QSAR Studies Of HCV NS5B Polymerase Inhibitors

Hepatitis C virus (HCV) is one of the major causative pathogens of chronic hepatitiswhich may lead to liver cirrhosis and hepatocellular carcinoma. Currently, there are noeffective and low side effects therapy agents. Therefore, it's urgent and important to developnew methods and drugs to cure the infection of HCV. The NS5B polymerase is a key enzymeof HCV replication. It is necessary for the synthesis of double-stranded viral RNA fromsingle-stranded viral RNA. Moreover, human cells don't express the enzymes which havesimilar function with NS5B polymerase. This makes NS5B polymerase inhibitors with goodselectivity. For these reasons, NS5B polymerase has become an important target fordeveloping anti-HCV agents, and has been widespread concern.In this work,2D-QSAR and3D-QSAR methods have been used for benzothiadiazinederivatives and indole-based C-3pyridone derivatives, respectively. A series of2D-QSARand3D-QSAR models which have a certain reliability and good predictive ability have beenbuilt. This provides a valuable theoretical reference for the design of highly active NS5Bpolymerase inhibitor. The research content is as follows:We used multiple-linear regression (MLR) method to build QSAR models for theinhibitors of each data set of benzothiadiazine derivatives, respectively. All the QSAR modelshave been internal and external validated. For the GT-1a, which built by pEC50, the high R²(0.89),Q_test²(0.82) and Q_LOO²(0.73) with the low SEE (0.31) and SEP (0.38) show a reliableand good predictive model. However, for the GT-1b, the model built by pIC50is better.Through the3D-QSAR study on the35newly synthesized indole-based C-3pyridonederivatives, we establish a better CoMFA model (Q²=0.427, R_pred²=0.882) than CoMSIAmodel (Q²=0.376). It shows that the CoMFA model has good predictive ability. The LowSEE value (0.212) and SEE value (0.473) with high F-test value (62.30) show the model isreliable. From the CoMFA model we can observe that the contribution of steric field greaterthan the contribution of the electrostatic field for the inhibitory activity. It illustrate that stericfield has more pronounced influence to the interaction of the inhibitor and protein moleculesthan that of the electrostatic field. Moreover, molecular descriptors MATS3e and G3m greatlyaffect the activities of inhibitors.