No abstract

The cDNA of rainbow trout estrogen receptor (rtER), highly and stably expressed in yeast, Sacchat-om j c e s c,erevisirie, was used to analyse the biological activity of the receptor. The rtER niRNA encoded a 65-kDa protein which was iininunorevealed by a specific antibody and migrated with the authentic rtER major protein form detected in trout liver. Yeast rtER bound estradiol with high affinity and the dissociation constant (KCI = 1.35 nM) was very similar to the value measured from trout liver extracts but 3-5-fold higher than the KtI found for human estrogen receptor (hER). This indicates therefore that the rtER has a lower estradiol affinity compared to the human receptor. While the hER K,, remained unchanged at both 4°C or 22"C, it was slightly modified at 30°C. The K,I measured for rtER at 22°C and 30°C were about 2-fold, and 12-fold higher, respectively, than the K,( obtained at 4°C suggesting an alteration of the rtER affinity for its ligand at elevated temperature. To examine the estrogen-receptor-mediated activation of transcription in yeast, reporter plasmids integrated or not in the yeast genome were used. The reporter genes consist of one, two, or three copies of estrogen-responsive elements (ERE) upstream of the yeast proximal CYCI or UKA3 promoters fused to the lacZ gene of Excherichia coli coding for p-galactosidase.The induction of /I-galactosidase activity for all reporter genes was strictly dependent on the presence of rtER and estrogens. The activation of transcription mediated by rtER responded in an estradiol-dosedependent manner as in animal cells. However, compared to hER, the estradiol concentration necessary to achieve maximal activation was 10-fold higher. This is probably a consequence of the lower estradiolaffinity for rtER compared to hER. The levels of induction of the reporter genes containing two or three ERE were \trongly enhanced compared to the one ERE construct. This is in agreement with the synergistic effect previously described for multiple ERE. The magnitudes of transcriptional induction mediated by rtER and hER were similar when the reporter gene containing three ERE was used but changed when the one ERE construct was used. In this case transcriptional activation mediated by rtER was 10-20-fold lower. This suggests that rtER requires protein/protein interaction for i t s stabilization on DNA. Antiestrogens were able to bind rtER and promote gene transcription. However, to produce effects comparable to those obtained with estrogens, much higher concentrations were required. This may imply nonetheless that antihormones were capable of provoking efficient interactions of rtER with the transcriptional machinery Keywonl.~; steroid action ; estrogen receptor; yeast expression; transcriptional activity.

Rainbow trout hepatocyte primary culture was used to test the influence of some xenobiotics on the expression of two genes implicated in reproduction, those for the estrogen implicated in reproduction, those for the estrogen receptor (ER) and vitellogenin (Vg). We showed that chlordecone, nonylphenol, a polychlorobiphenol (PCB) mixture (Aroclor 1245) and lindane were able to induce ER and Vg mRNA accumulation. Antiestrogens, 4-hydroxytamoxifen and ICI 164,384, prevented the effects of the xenobiotics, indicating that the induction of gene expression is mediated by the ER. Among these four xenobiotics, only chlordecone and nonylphenol were able to displace the binding of [3H]estradiol to ER-enriched COS-1 extracts, and to activate an estrogen-dependent reporter gene (ERE-TK-CAT) cotransfected with an expression vector containing ER cDNA. The results suggest that chlordecone and nonylphenol are direct inducers of rainbow trout ER and Vg gene expression, whereas PCBs and lindane act through their hepatic metabolites. Moreover, pentachlorophenol acts as an antagonist of the induction by estradiol of rainbow trout ER and Vg gene expression.

In the teleost fish, physiological and biochemical studies suggest that glucocorticoids regulate both salt balance and metabolic activities. In mammals, however, these functions are divided between glucocorticoids and mineralocorticoids. In mammals, separate receptors for these two classes of steroid hormone have been cloned and sequenced. To begin to understand the regulation in fish of the vital processes ascribed to glucocorticoids, we have cloned, sequenced, expressed, and studied the steroid-binding and transcriptional activation capabilities of the rainbow trout (Onchorhynchus mykiss) glucocorticoid receptor. Northern blot analysis shows a single rainbow trout GR messenger RNA species of 7.5 kilobases expressed in gill, intestine, skeletal muscle, kidney, and liver. The trout GR 2274-nucleotide coding sequence provides for a protein of 758 amino acids, with appropriate similarities to mammalian GR, with one striking exception. As in other members of the steroid/thyroid/retinoid receptor family, the DNA-binding domain contains two putative zinc fingers. These have high homology with those of other GRs. However, between the zinc fingers in the trout GR are found 9 more amino acids than are seen in mammalian GRs, raising questions as to the functional form of the fish, as opposed to the mammalian, GR. It has been proposed that as fish appear to use glucocorticoids for both metabolic and salt control, presumably through a single GR, GR would prove to be the evolutionary precursor to mammalian GR and mineralocorticoid receptor (MR). Computer analysis of the known sequences of GRs and MRs, however, suggests that the fish GR did not give rise to the MR of higher animals, but that both subfamilies of receptor arose from some earlier gene.

A relation between the chemical structure of a xenobiotic and its steroidal action has not yet been clearly established. Thus, it is not possible to define the estrogenic potency of different xenobiotics. An assessment may be accomplished by the use of different bioassays. We have previously developed a yeast system highly and stably expressing rainbow trout estrogen receptor (rtER) in order to analyze the biological activity of the receptor. The recombinant yeast system appears to be a reliable, rapid and sensitive bioassay for the screening and determination of the direct interaction between ER and estrogenic compounds. This system was used in parallel with a more elaborate biological system, trout hepatocyte aggregate cultures, to examine the estrogenic potency of a wide spectrum of chemicals commonly found in the environment. In hepatocyte cultures, the vitellogenin gene whose expression is principally dependent upon estradiol was used as a biomarker. Moreover, competitive binding assays were performed to determine direct interaction between rtER and xenobiotics. In our study, 50%of the 49 chemical compounds tested exhibited estrogenic activity in the two bioassays: the herbicide diclofop-methyl; the fungicides biphenyl, dodemorph, and triadimefon; the insecticides lindane, methyl parathion, chlordecone, dieldrin, and endosulfan; polychlorinated biphenyl mixtures; the plasticizers or detergents alkylphenols and phthalates; and phytoestrogens. To investigate further biphenyl estrogenic activity, its principal metabolites were also tested in both bioassays. Among these estrogenic compounds, 70% were able to activate rtER in yeast and hepatocytes with variable induction levels according to the system. Nevertheless, 30% of these estrogenic compounds exhibited estrogenic activity in only one of the bioassays, suggesting the implication of metabolites or different pathways in the activation of gene transcription. This paper shows that it is important to combine in vivo bioassays with in vitro approaches to elucidate the mechanism of xenoestrogen actions.

The estrogen-binding region of the cDNA for chicken ER reveals a mRNA of 3.5 kilobases (kb) in rainbow trout liver. The level of this messenger, which is very low in the liver of naive male animals, can be increased by estrogen stimulation. With this chicken probe, we have isolated a clone from a lambda gt10 trout liver cDNA library. The partial cDNA sequence, which encompasses most of the coding region, shows two domains of striking amino acid homology with human, avian, and Xenopus estrogen receptors (ERs) (DNA binding region: 90%, Hormone binding region: 60%). With this specific probe rainbow trout ER, we detected another messenger (4.5 kb) that is less expressed than the 3.5 kb messenger. The kinetics of stimulation of the two messengers is compared with the kinetics of accumulation of vitellogenin mRNA after E2 administration. This report constitutes the first identification of ER mRNA from a fish.

SummaryWe have previously described the cloning, sequencing and in vitro expression of a full-length rainbow trout estrogen receptor cDNA (rtER cDNA). This full cDNA randomly labelled was used to study the estrogen induction of hepatic rtER mRNA in correlation with vitellogenin (Vg) mRNA in different physiolo~cal situations.In this paper, we show that in the liver two mRNA species are under hormonal control and their level increases about g-fold after estrogen stimulation.These two mRNAs are expressed and induced in the liver as early as the hatching stage in correlation with the expression of Vg mRNA. A long-term analysis of rtER mRNA after estradiol (E2) injection shows a transient induction of the nuclear ER and its mRNA which recover to the basal level after 2 weeks. Nevertheless, a memory effect was observed on the expression of the Vg gene which does not appear to be directly related to the estrogen receptor level.

This study was conducted to determine if the cadmium-mediated inhibition of vitellogenesis observed in fish collected from contaminated areas or undergoing experimental exposure to cadmium correlated with modification in the transcriptional activity of the estrogen receptor. A recombinant yeast system expressing rainbow trout (Oncorhynchus mykiss) estradiol receptor or human estradiol receptor was used to evaluate the direct effect of cadmium exposure on estradiol receptor transcriptional activity. In recombinant yeast, cadmium reduced the estradiol-stimulated transcription of an estrogen-responsive reporter gene. In vitro-binding assays indicated that cadmium did not affect ligand binding to the receptor. Yeast one- and two-hybrid assays showed that estradiol-induced conformational changes and receptor dimerization were not affected by cadmium; conversely, DNA binding of the estradiol receptor to its cognate element was dramatically reduced in gel retardation assay. This study provides mechanistic data supporting the idea that cadmium is an important endocrine disrupter through a direct effect on estradiol receptor transcriptional activity and may affect a number of estrogen signaling pathways.