Influence of the N-1 alkyl chain length of cannabimimetic indoles upon CB₁ and CB₂ receptor binding

Abstract

The N-1 alkyl side chain of the aminoalkylindole analogues (AAI) has been implicated as one of a three-point interaction with the cannabinoid CB₁ receptor. In this study, the morpholinoethyl of WIN 55,212-2 was replaced with carbon chains of varying lengths ranging from a methyl to heptyl group. Additional groups were added to the naphthoyl and the C2 positions of the molecule. These structural changes revealed that high affinity binding to the CB₁ and CB₂ receptors requires an alkyl chain length of at least three carbons with optimum binding to both receptors occurring with a five carbon side chain. An alkyl chain of 3–6 carbons is sufficient for high affinity binding; however, extension of the chain to a heptyl group results in a dramatic decrease in binding at both receptors. The unique structure of the cannabimimetic indoles provides a useful tool to define the ligand-receptor interaction at both cannabinoid receptors and to refine proposed pharmacophore models.

Introduction

In the 35 years since the isolation and elucidation of the structure of Δ⁹-THC (Gaoni and Mechoulam, 1964) considerable effort has gone into modifying the structure of cannabinoids as well as in developing compounds structurally diverse from the classical tricyclic structures. Several nontraditional cannabinoids have been discovered, including a series of 3-arylcyclohexanols such as CP-55 940, analogues of the endogenous ligand anandamide, and various aminoalkylindole (AAI) compounds (D’Ambra et al., 1992, Melvin et al., 1995; for review see Martin et al., 1995, Khanolkar and Makriyannis, 1999). According to the three-point interaction receptor model (Fig. 1), the hydroxyl group at C-1, the lipophilic side chain at C-3 and the orientation of the C-9 substituent (Edery et al., 1971, Binder and Franke, 1982, Razdan, 1986, Thomas et al., 1991) are essential to the bioactivity of Δ⁹-THC. For purposes of aligning the prototypical AAI, WIN 55,212-2, and Δ⁹-THC in a common pharmacophore, it has been proposed that the corresponding overlapping regions of the two respective molecules are the naphthyl ring and the cyclohexene ring, the carbonyl oxygen and the phenolic hydroxyl, and the morpholine unit and the C-3 pentyl side chain as depicted in Fig. 1 (Huffman et al., 1994).

Although WIN 55,212-2, related AAIs and other indole-derived compounds bear no obvious structural similarities to traditional cannabinoids, they have been shown to bind to brain cannabinoid receptors (CB₁). The AAIs have been shown to produce a profile of behavioral effects characteristic of those observed with Δ⁹-THC and other classical and bicyclic cannabinoids that include antinociception, ring immobility, suppression of spontaneous activity, and hypothermia in mice (Compton et al., 1992). These pharmacological effects of Δ⁹-THC and WIN 55,212-2 are blocked by the CB₁ receptor antagonist, SR 141716A (Rinaldi-Carmona et al., 1994, Perio et al., 1996). Cannabimimetic indoles are also associated with potent activity at inhibiting electrically induced contractions of the mouse vas deferens (Pertwee et al., 1995). In autoradiographic studies, the distribution of AAI binding sites was similar to that reported for classically identified cannabinoid binding sites (Jansen et al., 1992).

The cDNA corresponding to the central (CB₁) receptor was cloned, isolated, and identified as a member of the G-protein linked super family of receptors (Matsuda et al., 1990). In vitro labeling of sections of the adult human brain with [³H] CP-55 940, a high affinity ligand, followed by quantitative receptor autoradiography revealed a heterogenous distribution of cannabinoid receptors throughout the brain (Herkenham et al., 1990, Devane et al., 1992). The cloning and isolation of the CB₂ receptor from cDNA of the human promyelocytic leukemic HL60 cells by Munro et al. (1993) was followed by the determination of the distribution of these receptors in the cells of the immune system. The protein product from this clone showed 44% amino acid identity to the human CB₁ receptor with the degree of identity rising to 68% for the transmembrane regions thought to be involved in ligand specificity (Munro et al., 1993). Transfection of the cDNA expression vector into Chinese hamster ovary cells and consequent binding assays allows the determination of ligands which are selective for CB₁ or CB₂ receptors (Munro et al., 1993, Huffman et al., 1996, Showalter et al., 1996).

The discovery that cannabinoids exert their pharmacological actions via at least two types of receptors, along with the discovery of endogenous cannabinoid ligands, prompted the development of a new generation of cannabimimetic analogues for probing differences in the pharmacophore of these receptor subtypes. These analogues are essential in elucidating the physiological and pharmacological role of cannabinoid receptors as well as the different structural features necessary for binding to either the CB₁ or the CB₂ receptor. In previous studies, we demonstrated that elimination of the oxygen bridge and aminoalkyl groups in WIN 55,212-2 resulted in indoles that were equally effective as cannabinoid agonists (Huffman et al., 1994, Pertwee et al., 1995, Wiley et al., 1998). The objective of the present investigation was to characterize the consequences of selected structural alterations at key positions of indoles and to determine whether receptor subtype selectivity could be achieved.