Evaluation of the rodent Hershberger bioassay: testing of coded chemicals and supplementary molecular-biological and biochemical investigations.

Under the auspices of the Organization for Economic Cooperation and Development (OECD) the Hershberger assay is being validated as an in vivo screen for compounds with (anti)androgenic potential. We participated in the final activity, the testing of coded chemicals. Test compounds included trenbolone (TREN; 1.5, 40 mg/kg), testosterone propionate (TP; 0.4 mg/kg), flutamide (FLUT; 3mg/kg), linuron (LIN; 10, 100mg/kg), 1,1-bis-(4-chlorophenyl)-2,2-dichloroethylene (p,p'-DDE; 16, 160 mg/kg), and two negative reference substances, i.e., compounds not considered to affect androgen-sensitive tissue weights (ASTWs) in the Hershberger assay, namely 4-nonylphenol (NP; 160 mg/kg) and 2,4-dinitrophenol (DNP; 10mg/kg); TREN, LIN, p,p'-DDE, NP, and DNP being used under code. Compounds were administered for 10 days by oral intubation or subcutaneous injection (TP). Additional investigations not mandatorily requested by OECD included organ gravimetry of the liver, gene expression analysis in prostate using quantitative RT PCR for prostate specific binding protein polypeptide C3 (PBPC3) and ornithine decarboxylase 1 (ODC1) and determination of testosterone metabolizing and phase II conjugating enzymes in the liver. After submission of all study reports to OECD by participants uncoding revealed the following results: (A) When assessing androgenic potential in castrated rats, administration of TREN increased the weights of ventral prostate (VP), seminal vesicles (SV), glans penis, levator ani and bulbocavernosus muscles, and Cowper's glands at the high dose. A similar or stronger (VP, SV) increase of ASTWs was observed for TP; NP and DNP were ineffective. TREN dose-dependently increased gene expression of ODC1 and PBPC3, TP induced expression of these genes even more strongly (almost) to the level of untreated intact animals, whereas NP and DNP were inactive. Liver enzyme activities depending on physiological androgen levels were lower in castrated than in intact rats and could not be restored by androgen treatment. (B) When assessing antiandrogenic potential in TP-supplemented castrated rats, administration of LIN and p,p'-DDE decreased ASTWs only at the high dose. FLUT even more effectively decreased ASTWs, NP and DNP were again without effect. Decreases in androgen-responsive gene expression in the prostate corresponding to the organ weight changes were only observed for p,p'-DDE (high dose) and flutamide (PBPC3 only). p,p'-DDE dose-dependently induced liver weights and most liver enzyme activities including androgen-dependent ones. Our study accurately reproduced ASTW changes obtained in previous studies also under code suggesting that the Hershberger assay is a robust tool to screen for an (anti)androgenic potential. Assessment of ODC1 and PBPC3 gene expression in prostate, however, may only represent a sensitive tool for the detection of an androgenic potential. Finally, p,p'-DDE may affect ASTWs by several mechanisms including enhanced testosterone metabolism.