The role of structural dynamics and flexibility of ligands: Application to MT1 receptors

Melatonin (N-acetyl-5-methoxytryptamine) is a pineal hormone of darkness responsible for regulation of circadian rhythms. It is responsible for binding and activation of G protein coupled MT1 and MT2 receptors. Many diseases like sleep disorders, aging, hypnotic agents, autism, cancer, Alzheimer's are linked to melatonin and its receptors. Several MT2-selective agonists and antagonists have been proposed and developed, but no considerate development has been seen in the case of MT1 receptors. There is a critical need for development of MT1 agonists and to study their specific effects on MT1 receptor. Our hypothesis is that (a) compounds with pharmacophoric features of melatonin will activate melatonin receptors in a non-selective manner (b) specific structural features that encompass melatonin binding site in MT receptors is essential to confer selectivity.;A Three-Dimensional Quantitative Structure - Activity Relationship (3D-QSAR) model was built using 29 compounds as a training set. A test set comprising of 150 compounds from GLIDA: GPCR ligand database was used to evaluate the 3D-QSAR model which showed scores of (n=29, r2 =0.852). Top ranked MT1 ligands from this 3D QSAR model were also good MT2 ligands, confirming our initial hypothesis.;In an effort to identify structural features encompassing MT binding sites, homology models of MT1 receptor were constructed using Rhodopsin and B-adrenergic receptors as templates. Analyzing the docking poses of melatonin within the active site indicated that structural movements of transmembrane helices III and V to orient the binding site residues H195, S110 and S114 was important for optimal binding of Melatonin. Virtual screening of melatonin and luzindole against an ensemble of 7644 MT1 receptor models generated by providing helix III and V with a combination of translational and rotational degrees of freedom has identified several key residues that are important for melatonin ligand binding in addition to H195, S110 and S114. A combination of our QSAR and virtual screening using the ensemble of theoretical models can provide important structural insights and possibly accelerate the discovery of a MT1 selective ligand.