Virtual Screening And Biological Evaluation Of Natural Inhibitors Of Phosphodiesterase And Xanthine Oxidase

Natural products have always been an important source of small molecule drug discovery with diverse,novel structure and high percentage of biological activity,but their small sample size and complex structure have hindered the speed of lead compound discovery to some extent.Computer Aided Drug Design(CADD)technology is based on molecular mechanics,molecular dynamics,quantum mechanics and other theoretical methods to simulate and predict the properties of protein targets and active small molecules for the screening and optimization of lead compounds,such as AriceptZ for the treatment of Alzheimer’s disease,Indinavir for the treatment of CADD has been widely used to improve the speed of new drug development.In this thesis,we used CADD technology combined with deep learning methods to screen for natural inhibitors of phosphodiesterase and xanthine oxidase,and verified the activity of the compounds using enzyme activity assays and enzyme-linked immunoassays.The development of multi-targeted drugs has become one of the research trends in recent years.The specific cAMP hydrolase phosphodiesterase 4(PDE4)and phosphodiesterase 7(PDE7)belong to the same phosphodiesterase superfamily and have diverse biological functions.diseases.It has been found that PDE4 is one of the important targets in chronic inflammation-related diseases,but excessive activity of PDE4 inhibitors can lead to adverse effects such as nausea,vomiting and diarrhea.dual-target inhibitors of PDE4/7 inhibit both PDE4 and PDE7,which can effectively reduce the occurrence of side effects while maintaining the efficacy of the drug,and have the effect of potentiation and toxicity reduction.Therefore,the development of efficient and safe dual-target inhibitors of PDE4 and PDE7 is of great value for the treatment of various diseases related to the immune and inflammatory systems.Xanthine oxidase(XO)catalyzes the synthesis of uric acid from hypoxanthine or xanthine and is therefore a potential target for the treatment of hyperuricemia(HUA).Natural XO inhibitors can reduce uric acid production and are of great value for drug development against hyperuricemia.In this study,a virtual screening of natural XO inhibitors was performed using pharmacophore-based techniques,combined with molecular docking techniques to improve screening accuracy and binding pattern analysis of compounds with high scoring,with the aim of identifying potential natural XO inhibitors.Three main aspects were accomplished in this thesis:(1)Virtual screening by molecular docking and small molecule similarity to obtain 16 potential natural inhibitors of PDE4/7,and molecular dynamics simulations of the complexes of PDE4/7 proteins with candidate compounds to verify the stability of the binding mode.Six of the 16 compounds obtained from the screening(anwulignan,broussochalcone B,emodin,β-lapachone,purpurin,and rhamnazin)were found to have PDE4/7 inhibitory effects and to reduce the levels of inflammatory factors NO,IL-6,and TNF-α by PDE enzyme activity assays and RAW264.7 macrophage inflammatory factorrelated assays;(2)To construct a PDE4/7 inhibitor screening system using an Artificial Neural Network(ANN)based machine learning method to further improve the quality and efficiency of the screening;(3)Virtual screening of xanthine oxidase inhibitors using a pharmacophore modeling approach and classification of compounds according to the source of the herbs.In this study,six PDE4/7 dual inhibitors were obtained based on the virtual screening of natural product databases TCMD and TCMSP and in vitro activity validation,and further animal validation will be completed subsequently.A preliminary PDE4/7 dual inhibitor screening system was constructed by deep learning to lay the foundation for a larger scale compound screening work.The screening of xanthine oxidase inhibitors by pharmacophore model will provide a basis for exploring novel XO inhibitor lead compounds.