The Peroxisome Proliferator Activated Receptor γ (PPARγ) is a nuclear receptor that acts as the master regulator of adipogenesis and lipid metabolism and is involved in type 2 diabetes mellitus (T2DM), cardiovascular disease (CVD), cancer and inflammation. Drugs that target PPARγ for treatment of T2DM improve insulin sensitivity, however these drugs also cause severe side effects including weight gain, decreased bone density and increased risk of CVD. These effects are caused by supraphysiological activation of PPARγ beyond what endogenous ligands can achieve; hence ligands that induce a similar or decreased level of transactivation compared to endogenous ligands (e.g. partial agonists, antagonists and inverse agonists) may present a novel treatment without adverse side effects.
AI was used to identify novel compounds predicted to bind to the PPARγ ligand binding domain (LBD) to regulate its activity. We selected the top 80 compounds and tested them in vitro to investigate binding affinity for PPARγ using fluorescence polarisation assays with a probe that binds competitively to the PPARγ LBD, allowing for detection of both agonists and antagonists. We then tested these compounds in cell-based luciferase reporter assays to determine their effects on the transcriptional activity of PPARγ. Several compounds were identified which displayed high affinity for PPARγ and lower than 30% activation compared to a full agonist.
To guide optimisation of hits from our drug screen, we investigated the structure activity relationship of PPARγ via generating several PPARγ LBD constructs with a variety of point mutations including structure-based alanine scanning mutants focused on ligand binding, and clinically relevant mutants associated with disease.
Cell-based reporter assays demonstrated that most mutations decreased transactivation potential, but several displayed higher maximum transactivation. In vitro binding assays were used to determine the effect of the point mutations on the binding affinity of a known high affinity ligand, correlating changes in structure with ligand binding and transactivation.
These data will inform the optimisation of hits from the initial drug screen to increase the potency of our ligands while minimising interactions which lead to full agonism. Once further optimised, these compounds may present novel drug leads for the treatment of T2DM.