Identification of novel small molecule TGF-β antagonists using structure-based drug design

Aberrant transforming growth factor-β (TGF-β) signalling has been associated with a number of disease pathologies, such as the development of fibrosis in the heart, lung and liver, cardiovascular disease and cancer, hence the TGF-β pathway represents a promising target for a variety of diseases. However, highly specific ways to inhibit TGF-β signalling need to be developed to prevent cross-talk with related receptors and minimise unwanted side effects. We have used used virtual screening and molecular docking to identify small molecule inhibitors of TGF-β binding to TßRII. The crystal structure of TGF-β3 in complex with the extracellular domain of the type II TGF-β receptor was taken as a starting point for molecular docking and we developed a structure-based pharmacophore model to identify compounds that competitively inhibit the binding of TGF-β to TβRII and antogonize TGF-β signalling. We have experimentally tested 67 molecules suggested by in silico screening and similarity searching for their ability to inhibit TGF-β signalling in TGF-β-dependent luciferase assays in vitro and the molecule with the strongest inhibition had an IC50 of 18 μM. These compounds were selected to bind to the SS1 subsite (composed of F30, C31, D32, I50, T51 S52, I53, C54 and E55) of TßRII and all share the general property of being aromatic and fairly flat. Molecular dynamics simulations confirmed that this was the most likely binding mode. The computational methods used and the hits identified in this study provide an excellent guide to medicinal chemistry efforts to design tighter binding molecules to disrupt the TGF-β/TßRII interaction.