CoMFA, CoMSIA, molecular docking and MOLCAD studies of pyrimidinone derivatives to design novel and selective tankyrase inhibitors

Abstract

Human Tankyrases (TNKS) are the potential and emerging targets for the treatment of many cancers like colon, breast, bladder, lung, gastric, pancreatic adenocarcinoma etc. Inhibition of Tankyrase results in inhibition of Wnt Signaling and indirect inhibition of telomerase actions which ultimately leads to cell growth inhibition and cell death respectively. Currently, all the tankyrase inhibitors are at early stage of discovery. Thus, there is an urge of novel lead molecules as tankyrase inhibitors. To design such novel leads, initially we performed 3D-QSAR studies on a series of 42 reported pyrimidinone derivatives using different alignment methods. Best significant CoMFA and CoMSIA models were obtained using 31 molecules in training set while 11 in test set using distill alignment. CoMFA model gave significant q2, r2 and r2 pred values of 0.509, 0.811 and 0.591 respectively while optimized CoMSIA model (SEDA) outperformed with 0.605 q2, 0.863 r2 and 0.530 r2 pred. To further validate the best CoMFA and CoMSIA model, it was used to predict the activity of external dataset and it successfully predicted the activity very close to the actual values with residual values less than 1. To support the QSAR results, molecular docking study was performed which predicted the binding mode of one of the most potent compound 42 of the training set with TNKS2. MOLCAD analysis provided the significant insight of various surface physico chemical map of the TNKS2 binding site. All these results revealed many helpful structural insights to improve the activity and selectivity of newly designed pyrimidinone derivatives as Tankyrase inhibitors and based on that, 20 novel molecules 54e73 were designed and their activities were predicted. Compounds 68e73 showed better predicted activities than the compound 1 and 42 of the dataset.