Bisphenol A was first synthesized in 1891 as a synthetic estrogen, long before industry discovered its usefulness in polycarbonate plastics and epoxy resins. Today it is produced in quantities of millions of tons per year and appears in food containers, thermal receipts, dental sealants, medical equipment, and personal care products. Because it leaches from these materials at every stage from manufacturing to disposal, humans are exposed continuously through food, dust inhalation, and skin contact. Researchers have detected the compound in serum, urine, breast milk, placental tissue, amniotic fluid, follicular fluid, and umbilical cord blood, confirming that it crosses the placenta and reaches the developing fetus.

The reason bisphenol A interferes with hormones is structural: its two phenolic rings echo the shape of 17β-estradiol, the body’s principal estrogen, allowing it to dock into the same pocket on estrogen receptor alpha. Once bound, the chemical-receptor complex moves into the nucleus, where it can bind to estrogen response elements on DNA and recruit co-activator proteins to switch genes on or off. Depending on dose, tissue, and receptor subtype, bisphenol A can either mimic or block estrogenic signaling, disrupting pathways governing reproduction, immunity, and brain development. When regulators restricted its use, manufacturers turned to structural cousins such as bisphenol S, bisphenol F, and bisphenol AF, but mounting evidence indicates many of these substitutes carry estrogenic or anti-estrogenic activity equal to or exceeding the original.

Testing every new chemical with traditional animal studies is impractical, so computational screening has become a central tool in endocrine risk assessment. In the new study, Satpathy assembled a panel of twenty endocrine-disrupting bisphenols and evaluated how each one binds the ligand-binding domain of human estrogen receptor alpha, using the high-resolution crystal structure PDB 3ERT as the starting point. Missing residues in the crystal structure were rebuilt with homology modeling, and the refined model passed rigorous validation checks, including an ERRAT quality factor above 95 percent and a Ramachandran plot showing no residues in disallowed regions.

The docking protocol itself was put through its paces before any bisphenols were scored. The co-crystallized ligand 4-hydroxytamoxifen was removed and re-docked, reproducing its crystallographic pose with an acceptable deviation, and caffeine was included as a negative control, binding weakly as expected for a compound with no known estrogen receptor affinity. Two independent docking engines, MzDOCK and AutoDock Vina, produced consistent rankings. Against this validated backdrop, one substitute stood out dramatically: bisphenol PH, or BPPH, achieved a docking score of −11.4 kcal/mol, comfortably stronger than 17β-estradiol itself at −9.9 kcal/mol.

Interaction analysis revealed why. All three reference ligands, bisphenol A, bisphenol PH, and estradiol, occupy the canonical estrogen-binding cavity and make contact with the same conserved residues, including Met343, Leu346, Leu387, Met388, Leu391, Arg394, and Phe404. Binding is dominated by hydrophobic contacts, with only a few hydrogen bonds anchoring each ligand. The phenolic hydroxyl groups of estradiol and bisphenol PH both form stable hydrogen bonds with Glu353, an interaction thought to be important for receptor activation, while the bulkier aromatic system of bisphenol PH adds extra hydrophobic and pi-stacking interactions that appear to lock it more firmly into the pocket.

Docking scores alone can mislead, so the study turned to molecular dynamics simulations, which track every atom of the protein-ligand complex as it moves in a bath of water and salt at physiological temperature. Three independent 50-nanosecond simulations of the bisphenol PH complex, run with different random seeds, all showed the same picture: the protein backbone settled quickly, the ligand remained firmly in place, and the hydrogen bonds it formed persisted throughout. An extended 100-nanosecond simulation confirmed that both the unbound receptor and the bisphenol PH complex reached stable equilibrium after roughly 20 nanoseconds, with the ligand maintaining its key contacts for the remainder of the run.

Free-energy calculations using the MM-PBSA method, which combines molecular mechanics energies with a model of solvation, ranked the complexes in the same order as the docking scores. Bisphenol PH showed the strongest predicted binding at +112.38 kJ/mol on the YASARA scale, where larger positive values indicate stronger binding, followed by estradiol at +86.63 kJ/mol and bisphenol A at +72.06 kJ/mol. The convergence of two independent computational approaches on the same ranking strengthens the case that bisphenol PH forms an unusually stable complex with the receptor.

The most striking results concern helix-12, a short helix at the receptor’s C-terminus that acts as the molecular switch for estrogen signaling. When estradiol binds, helix-12 folds closed over the ligand pocket, creating the AF-2 groove where co-activator proteins attach; antagonists displace the helix and block activation. Structural superposition showed that bisphenol PH holds helix-12 in an agonist-like position nearly identical to estradiol, with a helix-specific deviation of just 1.326 angstroms, while bisphenol A shifted the helix considerably more, at 2.187 angstroms. Dynamic cross-correlation analysis then revealed something subtler: estradiol coordinates helix-12 with helix-3 through strong anti-correlated motion, bisphenol PH partially preserves this coupling, and bisphenol A nearly abolishes it, decoupling the helix from the receptor core altogether.

That graded pattern, from full agonist-like coordination through partial mimicry to near-complete decoupling, suggests that bisphenols do not simply act as weak copies of estrogen. Instead, each compound imposes its own dynamic signature on the receptor, potentially altering which co-regulators are recruited and which genes are ultimately expressed. The study’s author notes that computational results are predictive rather than definitive: the simulations cover only the ligand-binding domain, run for at most 100 nanoseconds, and will need microsecond-scale trajectories, mutant-receptor comparisons, and laboratory validation to confirm biological activity. Still, the message for toxicology is clear. Judging endocrine disruptors by static binding affinity alone misses the dynamic story, and the substitutes marketed as safer alternatives may disturb estrogen receptor signaling through mechanistic pathways all their own.

Subject of Research: Computational analysis of how bisphenol A and its analogues bind and allosterically modulate estrogen receptor alpha

Article Title: Structural and dynamic characterization of estrogen receptor-α binding by bisphenol A and its analogues

Article References: Satpathy, R. (2026). Structural and dynamic characterization of estrogen receptor-α binding by bisphenol A and its analogues. Discover Chemistry, 3(1), Article 462. https://doi.org/10.1007/s44371-026-00911-2

Image Credits: AI Generated

DOI: 10.1007/s44371-026-00911-2

Keywords: bisphenol A, estrogen receptor alpha, endocrine disruptors, molecular docking, molecular dynamics simulation, MM-PBSA, helix-12, bisphenol PH, 17β-estradiol, ligand-receptor interaction, allosteric signaling, toxicology