Lisa A. Bonner

Based on the structure of the superpotent 5-HT2A agonist 2-(4-bromo-2,5-dimethoxyphenyl)-N-[(2-methoxyphenyl)methyl]ethanamine, which consists of a ring-substituted phenethylamine skeleton modified with an N-benzyl group, we designed and synthesized a small library of constrained analogues to identify the optimal arrangement of the pharmacophoric elements of the ligand. Structures consisted of diversely substituted tetrahydroisoquinolines, piperidines, and one benzazepine. Based on the structure of (S,S)-9b, which showed the highest affinity of the series, we propose an optimal binding conformation. (S,S)-9b also displayed 124-fold selectivity for the 5-HT2A over the 5-HT2C receptor, making it the most selective 5-HT2A receptor agonist ligand currently known.

To refine further the structure-activity relationships of D-1 dopamine receptor agonists, we investigated the roles of three conserved serine residues [Ser198(5.42), Ser199(5.43), and Ser202(5.46)] in agonist binding and receptor activation. These transmembrane domain 5 (TM5) residues are believed to engage catechol ligands through polar interactions. We stably expressed wild-type or mutant (S198A, S199A, and S202A) D-1 receptors in human embryonic kidney cells. These receptors were expressed at similar levels (approximately 2000 fmol/mg) and bound the radioligand [H-3]R(+)-7-chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine (SCH 23390), although S198A and S199A displayed significant losses of affinity compared with that for wild-type receptors. The endogenous agonist, dopamine, had losses of potency at each of the mutant receptors. We tested cyclohexyl-substituted isochroman, carbocyclic, and chroman bicyclic dopamine analogs and found that the mutations affected the chroman to a lesser extent than the other compounds. These results support our hypothesis that the decreased D-1 activity of chroman analogs results from a ligand intramolecular hydrogen bond that impairs the ability of the catechol to engage the receptor. Sensitivities of these rigid catechol agonists to the effects of the serine mutations were dependent on ligand geometry, particularly with respect to the rotameric conformation of the ethylamine side chain and the distance between the amino group and each catechol hydroxyl. Functional experiments in striatal tissue suggest that the ability to engage TM5 serines is largely correlated with agonist efficacy for cAMP stimulation. These results provide a new understanding of the complexities of D-1 ligand recognition and agonist activation and have implications for the design of rigid catechol ligands.

A novel class of isochroman dopamine analogues, originally reported by Abbott Laboratories, have > 100-fold selectivity for D-1-like over D-2-like receptors. We synthesized a parallel series of chroman compounds and showed that repositioning the oxygen atom in the heterocyclic ring decreases potency and confers D-2-like receptor selectivity to these compounds. In silico modeling supports the hypothesis that the altered pharmacology for the chroman series is due to potential intramolecular hydrogen bonding between the oxygen in the chroman ring and the meta-hydroxy group of the catechol moiety. This interaction realigns the catechol hydroxy groups and disrupts key interactions between these ligands and critical serine residues in TM5 of the D-1-like receptors. This hypothesis was tested by the synthesis and pharmacological evaluation of a parallel series of carbocyclic compounds. Our results suggest that if the potential for intramolecular hydrogen bonding is removed, D-1-like receptor potency and selectivity are restored.

The octahydrobenzo[h] isoquinoline scaffold is of interest as a conformationally-restricted phenethylamine that may be useful for constructing biologically active products. Surprisingly, however, no tractable synthesis of this ring system has been reported. We now describe a facile method for obtaining this framework, and illustrate that our approach is easily amenable to substitutions at the 5-position. Importantly, we demonstrate that the 7,8-dihydroxy-5-phenyl-substituted ligand is an extremely potent, high-affinity, full D-1 dopamine receptor-selective agonist. (C) 2010 Elsevier Ltd. All rights reserved.

In this study, we employed site‐directed mutagenesis to explore the complex hydrogen bonding interactions that are critical for D1 dopamine receptor activation. We stably expressed wild‐type and three transmembrane 5 serine to alanine mutants (S198(5.42)A, S199(5.43)A, and S202(5.46)A) in HEK 293 cells, and evaluated binding affinity and potency of cAMP production for many structurally diverse ligands. Each of these cell lines expressed similar levels of receptor (approximately 2,000 fmol/mg), and bound [3H]SCH‐23390. The radioligand exhibited significant losses of affinity, however, at S198A and S199A. Dopamine, tetracyclics (e.g. dihydrexidine) and cyclohexyl‐substituted bicyclics similarly suffered losses of affinity and potency at the mutant receptors, where S202A produced the largest disruption of binding and activity. Interestingly, the phenylbenzazepines (e.g. SKF‐38393) demonstrated reduced affinity and potency for each mutant, but were least affected by S202A. Uniquely, the non‐catechol ergolines (e.g. CY 208–243) were more potent at S198A, whereas S199A and S202A produced relatively minor losses of affinity and potency. These results demonstrate that these structurally diverse ligands uniquely interact with the D1 binding pocket, and present a new understanding of the complexities of D1‐ligand recognition and activation. This work was supported by NIH grants MH42705 and MH60397.