Applications Of Computational Alanine Scanning Based On MM/GBSA_IE Method

Due to the rapid development of computer power,theoretical calculation methods play more and more functions in many scientific research fields such as physics,chemistry,and biology.Among them,molecular modeling methods such as molecular dynamics simulations,molecular docking simulations,and quantum mechanical calculations play increasingly important roles in drug discovery and design.Relying on various computational biology methods,we can explore the physical and chemical reactions that occur in living organisms,interpret the mechanisms of biochemical reactions,understand the causes of diseases,and provide theoretical bases for the discovery and design of novel drugs.Although various molecular modeling methods have been greatly developed and widely used,there are still many limitations.For example,the calculation of binding free energies,especially the calculation of entropies,solvent models,and the calculation of complex reaction mechanisms still have much room for improvement.In order to efficiently and reliably calculate the binding free energy of large biomolecular systems such as protein-protein and protein-ligand interactions,our group recently developed an Interaction Entropy(IE)method for calculating the entropy contributions in binding free energies.Combining with MM/GBSA,the MM/GBSA_IE method was applied to calculate the binding free energy difference of the computational alanine scanning of protein interaction systems.Because of the high efficiency and accuracy,the MM/GBSA_IE method which combines MM/GBSA and IE methods shows broad application prospects.In this thesis,we apply the MM/GBSA_IE method on several protein-protein and protein-ligand interactions which have important biological functions and significance and analyze their binding mechanisms.In the second chapter of this thesis,we apply the MM/GBSA_IE method-based computational alanine scanning to p53-MDM2 and the related protein-protein interactions and compare the calculated binding free energy differences with experimental data.The result shows that the MM/GBSA_IE method is more accurate and reliable than the traditional MM/GBSA method.While experimental alanine scanning has not been performed on the MDM2 and MDMX proteins,the hot spots of the MDM2 and MDMX proteins are predicted and the binding mechanisms of p53-MDM2 and the related protein-protein interactions are further analyzed.In Chapter 3,we use the computational alanine scanning to predict the hot spots in the PD-1/PD-L1 protein-protein interaction.The result reveals a hydrophobic core in the PD-1/PD-L1 interface which consisted of hot spots.Combining with hydrogen bond analysis,we find a hydrogen bond network formed by the side chain of hot spots and the backbone of null spots that surrounding the hydrophobic core.Several other hot spots detection methods are used to verify the hot spots that predicted by the MM/GBSA_IE method and two alanine hot spots that cannot be detected by the MM/GBSA_IE method are predicted.In Chapter 4,we use molecular docking simulations and the computational alanine scanning to compare the binding modes of small molecule inhibitors of PRC2 which target the EED subunit protein of the core PRC2 complex.The docking result shows that the binding modes of the MAK683 and EED226 small molecules are very close.The computational alanine scanning shows that the binding free energy profiles of ligand/EED interactions are significantly different even if the ligands have similar binding modes.Since the MAK683 ligand combines the advantages of the EED226 and A-395 ligands,it exhibits a stronger binding affinity to EED.This result indicates that MAK683 is the most promising PRC2 inhibitor.