Computational Screening of Phthalate Monoesters for Binding to PPARγ

Phthalate esters are ubiquitous environmental contaminants that interact with peroxisome proliferator-activated receptors (PPARs), a family of nuclear receptors. Molecular docking and free energy calculations were performed in an effort to identify novel phthalate ligands of PPARγ, a subtype expressed in a wide range of human tissues. The method was validated using several agonists and partial agonists of PPARγ, whose binding orientations were correctly reproduced; however, reduced accuracy in docking was observed with ligands of increasing size and flexibility. Improved results were obtained by introduction of a more accurate scoring function based on the all-atom molecular mechanics potential CHARMM and a generalized Born/surface area solvation term ACE (analytical continuum electrostatics). Comparison of the lowest CHARMM/ACE energy of each phthalate vs the logarithm of the experimentally determined EC₅₀ value for PPARγ trans-activation yielded a good correlation (R² = 0.82). Thus, we can reliably distinguish phthalates that bind and activate PPARγ from those that do not, with the computational method predicting relative PPARγ binding activities with some degree of accuracy. We have applied this method to screen a series of 73 mono-ortho-phthalate esters listed in the Available Chemicals Directory. Several putative PPARγ binding phthalates were identified, including compounds that are known PPARγ agonists. These findings support the use of computational methods to identify environmental chemicals that warrant further experimental evaluation for PPAR binding and trans-activation potential in cell-based models.