You are here:
Evaluation of Alternative Estrogen Activity Assays for New Computational Models to Support the Endocrine Disruption Screening Program
Citation:
Sadrpour, P., S. Lynn, R. Judson, AND S. Simmons. Evaluation of Alternative Estrogen Activity Assays for New Computational Models to Support the Endocrine Disruption Screening Program. SOT Conference 2024: A New Approach Method (NAM) to Screen for the Impact of Endogenous Stress on Chemical Toxicity, Salt Lake City, UT, March 10 - 14, 2024. https://doi.org/10.23645/epacomptox.25556079
Impact/Purpose:
Presentation to SOT Conference 2024: A New Approach Method (NAM) to Screen for the Impact of Endogenous Stress on Chemical Toxicity
Description:
Background and Purpose The Endocrine Disruption Screening Program (EDSP) was launched by EPA to screen pesticides and other chemicals for the potential to perturb the estrogen, androgen, and thyroid pathways in humans and wildlife. Recently, EPA has worked to develop in vitro high throughput assays and computational models that could be alternatives for EDSP Tier 1 in vivo assays. A computational model to predict estrogen receptor (ER) activity was developed using 18 ToxCast assays and has been proposed as an alternative for the EDSP Tier 1 uterotrophic assay. Subsequent analyses demonstrated that a model using as few as four high quality assays could provide comparable predictivity, provided those assays covered the key events in ER pathway activation: (1) ligand binding, (2) dimerization, (3) transcriptional activation (TA), and (4) proliferation. Some of the original ToxCast ER assays used to build the streamlined model are either no longer available or have other barriers that prevent future use. The purpose of this study is to evaluate widely available ER assays for dimerization and TA to replace the original ToxCast assays in a new predictive ER model that is more amenable to implementation and adoption. Methods Four ER TA assays (T47D-KBluc, hERα_HeLa_9903, ERα_CALUX and VM7Luc4E2) and one ER dimerization assay (ER12zf2) were optimized for 384-well format screening and evaluated for performance using 31 reference chemicals listed in OECD Test Guideline 455. Seeding densities, serum (FBS) concentrations in the assay medium and chemical exposure times were optimized prior to reference chemical screening. Reference chemicals were tested at 0.1 pM to 100 µM under optimized conditions for each assay. Data were fit to one of three models (Hill, gain-loss, or constant) and half-maximal response concentration (AC50) values were derived using tcpl R package. The receiver operating characteristic (ROC) was used to optimize activity thresholds. Balanced accuracies (BA) were calculated for each candidate assay and the corresponding ToxCast assay using either the reference activities listed by OECD or those for the uterotrophic assay. Potency concordance was calculated using Pearson correlations of the AC50 values of the candidate assays to those from the corresponding ToxCast assay. Results Higher serum concentrations in assay medium and a 24-hour chemical exposure enhanced the performance of each of the TA assays, while a 2% FBS assay medium and 6-hour exposure were optimal for the ER12zf2 dimerization assay. ROC determined a peak activity threshold of 25% Estradiol response for all five assays and was used to delineate active chemicals. The T47D-KBluc, hERα_HeLa_9903, and ERα_CALUX assays predicted OECD reference chemical activity with BA of 98%, 89% and 100%, respectively, compared to ATG_ER_TRANS (ATG), the best performing ToxCast TA assay at 95%. A subset of 16 uterotrophic reference chemicals was predicted with 100% accuracy by the ATG assay whereas each of the candidate TA assays miscategorized two phthalates as active, lowering their BA to 75%. The ER12zf2 assay predicted OECD reference chemical activity with a BA of 89%, compared to 88% for OT_ER_ERaERb_1440 (OT), the best performing ToxCast dimerization assay. Moreover, ER12zf2 was more predictive of the uterotrophic reference chemicals (75% vs. 58%). Potency correlations of the candidate TA assays compared to ATG varied: 0.89, 0.65 and 0.93 for T47D-KBluc, hERα_HeLa_9903, and ERα_CALUX, respectively. All candidate TAs produced lower AC50 values than ATG for six high potency chemicals: 17α- and 17β- estradiol, diethylstilbestrol, estrone, meso-hexestrol, and kepone. The potency correlation of the OT assay and ER12zf2 was 0.92. ER12zf2 produced AC50 values at least 10 times lower than OT for all but one reference chemical, and these . . .
URLs/Downloads:
DOI: Evaluation of Alternative Estrogen Activity Assays for New Computational Models to Support the Endocrine Disruption Screening Program
PARISA_SOT_FINAL.PDF (PDF, NA pp, 1372.382 KB, about PDF)