Application of the Generalized Concentration Addition Model to Predict In Vitro Responses of Tertiary Mixtures of Glucocorticoid Ligands@

Prescription and endogenous glucocorticoid ligands are frequently reported in environmental surface water, sewage treatment plant influent, wastewater effluent, hospital effluent, and industrial effluent due to the inability of wastewater treatment plants to effectively eliminate these bioactive compounds, leading to potentially harmful exposures. As a result, it is increasingly imperative to predict adverse outcomes of exposure and expedite water quality testing. An emerging focus of toxicological research aims to improve these methods by developing computational models to predict joint-responses of chemical mixtures. In this study, we expand the application of the generalized concentration addition model and demonstrate its ability to accurately predict the summed response produced by tertiary mixtures containing full and partial agonists. This broadened application could save time, resources, and costs associated with experimental studies to determine joint-responses of more complex mixtures accurately representing environmental concentrations.

Impact/Purpose

Glucocorticoid receptor (GR) ligands are frequently detected in surface water and wastewater effluent in complex mixtures. While individual glucocorticoid environmental concentrations may be low, the combined total of potential mixtures may cause adverse effects, emphasizing the need to study complex mixture interactions. We previously demonstrated the generalized concentration addition (GCA) model most accurately predicts in vitro responses of binary GR ligand mixtures with a single partial agonist. Herein, we applied the GCA model to non-equipotent tertiary mixtures containing full and partial GR agonists. CV1-hGR cells were treated with individual compounds or non-equipotent tertiary mixtures and assessed for potency and relative efficacy. In mixtures containing greater relative concentrations of partial agonists, the maximal response observed decreased at the highest mixture concentrations tested and increased at lower concentrations tested. GCA, but not concentration addition (CA), model predictions fell within 95% confidence intervals of in vitro responses for tertiary mixtures containing at least one full agonist and one partial agonist. Furthermore, GCA model predictions fell within the same statistical distribution as the observed results, supporting the GCA model as an effective predictor of GR activation. However, neither the GCA model, CA model, nor a derived homodimer composite function provided predictions within the 95% confidence interval of observed results for tertiary mixtures containing only partial agonists. Environmental mixtures are complex, potentially containing multiple partial agonists. Therefore, the expanded application of the GCA model may offer additional insight into GR chemical mixture interactions, supporting future water quality screening and establishing effects-based limits for routine monitoring.

Citation

Green, E., M. Cardon, N. Evans, AND E. MedlockKakaley. Application of the Generalized Concentration Addition Model to Predict In Vitro Responses of Tertiary Mixtures of Glucocorticoid Ligands@. Society of Toxicology, San Diego, California, March 27 - 31, 2022.