Virtual screening against p50 NF-κB transcription factor for the identification of inhibitors of the NF-κB-DNA interaction and expression of NF-κB upregulated genes

Virtual screening against NF-κB p50 using docking simulations was applied by starting from a three-dimensional (3D) database containing more than 4.6 million commercially available structures. This database was filtered by specifying a subset of commercially available compounds sharing a (2E,Z)-3-(2-hydroxyphenyl)-2-propenoate substructure and relevant druglike properties. Docking to p50 NF-κB was performed with a test set of six known inhibitors of NF-κB-DNA interactions. In agreement with docking results, the highest-scored compound displayed a high level of inhibitory activity in electrophoretic mobility shift assay (EMSA) experiments (inhibition of NF-κB-DNA interactions) and on biological functions dependent on NF-κB activity (inhibition of IL-8 gene expression in cystic fibrosis IB3-1 cells). We found that this in silico screening approach is suitable for the identification of low-molecular-weight compounds that inhibit NF-κB-DNA interactions and NF-κB-dependent functions. Information deduced from the discovery of the new lead compound and its binding mode could result in further lead optimization resulting in more potent NF-κB inhibitors. © 2009 Wiley-VCH Verlag GmbH & Co. KGaA.