Journal of Computer-Aided Molecular Design, 11 (1997) 278–292. *To whom correspondence should be addressed. © 1997 Kluwer Academic Publishers. Printed in The Netherlands. J-CAMD 393 Structure–activity relationships of cannabinoids: A joint CoMFA and pseudoreceptor modelling study Silke Schmetzer a , Paulette Greenidge b , Karl-Artur Kovar a, *, Meike Schulze-Alexandru a and Gerd Folkers b a Institute of Pharmacy, University of Tübingen, Auf der Morgenstelle 8, D-72076 Tübingen, Germany b Department of Pharmacy, ETH Zürich, CH-8057 Zürich, Switzerland Received 6 May 1996 Accepted 23 December 1996 Keywords: ∆ 9 -Tetrahydrocannabinol; Pharmacophore; Molecular modelling; CADD; YAK Summary A cannabinoid pseudoreceptor model for the CB1-receptor has been constructed for 31 cannabinoids using the molecular modelling software YAK. Additionally, two CoMFA studies were performed on these ligands, the first of which was conducted prior to the building of the pseudoreceptor. Its pharma- cophore is identical with the initial superposition of ligands used for pseudoreceptor construction. In contrast, the ligand alignment for the second CoMFA study was taken directly from the final canna- binoid pseudoreceptor model. This altered alignment gives markedly improved cross-validated r 2 values as compared to those obtained from the original alignment with r 2 cross values of 0.79 and 0.63, respective- ly, for five components. However, the pharmacophore alignment has the better predictive ability. Both the CoMFA and pseudoreceptor methods predict the free energy of binding of test ligands well. Introduction Cannabis preparations, such as marihuana and hash- ish, have been used for centuries on account of both their psychotropic and pharmacological effects. (6aR,10aR)-∆ 9 - Tetrahydrocannabinol (1, ∆ 9 -THC, Fig. 1), the psycho- tropic component of cannabis preparations [1], also ex- hibits anti-emetic, analgesic, muscle-relaxing, blood-pres- sure-reducing, bronchodilating effects and reduces patho- logically elevated intraocular pressure (glaucoma) [2–6]. In all probability, the analgesic and the psychotropic effects are mediated through the central cannabinoid receptor (CB1); however, almost nothing is known, as yet, concerning signal transduction pathways of the other pharmacological effects [2,7,8]. The pharmacological effects have led to the synthesis of other cannabinoids. The synthetic analogues are referred to as non-classical cannabinoids and differ from the classical cannabinoids in the absence of the tetrahydropyran ring [9,10]. The essential structural prerequisites for ligand–receptor inter- action of the cannabinoids are the free phenolic hydroxyl group [11] and the alkyl side chain at ring A [2]. Exten- sion and branching of the pentyl side chain of ∆ 9 -THC yields the dimethylheptyl (DMH) derivative with a greatly increased affinity for the cannabinoid receptors [2,12]. Within the non-classical cannabinoids, increasing the length of the DMH side chain to 9–11 C atoms and shor- tening it to ≤6 C atoms leads to a reduction of the affin- ity [13]. While an alcoholic hydroxyl group at C-9 or C- 11 is not essential, although it increases the activity in the case of classical cannabinoids, structure–activity relation- ship studies reveal that the hydroxyl group at ring C is essential for the analgesic activity in the case of non-clas- sical cannabinoids [14]. Furthermore, Reggio et al. [15] have reported a receptor essential volume near the top of the C ring in the bottom face of the molecule. Although the pharmacological activity and the structure–effect relationships are well documented [14,16], almost nothing is known about the ligand–receptor interaction at the molecular level. The identification [17] and cloning [18] of the cannabinoid receptor in the brain (CB1) and the characterization of the peripheral cannabinoid receptor (CB2) [19] have provided a new impulse for the investiga- tion of the molecular basis of the ligand–receptor interac-