Molecular Design And QSAR Research On HIV-1 Reverse Transcriptase Inhibitors

Acquired immunodeficiency syndrome (AIDS) is caused by human immunodeficiency virus (HIV), a chronic retrovirus. Since the first AIDS patient was found in the United States of America in 1981, the persons infected with HIV increased with years. For example, in 2009, there are a total of 2.9 million people infected with HIV. Faced with increasingly severe AIDS infection situation, research and development of new highly potent anti-HIV drugs is urgently demanded. In recent years, non-nucleoside reverse transcriptase inhibitor (NNRTI) has been attracted much attention due to its efficiency, low toxicity, and high structural diversityIn this paper, molecular docking was used to explore the binding modes of N-dihydro-alkoxy-benzyl-oxopyrimidines (N-DABO) derivatives, S- dihydro-alkoxy-benzyl-oxopyrimidines (S-DABO) derivatives, diaryltriazines (DATA) derivatives, 2-amino-6-aryl thio/sulfonyl/sulfinyl benzonitriles (AASB) derivatives and benzophenone derivatives and thiazolidenebenzenesulfonamide derivatives of 5 NNRTI. Based on the docking results, comparative molecular field analysis (CoMFA), comparative molecular similarity indices analysis (CoMSIA), hologram QSAR (HQSAR), and Topomer CoMFA methods were performed to illustrate the quantitative relationships between structures and activities. Finally, two methods of de novo drug design, i.e. LeapFrog and EA-Inventor, were employed to discover new potent non-nucleoside reverse transcriptase inhibitors. The main results are as follows:â‘ N-DABO derivatives: Results of molecular docking show that the docking scores (G_Score) are significantly correlatted with the pEC50 with r=-0.631 and P <0.0001. The results also show that hydrogen-bond interactions between the ligands and residues Lys101 of the receptor, the hydrophobic interactions between ligands and the bottom of the binding pocket, and electrostatic interactions are the main factors affecting the affinity activity. Based on the docking conformations of the 46 samples, CoMFA and CoMSIA modelings were performed. An optimal CoMSIA model was obtained, which consists of hydrogen-bond acceptor, hydrophobic and steric fields. The principal components, r², q², and r_pred² of that optimal model are 4, 0.862, 0.532 and 0.829, respectively. Using samples 35 with highest activity as a template, LeapFrog was used for the de novo molecular design. Total 9 new compounds were obtained, the binding energies and docking scores (G_Score) of which are better than those of the template respectively.â‘¡S-DABO derivatives: Molecular docking results show that the docking scores (G_Score) are significantly correlatted with the pEC50 with r = -0.6445 and P <0.0001. The hydrogen-bonds interactions and the hydrophobic interactions between ligand and the bottom of the binding pocket are main factors affecting ligand binding affinities. By CoMFA, CoMSIA, and HQSAR modelings, 5 QSAR models were established. The q², r², and r_pred² of these 5 models are between 0.52⁰.63, 0.69⁰.98 and 0.56⁰.71, respectively. Taken sample 6l with the highest activity as a template, 5 new compounds were obtained by the EA-Inventor method. The docking scores and predicted activities of these 5 new compounds are better than those of the template respectively.â‘¢DATAs derivatives: Molecular docking results show that the hydrogen-bond, hydrophobic, and van der Waals interactions are important factors which affect the activity. The steric hindrance around R₄ substituent also has an important influence on the activity. 9 QSAR models with strong predictive abilities were established by using CoMFA, CoMSIA and HQSAR methods. The q², r², and r_pred² of these 9 models are between 0.62⁰.82,0.77⁰.97 and 0.66⁰.95, respectively. Based on the best CoMFA model, LeapFrog was carried out by using compound 9h as a template. 7 new compounds were obtained with better binding energies and predicted activities than those of the template respectively.â‘£AASB derivatives: Molecular docking results show there is a significant linear relationship between activities and docking scores with r = -0.4680, P <0.001 (G_Score) and r = 0.5456, P <0.0001 (Total_Score). Hydrogen-bond, hydrophobic, and electrostatic interactions are the main factors affecting the bind activities. 4 QSAR models were established by CoMFA and CoMSIA methods, of which q² and r² are above 0.5 and 0.8, respectively. Taken sample 55 with the highest activity as a template, 5 new compounds were obtained by the EA-Inventor method. The docking scores (Total_Score) and predicted activitis of these 5 new compounds are better than those of the template respectively.⑤Benzophenone and thiazoline acid derivatives: Results of molecular docking show that the biological activities are significantly correlated with the Total_Scores (r = 0.7576 and P < 0.0001). The most important interactions between ligand and the K103N mutant are hydrogen-bond and hydrophobic interactions. 7 QSAR model were established by HQSAR, CoMFA, CoMSIA and Topomer CoMFA, of which the q² and r_pred² were all above 0.5 and 0.7, respectively. Taken Compound 30 as a template, 4 new compounds were obtained by EA-Inventor method. The docking scores (Total_Score) and predicted activities of these 4 molecules are better than those of the template respectively.