Identify of promising isoquinolone JNK1 inhibitors by combined application of 3D-QSAR, molecular docking and molecular dynamics simulation approaches

Abstract JNK1 has been suggested to play a key mediator in metabolic disorders, especially been considered to be a drug target for treatment of type-2 diabetes in recent years. Therefore, inhibiting the activity of JNK1 is a promising way for drug development. In the study, in order to design and identify the potential compounds targeting JNK1, a series of isoquinolone analogs were used to develop 3D-QSAR models. Here, two methods of docking-based and minimization-based were adopted for generation of molecular poses and the subsequent model construction. CoMFA model #5 and CoMSIA model #6 based on docking method were selected due to their informative and satisfactory internal and external validation results. An in-house library containing 113 new designs was built based on contour maps analyzes and then designated for screening. Compound 20JX82 was focused for its preferable predicted activity and docking score. Detailed interactions with JNK1 of compound 20JX82 was investigated via deep analyze of docking results and molecular dynamics simulation. Strong interactions and extra hydrogen bonds which were consistent with the summarized design principles were observed. From MD analyzes, greater stability of the 20JX82 bound complex was found when using sole protein as reference. Thus, compound 20JX82 was successfully identified as promising JNK1 inhibitor.