The blockage of the hERG potassium channel by a wide number of diverse compounds has become a major pharmacological safety concern as it can lead to sudden cardiac death. In silico models can be potent tools to screen out potential hERG blockers as early as possible during the drug-discovery process. In this study, predictive models developed using the recursive partitioning method and created using diverse datasets from 203 molecules tested on the hERG channel are described. The first model was built with hERG compounds grouped into two classes, with a separation limit set at an IC50 value of 1 microm, and reaches an overall accuracy of 81%. The misclassification of molecules having a range of activity between 1 and 10 microM led to the generation of a tri-class model able to correctly classify high, moderate, and weak hERG blockers with an overall accuracy of 90%. Another model, constructed with the high and weak hERG-blocker categories, successfully increases the accuracy to 96%. The results reported herein indicate that a combination of precise, knowledge management resources and powerful modeling tools are invaluable to assessing potential cardiotoxic side effects related to hERG blockage.

Three-dimensional models of ligand-receptor complexes based on site-directed mutagenesis experiments of the monoamine G protein-coupled receptors reveal the existence of three distinct drug binding sites inside the receptors. Here, we develop this``three-site'' hypothesis and outline its implications for the modular design of ligands for monoamine GPCRs. Molecular models of receptor-ligand complexes are built for the 5-HT 1A receptor where mutagenesis studies map three spatially distinct binding regions which correspond to the binding sites of thè`s mall, one site-®lling'' ligands 5-HT, propranolol and 8-OH-DPAT, respectively. The models of the 5-HT 1A ligandreceptor complexes provide a frame for the discussion of other ligand-receptor interactions, including a 1 and b 2 adrenoceptors, D 1 and D 2 dopamine, and 5-HT 1D and 5-HT 2A receptors, where mutagenesis and modelling studies also showed occupation of the corresponding three binding locations. All three binding sites are located within the highly conserved seven helix transmembrane domain of the receptor and overlap partially at the prominent Asp residue in TM3 which constitutes the benchmark anchor site for monoamine ligands. The analysis of the sequence similarity, for each binding site, among the monoamine GPCR superfamily shows that the three loci display different degrees of evolutionary conservation. This result suggests different roles for each of the binding sites in intrinsic receptor functions and provides additional insights for the design of ligand functionality and selectivity. The existence of three distinct binding sites is also re¯ected by the architecture of known high af®nity ligands which crosslink two or three``one site-®lling'' fragments around a basic amino group. Typical ligands reported in the CipslineyMDDR portfolio illustrate this point despite the occasional dif®culty of attributing the individual ligand fragments to a speci®c receptor site. The database exploration illustrates the binding site promiscuity of some fragments which is particularly evident for symmetrical ligands and which has implications for 3D QSAR methods dependent on alignments. We propose to generate by deconvolution of known ligands three distinct databases of site-speci®c bioisosters which should provide keystones for the design of novel recomposed monoamine GPCR ligands. The systematic exploration of the``three site'' hypothesis should open novel perspectives for the understanding of ligand recognition for this class of therapeutically important receptors.

Conformational analysis was used to characterize the agonist pharmacophore for melatonin sheep brain receptor recognition and activation. The molecular geometry shared by all conformations of the selected active ligands was determined. Assuming that all the compounds interact at the same binding site at the receptor level, 2-iodomelatonin pharmacophoric conformation served as a template for the superimposition of 64 structurally heterogeneous agonists constituting the training set used to perform a three-dimensional quantitative structure-activity relationship study via the comparative molecular field analysis method. A statistically significant model was obtained for the totality of the compounds (n = 64, q2 = 0.62, N = 6, r2 = 0.96, s = 0.28, F = 249) with steric, electrostatic, and lipophilic relative contributions of 28%, 35%, and 37%, respectively. The predictive power of the proposed model was discerned by successfully testing the 78 agonist ligands constituting the test set. The model so obtained and validated brings important structural insights to aid the design of novel melatoninergic agonist ligands prior to their synthesis.