Theoretical Calculation And Design Of Protein-Substrate Interaction

Interactions between protein and substrate are of great significance in biological processes,and are also closely related to many human diseases.Quantitative understanding of the binding energetics by calculating the interaction energy between the protein and the substrate can deepen our understanding of the interaction mechanism.However,accurate computation of binding free energy remains a major challenge in computational biology.In this thesis,we used alanine scanning and interaction entropy(AS-IE)method to calculate the binding free energy between protein and substrate.Histone methylation as an epigenetic way of transcriptional regulation of genes,usually occurs at the N-terminal arginine or lysine residues of histones H3 and H4,and is catalyzed by histone methyltransferases.The modified enzymes involved in histone methylation can be classified into three types: “Writers”,“Erasers” and “Readers”,we have studied the recognition mechanism of histone methylation.In the first part of the thesis,we have studied PHF1 and its homologous protein MTF2,which could recognize K36me3 on histone H3,using the AS-IE method to calculate and analyze the interaction energy between the protein and peptide.The obtained results could deepen our understanding of the recognition of histone methylation modifications.Lipase is a surface enzyme that widely exists in organisms.Its function is to catalyze the hydrolysis of fat,as well as esterification and transesterification.An important use of lipase in the industry is the enrichment of polyunsaturated fatty acids from fish and vegetable oils.Polyunsaturated fatty acids such as eicosapentaenoic acid and docosahexaenoic acid are beneficial to human health,including lowering cholesterol,promoting brain development,and improving cardiovascular circulation.In this study,lipase Lip K107 was designed to improve its activity to polyunsaturated fatty acids.First,we constructed a reasonable complex structure of lipase Lip K107 and DHA/EPA substrates through molecular dynamics simulation and molecular docking.The residues around the substrate were then designed with Rosetta,taking into account the probability of 20 natural amino acids at each position predicted from a deep-learning neural-network.The substrate binding energies of the designed proteins were calculated using the MM/GBSA method.A number of designs were finally selected for experimental verifications by considering the Rosetta score and the binding free energy.