Molecular Modeling Study for the Design of Novel Peroxisome Proliferator-Activated Receptor Gamma Agonists using 3D-QSAR and Molecular Docking

• Published in: International journal of molecular sciences 2018-02, Vol.19 (2), p.630
• Main Authors: Jian, Yaning, He, Yuyu, Yang, Jingjing, Han, Wei, Zhai, Xifeng, Zhao, Ye, Li, Yang
• Format: Article
• Language: English
• Subjects: 3D-QSAR; Biological activity; CoMFA; CoMSIA; Diabetes; Diabetes mellitus; Diabetes mellitus (non-insulin dependent); Hydrogen bonds; Mathematical models; Molecular docking; Molecular modelling; Molecular structure; N-benzylbenzamide derivatives; PPARγ; Structure-activity relationships
• ISSN: 1422-0067; 1661-6596; 1422-0067
• DOI: 10.3390/ijms19020630

Type 2 diabetes is becoming a global pandemic disease. As an important target for the generation and development of diabetes mellitus, peroxisome proliferator-activated receptor γ (PPARγ) has been widely studied. PPARγ agonists have been designed as potential anti-diabetic agents. The advanced development of PPARγ agonists represents a valuable research tool for diabetes therapy. To explore the structural requirements of PPARγ agonists, three-dimensional quantitative structure-activity relationship (3D-QSAR) and molecular docking studies were performed on a series of -benzylbenzamide derivatives employing comparative molecular field analysis (CoMFA), comparative molecular similarity indices analysis (CoMSIA), and surflex-dock techniques. The generated models of CoMFA and CoMSIA exhibited a high cross-validation coefficient ( ²) of 0.75 and 0.551, and a non-cross-validation coefficient ( ²) of 0.958 and 0.912, respectively. The predictive ability of the models was validated using external validation with predictive factor ( ² ) of 0.722 and 0.682, respectively. These results indicate that the model has high statistical reliability and good predictive power. The probable binding modes of the best active compounds with PPARγ active site were analyzed, and the residues His323, Tyr473, Ser289 and Ser342 were found to have hydrogen bond interactions. Based on the analysis of molecular docking results, and the 3D contour maps generated from CoMFA and CoMSIA models, the key structural features of PPARγ agonists responsible for biological activity could be determined, and several new molecules, with potentially higher predicted activity, were designed thereafter. This work may provide valuable information in further optimization of -benzylbenzamide derivatives as PPARγ agonists.