| Summary: | 碩士 === 國立臺灣大學 === 生醫電子與資訊學研究所 === 98 === Vascular endothelial growth factor receptor 2 and 3 (VEGFR-2/3) are important protein kinases targets for angiogenesis and lymphangiogenesis of cancer treatments. Most “Type I kinase inhibitors” are ATP competitive while “Type II kinase inhibitors” target the inactive form of kinase by interacting with a hydrophobic allosteric site created by the outstretched DFG domain. Designing compounds with good inhibition, selectivity, specificity and solubility is challenging for kinase inhibitors. It would be beneficial to start with a known drug and use its physicochemical properties and binding mode as a template.
In this study, we developed a fragment-based de novo design strategy to design novel type II VEGFR-2/3 inhibitors based on sorafenib (Nexavar®) , a known Type II kinase inhibitor. Three steps were employed in the design of new VEGFR-2/3 inhibitors. First, the inactive VEGFR-2/3 binding pocket, where sorafenib binds, was divided into an allosteric site, linker spacer, and ATP binding site. Second, potential active site fragments were selected from the Scifinder database and used to flood the binding site. Eight docked poses for each fragment were selected from the ensemble, located in the ATP binding and the allosteric site, and each fragment was prioritized based on a group efficiency score and selected via drug-likeness filters. The final step generates a series of new compounds by linking fragments that have an average of three alternative connections per linker. The new compound was prescribed a bound conformation similar to that of the known ligand (preserving key receptor - ligand interactions) and analyzed. To test the performance of our method, we were able to re-generate sorafenib and its analogues and another type II inhibitor – nilotinib (Tasigna®) automatically from a total combination of 1330 fragments for each segment of the binding site; eight docked poses for each fragment and average 3 linker attachment points for each linker.
|
|---|
