N-(piperidin-1-yl)-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide (SR141716; 1) is a potent and selective antagonist for the CB1 cannabinoid receptor. Using the AM1 molecular orbital method, conformational analysis of 1 around the pyrazole C3 substituent identified four distinct conformations designated Tg, Ts, Cg, and Cs. The energetic stability of these conformers followed the order Tg > Cg > Ts > Cs for the neutral (unprotonated) form of 1 and Ts > Tg > Cs > Cg for its piperidine N-protonated form. Unified pharmacophore models for the CB1 receptor ligands were developed by incorporating the protonated form of 1 into the superimposition model for the cannabinoid agonists 4-[4-(1,1-dimethylheptyl)-2-hydroxyphenyl]perhydro-2alpha,6beta-dihydroxynaphthalene (CP55244; 2) and the protonated form of (R)-[2,3-dihydro-5-methyl-3-[(4-morpholinyl)methyl]pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl](1-naphthalenyl)methanone (WIN55212-2; 3) reported previously (Shim et al. In Rational Drug Design Symposium Series; Parrill, A. L., Reddy, M. R., Eds.; American Chemical Society: Washington, DC, 1999; pp 165-184). Values of K(i) for 1 and a series of 31 structural analogues were determined from radioligand binding analyses by competitive displacement of [3H]CP55940 from cannabinoid receptors in a rat brain membrane preparation. Comparative molecular field analysis (CoMFA) was employed to construct three-dimensional (3D)-quantitative structure-activity relationship (QSAR) models for this data set as unprotonated species assuming the Tg, Cg, and Ts conformers and for the protonated species assuming the Ts, Tg, and Cs conformers. Values of the conventional r2 and cross-validated r2 (r(cv)2) associated with these CoMFA models exceeded the threshold for statistical robustness (r2 > or = 0.90) and internal predictive ability (r(cv)2 > or = 0.50) in each of these six cases except for the protonated species assuming the Tg conformer (i.e., r2 = 0.97; r(cv)2 = 0.36). Results from conformational analyses, superimposition models, and 3D-QSAR models suggest that the N1 aromatic ring moiety of 1 dominates the steric binding interaction with the receptor in much the same way as does the C3 alkyl side chain of cannabinoid agonists and the C3 aroyl ring of the aminoalkylindole agonists. We also determined that several of the conformers considered in this study possess the proper spatial orientation and distinct electrostatic character to bind to the CB1 receptor. We propose that the unique region in space occupied by the C5 aromatic ring of 1 might contribute to conferring antagonist activity. We further propose that the pyrazole C3 substituent of 1 might contribute to conferring either neutral antagonist or inverse agonist activity, depending upon the interaction with the receptor.