A conserved region in the hormone‐dependent activation domain AF2 of nuclear receptors plays an important role in transcriptional activation. We have characterized a novel nuclear protein, RIP140, that specifically interacts in vitro with this domain of the estrogen receptor. This interaction was increased by estrogen, but not by anti‐estrogens and the in vitro binding capacity of mutant receptors correlates with their ability to stimulate transcription. RIP140 also interacts with estrogen receptor in intact cells and modulates its transcriptional activity in the presence of estrogen, but not the anti‐estrogen 4‐hydroxytamoxifen. In view of its widespread expression in mammalian cells, RIP140 may interact with other members of the superfamily of nuclear receptors and thereby act as a potential co‐activator of hormone‐regulated gene transcription.

Peroxisome proliferator-activated receptors (PPARs) are nuclear hormone receptors controlling the expression of genes involved in lipid homeostasis. PPARs activate gene transcription in response to a variety of compounds including hypolipidemic drugs as well as natural fatty acids. From the plethora of PPAR activators, Scatchard analysis of receptor-ligand interactions has thus far identified only four ligands. These are the chemotactic agent leukotriene B4 and the hypolipidemic drug Wy 14,643 for the alpha-subtype and a prostaglandin J2 metabolite and synthetic antidiabetic thiazolidinediones for the gamma-subtype. Based on the hypothesis that ligand binding to PPAR would induce interactions of the receptor with transcriptional coactivators, we have developed a novel ligand sensor assay, termed coactivator-dependent receptor ligand assay (CARLA). With CARLA we have screened several natural and synthetic candidate ligands and have identified naturally occurring fatty acids and metabolites as well as hypolipidemic drugs as bona fide ligands of the three PPAR subtypes from Xenopus laevis. Our results suggest that PPARs, by their ability to interact with a number of structurally diverse compounds, have acquired unique ligand-binding properties among the superfamily of nuclear receptors that are compatible with their biological activity.

We have characterized two distinct binding sites, called site 1 and site 2, in the nuclear protein RIP-140 which interact with the ligand binding domain of the estrogen receptor both in solution and when the receptor is bound to DNA. Both sites are capable of independently interacting with other nuclear receptors, including the thyroid hormone and retinoic acid receptors, but they are not identical since the interaction with retinoid X receptor is mediated primarily by site 1. The interaction is enhanced by agonists but not by antagonists, and the in vitro binding activities to a number of mutant receptors correlate with their abilities to stimulate transcription in vivo. When RIP-140 is fused to heterologous DNA binding domains, it is able to stimulate the transcription of reporter genes in both yeast and mammalian cells. Thus, RIP-140 is likely to function as a bridging protein between receptors and the basal transcription machinery and thereby stimulate the transcription of target genes.The nuclear receptor superfamily of transcription factors binds to DNA and activates or represses the transcription of genes in higher organisms (31,35). The activities of some of these receptors depend on the binding of hormonal ligands, including steroids, retinoids, and thyroid hormone, but the activating ligand has yet to be identified for the majority of them. Nevertheless, the modular structure of the entire superfamily seems to be conserved, since they all consist of three structural domains: an N-terminal domain containing an activation function, AF-1; a DNA binding domain; and a C-terminal ligand binding domain containing an additional activation function, AF-2 (6, 13). The activities of AF-1 and AF-2 depend on the promoter and cell type, and in some cases, both are required for full transcriptional stimulation (45). While the sequence of the N-terminal activation domain, AF-1, varies considerably in different nuclear receptors, that for AF-2 contains a highly conserved C-terminal amphipathic ␣-helix, which is essential for ligand-dependent transcriptional activity (4,11,12,40).The ability of nuclear receptors to stimulate transcription is likely to involve the recruitment of the basal transcription machinery into a preinitiation complex (16, 37). Although receptors bind directly with a number of basal transcription factors in vitro, including the TATA box-binding protein (41), TFIIB (3,19), and human TAF II 30 (20), the interactions are unaffected by ligand binding or by mutations in the AF-2 amphipathic ␣-helix that abolish transcriptional activity, suggesting that receptors are likely to interact with alternative proteins. Furthermore, the observation that AF-2 activity can be inhibited by overexpressing the hormone binding domain in squelching experiments (43) suggests that AF-2 is likely to interact with target proteins that are distinct from basal transcription factors. Several candidate target proteins have been identified; RIP-140 and RIP-160 (8, 9), ERAP-140 and ERAP-160 (17), TIF-1 (28), a number of isoforms...

The calbindin-D9k (CaBP9k) gene is mainly expressed in differentiated duodenal epithelial cells and is used as a model for studying the molecular mechanisms of intestine-specific transcription. The gene has been cloned, two major DNase-I-hypersensitive sites in the duodenum have been described, and a vitamin-D-response element has been identified. We have now analysed the transcription factors and regulatory sequences involved in the transcription of the CaBP9k gene in the intestine in ex vivo and in vitro experiments. Transfection experiments in intestinal (CaCo-2) and non-intestinal (HeLa) cell lines defined two regions in the 5'-flanking sequences of the rat CaBP9k gene. A minimal proximal region (-117 to +20) promoted transcription in both intestinal expressing and non-expressing cell lines. Tissue specificity was conferred by the sequences situated further upstream, which are responsible for complete repression in the non-intestinal cells. Intestinal transcription was specified by the proximal region, containing a specialized TATA box, and a distal region, which contains a previously described intestinal DNase-I-hypersensitive site. In vitro DNase I footprinting, electrophoretic mobility shift assays and antibody supershift assays were used to examine the factors bound to the proximal promoter region (-800 to +80 bp). Rat duodenal nuclear extracts protected 12 sites. Some of them appear to be binding sites for ubiquitous (nuclear factor 1) or hepatic-enriched sites (hepatocyte nuclear factors 1 and 4, enhancer binding protein a and p) factors. DNA binding studies and transfection experiments indicated that an intestine-specific transcription factor, caudal homeobox-2, binds to the TATA box of the rat CaBP9k gene. These data contribute to our understanding of the control of the intestinal transcription of the CaBP9k gene and demonstrate that several trans-acting factors, other than the vitamin D receptor, may be factors for intestine-specific CaBP9k gene expression.Keywords: intestine ; homeodomain protein ; transcription ; calbindin-D9k gene promoter; TATA box, Relatively few of the genes expressed in the intestine have been analysed and the way in which gene transcription is regulated in the intestine is not clearly understood. Hence, the characterization of the elements required for the expression of a gene in the intestine, particularly in enterocytes, is a necessary first step in the development of molecular tools for use in the intestine. The transcription factors implicated in the expression of few intestinal genes have been identified, including those for Correspondence to M. Thomasset, lNSERM U120, HBpital Robert

Progesterone modulates estrogen-stimulated responses in the uterus. Calbindin-D 9k (CaBP9k), a 17 beta-estradiol-responsive gene expressed in the uterus, was used as a marker to examine the interactions between endogenous progesterone and estradiol in the rat. The variations in uterine CaBP9k messenger RNAs (mRNAs) during the rat estrous cycle indicated that CaBP9k gene expression was greatest during the estrogen-dominated phases (proestrus and estrus) and became totally repressed during diestrus, when progesterone predominates. Estradiol was found to be the major controlling factor of CaBP9k gene expression in vivo, progesterone antagonizing estrogen-induced CaBP9k gene expression. The inhibitory role of progesterone was further examined in two experiments. Mature cyclic rats were injected with the progesterone antagonist RU486 before the progesterone surge of proestrus, and the estrous cycle was mimicked in ovariectomized rats by sequential injections of estrogen and progestin. Progesterone did not appear to be involved in the rapid decrease in CaBP9k mRNA during estrus but was implicated in the down-regulation of the estrogen-stimulated CaBP9k gene expression at the end of estrus and during diestrus. This delayed effect of progesterone was confirmed in the ovariectomized rat model. CaBP9k mRNA accumulation in estrogen-primed ovariectomized rats was suppressed by estrogen followed 1 h later by the progesterone agonist R5020. This effect occurred more than 24 h after progestin treatment. The inhibition of the estrogen-induced CaBP9k gene expression in the rat uterus by progesterone is certainly mediated by the progesterone receptor, because progesterone had no effect without estrogen priming or when the antagonist RU486 was used. The delayed progesterone effect probably does not involve depletion of nuclear estrogen receptors, the major rapid mechanism proposed for estrogen inhibition by progesterone in the rodent uterus, or control of estrogen receptor synthesis, as shown by Northern blot analysis of estrogen receptor mRNA.

The 9 kilodalton vitamin D-dependent calcium-binding protein (CaBP9k), calbindin-D9k, is expressed in the intestine and uterus of mammals. Rat intestinal CaBP9k is a well documented expression of the steroid hormone like action of 1,25-dihydroxyvitamin D3 (1,25-(OH)2D3). In contrast exogenous 1,25-(OH)2D3 does not affect the concentration of uterine CaBP9k which is dependent on estrogen. We have analyzed the effect of 17 beta-estradiol on the regulation of CaBP9k gene expression in the uterus of mature ovariectomized rats. CaBP9k mRNA is undetectable in the uterus of mature ovariectomized rats. A single dose of 17 beta-estradiol results in a detectable level of CaBP9k mRNA at 1 h and a significant increase 3 h after injection. The maximal CaBP9k mRNA level is reached 6 to 12 h post injection. These results show that 17 beta-estradiol increases CaBP9k production by increasing CaBP9k gene transcription. Chronic 17 beta-estradiol administration results in a plateau of CaBP9k mRNA but in a large increase in CaBP9k concentration. The kinetic response to a single estradiol injection was similar in immature rats. This result shows that no cellular differentiation is required for the control of CaBP9k gene expression by 17 beta-estradiol. The uterine cells of immature rats are already competent to respond optimally to estradiol. There is a single 0.5 kilobase CaBP9k gene transcript in the rat duodenum. In contrast there are two estrogen-inducible CaBP9k mRNA species in the uterus of both the mature ovariectomized and the immature rats. The smallest species corresponds to the duodenal CaBP9k mRNA species, while the larger species is at least 50 nucleotides larger. However, a unique CaBP9k identical to that in the duodenum is expressed in the uterus.

The calbindin D9k (CaBP9k) gene is under strict estrogen control in the rat uterus. This tissue contains two CaBP9k messenger RNA (mRNA) species. We have used primer extension analysis, reverse transcriptase associated with polymerase chain reaction, and RNase H digestion to show that these two mRNA species have the same structural features, including 5'- and 3'-ends, and poly(A) tail length. Our results suggest that the difference in electrophoretic mobilities of the two mRNA species might be due to interaction with another factor. We also analyzed the imperfect estrogen-responsive element (ERE) present on the first 5'-splice site of the rat CaBP9k gene. The oligonucleotide corresponding to the CaBP9k ERE was cloned in the plasmid pBLCAT2 (where the thymidine kinase promoter governs the expression of the chloramphenicol acetyl transferase gene) and transfected into MCF7 cells. This CaBP9k ERE was found to be a hormone-inducible enhancer that worked in an orientation-independent manner on a heterologous promoter and was functional at physiological hormone concentrations. One CaBP9k ERE conferred only weak (about 2-fold) estrogen induction, but two EREs cloned in tandem were strongly synergistic (14- to 16-fold). The CaBP9k ERE also bound to the partially purified estrogen receptor (ER) and to ER expressed in COS cells by gel shift assay. Methylation interference showed that all the guanine residues in both half-sites of the CaBP9k ERE were protected by ER binding. Thus, ER binds to the CaBP9k ERE in a way similar to other EREs. The gel shift assay results indicate that the strong synergistic effect of two EREs cloned in tandem is not due to cooperative binding between the two elements. As the CaBP9k gene is under strong estrogenic control in the uterus in vivo, the imperfect CaBP9k ERE may cooperate with another trans-acting factor to become fully efficient.

Progesterone modulates estrogen-stimulated responses in the uterus. Calbindin-D 9k (CaBP9k), a 17 beta-estradiol-responsive gene expressed in the uterus, was used as a marker to examine the interactions between endogenous progesterone and estradiol in the rat. The variations in uterine CaBP9k messenger RNAs (mRNAs) during the rat estrous cycle indicated that CaBP9k gene expression was greatest during the estrogen-dominated phases (proestrus and estrus) and became totally repressed during diestrus, when progesterone predominates. Estradiol was found to be the major controlling factor of CaBP9k gene expression in vivo, progesterone antagonizing estrogen-induced CaBP9k gene expression. The inhibitory role of progesterone was further examined in two experiments. Mature cyclic rats were injected with the progesterone antagonist RU486 before the progesterone surge of proestrus, and the estrous cycle was mimicked in ovariectomized rats by sequential injections of estrogen and progestin. Progesterone did not appear to be involved in the rapid decrease in CaBP9k mRNA during estrus but was implicated in the down-regulation of the estrogen-stimulated CaBP9k gene expression at the end of estrus and during diestrus. This delayed effect of progesterone was confirmed in the ovariectomized rat model. CaBP9k mRNA accumulation in estrogen-primed ovariectomized rats was suppressed by estrogen followed 1 h later by the progesterone agonist R5020. This effect occurred more than 24 h after progestin treatment. The inhibition of the estrogen-induced CaBP9k gene expression in the rat uterus by progesterone is certainly mediated by the progesterone receptor, because progesterone had no effect without estrogen priming or when the antagonist RU486 was used. The delayed progesterone effect probably does not involve depletion of nuclear estrogen receptors, the major rapid mechanism proposed for estrogen inhibition by progesterone in the rodent uterus, or control of estrogen receptor synthesis, as shown by Northern blot analysis of estrogen receptor mRNA.