Robust Array-Based Coregulator Binding Assay Predicting ERα-Agonist Potency and Generating Binding Profiles Reflecting Ligand Structure

Testing chemicals for their endocrine-disrupting potential, including interference with estrogen receptor (ER) signaling, is an important aspect of chemical safety testing. Because of the practical drawbacks of animal testing, the development of in vitro alternatives for the uterotrophic assay and other in vivo (anti)estrogenicity tests has high priority. It was previously demonstrated that an in vitro assay that profiles ligand-induced binding of ER alpha to a microarray of coregulator-derived peptides might be a valuable candidate for a panel of in vitro assays aiming at an ultimate replacement of the uterotrophic assay. In the present study, the reproducibility and robustness of this coregulator binding assay was determined by measuring the binding profiles of 14 model compounds that are recommended by the Office of Prevention, Pesticides and Toxic Substances for testing laboratory proficiency in estrogen receptor transactivation assays. With a median coefficient of variation of 5.0% and excellent correlation (R-2 = 0.993) between duplicate measurements, the reproducibility of the ER alpha-coregulator binding assay was better than the reproducibility of other commonly used in vitro ER functional assays. In addition, the coregulator binding assay is correctly predicting the estrogenicity for 13 out of 14 compounds tested. When the potency of the ER-agonists to induce ER alpha-coregulator binding was compared to their ER binding affinity, their ranking was similar, and the correlation between the EC50 values was excellent (R-2 = 0.96), as was the correlation with their potency in a transactivation assay (R-2 = 0.94). Moreover, when the ER alpha-coregulator binding profiles were hierarchically clustered using Euclidian cluster distance, the structurally related compounds were found to cluster together, whereas the steroid test compounds having an aromatic A-ring were separated from those with a cyclohexene A-ring. We concluded that this assay is capable of distinguishing ER alpha agonists and antagonists and that it even reflects the structural similarity of ER alpha agonists, indicating a potential to achieve identification and classification of ER alpha endocrine disruptors with high fidelity.