Discovery of New non-steroidal selective glucocorticoid receptor agonists

Abstract

Glucocorticoids (GCs) are widely used as potent anti-inflammatory drugs; however, GC therapy is often accompanied by adverse side effects. The anti-inflammatory action of GCs is exerted through the glucocorticoid receptor (GR) in part by antagonizing the pro-inflammatory nuclear factor k B (NF-kB) whereas the majority of side effects are assumed to be mediated by transactivation of GR target genes. We set out to identify novel non-steroidal selective GR agonists (SEGRA) favoring transrepression of NF-kB target genes over transactivation of genes associated with undesirable effects. Our virtual screening protocol was driven by a pharmacophore model based on a pyrrolidinone amide analogue (named as ‘compound 12′ in Biggadike et al 2009, PNAS USA 106, 18,114) bound to the extended binding pocket of the GR ligand binding domain (GR-LBD). Ambinter library (7.8 million compounds) was queried by our validated pharmacophore hypothesis and the prioritized compounds were biologically evaluated using a series of well-established screening assays. Two structurally similar hits (1 and 13) were identified that bind to GR, induce its translocation to the nucleus, do not mediate transactivation of GR target genes whereas partially repress a number of pro-inflammatory NF-kB target genes, in a GR-dependent manner. Explanatory molecular dynamics (MD) calculations could detail the per-residue interactions accounting for the binding of 1 and 13 to the extended binding pocket of GR. The discovered 1,3-benzothiazole analogs introduce a new class of genuine SEGRA paving the way for hit-to-lead optimization.