Virtual screening for the identification of novel nonsteroidal glucocorticoid modulators

In this work, we describe the first application of ligand-based drug design (LBDD) to the derivation of a predictive pharmacophore for the human glucocorticoid receptor (hGR). Creation of a four feature pharmacophore in Catalyst was subsequently validated through a virtual screen of 264000 commercially available compounds. From a selected hit list of 11 diverse compounds, two nonsteroidal molecules demonstrated low micromolar activity against hGR as validated through fluorescence polarization competitive assay. Additionally, these compounds were tested for their trans-repression potential by their ability to inhibit IL-1 induced, IL-6 expression in the human A549 lung epithelial cell line. Co-treatment of A549 with 21 (MDG169) (10 microM) in combination with dexamethasone showed an improved inhibitory effect when compared to dexamethasone alone with the cooperative effect being dependent on the dexamethasone dose. Putative binding orientations in the hGR ligand binding domain crystal structure are presented. These compounds represent novel nonsteroidal hGR modulating scaffolds, rationally identified through ligand-focused computational modeling.