Citation:

MedlockKakaley, E., M. Cardon, E. Gray, P. Hartig, AND V. Wilson. Generalized concentration addition mixtures model predicts glucocorticoid receptor activation by environmental full and partial agonists. Society of Environmental Toxicology and Chemistry Annual Meeting, Sacramento, California, November 04 - 08, 2018.

Impact/Purpose:

We, and others, have detected glucocorticoid receptor (GR) agonist activity in waste and surface waters domestically and around the world. This study describes the characterization of environmentally relevant GR ligands in a cell-based detection method. Herein, we illustrate the variety of responses obtained from the selected suite of chemicals and the complexities of modeling specific GR agonists in environmental mixtures.

Description:

Abstract describing study design, results and conclusions will be presented in poster or platform format to attendees of the 2018 Society of Environmental Toxicology and Chemistry annual meeting pending acquisition of travel funds. Abstract: Endocrine disrupting compounds are pervasive in surface and waste waters worldwide. We, and others, have detected complex mixtures of glucocorticoid receptor (GR) agonists using both chemical analysis and non-targeted effects-based methods in recent water quality surveys. Although the GR is ubiquitously expressed throughout tissues of aquatic species and has been implicated in growth, reproduction, metabolism, and osmoregulation, the way in which a mixture contributes to a single biological endpoint remains unknown. Therefore, we characterized environmentally relevant GR agonists using a CV1 cell line transcriptional activation assay. Cells were treated with individual GR ligands, a fixed ratio mixture of full and partial agonists, or using a two-chemical matrix design with full and partial agonists. Individual agonist efficacy (full vs. partial agonism) varied and potency (EC50) ranged over several orders of magnitude, 48.09 to 102.5% and 1.278 x 10-10 to 3.93 x 10-8 M, respectfully. Concentration addition (CA) and response addition (RA) mixture models predicted concentration response curves within the observed 95% prediction interval bands (EC50(OBS) = 2.341e-10 M, 95% CI 2.114e-10 to 2.598e-10 M) for an equipotent mixture of 12 full agonists. However, for equipotent mixtures containing partial agonists (21-hydroxyprogesterone or corticosterone), the observed maximum efficacies were significantly lower than the CA and RA model-predicted maximum efficacies. Therefore, the generalized concentration addition (GCA) mixtures model, which accounts for agonists with <100% reference agonist efficacy, was applied to predict non-equipotent mixtures containing full and partial GR agonists. The GCA modeled responses of matrix exposures for partial agonist + full agonist, and full agonist + full agonist, mixtures were identified as part of the same distribution as observed responses (p < 0.05). The GCA model resulted in the best overall fit for mixtures containing both full and/or partial GR agonists compared to the models without efficacy variables. Elucidating the mechanistic basis of GR activation by environmental ligands will not only aid in characterizing environmental mixture samples using this effects-based approach, but will also provide a foundation for future studies in predicting and interpreting potential phenotypic adverse outcomes when the approach is applied as a chemical screening tool. Abstract does not necessarily reflect USEPA views or policy.

Record Details: