Year of Award
Doctor of Philosophy (PhD)
Department of Chemistry.
Ma, Edmond Dik Lung
Computer simulation, Enzyme inhibitors, Genetic regulation, Molecular dynamics
Molecular docking technologies enable the extraordinary structural diversity of natural products to be harnessed in an efficient manner. In this thesis, in silico techniques were used to discover and develop c-myc oncogene down-regulators and JAK2 inhibitors from databases of natural products and approved drugs. In Chapter 1, current literature on the use of molecular docking in virtual screening for the identification of bioactive molecules from natural product databases are reviewed. Chapter 2 provides an overview of the experiments performed during the course of this work, including molecular docking, PCR stop assay, absorption spectroscopy, CD spectroscopy, FID assay, mass spectrometry, SPR spectroscopy, ELISA assay, MTT assay, luciferase assay and Western blot analysis. The mechanisms, applications and protocols of these experiments are detailed. A unique intramolecular G-quadruplex c-myc NHE IIIi loop isomer model developed by our group was employed to design and screen 30 flavone derivatives in silico. The highest-scoring flavone derivatives 3.5, 3.6 and 3.7 containing cationic pyridinium side chains that could interact with the G-quadruplex grooves were synthesized. The flavone derivatives could stabilize the c-myc G-quadruplex in the PCR-stop assay, and induce the G-quadruplex structure in guanine-rich sequences as revealed by CD spectroscopy. The binding affinity of the derivatives towards various DNA structures was examined using UV-visible spectroscopy. The most promising derivative 3.7 was further subjected to surface plasmon resonance spectroscopy, in silico molecular modeling and luciferase reporter assay to determine its selectivity, binding interaction mode and c-myc G-quadruplex promoter inhibitory activity in cancer cells. This compound also displayed promising cytotoxic behavior against human cancer cell lines. This part of work is detailed in chapter 3. Chapter 4 describes the application of computer-aided techniques for the repurposing of FDA-approved drugs as c-myc oncogene G-quadruplex stabilizers. Methylene blue (MB) emerged as a promising scaffold after virtual screening of 3,000 FDA-approved drugs. A structure-based lead optimization approach was used to generate and screen 50 MB derivatives, containing side chains that could interact with the G-quadruplex grooves, in silico. The highest-scoring compounds 4.10, 4.11 and 4.12 were synthesized and their ability to interact with the c-myc G-quadruplex was investigated using FID assay. The most promising compound 4.11 stabilized c-myc G-quadruplex DNA in a PCR-stop assay. The selectivity of 4.11 for the c-myc G-quadruplex over duplex DNA and other G-quadruplexes was demonstrated using UV-visible spectroscopy and mass spectrometry. Compound 4.11 could induce or stabilize c-myc G-quadruplex formation in cellular models, and displayed higher cytotoxicity against human hepatocarcinoma cells compared to the parent compound, MB. The application of the DOLPHIN kinase model to discover natural product scaffolds as Type II JAK2 inhibitors is presented in chapter 5. Amentoflavone, a biflavonoid from the Chinese plant Gingko biloba, emerged as a promising candidate after biological verification of the hit structures. Amentoflavone was optimized in silico, and the top scoring derivatives were synthesized. The activity of the amentoflavone analogues against JAK2 activity in HEL cells was evaluated using a Western blot assay. Two derivatives, 5.3 and 5.7, showed low-micromolar activity against JAK2 phosphorylation in cellulo. Analogue 5.2 inhibited total JAK2 content in HEL cells and also displayed potent anti-proliferative activity against HEL cells in the MTT assay. This chapter also describes the total synthesis of amentoflavone.
Yang, Hui, "Structure-based discovery and development of c-myc down-regulators and JAK2 inhibitors" (2013). Open Access Theses and Dissertations. 17.
Large files may be slow to open. For best results, right-click and select "save as..."