Evaluation of androgen assay results using a curated Hershberger database

• Published in: Reproductive toxicology (Elmsford, N.Y.) N.Y.), 2018-10, Vol.81, p.272-280
• Main Authors: Kleinstreuer, N.C., Browne, P., Chang, X., Judson, R., Casey, W., Ceger, P., Deisenroth, C., Baker, N., Markey, K., Thomas, R.S.
• Format: Article
• Language: English
• Subjects: Androgen Antagonists - toxicity; Androgen receptor; Androgenic; Androgens - toxicity; Animals; Anti-androgenic; Biological Assay; Databases, Factual; Endocrine disruptor; Hershberger; High-throughput screening; In vitro; Male; Models, Biological; Rats; Receptors, Androgen - metabolism; Reproducibility of Results; ToxCast/Tox21
• ISSN: 0890-6238; 1873-1708
• DOI: 10.1016/j.reprotox.2018.08.017

[Display omitted] •We used 39 reference chemicals with reproducible AR pathway effects in rodents to evaluate the ToxCast/Tox 21 AR model based on 11 in vitro assays.•Agreement was 66% (19/29), with ten inconclusive in vitro model chemicals; most discrepancies were explained by in vitro to in vivo extrapolation.•In vitro false negatives were low potency chemicals in vivo; lowest effect levels were above doses estimated from top in vitro tested concentration.•The AR model had 100% positive predictive value, i.e. no false positives and chemicals with conclusive results were consistently positive in vivo.•The AR model is a critical tool for rapidly screening chemicals and anchoring adverse effects in intact animals to an endocrine mode of action. A set of 39 reference chemicals with reproducible androgen pathway effects in vivo, identified in the companion manuscript [1], were used to interrogate the performance of the ToxCast/Tox 21 androgen receptor (AR) model based on 11 high throughput assays. Cytotoxicity data and specificity confirmation assays were used to distinguish assay loss-of-function from true antagonistic signaling suppression. Overall agreement was 66% (19/29), with ten additional inconclusive chemicals. Most discrepancies were explained using in vitro to in vivo extrapolation to estimate equivalent administered doses. The AR model had 100% positive predictive value for the in vivo response, i.e. there were no false positives, and chemicals with conclusive AR model results (agonist or antagonist) were consistently positive in vivo. Considering the lack of reproducibility of the in vivo Hershberger assay, the in vitro AR model may better predict specific AR interaction and can rapidly and cost-effectively screen thousands of chemicals without using animals.