Mutagenesis Analysis Reveals Distinct Amino Acids of the Human Serotonin 5-HT2C Receptor Underlying the Pharmacology of Distinct Ligands.

While exploring the structure-activity relationship of 4-phenyl-2-dimethylaminotetralins (PATs) at serotonin 5-HT2C receptors, we discovered that relatively minor modification of PAT chemistry impacts function at 5-HT2C receptors. In HEK293 cells expressing human 5-HT2C-INI receptors, for example, (-)-trans-3'-Br-PAT and (-)-trans-3'-Cl-PAT are agonists regarding Gαq-inositol phosphate signaling, whereas (-)-trans-3'-CF3-PAT is an inverse agonist. To investigate the ligand-receptor interactions that govern this change in function, we performed site-directed mutagenesis of 14 amino acids of the 5-HT2C receptor based on molecular modeling and reported G protein-coupled receptor crystal structures, followed by molecular pharmacology studies. We found that S3.36, T3.37, and F5.47 in the orthosteric binding pocket are critical for affinity (Ki) of all PATs tested, we also found that F6.44, M6.47, C7.45, and S7.46 are primarily involved in regulating EC/IC50 functional potencies of PATs. We discovered that when residue S5.43, N6.55, or both are mutated to alanine, (-)-trans-3'-CF3-PAT switches from inverse agonist to agonist function, and when N6.55 is mutated to leucine, (-)-trans-3'-Br-PAT switches from agonist to inverse agonist function. Notably, most point-mutations that affected PAT pharmacology did not significantly alter affinity (KD) of the antagonist radioligand [3H]mesulergine, but every mutation tested negatively impacted serotonin binding. Also, amino acid mutations differentially affected the pharmacology of other commercially available 5-HT2C ligands tested. Collectively, the data show that functional outcomes shared by different ligands are mediated by different amino acids and that some 5-HT2C receptor residues important for pharmacology of one ligand are not necessarily important for another ligand.

Novel 4-substituted-N,N-dimethyltetrahydronaphthalen-2-amines: synthesis, affinity, and in silico docking studies at serotonin 5-HT2-type and histamine H1 G protein-coupled receptors.

Syntheses were undertaken of derivatives of (2S,4R)-(-)-trans-4-phenyl-N,N-dimethyl-1,2,3,4-tetrahydronaphthalen-2-amine (4-phenyl-2-dimethylaminotetralin, PAT), a stereospecific agonist at the serotonin 5-HT2C G protein-coupled receptor (GPCR), with inverse agonist activity at 5-HT2A and 5-HT2B GPCRs. Molecular changes were made at the PAT C(4)-position, while preserving N,N-dimethyl substitution at the 2-position as well as trans-stereochemistry, structural features previously shown to be optimal for 5-HT2 binding. Affinities of analogs were determined at recombinant human 5-HT2 GPCRs in comparison to the phylogenetically closely-related histamine H1 GPCR, and in silico ligand docking studies were conducted at receptor molecular models to help interpret pharmacological results and guide future ligand design. In most cases, C(4)-substituted PAT analogs exhibited the same stereoselectivity ([-]-trans>[+]-trans) as the parent PAT across 5-HT2 and H1 GPCRs, albeit, with variable receptor selectivity. 4-(4'-substituted)-PAT analogs, however, demonstrated reversed stereoselectivity ([2S,4R]-[+]-trans>[2S,4R]-[-]-trans), with absolute configuration confirmed by single X-ray crystallographic data for the 4-(4'-Cl)-PAT analog. Pharmacological affinity results and computational results herein support further PAT drug development studies and provide a basis for predicting and interpreting translational results, including, for (+)-trans-4-(4'-Cl)-PAT and (-)-trans-4-(3'-Br)-PAT that were previously shown to be more potent and efficacious than their corresponding enantiomers in rodent models of psychoses, psychostimulant-induced behaviors, and compulsive feeding ('binge-eating').

Aromatic interactions impact ligand binding and function at serotonin 5-HT2C G protein-coupled receptors: Receptor homology modeling, ligand docking, and molecular dynamics results validated by experimental studies.

The serotonin (5-hydroxytryptamine, 5-HT) 5-HT2 G protein-coupled receptor (GPCR) family consists of types 2A, 2B, and 2C that share ~75% transmembrane (TM) sequence identity. Agonists for 5-HT2C receptors are under development for psychoses, whereas, at 5-HT2A receptors, antipsychotic effects are associated with antagonists-in fact, 5-HT2A agonists can cause hallucinations and 5-HT2B agonists cause cardiotoxicity. It is known that 5-HT2A TM6 residues W6.48, F6.51, and F6.52 impact ligand binding and function, however, ligand interactions with these residues at the 5-HT2C receptor has not been reported. To predict and validate molecular determinants for 5-HT2C-specific activation, results from receptor homology modeling, ligand docking, and molecular dynamics (MD) simulation studies were compared with experimental results for ligand binding and function at wild type and W6.48A, F6.51A, and F6.52A point-mutated 5-HT2C receptors.

Molecular pharmacology and ligand docking studies reveal a single amino acid difference between mouse and human serotonin 5-HT2A receptors that impacts behavioral translation of novel 4-phenyl-2-dimethylaminotetralin ligands.

During translational studies to develop 4-phenyl-2-dimethylaminotetralin (PAT) compounds for neuropsychiatric disorders, the (2R,4S)-trans-(+)- and (2S,4R)-trans-(-)-enantiomers of the analog 6-hydroxy-7-chloro-PAT (6-OH-7-Cl-PAT) demonstrated unusual pharmacology at serotonin (5-HT) 5-HT2 G protein-coupled receptors (GPCRs). The enantiomers had similar affinities (Ki) at human (h) 5-HT2A receptors (≈ 70 nM). In an in vivo mouse model of 5-HT2A receptor activation [(±)-(2,5)-dimethoxy-4-iodoamphetamine (DOI)-elicited head twitch], however, (-)-6-OH-7-Cl-PAT was about 5-fold more potent than the (+)-enantiomer at attenuating the DOI-elicited response. It was discovered that (+)-6-OH-7-Cl-PAT (only) had ≈ 40-fold-lower affinity at mouse (m) compared with h5-HT2A receptors. Molecular modeling and computational ligand docking studies indicated that the 6-OH moiety of (+)- but not (-)-6-OH-7-Cl-PAT could form a hydrogen bond with serine residue 5.46 of the h5-HT2A receptor. The m5-HT2A as well as m5-HT2B, h5-HT2B, m5-HT2C, and h5-HT2C receptors have alanine at position 5.46, obviating this interaction; (+)-6-OH-7-Cl-PAT also showed ≈ 50-fold lower affinity than (-)-6-OH-7-Cl-PAT at m5-HT2C and h5-HT2C receptors. Mutagenesis studies confirmed that 5-HT2A S5.46 is critical for (+)- but not (-)-6-OH-7-Cl-PAT binding, as well as function. The (+)-6-OH-7-Cl-PAT enantiomer showed partial agonist effects at h5-HT2A wild-type (WT) and m5-HT2A A5.46S point-mutated receptors but did not activate m5-HT2A WT and h5-HT2A S5.46A point-mutated receptors, or h5-HT2B, h5-HT2C, and m5-HT2C receptors; (-)-6-OH-7-Cl-PAT did not activate any of the 5-HT2 receptors. Experiments also included the (2R,4S)-trans-(+)- and (2S,4R)-trans-(-)-enantiomers of 6-methoxy-7-chloro-PAT to validate hydrogen bonding interactions proposed for the corresponding 6-OH analogs. Results indicate that PAT ligand three-dimensional structure impacts target receptor binding and translational outcomes, supporting the hypothesis that GPCR ligand structure governs orthosteric binding pocket molecular determinants and resulting pharmacology.

Molecular determinants of ligand binding at the human histamine H1 receptor: Site-directed mutagenesis results analyzed with ligand docking and molecular dynamics studies at H1 homology and crystal structure models.

The human histamine H1 G-protein coupled receptor (GPCR) is an important drug target for inflammatory, sleep, and other neuropsychiatric disorders. To delineate molecular determinants for ligand binding for drug discovery purposes, human H1 receptor models were built by homology to the crystal structure of the human ß2 adrenoceptor (ß2AR) and from the recently reported crystal structure of the human H1 receptor complex with doxepin at 3.1 Å (PDB code 3RZE). Ligand affinity of histamine and the H1 antagonists mepyramine and (2S, 4R)-(-)-trans-4-phenyl-2-N, N-dimethylaminotetralin (PAT) at wild type and point-mutated (D3.32A, Y3.33A, W4.56A, F5.47A, W6.48A, Y6.51A, F6.52A, F6.55A, Y7.43A) human H1 receptors were determined experimentally and results analyzed by ligand docking and molecular dynamic studies at WT and point-mutated H1 receptor models. Differences in ligand binding affinities correlated to differences in ligand binding modes at models built according to homology or crystal structure, indicating, both models are accurate templates for predicting ligand affinity for H1 drug design.

Molecular Determinants for Ligand Binding at Serotonin 5-HT2A and 5-HT2C GPCRs: Experimental Affinity Results Analyzed by Molecular Modeling and Ligand Docking Studies.

Ligands that activate the serotonin 5-HT2C G protein-coupled receptor (GPCR) may be therapeutic for psychoses, addiction, and other neuropsychiatric disorders. Ligands that are antagonists at the closely related 5-HT2A GPCR also may treat neuropsychiatric disorders; in contrast, 5-HT2A activation may cause hallucinations. 5-HT2C-specific agonist drug design is challenging because 5-HT2 GPCRs share 80% transmembrane (TM) homology, same second messenger signaling, and no crystal structures are reported. To help delineate molecular determinants underlying differential binding and activation of 5-HT2 GPCRs, 5-HT2A, and 5-HT2C homology models were built from the ß2-adrenergic GPCR crystal structure and equilibrated in a lipid phosphatidyl choline bilayer performing molecular dynamics simulations. Ligand docking studies at the 5-HT2 receptor models were conducted with the (2R, 4S)- and (2S, 4R)-enantiomers of the novel 5-HT2C agonist/5-HT2A/2B antagonist trans-4-phenyl-N,N-dimethyl-2-aminotetralin (PAT) and its 4'-chlorophenyl congners. Results indicate PAT-5-HT2 molecular interactions especially in TM domain V are important for the (2R, 4S) enantiomer, whereas, TM domain VI and VII interactions are more important for the (2S, 4R) enantiomer.

Drug discovery targeting human 5-HT(2C) receptors: residues S3.36 and Y7.43 impact ligand-binding pocket structure via hydrogen bond formation.

Specific activation of serotonin (5-HT) 5-HT(2C) G protein-coupled receptors may be therapeutic for obesity and neuropsychiatric disorders. Mutagenesis coupled with computational and molecular modeling experiments based on the human ß2 adrenergic receptor structure was employed to delineate the interactions of different ligands at human 5-HT(2C) residues D3.32, S3.36 and Y7.43. No binding of the tertiary amine radioligand ([³H]-mesulergine) could be detected when the 5-HT(2C) D3.32 residue was mutated to alanine (D3.32A). The S3.36A point-mutation greatly reduced affinity of primary amine ligands, modestly reduced affinity of a secondary amine, and except for the 5-HT(2C)-specific agonist N(CH3)2-PAT, affinity of tertiary amines was unaffected. Molecular modeling results indicated that the primary amines form hydrogen bonds with the S3.36 residue, whereas, with the exception of N(CH3)2-PAT, tertiary amines do not interact considerably with this residue. The Y7.43A point-mutation greatly reduced affinity of 5-HT, yet reduced to a lesser extent the affinity of tryptamine that lacks the 5-hydroxy moiety present in 5-HT; modeling results indicated that the 5-HT 5-hydroxy moiety hydrogen bonds with Y7.43 at the 5-HT(2C) receptor. Additional modeling results showed that 5-HT induced a hydrogen bond between Y7.43 and D3.32. Finally, modeling results revealed two low-energy binding modes for 5-HT in the 5-HT(2C) binding pocket, supporting the concept that multiple agonist binding modes may stabilize different receptor active conformations to influence signaling. Ligand potencies for modulating WT and point-mutated 5-HT(2C) receptor-mediated phospholipase C activity were in accordance with the affinity data. Ligand efficacies, however, were altered considerably by the S3.36A mutation only.