3D-QSAR studies on the inhibitors of AP-1 and NF-κB mediated transcriptional activation

Comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) were performed on a series of 68 inhibitors of AP-1 and NF-κB mediated transcriptional activations. The CoMFA model produced statistically significant results with the cross-validated q² of 0.594 and the conventional correlation coefficient r² of 0.968. The best CoMSIA model was obtained by the combination use of steric, electrostatic, hydrogen-bond donor and acceptor fields. The corresponding q² and r² of CoMSIA model were 0.703 and 0.932, respectively. From the cross-validated results, it can be seen that the CoMSIA model has a better predictive ability than CoMFA model due to the importance of the hydrogen bonds for the activity of these inhibitors. The predictive abilities of the two models were further validated by a test set of 15 compounds. The models gave predicted correlation coefficient r_pred² of 0.891 for CoMFA model and 0.810 for CoMSIA model. Based on the above results, we identified the key structural features that may help to design potent inhibitors with improved activities: (1) the NH linker at the position R₄ acts as important hydrogen-bond donor and any group on phenyl or 2-thienyl ring of R₁ substituent decreases inhibitory activity, (2)further structural modification of compound 50 on the phenyl ring of the quinazoline ring considering steric, electrostatic and hydrogen-bond acceptor properties will influence the inhibitory activity.