Font Size: a A A

Quantitative Prediction Of Dissociation Rate Constants Of HIV-1 Protease Inhibitors And Dissociation Mechanism Reasearch

Posted on:2018-05-01Degree:MasterType:Thesis
Country:ChinaCandidate:S J QuFull Text:PDF
GTID:2334330536968946Subject:Pharmacy
Abstract/Summary:PDF Full Text Request
Accumulated evidences suggest that in vivo biological potency of a drug is strongly correlated with the binding/unbinding kinetics.However,few predictive models have been developed for the quantitative predictions of the association and dissociation rate constants of drugs.In this paper,the computational techniques,such as quantitative structure-activity relationship(QSAR),molecular docking,and molecular dynamics(MD)simulation were employed to quantitativly predict the thermodynamic and kinetic properties of HIV-1 protease inhibitors and to understand the dissociation mechanism of HIV-1 protease inhibitors(PIs).The results can provide important references for drug develepment and dissociation mechanisms of HIV-1 PIs.The main results are as follows:(1)A total of 37 HIV-1 PIs,characterized by three-dimensional(3D)molecular interaction field descriptors(Volsurf)and physicochemical descriptors,were used to construct the quantitative prediction models of the dissociation rate constants(-log(koff)),association rate constants(log(kon))and equilibrium dissociation constants(-log(KD))by partial least squares(PLS)and support vector machine(SVM)methods,respectively.The results showed that the QSAR models established by Volsurf descriptors are reasonable,inter-consistent,and highly predictive for both the kinetic and thermodynamic properties of HIV-1 PIs.Further analysis showed that 9 Volsurf descriptors including BV12-OH2,BV31-OH2,Cw5-OH2,ID6-DRY,D23-DRY,HB1-O,W6,CP and D23-OH2 are molecular determinants for the kinetic and thermodynamic properties of the HIV-1 PIs.(2)The 37 HIV-1 PIs were docked into crystal structure of HIV-1 protease(PDB ID: 1AJX)by using Surflex-Dock method.The docking results were evaluated by 5 scoring functions.Unfortunately,no significant correlation was observed between the 5 docking scores and equilibrium dissociation constants.The possible reasons are as follows: 1)the conformation flexibilities of the flap region of HIV-1 protease were considered insufficiently in Surflex-Dock;2)the 5 scoring functions can not predict accurately the hydrophobic interactions between the inhibitors and HIV-1 protease.(3)Molecular dynamics(MD)was performed based on the docking conformations of 37 HIV-1 PIs.The results showed that the 37 HIV-1 protease complexes reached equilibrium states after 10 ns MD simulations.For 7 clinical drugs,the lowest-energy conformations among the last 2ns MD trajectorys showed that the H-bond interactions between PIs with Asp25,Asp29 and Asp30,and the hydrophobic interactions with Leu23,Ala28,Val32,Ile47,Ile50,Phe53,Pro81,Val82,Ile84,Leu23’,Ala28’,Ile47’,Ile50’,Pro81’ Val82’,and Ile84’ are determinant factors for binding affinities and structural stabilities.In addition,the hydrophobic interactions between PIs and Ile47,Gly48,Gly49,Ile50,Ile47’,Gly48’,Gly49’,Ile50’ ect.play important roles in maintaining a closed state of the flap region of HIV-1 protease.(4)According to the results of MD simulations,the lowest-energy conformations of the 37 HIV-1 PIs complexes were used as starting comformations of steered molecular dynamics simulations(SMDs).A total of 20 conformations in SMD trajectories were extracted to calculate electrostatic and Van der Waals interactions between the 37 HIV-1 PIs and the important residues in the dissociation pathway.Based on the 2880 fingerprint descriptors obtained,PLS combined by forward variable selections was used to establish prediction models of the dissociation rate constants(-log(koff))of HIV-1 PIs.For the optimal model with 5 descriptors,the R2,3-fold Q2,and R2pred were 0.749,0.742,and 0.834,respectively.In order to further validate the resulting model,30 repeating modelings and Y random purmutation test were performed.The results showed that the model has good predictive capability.Further analysis showed that Asp25,Ile47,Ile50 in chain A,and Ile47’,Ile50’ in chain B are key residues affecting the dissociation rates of the HIV-1 PIs.Moreover,the conformational changes of Ile47,Ile47’,Ile50,Ile50’,and Thr80 residues also have important influence on the dissociation processes of the HIV-1 PIs.
Keywords/Search Tags:HIV-1 protease, Inhibitors, Dissociation rate constants, Quantitative structure-activity relationship, Steered molecular dynamics
PDF Full Text Request
Related items