Effects of antiandrogenic progestins, chlormadinone and cyproterone acetate, and the estrogen 17α-ethinylestradiol (EE2), and their mixtures: Transactivation with human and rainbowfish hormone receptors and transcriptional effects in zebrafish (Danio rerio) eleuthero-embryos.

Synthetic progestins act as endocrine disrupters in fish but their risk to the environment is not sufficiently known. Here, we focused on an unexplored antiandrogenic progestin, chlormadinone acetate (CMA), and the antiandrogenic progestin cyproterone acetate (CPA). The aim was to evaluate whether their in vitro interaction with human and rainbowfish (Melanotaenia fluviatilis) sex hormone receptors is similar. Furthermore, we investigated their activity in zebrafish (Danio rerio) eleuthero-embryos. First, we studied agonistic and antagonistic activities of CMA, CPA, and 17α-ethinylestradiol (EE2), in recombinant yeast expressing either the human progesterone (PGR), androgen (AR), or estrogen receptor. The same compounds were also investigated in vitro in a stable transfection cell system expressing rainbowfish nuclear steroid receptors. For human receptors, both progestins exhibited progestogenic, androgenic and antiestrogenic activity with no antiandrogenic or estrogenic activity. In contrast, interactions with rainbowfish receptors showed no progestogenic, but antiandrogenic, antiglucocorticoid, and some antiestrogenic activity. Thus, interaction with and transactivation of human and rainbowfish PGR and AR were distinctly different. Second, we analyzed transcriptional alterations in zebrafish eleuthero-embryos at 96 and 144h post fertilization after exposure to CPA, CMA, EE2, and binary mixtures of CMA and CPA with EE2, mimicking the use in oral contraceptives. CMA led to slight down-regulation of the ar transcript, while CPA down-regulated ar and pgr transcripts. EE2 exposure resulted in significant transcriptional alterations of several genes, including esr1, pgr, vtg1, cyp19b, and gonadotropins (fshb, lhb). The mixture activity of CMA and EE2 followed the independent action model, while CPA and EE2 mixtures showed additive action in transcriptional alterations. Third, we analyzed the interactions of binary mixtures of CMA and CPA, and of CMA and EE2 for their joint activity in vitro and in eleuthero-embryos. Both mixtures behaved according to the concentration addition model in their in vitro interaction with human and rainbowfish receptors, often showing antagonism. In zebrafish eleuthero-embryos, binary mixtures of CMA and EE2 showed the same expression patterns as EE2 alone, indicating an independent action in vivo. Our study demonstrates that CMA and CPA interact distinctly with human and rainbowfish receptors, suggesting that activities of these and possibly additional environmental steroids determined with yeast expressing human receptors cannot simply be translated to fish. The lack of agonistic activities of both progestins to rainbowfish PGR and AR is the probable reason for the low activity found in zebrafish eleuthero-embryos.