Molecular Modeling And Virtual Screening Studies Of Aromatase Inhibitors And B-Raf Kinase Inhibitors

With the development of cell molecular biology, molecular targeted therapy of tumor has become an important area of anticancer drugs research recently. Aromatase is a late-limiting enzyme which can catalyze the aromatization of androgens to estrogens in human body, and it belongs to cytochrome P450 enzyme system. By inhibiting the activity of aromatase, estrogens in body can be controlled at a low level, which can offer a treatment of breast cancer and other estrogen-dependent disease. Protein kinase inhibitor is a quite effective targeted therapy drug, which can affect tumor cell survival, proliferation and development of diseases by blocking intracellular molecular, pathway, and B-Raf kinase has recently emerged as one of most attractive anticancer targets. Nowadays, despite the success of aromatase inhibitors and B-Raf kinase inhibitors in market, they still have some side effects and can develop cellular resistance after a period time of usage. Therefore, it is still important to search and develop the above inhibitors with new skeleton, high inhibitory activity and lower side effects.Computer-aided drug design （CADD） is a method that theoretically design and discover new drug molecules by simulating the interaction between drug molecules and targets, or by analyzing the relationship between drug molecular structures and their activities, based on various molecular simulation technologies and mathematical statistics methods. CADD method can improve the efficiency of drug development greatly, so it has been an important tool in the recent drug development. Various CADD methods were used to study the aromatase inhibitors and B-Raf kinase inhibitors in this dissertation.In chapter 1, the aromatase, B-Raf kinase and their inhibitors were introduced in brief. CADD methods were summarized in general, and the methods used in this dissertation were presented in detail.In chapter 2, the studies on 3D-QSAR, pharmacophore modeling and virtual screening were performed for the 66 new-synthesized steroidal aromatase inhibitors （SAIs）. Firstly, the 3D-QSAR model was built on the basis of the molecular common structure alignment, which shows good statistical results:q2=0.636, r2ncv=0.988, r2pred=0.658 for CoMFA and q2=0.843, r2ncv=0.989, r2pred=0.601 for CoMSIA. By analyzing the contour maps, the requirements for structural modification of SAIs were obtained. Secondly, pharmacophore models were derived from SAIs and the obtained best pharmacophore model （Model₀4） includes two acceptor atoms and four hydrophobic centers. Finally, we use the best pharmacophore model and 3D-QSAR model to screen the NCI2000 database, and obtain two best hit compounds, which are expected to design potent and novel SAIs.In chapter 3, the studies on pharmacophore modeling,3D-QSAR and virtual screening were carried out for the 84 new-synthesized non-steroidal aromatase inhibitors （NSAIs）. First, the best pharmacophore model （Model₀8） was obtained from NSAIs, which contains one acceptor atom, one donor atom, and two hydrophobes. Next, the 3D-QSAR studies on NSAIs were performed on both pharmacophore and docking alignments, and the models on pharmacophore alignment shows better statistical results: q2=0.634, r2ncv=0.986, r2pred=0.737 for CoMFA and q2=0.668, r2ncv=0.962, r2pred=0.708 for CoMSIA. The requirements for structural modification of NSAIs were obtained by analyses of the contour maps. At last, a combination of the obtained pharmacophore model, molecular docking and 3D-QSAR model was used to screen the NCI2000 database, and six best hit compounds with different skeletons were achieved, which are expected to develop potent and novel NSAIs.In chapter 4, the studies on 2D,3D-QSAR, pharmacophore modeling, and virtual screening were carried out for the 39 new-synthesized imidazopyridine B-Raf inhibitors （BRIs）. Firstly, an acceptable 2D-QSAR model （r=0.677） was obtained, which was built by using a combination of quantum chemistry, molecular mechanics and statistics. Next, the CoMFA and CoMSIA models of 3D-QSAR were built on the basis of both pharmacophore and docking alignments respectively. The obtained CoMSIA model based on the pharmacophore shows the best result:q2=0.621, r2ncv= 0.996, r2pred=0.885, and the requirements for structural modification of imidazopyridine BRIs were obtained by analyzing the contour maps. In addition, a combination of the pharmacophore model, molecular docking and 3D-QSAR model was used to screen the NCI2000 and ZINC database respectively, and three best hits from NCI2000 and twenty best hits from ZINC with different skeletons were obtained, which provides useful information to develop potent and novel BRIs.In chapter 5, an investigation of molecular docking, molecular dynamic （MD） simulation and binding free energy （ΔGbind） calculation between B-Raf kinase and imidazopyridine inhibitors were carried out, and the interaction mechanism between B-Raf kinase and imidazopyridine inhibitors was explored as well. The result of molecular docking is not in accordance with the inhibitory activity completely, while the results of both MD simulation and ΔGbind calculation are consistent with inhibitory activity completely, which indicates that MD simulation and ΔGbind calculation results can reveal the binding mode between receptor and ligand more accurately than molecular docking result does.In chapter 6, the biological activities of ten BRIs hit compounds screened from ZINC commercial database were tested, and six of ten hit compounds show good inhibitory activities. In addition, MD simulation and ΔGbind calculation of the ten complexes between B-Raf kinase and each hit compound were carried out respectively, and the results obtained from theoretical calculation are in good accordance with the data derived from inhibitory activity test. Both experimental and theoretical results show that the virtual screening strategy we used, a combination of pharmacophore model, molecular docking and 3D-QSAR model, is feasible.