Virtual identification of novel PPARα/γ dual agonists by 3D-QSAR, molecule docking and molecular dynamics studies.

Peroxisome proliferator-activated receptors (PPARs) are considered important targets for the treatment of Type 2 diabetes (T2DM). To accelerate the discovery of PPAR α/γ dual agonists, the comparative molecular field analysis (CoMFA) were performed for PPARα and PPARγ, respectively. Based on the molecular alignment, highly predictive CoMFA model for PPARα was obtained with a cross-validated value of 0.741 and a conventional of 0.975 in the non-cross-validated partial least-squares (PLS) analysis, while the CoMFA model for PPARγ with a better predictive ability was shown with and values of 0.557 and 0.996, respectively. Contour maps derived from the 3D-QSAR models provided information on main factors towards the activity. Then, we carried out structural optimization and designed several new compounds to improve the predicted biological activity. To investigate the binding modes of the predicted compounds in the active site of PPARα/γ, a molecular docking simulation was carried out. Molecular dynamic (MD) simulations indicated that the predicted ligands were stable in the active site of PPARα/γ. Therefore, combination of the CoMFA and structure-based drug design results could be used for further structural alteration and synthesis and development of novel and potent dual agonists. AbbreviationsDMdiabetes mellitusT2DMtype 2 diabetesPPARsperoxisome proliferator-activated receptorsLBDDligand based drug design3D-QSARthree-dimensional quantitative structure activity relationshipCoMFAcomparative molecular field analysisPLSpartial least squareLOOleave-one-outcross-validated correlation coefficientONCoptimal number of principal componentsnon-cross-validated correlation coefficientSEEstandard error of estimatethe Fischer ratiopredictive correlation coefficientDBDDNA binding domainMDmolecular dynamicsRMSDroot-mean-square deviationRMSFroot mean square fluctuationsCommunicated by Ramaswamy H. Sarma.