Search

Currently available progesterone (P4) receptor (PR) antagonists, such as mifepristone (RU486), lack specificity and display partial agonist properties, leading to potential drawbacks in their clinical use. Recent x-ray crystallographic studies have identified key contacts involved in the binding of agonists and antagonists with PR opening the way for a new rational strategy for inactivating PR. We report here the synthesis and characterization of a novel class of PR antagonists (APRn) designed from such studies. The lead molecule, the homosteroid APR19, displays in vivo endometrial anti-P4 activity. APR19 inhibits P4-induced PR recruitment and transactivation from synthetic and endogenous gene promoters. Importantly, it exhibits high PR selectivity with respect to other steroid hormone receptors and is devoid of any partial agonist activity on PR target gene transcription. Two-hybrid and immunostaining experiments reveal that APR19-bound PR is unable to interact with either steroid receptor coactivators 1 and 2 (SRC1 and SCR2) or nuclear receptor corepressor (NcoR) and silencing mediator of retinoid acid and thyroid hormone receptor (SMRT), in contrast to RU486-PR complexes. APR19 also inhibits agonist-induced phosphorylation of serine 294 regulating PR transcriptional activity and turnover kinetics. In silico docking studies based on the crystal structure of the PR ligand-binding domain show that, in contrast to P4, APR19 does not establish stabilizing hydrogen bonds with the ligand-binding cavity, resulting in an unstable ligand-receptor complex. Altogether, these properties highly distinguish APR19 from RU486 and likely its derivatives, suggesting that it belongs to a new class of pure antiprogestins that inactivate PR by a passive mechanism. These specific PR antagonists open new perspectives for long-term hormonal therapy.

Type 1 pseudohypoaldosteronism (PHA1), a primary form of mineralocorticoid resistance, is due to inactivating mutations of the NR3C2 gene, coding for the mineralocorticoid receptor (MR).The objective of the study was to assess whether different NR3C2 mutations have distinct effects on the pattern of MR-dependent transcriptional regulation of aldosterone-regulated genes.Four MR mutations affecting residues in the ligand binding domain, identified in families with PHA1, were tested. MR proteins generated by site-directed mutagenesis were analyzed for their binding to aldosterone and were transiently transfected into renal cells to explore the functional effects on the transcriptional activity of the receptors by cis-trans-cotransactivation assays and by measuring the induction of endogenous gene transcription.Binding assays showed very low or absent aldosterone binding for mutants MR(877Pro), MR(848Pro), and MR(947stop) and decreased affinity for aldosterone of MR(843Pro). Compared with wild-type MR, the mutations p.Leu843Pro and p.Leu877Pro displayed half-maximal aldosterone-dependent transactivation of reporter genes driven by mouse mammary tumor virus or glucocorticoid response element-2 dependent promoters, whereas MR(848Pro) and MR(947stop) nearly or completely lost transcriptional activity. Although MR(848Pro) and MR(947stop) were also incapable of inducing aldosterone-dependent gene expression of endogenous sgk1, GILZ, NDRG2, and SCNN1A, MR(843Pro) retained complete transcriptional activity on sgk1 and GILZ gene expression, and MR(877Pro) negatively affected the expression of sgk1, NDRG2, and SCNN1A.Our data demonstrate that MR mutations differentially affect individual gene expression in a promoter-dependent manner. Investigation of differential gene expression profiles in PHA1 may allow a better understanding of the molecular substrate of phenotypic variability and to elucidate pathogenic mechanisms underlying the disease.

A gain of function mutation resulting in the substitution of leucine for serine at codon 810 (S810L) in the human mineralocorticoid receptor (MR) is responsible for early-onset hypertension that is exacerbated in pregnancy. All steroids, including progesterone, that display antagonist properties when bound to the wild-type MR are able to activate the mutant receptor (MR(L810)). These findings suggest that progesterone may contribute to the dramatic aggravation of hypertension in MR(L810) carriers during pregnancy. However, the steroid(s) responsible for hypertension in MR(L810) carriers (men and nonpregnant women) has not yet been identified. Here we show that cortisone and 11-dehydrocorticosterone, the main cortisol and corticosterone metabolites produced in the distal nephron, where sodium reabsorption stimulated by aldosterone takes place, bind with high affinity to MR(L810). The potency with which cortisone and 11-dehydrocorticosterone bind to the mutant MR contrasts sharply with their low wild-type MR-binding capacity. In addition, cotransfection assays demonstrate that cortisone and 11-dehydrocorticosterone are potent activators of the MR(L810) trans-activation function. Because the plasma concentration of cortisol in humans is about 30-fold higher than that of corticosterone, these findings strongly suggest that cortisone is one of the endogenous steroids responsible for early-onset hypertension in men and nonpregnant women carrying the MR(L810) mutation.

The crystal structures of ligand-free and agonist-associated ligand-binding domain (LBD) of nuclear receptors (NRs) reveal that the amphipathic helix H12 is folded back toward the LBD core in the agonist-associated conformation, allowing the binding of coactivators. We used alanine scanning mutagenesis to explore the role of the residues of the loop connecting H11 and H12 in the activation of the human mineralocorticoid receptor (hMR), a member of the NRs family. H950A retained the ligand binding and transcriptional activities of the wild-type receptor and interacted with coactivators. In contrast F956A had no receptor functions. Aldosterone bound to the mutant hMRs (L952A, K953A, V954A, E955A, P957A) with nearly the same affinity as to the wild-type receptor and caused a receptor conformational change in these mutant hMRs as it does for the wild-type receptor. But the aldosterone-induced transcriptional activity of the mutant hMRs was lower (L952A, E955A, P957A) than that of the wild-type receptor or completely abolished (K953A, V954A) and their interaction with coactivators was impaired (E955A) or suppressed (L952A, K953A, V954A, P957A). In the light of a hMR-LBD model based on the structure of the progesterone-associated receptor-LBD, we propose that the integrity of the H11-H12 loop is crucial for folding the receptor into a ligand-binding competent state and for establishing the network of contacts that stabilize the active receptor conformation.

The effects of aldosterone are mediated by the mineralocorticoid receptor (MR), a ligand-dependent transcription factor. We investigated the structural determinants for ligand binding to the receptor using a series of human MR (hMR) deletion mutants. These proteins were produced in vitro in rabbit reticulocyte lysate and analyzed for their ability to bind agonists, antagonists, and the heat shock protein hsp90, which is a prerequisite for ligand binding to hMR. Studies on N terminus-truncated hMRs showed that the ligand-binding domain (LBD: amino acids 734-984) has a lower affinity for aldosterone than the entire receptor [dissociation constant (Kd) 2.9 vs. 0.47 nM] and does not interact with hsp90. Addition of the five-amino acid sequence (729-733) upstream from the LBD is necessary for interaction with hsp90, but a larger region is needed for high aldosterone affinity. Deletions at the C-terminal end of the hMR greatly reduced both agonist and antagonist binding: deletion of the last three amino acids reduced the affinity for aldosterone to 1/20 that of the entire protein, and deletion of the last four amino acids completely abolished binding, although the interaction with hsp90 was not affected. These effects can be explained by misfolding of the receptor, since limited proteolysis assays showed that deletions at the C-terminal end of hMR affect the accessibility of the cleavage sites within the DNA-binding domain and the N-terminal part of the hinge region to trypsin. Thus, our results support the idea that a short sequence upstream of the LBD is essential for the interaction of hMR with hsp90 and that the C terminus of hMR and hsp90 are both essential for folding of the receptor in a high-affinity hormone-binding state.

To investigate the mechanisms involved in the in vivo aldosterone selectivity of the mineralocorticoid receptor (MR), we studied the respective contribution of the receptor and the enzyme 11 beta-hydroxysteroid dehydrogenase (11HSD), which converts glucocorticoids into inactive metabolites. Using a cotransfection assay in CV-1 cells, aldosterone activated mouse mammary tumor virus promoter through human MR (hMR) with an ED50 of 0.01 nM. An at least 100-fold higher concentration of cortisol (F), corticosterone (B), or dexamethasone was required to obtain half-maximum transactivation, indicating a functional preference of hMR for aldosterone over glucocorticoids. The catalytic activity of 11HSD was analyzed using HPLC by measuring the tritiated metabolites produced in CV-1 and COS cells. Both cell types displayed a significant dehydrogenase activity (20 fmol/10 min.10(3) cells) inhibitable by carbenoxolone, but no detectable reductase activity. In this model, B was more rapidly metabolized than F. Carbenoxolone treatment of hMR-transfected CV-1 cells did not result in a shift of the dose-response transactivation curves of B and F toward lower concentrations, ruling out the implication of 11HSD in the aldosterone MR selectivity of these conditions. Despite similar affinity constants of aldosterone and glucocorticoids for the hMR, kinetic experiments showed that the off-rate of aldosterone from hMR was 5 times lower than that of glucocorticoids, pointing to an intrinsic discriminating property of the receptor. Therefore, we propose that in addition to 11HSD, MR plays an active role in the mechanism of aldosterone selectivity.

The ability of the glucocorticosteroid receptor to bind mineralocorticosteroids suggests that spironolactone, a potent aldosterone antagonist, may also interact with the glucocorticosteroid receptor, resulting in an agonist or antagonist glucocorticosteroid activity. We have investigated the effect of this drug on the activity of the glucocorticosteroid-regulated mouse mammary tumor virus (MMTV) promoter. For these studies we used the mouse fibroblast cell line 1471.1. It contains about 200 copies of a permanently established chimeric DNA construct comprising a transcription unit [MMTV long terminal repeat (LTR)] driving the reporter gene chloramphenicol acetyltransferase linked to the 69% transforming fragment of the bovine papilloma virus genome. This cell line has a high level of glucocorticosteroid receptor (1200 fmol/mg protein) and no detectable mineralocorticosteroid receptor. Competition experiments showed a binding of spironolactone to glucocorticosteroid receptor, with an affinity 50-fold lower than that of dexamethasone. In these cells, spironolactone behaves as an antiglucocorticosteroid, inhibiting in a dose-dependent fashion dexamethasone-induced chloramphenicol acetyltransferase activity, with an ED50 of 8 microM. The absence of agonist activity, even at a high concentration of this compound (10 microM), demonstrates that spironolactone is a pure antiglucocorticosteroid in this cell line. MMTV LTR DNase-I hypersensitivity studies demonstrated that spironolactone, when administered in combination with dexamethasone, inhibits formation of the hormone-induced hypersensitive site located about 160 basepairs up-stream of the MMTV cap site. Furthermore, spironolactone alone failed to induce this DNase-I-hypersensitive site, suggesting that the antagonist-receptor complex does not interact productively with MMTV LTR chromatin.

Two new aldosterone antagonists, K-prorenoate [potassium 3(17 beta-hydroxy-6 beta, 7 beta-methylen-3-oxo-4-androsten-17 alpha-yl)propionate] and prorenone [3(17 beta-hydroxy-6 beta, 7 beta-methylen-3-oxo-4-androsten-17 alpha-yl) propionic acid gamma-lactone], its lactonic form, were studied in rat kidney using in vitro systems. Study of [3H]prorenone binding by a recently developed computer method indicated a high affinity, low capacity class of sites which are, seemingly, mineralocorticoid receptors. In competition experiments performed on [3H]aldosterone- and [3H]dexamethasone-binding sites, prorenone appeared to be a good competitor for mineralocorticoid-binding sites and a poor competitor for glucocorticoid-binding sites. The specificity of this molecule was further confirmed by its poor ability to displace [3H]dihydrotestosterone from rat prostate androgenic receptors compared to spironolactone [3-(3-oxo-7 alpha-acetylthio-17 beta-hydroxy-4-androsten-17 alpha-yl) propionic acid gamma-lactone]. In the same experiments, K-prorenoate demonstrated a very low affinity for the two types of receptors. The behavior of [3H]prorenone cytosolic complex was also studied in kidney mince experiments, which showed that the [3H]prorenone complex was not able to translocate into the nucleus. Prorenone inhibited the binding of [3H]aldosterone to the receptor and, consequently, the nuclear binding of aldosterone was not observed.

The disposition of aldosterone radiometabolites in rats has been studied following iv [3H]aldosterone administration. Of injected [3H]-aldosterone, 0.31% is recovered in 24 h urine as free aldosterone and 0.08% as acid-labile conjugate. A simple, sensitive and reliable radioimmunoassay of free aldosterone has been developed and the effect of acute oral potassium loading (171, 513 or 769 mueq of KCl/100 g body weight) on 4 h aldosterone excretion, plasma renin concentration and sodium and potassium balance has been investigated. There was a positive correlation between log urinary aldosterone and potassium load (r = 0.92, P less than 0.001). Potassium induced a natriuresis which was correlated directly with the dose of potassium administered (r = 0.89, P less than 0.001). Ptasssium loading also increased plasma renin concentration which was correlated with the sodium excretion rate (r = 0.64, P less than 0.01). Prevention of a negative sodium balance during the 769 mueq potassium load was obtained by administration of 513 mueq sodium. In this experiment, plasma renin concentration increased little, whereas the aldosterone excretion rate was as high as during the 769 mueq potassium load without sodium addition.