Molecular Dynamics Simulations Of Protein-Ligand Interactions In AR And C-Met Kinase

With the in-depth study on tumor molecular biology,we have gained a deep understanding of the molecular mechanisms of tumor initiation and progression.Molecular targeted anti-tumor drugs offer new opportunities for the treatment of maigant tumors.Drugs are specially designed and developed to interfere with associated signal transduction pathways in tumor cells by targeting abnormal proteins or genes("molecular targets")especially if they are known to be involved in the progression and spread of tumors.Both androgen receptor(AR)and mesenchymal-epithelial transition factor(c-Met)are important targets for the design and development of anti-tumor drugs.In most instances,the prolonged use of molecular targeted anti-tumor drugs may result in reduced efficacy or even loss of clinical therapeutic effect.From a macroscopic point of view,drug resistance is generally related to a decrease in the affinity of the drug to the mutated target.Therefore,the studies of the interaction modes between drugs and drug targets as well as the drug resistance mechanisms may provide valuable clues for the design and development of new drug candidates.Abnormal activation of the AR signaling pathway plays an important physiological role in the development and progression of both early prostate cancer and relapsing castration-resistant prostate cancer.Up to now,the crystallographic structures of the wild-type AR in complex with its antagonists have not been experimentally determined,which significantly obstructs the investigation of the atomic level interaction mechanisms between AR and its antagonists as well as the regulation mechanisms between antagonists and AR coactivators.In the second chapter,molecular dynamics(MD)simulations have been employed to explore the conformational space of AR in complex with its antagonist and search for the antagonistic forms of AR.We also investigated the effect of the binding of different types of ligands,including the agonist(DHT),antagonist(HFT)and coactivator(SRC3),on the conformations of the two binding sites of AR(LBP and AF2).The MM/GBSA calculation results illustrated that the association of the coactivator SRC3 with AF2 would enhance the binding affinity of the agonist DHT to AR,while not facilitate the binding of the antagonist HFT.Meanwhile,the coactivator-antagonist interactions would be mediated by a SRC3?H3/H4?L1-3/H5?HFT allosteric pathwayClinical studies have found that the HGF/c-Met signal pathway is abnormally activated in a variety of solid tumors.The development of small molecule inhibitors targeting c-Met kinase has become one of the research hotspots of anti-tumor drugs.However,both the multi-target tyrosine kinase inhibitor(Crizotinib)and the potent selective c-Met kinase inhibitor(SAR125844)exhibit the reduced inhibitory activities against the c-Met Y1230H mutant.In the third chapter,the MD simulations,MM/GBSA binding free energy calculations and free energy decomposition were carried out to uncover the molecular principle of the interactions between the wild-type c-Met kinase or Y1230H mutant and small molecule inhibitors(SAR125844 and Crizotinib).The unique binding modes with the triazolopyridazine and benzothiazole groups of SARI 25844 bound to the hinge and A-loop regions of c-Met resulted in the exquisitely c-Met selectivity of SAR125844.In comparison to the wild-type c-Met kinase,the outward dislodgment of the A-loop region of the Y1230H mutant led to the expansion of the ATP binding pocket.The loss of the strong hydrophobic interactions and ?-?stacking interactions between the inhibitors and TYR1230 of the wild-type protein had direct impact on a decrease in the binding affinity of both Crizotinib and SARI 25844 to the Y1230H mutant.There are currently numerous MD simulation software packages in the computational biology field,but most of them are only using shell scripts and executed as command-line programs.In principle,those classical MD simulation packages operate in approximately the same way.The first step in any modeling project is to build the initial model structure whose rationality influences dramatically a successful or failed outcome of MD simulations.In the forth chapter,we developed an easy-to-use tool called MDBuilder to accomplish the system preparation task for biomolecular MD simulations.The functions fulfilled by MDBuilder included parsing lists of force field parameters and their associated topology terms,cleaning crystallographic structures of proteins,adding hydrogen atoms.detecting possible disulfide bonds,immersing solutes into a water box,adding counter-ions,finally generateing a topology file and a coordinate file of the system.MDBuilder implemented as a PyMOL plugin provided a user-friendly and flexible graphical user interface,which might help MD novices lower the learning curve.