Modulation of opioid receptor affinity and efficacy via N-substitution of 9β-hydroxy-5-(3-hydroxyphenyl)morphan: Synthesis and computer simulation study.

The enantiomers of a variety of N-alkyl-, N-aralkyl-, and N-cyclopropylalkyl-9β-hydroxy-5-(3-hydroxyphenyl)morphans were synthesized employing cyanogen bromide and KCO to improve the original N-demethylation procedure. Their binding affinity to the μ-, δ-, and κ-opioid receptors (ORs) was determined and functional (GTPS) assays were carried out on those with reasonable affinity. The 1R,5R,9S-enantiomers (1R,5R,9S)-(-)-5-(3-hydroxyphenyl)-2-(4-nitrophenethyl)-2-azabicyclo[3.3.1]nonan-9-ol (1R,5R,9S-16), (1R,5R,9S)-(-) 2-cinnamyl-5-(3-hydroxyphenyl)-2-azabicyclo[3.3.1]nonan-9-ol (1R,5R,9S-20), and (1R,5R,9S)-(-)-5-(3-hydroxyphenyl)-2-(4-(trifluoromethyl)phenethyl)-2-azabicyclo[3.3.1]nonan-9-ol (1R,5R,9S-15), had high affinity for the μ-opioid receptor (e.g., 1R,5R,9S-16: Ki=0.073, 0.74, and 1.99nM, respectively). The 1R,5R,9S-16 and 1R,5R,9S-15 were full, high efficacy μ-agonists (EC=0.74 and 18.5nM, respectively) and the former was found to be a partial agonist at δ-OR and an antagonist at κ-OR, while the latter was a partial agonist at δ-OR and κ-OR in the GTPS assay. The enantiomer of 1R,5R,9S-16, (+)-1S,5S,9R-16 was unusual, it had good affinity for the μ-OR (Ki=26.5nM) and was an efficacious μ-antagonist (Ke=29.1nM). Molecular dynamics simulations of the μ-OR were carried out with the 1R,5R,9S-16 μ-agonist and the previously synthesized (1R,5R,9S)-(-)-5-(9-hydroxy-5-(3-hydroxyphenyl-2-phenylethyl)-2-azabicyclo[3.3.1]nonane (1R,5R,9S-(-)-NIH 11289) to provide a structural basis for the observed high affinities and efficacies. The critical roles of both the 9β-OH and the p-nitro group are elucidated, with the latter forming direct, persistent hydrogen bonds with residues deep in the binding cavity, and the former interacting with specific residues via highly structured water bridges.