 5.27 µM) being the most potent conjugate. However, this compound also showed binding affinities to both CB receptor subtypes in competition binding assays. Hydrolysis of the ester group of <strong>97</strong> eliminated binding to CB₁ and CB₂, and slightly decreased the potency at GPR18 (<em>N</em>-arachidonoyl-L-tryptophan <strong>103</strong>, IC₅₀ = 9.05 µM). Thus, it was assumed that the carboxyl group plays an important role in the selectivity of <em>N</em>-acyltryptophan derivatives at cannabinoid-like receptors. <br /> The structure-activity relationships provided herein will serve as a basis for further development of more potent and selective antagonists, which will be useful as pharmacological tools for elucidating the pathophysiological roles of these two orphan GPCRs and to validate them as future drug targets.