The transcription of the progesterone receptor gene is induced by estrogens and decreased by progestins. Studies were performed to define the regions of the gene and the molecular mechanisms involved. No hormonal regulation could be observed using 5′ flanking regions of the gene up to −2762 in front of a heterologous gene. Estrogen and progestin regulation could be observed only when using fragments of the gene extending down to +788. Progressive deletions from the 5′ and 3′ ends, site‐directed mutagenesis and DNase protection experiments with purified estrogen receptor suggested that the biologically active estrogen responsive element (ERE) is present at +698/+723, overlapping the initiation of translation. An oligonucleotide was synthesized bearing this ERE and shown to impart estrogen inducibility to a heterologous gene. Its regulation by anti‐estrogens corresponded to that of the in situ progesterone receptor gene since tamoxifen was a partial agonist whereas ICI 164384 was a full antagonist. This ERE also mediated down‐regulation by progestins in the presence of the progesterone receptor, even though it has no progesterone receptor binding ability. DNase footprinting showed that this effect was not due to a decrease of estrogen receptor affinity for the ERE in the presence of progesterone receptor. Finally, use of deletion mutants of the progesterone receptor showed that the steroid binding and the DNA binding domains were necessary for down‐regulation whereas deletions of various parts of the N‐terminal domain were without effect.

The effects of ligand binding and receptor phosphorylation on the interaction of progesterone receptor with specific DNA sequences in the uteroglobin gene were studied by nitro‐cellulose filter binding and DNase I footprinting. High affinity sites were mapped upstream from the transcription start and in the first intron. They contained a common TGTTCACT sequence. These sites were occupied with similar affinity by the receptor, either in its free state, or complexed with the hormone or an antagonist (RU486); and also by receptor which had been phosphorylated in vivo in a hormone‐dependent manner. In all cases identical footprints were observed. These experiments led to the following conclusions. The hormone‐dependency of receptor binding to DNA or chromatin is observed in intact cells and in crude cellular extracts but not with purified receptor. Thus in situ, the unliganded receptor probably interacts with some nuclear component(s) which stabilizes it in a ‘non‐activated’ form (non‐chromatin and non‐DNA binding form). When isolated, the receptor may undergo activation, even in the absence of the hormone. Binding by receptor of an antihormone (and possibly receptor phosphorylation) exerts an effect on gene transcription through a mechanism which is different from (and probably follows) receptor interaction with the gene.

The mechanism of action of antisteroids is not understood and explanations of their antagonistic activity have been sought at all levels of hormone action. It has been proposed that antisteroids, after binding to receptor, trap it into a non-activated (non DNA-binding) form possibly through interaction with a heat-shock protein of relative molecular mass (Mr) 90,000 (90 K), or that the antisteroids provoke binding of receptor to nonspecific DNA sites but not to hormone responsive elements (HREs), or that the antisteroid-receptor complexes can bind to HREs but form abortive complexes that fail to regulate transcription. We have constructed a deleted cDNA encoding a mutant form of rabbit progesterone receptor which exhibits constitutive activity, that is, binds to HREs in the absence of hormone and thus bypasses the first two steps discussed above. Co-transfection experiments allowed the expression of both constitutive and wild-type receptors in the same recipient cells. Antiprogestin RU486-wild-type receptor complexes completely suppressed the activity of the constitutive receptor on a reporter gene, showing that the inhibition is at the level of their common responsive elements.

The hepatocyte nuclear factor (HNF) 4␣ gene possesses two promoters, proximal P1 and distal P2, whose use results in HNF4␣1 and HNF4␣7 transcripts, respectively. Both isoforms are expressed in the embryonic liver, whereas HNF4␣1 is almost exclusively in the adult liver. A 516-bp fragment, encompassing a DNase I-hypersensitive site associated with P2 activity that is still retained in adult liver, contains functional HNF1 and HNF6 binding sites and confers full promoter activity in transient transfections. We demonstrate a critical role of the Onecut factors in P2 regulation using site-directed mutagenesis and embryos doubly deficient for HNF6 and OC-2 that show reduced hepatic HNF4␣7 transcript levels. Transient transgenesis showed that a 4-kb promoter region is sufficient to drive expression of a reporter gene in the stomach, intestine, and pancreas, but not the liver, for which additional activating sequences may be required. Quantitative PCR analysis revealed that throughout liver development HNF4␣7 transcripts are lower than those of HNF4␣1. HNF4␣1 represses P2 activity in transfection assays and as deduced from an increase in P2-derived transcript levels in recombinant mice in which HNF4␣1 has been deleted and replaced by HNF4␣7. We conclude that although HNF6/OC-2 and perhaps HNF1 activate the P2 promoter in the embryo, increasing HNF4␣1 expression throughout development causes a switch to essentially exclusive P1 promoter activity in the adult liver.Hepatocyte nuclear factor (HNF) 1 4␣, an orphan member of the steroid/thyroid receptor superfamily, is highly expressed in the adult liver (1). During mouse development, it is one of the first liver-enriched transcription factors (LETF) to be expressed, HNF4␣ transcripts being detected at E (embryonic day) 4.5 in primitive endoderm, in the visceral endoderm (E 5.5), and in the liver bud (E 8.5) (2, 3). The crucial role of this factor in the embryo was demonstrated by the inactivation of the HNF4␣ gene, causing perigastrulation lethality (4) because of failure to activate visceral endoderm functions (5). When null embryos were transiently rescued by tetraploid complementation with wild-type morulae-derived visceral endoderm, HNF4␣ was dispensable for hepatic specification but not for activation of hepatic functions (6). In addition, liver-specific HNF4␣ deletion in the embryo (7) led to disorganization of liver architecture (8), and its forced expression in cultured hepatic cells resulted in the reexpression of some liver-specific genes (9) and restoration of epithelial morphology (10, 11). HNF4␣ plays a pivotal role, regulating expression of genes involved in nutrient metabolism and transport (for review, see Ref. 12), and its induced disruption in the adult liver provokes lethal defects in lipid and bile acid metabolism and ureagenesis (13,14). Hence, HNF4␣ is essential for both the induction and the maintenance of hepatic functions.The HNF4␣ nuclear receptor contains six domains (A-F) and possesses two activation functions (AF): AF1 corresponds to the 24 N-t...