Development and validation of a test for environmental estrogens: Checking xeno-estrogen activity by CXCL12 secretion in BREAST CANCER CELL LINES (CXCL-test).

Several methods have been developed to evaluate and quantify the effects of Endocrine disruptor chemicals (EDC). Nevertheless, most of these methods are time-consuming or not enough sensitive to detect EDC at the environmental range. To link the biological effect of tested EDC to natural protein secretion, we have developed a new screening method based on the secretion of the cytokine CXCL12 (or SDF-1, Stroma-cell Derived Factor 1), which plays a capital role in cell survival and migration. We have demonstrated that CXCL12 secretion is regulated by estrogenic compounds in a dose-dependent way in ER-positive breast cancer cell lines (MCF-7 and T47D). By combining cell culture and ELISA test, we used this up-regulation of CXCL12 secretion to test several major environmental contaminants. Our results showed that 17ß-estradiol (from 10(-11) M), 17α-ethynylestradiol (from 10(-12) M), genistein (from 10(-8) M) and bisphenol A (from 10(-6) M) dose-regulate CXCL12 secretion in T47D. In contrast, antiestrogens, raloxifen and 4-hydroxytamoxifen, had no effect on the CXCL12 secretion, but were able to inhibit E2 effect. Moreover, we used cell proliferation assays to evaluate the effect of these different compounds on the growth of T47D cells. We found strong correlation (P = 0.7) between proliferation and CXCL12 secretion. However CXCL12 secretion was as sensitive as cell proliferation assays but appeared more rapid. Thus, this bioassay named CXCL-test (for Checking Xeno-estrogen activity by CXCL12 secretion in breast cancer cell Lines) constitutes a fast and sensitive method for the detection of estrogenic compounds allowing in 14 h to achieve a detection limit of 10(-11) M of E2 (2.7 ng/L).

17Beta-oestradiol up-regulates the expression of a functional UDP-glucose dehydrogenase in articular chondrocytes: comparison with effects of cytokines and growth factors.

OBJECTIVES: To investigate the mechanisms by which cytokines and 17beta-oestradiol (17beta-E2) modulate gene expression and activity of uridine diphosphoglucose dehydrogenase (UGDH), a key enzyme of GAG synthesis in articular chondrocytes. METHODS: Rabbit articular chondrocytes (RAC) from 3-week-old animals were incubated for 24 h with TGF-beta, insulin like growth factor-I (IGF-I), IL-1beta, IL-6 and 17beta-E2. GAG synthesis was measured by [35S]-sulphate labelling and the expression of the UGDH gene was estimated by both real-time polymerase chain reaction and western blotting, whereas the enzyme activity was assayed by a spectrophotometric procedure. In addition, the transcriptional activity of several UGDH gene promoter constructs was determined in RAC transiently transfected with wild-type or deleted human oestrogen receptor-alpha gene (hER alpha66 or hER alpha46, respectively). RESULTS: 17Beta-E2 and its receptor hER alpha66 enhanced GAG neosynthesis in rabbit articular chondrocytes, as did TGF-beta1 whereas IL-1beta decreased this synthesis. 17Beta-E2 was found to exert positive regulatory effects at mRNA, protein and UGDH activity levels. In addition, the receptor hER alpha66, but not hER alpha46, increased the transcriptional activity of the UGDH gene. In contrast, no clear correlation between transcription, translation and activity of the UGDH was found under the effects of the cytokines studied. However, TGF-beta enhanced the enzyme activity, whereas IL-1beta, IL-6 and IGF-I were without significant effect. CONCLUSIONS: 17Beta-E2 enhanced GAG synthesis in chondrocytes via up-regulation of the UGDH gene expression and enzyme activity. These data provide insights into the molecular mechanisms involved in the regulation of the UGDH gene and offer new approaches to investigate its potential alteration in joint diseases.

11-deoxycorticosterone is a potent agonist of the rainbow trout (Oncorhynchus mykiss) mineralocorticoid receptor.

The teleost fish are thought to lack the mineralocorticoid hormone aldosterone but possess mineralocorticoid receptor (MR) homologs. Here we describe the characterization of two rainbow trout (Oncorhynchus mykiss) MRs, called rtMRa and rtMRb. The open reading frame of rtMRa cDNA encoded a protein of 1041 amino acids. The rtMRb predicted protein sequence is similar, differing in only 10 amino acids in the nonconserved A/B domain and lacking a three-amino acid insertion between the two zinc fingers of the C domain. Expression of rtMR mRNA (sum of both forms), measured in juvenile trout by real-time RT-PCR, shows that the transcripts are ubiquitous. Expression was significantly higher in brain than the other tissues studied (eye, trunk kidney, head kidney, gut, gills, liver, spleen, ovary, heart, white muscle, skin). Hormonal stimulation of receptor transactivation activity was studied in COS-7 cells transiently cotransfected with receptor cDNA and a mouse mammary tumor virus-luciferase reporter. The mineralocorticoids 11-deoxycorticosterone and aldosterone were more potent enhancers of rtMRa transcriptional activity (EC50 = 1.6 +/- 0.5 x 10(-10) and 1.1 +/- 0.4 x 10(-10) M, respectively) than the glucocorticoids cortisol and 11-deoxycortisol (EC50 = 1.1 +/- 0.3 x 10(-9) and 3.7 +/- 1.9 x 10(-9) M, respectively). A similar response was observed in transactivation assays with rtMRb. These results are discussed in the view of reported circulating levels of corticosteroids in trout.

ERalpha gene expression in human primary osteoblasts: evidence for the expression of two receptor proteins.

The beneficial influence of E2 in the maintenance of healthy bone is well recognized. However, the way in which the actions of this hormone are mediated is less clearly understood. Western blot analysis of ERalpha in osteoblasts clearly demonstrated that the well characterized 66-kDa ERalpha was only one of the ERalpha isoforms present. Here we describe a 46-kDa isoform of ERalpha, expressed at a level similar to the 66-kDa isoform, that is also present in human primary osteoblasts. This shorter isoform is generated by alternative splicing of an ERalpha gene product, which results in exon 1 being skipped with a start codon in exon 2 used to initiate translation of the protein. Consequently, the transactivation domain AF-1 of this ERalpha isoform is absent. Functional analysis revealed that human (h)ERalpha46 is able to heterodimerize with the full-length ERalpha and also with ERbeta. Further, a DNA-binding complex that corresponds to hERalpha46 is detectable in human osteoblasts. We have shown that hERalpha46 is a strong inhibitor of hERalpha66 when they are coexpressed in the human osteosarcoma cell line SaOs. As a functional consequence, proliferation of the transfected cells is inhibited when increasing amounts of hERalpha46 are cotransfected with hERalpha66. In addition to human bone, the expression of the alternatively spliced ERalpha mRNA variant is also detectable in bone of ERalpha knockout mice. These data suggest that, in osteoblasts, E2 can act in part through an ERalpha isoform that is markedly different from the 66-kDa receptor. The expression of two ERalpha protein isoforms may account, in part, for the differential action that estrogens and estrogen analogs have in different tissues. In particular, the current models of the action of estrogens should be reevaluated to take account of the presence of at least two ERalpha protein isoforms in bone and perhaps in other tissues.

Synergism between ERalpha transactivation function 1 (AF-1) and AF-2 mediated by steroid receptor coactivator protein-1: requirement for the AF-1 alpha-helical core and for a direct interaction between the N- and C-terminal domains.

The transcriptional activity of ERalpha (or NR3A1) after binding of ligand is mediated through synergistic action between activation functions (AFs) AF-1 and AF-2 and the transcriptional machinery. This is functionally achieved by bridging coactivators such as CEBP binding protein/p300 and members of the p160 subfamily such as steroid receptor coactivator protein-1 (SRC-1). We previously identified a conserved potential alpha-helical structure within the AF-1 functional core, and by evaluating point mutants of human ERalpha (hERalpha) within this region, we show that in transfection experiments this structure is required for synergism between SRC-1 and hERalpha. We report that the transcriptional synergism between AF-1 mutants and SRC-1 was abolished in AF-1-sensitive cells such as HepG2, whereas it was reduced by 50% in CHO-K1 cells, which have a mixed context that is sensitive to both the AF-1 and AF-2 regions of hERalpha. Glutathione-S-transferase pulldown assays demonstrate that the AF-1 core is able and sufficient for the hERalpha N-terminal region to interact with SRC-1. Interestingly, an enhancement of this recruitment in the presence of the hERalpha ligand-binding domain was observed, which was found to be dependent on a direct interaction between the N-terminal B domain and the ligand-binding domain. Another functional consequence of this physical interaction, which is promoted by both partial and full agonists of hERalpha, was an increase in the phosphorylation state of the N-terminal domain. Binding of 4-hydroxytamoxifen (OHT) to the hERalpha C-terminal region induced a functional AF-1 conformation in vitro through this N- and C-terminal interaction. The involvement of an SRC-1-mediated pathway in transactivation mediated by hERalpha AF-1 was further substantiated by transfection experiments using the OHTresponsive human C3 promoter, which showed that OHT-induced hERalpha AF-1 activity was enhanced by SRC-1 and required the AF-1 alpha-helical structure. In conclusion, we demonstrate that the synergism between AF-1 and AF-2 is mediated in part by a cooperative recruitment of SRC-1 by both the AF-1 alpha-helical core and AF-2 regions and that it is stabilized by a direct interaction between the B and C-terminal domains. This interaction of SRC-1 with the AF-1 alpha-helical core is essential for both E2- and OHT-induced ERalpha activity.

Tissue-specific expression of two structurally different estrogen receptor alpha isoforms along the female reproductive axis of an oviparous species, the rainbow trout.

In oviparous species, in addition to a full-length estrogen receptor alpha (ER alpha), another ER alpha isoform lacking the A domain and exhibiting a ligand-independent transactivation function has been consistently reported. Although both isoforms are expressed in the liver, their respective sites of expression in other potential target tissues are unknown. In contrast to the situation in Xenopus and chicken, the two isoforms of rainbow trout (Oncorhynchus mykiss) are generated from two classes of transcripts with different 5' untranslated sequences issued from the same gene by alternative splicing and promoter usage. The aim of this study was to take advantage of the unique organization of the rainbow trout ER alpha gene to investigate the tissue distribution of these two messenger species along the reproductive axis of female trout. The S1 nuclease assay and in situ hybridization were used, with probes specific for each of the transcripts. Reverse transcription polymerase chain reaction (RT-PCR) with primers specific for each of the isoforms also was performed. The data indicated that the full-length ER alpha is expressed in liver, brain, pituitary, and ovary, whereas expression of the isoform with the truncated A domain is restricted to the liver, demonstrating a tissue-specific expression of these two ER alpha isoforms. The presence of a short liver-specific isoform in oviparous species suggests its role in the development and/or maintenance of the unique function of the liver in the vitellogenesis process.

Distribution analysis of the two chicken estrogen receptor-alpha isoforms and their transcripts in the hypothalamus and anterior pituitary gland.

Estrogen plays a key role in the control of reproductive behavior and in the regulation of the neuroendocrine system. To elucidate the mechanisms by which it controls these functions it is important to understand how estrogenic effects are mediated. We have investigated the distribution of the two isoforms of the chicken estrogen receptor alpha (cER-alpha) protein; the previously characterized cER-alpha 66 and a new N-terminal truncated isoform, cER-alpha 61. Immunolocalization demonstrated the presence of cER-alpha 66 protein in hypothalamic areas, principally the nucleus septalis lateralis, bed nucleus striae terminalis medialis, nucleus preopticus medialis, and nucleus infundibuli hypothalami, and in the anterior pituitary gland. When the distribution of ER-alpha immunoreactive cells was compared using the antibodies H 222 (directed against the hormone-binding domain) and ER 221 (directed against the 21-amino acid N-terminus), no apparent differences could be detected. Because this immunocytochemical approach was not able to distinguish whether full-length cER-alpha 66 is the only isoform observed in the ER-positive regions or whether both cER-alpha receptor isoforms are present, SI nuclease assays were performed to compare the relative abundance in these regions of the two distinct classes of cER-alpha mRNA variants (A1-D and A2), which encode the cER-alpha 66 and cER-alpha 61 protein isoforms, respectively. In cockerels and hens, both variants of cER-alpha mRNA are expressed in the anterior pituitary gland and basal hypothalamus with a dominance of the mRNA that encodes cER-alpha 66, whereas the mRNA that encodes cER-alpha 61 was not detectable in the anterior hypothalamus. Therefore, because both receptor isoforms differ in their ability to modulate estrogen target gene expression in a promoter and cell type-specific manner, these differences may mediate the pleiotropic actions of estrogen in reproductive behavior and neuroendocrine functions.

Identification of a new isoform of the human estrogen receptor-alpha (hER-alpha) that is encoded by distinct transcripts and that is able to repress hER-alpha activation function 1.

A new isoform of the human estrogen receptor-alpha (hER-alpha) has been identified and characterized. This 46 kDa isoform (hERalpha46) lacks the N-terminal 173 amino acids present in the previously characterized 66 kDa isoform (hERalpha66). hERalpha46 is encoded by a new class of hER-alpha transcript that lacks the first coding exon (exon 1A) of the ER-alpha gene. We demonstrated that these Delta1A hER-alpha transcripts originate from the E and F hER-alpha promoters and are produced by the splicing of exon 1E directly to exon 2. Functional analysis of hERalpha46 showed that, in a cell context sensitive to the transactivation function AF-2, this receptor is an effective ligand-inducible transcription factor. In contrast, hERalpha46 is a powerful inhibitor of hERalpha66 in a cell context where the transactivating function of AF-1 predominates over AF-2. The mechanisms by which the AF-1 dominant-negative action is exerted may involve heterodimeri zation of the two receptor isoforms and/or direct competition for the ER-alpha DNA-binding site. hERalpha66/hERalpha46 ratios change with the cell growth status of the breast carcinoma cell line MCF7, suggesting a role of hERalpha46 in cellular proliferation.

The 3'-untranslated region of the human estrogen receptor alpha gene mediates rapid messenger ribonucleic acid turnover.

Human estrogen receptor-alpha messenger RNA (hERalpha mRNA) has a relatively short half-life, which was determined to be approximately 5 h in MCF-7 cell line after actinomycin D treatment. The 3'-untranslated region (3'UTR) of hERalpha mRNA was previously shown to completely down-regulate chloramphenicol acetyltransferase activity when present at the 3'-end of chloramphenicol acetyltransferase transcripts, suggesting a destabilizing function of the hERalpha 3'UTR sequence. Chimeric genes composed of a serum-inducible Fos promoter, GH-coding sequences, and different segments of the hERalpha complementary DNA 3'UTR sequence were used to confirm this hypothesis and to localize the RNA region responsible for the destabilizing effect. The presence of the complete hERalpha 3'UTR reduced the half-life of the reporter mRNA from more than 24 to 3 h. When the hERalpha 3'UTR was subdivided into four fragments (UTR1-4), one fragment, UTR2, retained the most ability to down-regulate the reporter mRNA (t1/2 = 4 h). A stretch of four AUUUA motifs within UTR2 was shown not to mediate mRNA destabilization. In contrast, further subdivision of the UTR2 into three parts (UTR2a-c) resulted in the loss of the destabilizing activity. Finally, recombination of two UTR2 subfragments (UTR2a and -b) partially restored this function, indicating a cooperative role among the three UTR2a-c subfragments in the process that leads to destabilization of the hERalpha transcript.

Tissue-specific expression of multiple mRNA variants of the mouse estrogen receptor alpha gene.

The isolation of five new mouse estrogen receptor alpha mRNA variants (mERalpha A, B, F1, F2 and H) is described. All of these transcripts are generated by alternative splicing and all encode the 66 kDa ERalpha protein that the previously identified mRNA C variant generates. However, these transcripts differ in their 5' untranslated regions. RT-PCR and S1 nuclease protection assays revealed a tissue- and sex-specific expression pattern of all variants. The C and F mRNA variants are the predominantly expressed mERalpha variants in mouse. The expression of mERalpha H mRNA is restricted to liver, although female mice produce around a five fold higher level of this transcript than males. Our results show that the mERalpha gene is a complex genomic unit in mice that exhibits alternative splicing which is regulated in a tissue-specific manner.

Transcriptional interference between glucocorticoid receptor and estradiol receptor mediates the inhibitory effect of cortisol on fish vitellogenesis.

In oviparous species, the synthesis of vitellogenin (Vg) takes place in the liver according to a strictly estrogen-dependent mechanism that first involves an up-regulation of the estrogen receptor (ER) by its own ligand. However, reports from the literature indicate that in trout stress or cortisol may cause a reduction of cytosolic E2-binding sites in the liver and a decrease in plasma Vg levels. To investigate the mechanisms underlying these effects, in vivo and in vitro experiments were designed in rainbow trout (Oncorhynchus mykiss). The results demonstrate that cortisol implanted into maturing females caused a marked decrease of rainbow trout ER (rtER) and rainbow trout Vg (rtVg) mRNA levels in the liver. In vitro experiments on hepatocyte aggregates also showed that dexamethasone (Dex) caused a strong decrease in the basal and E2-stimulated rtER mRNA and to a lesser extent rtVg mRNA. These effects were specific as no other hormones were able to mimic the inhibitory action of Dex. A study of rtER mRNA stability indicated that the effects of glucocorticoids are likely to take place at the transcriptional level. This was further indicated by transfection experiments in CHO-K(1) cells, which showed that rainbow trout glucocorticoid receptor (rtGR) strongly inhibited the E2-stimulated transcriptional activity of the rtER promoter. Taken together, these results indicate that the rtGR exerts a transcriptional interference on the expression of the rtER that may explain some of the negative effects of stress or cortisol on vitellogenesis.

Two estrogen receptor (ER) isoforms with different estrogen dependencies are generated from the trout ER gene.

A characteristic of all estrogen receptors (ER) cloned from fish to date is the lack of the first 37-42 N-terminal amino acids specific to the A domain. Here we report the isolation and characterization from trout ovary of a full-length complementary DNA (cDNA) clone encoding an N-terminal variant form of the rainbow trout ER (rtER). Sequence analysis of open reading frame of this cDNA predicts a 622-amino acid protein. The C-terminal region of this protein, from amino acid position 45 to the end, was very similar to the previously reported rtER (referred to as the short form, or rtER(S)). In contrast, this novel rtER cDNA (referred to as the long form, or rtER(L)) contains an additional in-frame ATG initiator codon that adds 45 residues to the N-terminal region of the protein. This new N-terminal region may represent the A domain of ER found in tetrapod species. The first 227 bp of this new cDNA were similar to the 3'-end intronic sequence of the rtER gene intron 1. These data together with S1 nuclease, primer extension, and RT-PCR experiments demonstrate that the rtER(L) represents a second isoform of rtER that arises from an alternative promoter within the first intron of the gene. Transcripts encoding both rtER forms were expressed in the liver. In vitro translation of the rtER(L) cDNA produced 2 proteins with molecular masses of 71 and 65 kDa, whereas rtER(S) cDNA produced 1 65-kDa protein. Interestingly, Western blot analysis with a specific antibody against the C-terminal region of rtER revealed 2 receptor forms of 65 and 71 kDa in trout liver nuclear extracts, in agreement with the presence of the 2 distinct classes of rtER messenger RNA in this tissue. Functional analysis of both rtER isoforms revealed that although rtER(S) consistently exhibited a basal (estrogen-independent) trans-activation activity that could be further increased in the presence of estrogens, the novel isoform rtER(L) is characterized by a strict estrogen-dependent transcriptional activity. These data suggest that the additional 45 residues at the N-terminal region of rtER(L) clearly modify the hormone-independent trans-activation function of the receptor.

Two functionally different protein isoforms are produced from the chicken estrogen receptor-alpha gene.

The existence of two forms of the chicken estrogen receptor-alpha protein (ER-alpha) in chicken tissues is demonstrated: the previously reported receptor (cER-alpha form I), which has a size of 66 kDa, and a new form (cER-alpha form II), which lacks the N-terminal 41 amino acids present in form I and thus gives rise to a protein of 61 kDa. Whereas the 66-kDa protein is the translation product of several cER-alpha mRNAs (A1-D), the cER-alpha protein isoform II is encoded by a new cER-alpha mRNA (A2), which is transcribed in vivo from a specific promoter that is located in the region of the previously assigned translation start site of the cER-alpha gene. SI nuclease mapping analysis reveals that cER-alpha mRNA A2 is liver enriched. The resulting cER-alpha forms I and II differ in their ability to modulate estrogen target gene expression in a promoter- and cell type-specific manner. Whereas cER-alpha form I activates or represses in a strictly E2-dependent manner, the truncated form is characterized by a partial transactivating or repressing activity in the absence of its ligand. Comparison of the N-terminal coding regions of different vertebrate ER-alpha reveal a conservation of the translation start methionine of the protein ER-alpha form II in other oviparous species but not in mammals. The expression of two classes of ER-alpha transcripts encoding the two ER-alpha receptor forms in the liver of Xenopus laevis and rainbow trout is demonstrated. Therefore, the existence of two functionally different protein isoforms produced from the ER-alpha gene is probably a common and specific feature in oviparous species.

Identification of differentially expressed 5'-end mRNA variants by an improved RACE technique (PEETA).

A rapid and efficient procedure is described for mapping and cloning the 5'-ends of mRNAs, including those generated from a unique gene by alternative splicing and promoter usage. This method involves reverse transcription of the targeted mRNAs from a long, highly labeled specific primer, resolution of the extension products on a DNA sequencing gel, elution and poly(dC) tailing of the single-stranded cDNAs of interest, amplification of these cDNAs by PCR using an oligo(dG) adapter-primer and a gene-specific primer and finally DNA sequencing of the subcloned PCR fragments. The overall method is called PEETA (primer extension, electrophoresis, elution, tailing and amplification).

Differentially expressed messenger RNA isoforms of the human estrogen receptor-alpha gene are generated by alternative splicing and promoter usage.

The isolation and characterization of several new human estrogen receptor-alpha (hERalpha) mRNAs are described. Together with those previously identified, they give rise to a total of six hERalpha mRNA isoforms (A-F hERalpha mRNAs). Produced from a single hERalpha gene by multiple promoter usage, all these transcripts encode a common protein but differ in their 5'-untranslated region as a consequence of alternative splicing of five upstream exons (1B-1F). RT-PCR and S1 nuclease mapping analysis of these different hERalpha mRNA isoforms revealed a differential pattern of expression of the hERalpha gene in human tissues and cell types. The A hERalpha mRNA is the main isoform detected in mammary glands or in the tumor cell lines derived from this tissue. In endometrium, the predominant forms are the A and C hERalpha mRNA isoforms, whereas the C and F hERalpha mRNA isoforms are the major forms detected in ovary. Finally, high levels of the E hERalpha mRNA isoform are restricted to the liver with an increased expression in females. Taken together, our results demonstrate that the hERalpha gene is a complex genomic unit exhibiting alternative splicing and promoter usage in a tissue-specific manner.

Identification of novel chicken estrogen receptor-alpha messenger ribonucleic acid isoforms generated by alternative splicing and promoter usage.

Using the rapid amplification of complementary DNA ends (RACE) methodology we have identified three new chicken estrogen receptor-alpha (cER alpha) messenger RNA (mRNA) variants in addition to the previously described form (isoform A). Whereas one of the new variants (isoform B) presents a 5'-extremity contiguous to the 5'-end of isoform A, the two other forms (isoforms C and D) are generated by alternative splicing of upstream exons (C and D) to a common site situated 70 nucleotides upstream of the translation start site in the previously assigned exon 1 (A). The 3'-end of exon 1C has been located at position -1334 upstream of the transcription start site of the A isoform (+1). Whereas the genomic location of exon 1D is unknown, 700 bp 5' to this exon were isolated by genomic walking, and their sequence was determined. The transcription start sites of the cER alpha mRNA isoforms were defined. In transfection experiments, the regions immediately upstream of the A-D cER alpha mRNA isoforms were shown to possess cell-specific promoter activities. Three of these promoters were down-regulated in the presence of estradiol and ER alpha protein. It is concluded, therefore, that the expression of the four different cER alpha mRNA isoforms is under the control of four different promoters. Finally, RT-PCR, S1 nuclease mapping, and primer extension analysis of these different cER alpha mRNA isoforms revealed a differential pattern of expression of the cER alpha gene in chicken tissues. Together, the results suggest that alternative 5'-splicing and promoter usage may be mechanisms used to modulate the levels of expression of the chicken ER alpha gene in a tissue-specific and/or developmental stage-specific manner.

Transcriptional regulation of expression of the rainbow trout albumin gene by estrogen.

Estrogens modulate the expression of many liver-specific genes in oviparous species. For instance, expression of the estrogen receptor and vitellogenin genes is strongly up-regulated by estradiol in rainbow trout liver. Using hepatocyte primary cultures, we demonstrate that trout albumin (Alb) gene is also regulated by this hormone. Indeed, treatment of hepatocytes with 1 microM estradiol led, after 24 h, to a dramatic decrease in Alb mRNA level. To investigate the mechanism of this down-regulation, run-off experiments were performed and mRNA half-lives were determined in the presence and absence of estradiol. The results show that the down-regulation of Alb mRNA expression by estrogens occurs only at the transcriptional level.

Systematic genomic screening and analysis of mRNA in untranslated regions and mRNA precursors: combining experimental and computational approaches.

MOTIVATION: The untranslated regions (UTRs) of mRNA upstream (5'UTR) and downstream (3'UTR) of the open reading frame, as well as the mRNA precursor, carry important regulatory sequences. To reveal unidentified regulatory signals, we combine information from experiments with computational approaches. Depending on available knowledge, three different strategies are employed. RESULTS: Searching with a consensus template, new RNAs with regulatory RNA elements can be identified in genomic screens. By this approach, we identify new candidate regulatory motifs resembling iron-responsive elements in the 5'UTRs of HemA, FepB and FrdB mRNA from Escherichia coli. If an RNA element is not yet defined, it may be analyzed by combining results from SELEX (selective enrichment of ligands by exponential amplification) and a search of databases from RNA or genomic sequences. A cleavage stimulating factor (CstF) binding element 3 of the polyadenylation site in the mRNA precursor serves as a test example. Alternatively, the regulatory RNA element may be found by studying different RNA foldings and their correlation with simple experimental tests. We delineate a novel instability element in the 3'UTR of the estrogen receptor mRNA in this way. AVAILABILITY: Strategy, methods and programs are available on request from T.Dandekar. CONTACT: dandekar@embl-heidelberg.de

Differential regulation of two genes implicated in fish reproduction: vitellogenin and estrogen receptor genes.

In rainbow trout as well as in other species, variability of estrogen receptor (ER) gene expression according to the cell type and the physiological state reflects a differential cell and gene sensitivity to estrogen. We previously demonstrated that expression of the rainbow trout estrogen receptor (rtER) and vitellogenin (Vg) genes were induced differently by estrogens in rainbow trout liver. Therefore, these two genes offered a suitable model to study the influence of ER concentration on gene transcriptional activities. In the present study we show that the transcription rate of rtER and Vg genes during an estrogenic treatment are affected differently by variation of cellular ER concentration. We demonstrate that rtER gene exhibits a low threshold response to loaded estrogen receptor, and increasing ER amounts do not affect the transcriptional response of this gene during an estrogenic stimulation. On the contrary, Vg gene expression requires the presence of a higher loaded estrogen receptor level to be induced, and its transcriptional response is directly proportional to the amount of synthesised ER.