Citation:

Gray, E., J. Conley, C. Lambright, N. Evans, V. Wilson, AND p. foster. Biological Relevance of Key Events (KE) in utero in The Androgen Adverse Outcome Pathway Network (AOPn) to Adverse Effects in F1 Male Rats. SOT ANNUAL MEETING, SAN ANTONIO, TX, March 10 - 16, 2018.

Impact/Purpose:

IMPACT: WE HAVE IDENTIFIED KEY EVENTS IN SHORT-TERM STUDIES OF TWO PATHWAYS RELATED TO THE DEVELOPMENT OF MALE REPRODUCTIVE TRACT MALFORMATIONS THAT CAN BE USE TO PREDICT THE IN UTERO DOSES THAT CAUSE ADVERSE EFFECTS IN LONG TERM STUDIES. THIS ALLOWS RISK ASSESSORS TO CHARACTERIZE THE HAZARD FOR THESE TYPES OF EFFECTS WITH FAR FEWER ANIMALS WHILE STILL PROVIDING HIGH QUALITY DATA

Description:

We are conducting studies to evaluate the biological relevance of changes in KEs and molecular initiating events (MIE) in AOPs to determine if these can accurately predict of the dose levels of chemicals that disrupt the androgen signaling pathway in utero. Herein, we focus on chemicals that disrupt fetal testis endocrine function and androgen receptor (AR) antagonists. Our objective is to determine if KE ED50s or MIE EC50s can be used to predict the doses that produce adverse effects after exposure during sexual differentiation. While several MIEs for the phthalate AOP have been proposed, the data indicate that PPAR alpha agonists (pirinixic acid, clobirate and FRD-903 (GenX)), and COX inhibition (Paracetamol), in contrast with the phthalate esters (PE), do not reduce fetal testis testosterone production (TPROD) or gene expression profiles. Since the MIE is unknown, we have developed a statistical model of the relationships between the KEs with the adverse effects of PEs. This model uses data from several different PEs (>8) and compares reductions in fetal TPROD and testis gene expression with the doses that produce reproductive malformations. For AR antagonists, statistical analyses of data for> 50 chemicals indicate that the in vitro EC50 values for AR antagonism are poorly correlated the ED50 values for effects on male rat reproductive development (R2 less than 10%). For example, potency rankings of the in vitro EC50s with the in vivo ED50s for BPC, BPA, vinclozolin and pyrfiluquinazon are negatively correlated. In contrast, the ED50 values for the inhibition of androgen-dependent tissue growth in short-term in vivo assays are highly correlated (R2>85%) with the ability of the chemical to disrupt androgen-dependent reproductive tract development in utero. While at present we are unable to entirely replace animal use to interrogate the in utero effects of chemicals using the MIEs for PEs or AR antagonists, these models enable us to significantly reduce animals use by using two short term assays to identify points of departure for hazard identification. However, these KEs would not be able to predict the adverse effects of chemicals that disrupt this AOPn via different AOPs. Further research is needed to identify as yet unquantified AOPs in the AOPn to improve the predictive ability across all MIEs and to integrate these into a single predictive statistical model. This abstract does not reflect Agency Policy.

Record Details: