Bisphenol A (BPA) is a widely used plasticizer and endocrine disruptor that binds nuclear receptors such as the estrogen receptor α (ERα) and estrogen-related receptor γ (ERRγ), altering hormonal signaling even at low-exposure levels. Lignin-derived bisphenols, renewable aromatic dimers from biomass, have been proposed as less bioactive BPA alternatives. Accurately predicting their estrogenic activity by molecular simulation is challenging because ERα and ERRγ function as homodimers whose two ligand-binding pockets can exhibit distinct binding preferences and complex sampling behavior. To address this complexity, we compared workflows for absolute and relative binding free-energy calculations that incorporate enhanced sampling, including conventional free-energy perturbation (FEP), replica-exchange thermodynamic integration (RETI), and interleaved double-wide sampling free-energy perturbation (IDWS-FEP) to assess how effectively each approach captures the pocket-specific heterogeneity that arises in homodimeric receptors. This study provides guidance for selecting reliable binding free-energy workflows for homodimeric estrogen receptors and supports future computational screening of lignin-derived bisphenols as renewable, potentially safer BPA substitutes.
Lin et al. (Sun,) studied this question.