Mapping the Allosteric Landscape of PPAR γ : a Markov State Modeling and Energetic Analysis Approach

Peroxisome proliferator-activated receptor γ (PPARγ) is a nuclear receptor whose functional versatility plays a central role in metabolic regulation due to its ligand-dependent conformational dynamics. However, the molecular mechanism through which different ligand classes selectively modulate the conformational landscape of PPARγ remains poorly understood. In this study, we applied an integrative computational approach combining extensive molecular dynamics (MD) simulations, Markov state modeling (MSM), and binding free energy calculations to dissect the dynamic allostery underlying PPARγ activation and inhibition. Our simulations revealed that ligand binding reshapes the conformational landscape of PPARγ by selectively altering dynamic hubs within the R2 (Gly284-Gln294) and R3 (Ile341-Gln345) regions. Energetic analysis revealed a thermodynamic hierarchy of ligand efficacy, with antagonists exhibiting the highest binding affinity (−76.19 kcal·mol–1), driven predominantly by hydrophobic interactions that sterically restrict Helix 12 (H12) mobility. Key residues, particularly Arg288 and Ile341, were identified as critical nodes within the allosteric network. To prospectively validate this mechanistic model, we conducted an MSM-guided virtual screening of the ZINC20 database. This yielded two natural compounds, ZINC000000834437 and ZINC000008952648, predicted by our model to stabilize distinct receptor conformations associated with antagonist-like and partial agonist-like states, respectively. Overall, this work provides atomistic insights into the conformational selection mechanism of PPARγ, and presents a transferable computational framework for probing allosteric regulation in nuclear receptors, with direct implications for the design of selective modulators.