Homology model for the ligand-binding domain of the human estrogen receptor

We have modeled the ligand-binding domain (LBD) of the human estrogen receptor protein (hER) by homology to the known crystal structure of the LBD of the alpha isoform of human retinoate-X receptor (hRX). Alignment of hER with members of the nuclear receptor superfamily defined probable secondary structures which we used to constrain backbone torsion angles and hydrogen bonds. From published studies we identified key interactions between hER and estradiol to use to dock the hormone in its ligand-binding pocket. Since the hRX crystal structure corresponds to the unliganded form of the LBD, we adopted the "mousetrap" mechanism proposed by Renaud et al to predict the structure of the E2-bound hER. Refinement by molecular dynamics and energy minimization gave a model which matches well the known facts about the estradiol phamacophore. It also provides a possible explanation for how hER discriminates between estradiol and testosterone.     

The key to the antiestrogenic mechanism of raloxifene is amino acid 351 (aspartate) in the estrogen receptor

The crystallization of the ligand-binding domain (LBD) of the estrogen receptor (ER) with 17beta-estradiol and raloxifene [A. M. Brzozowski et al., Nature (Lond.), 389: 753-758, 1997] now provides a molecular basis for the biological activity of complexes as either agonists or antagonists. It is well established that the critical structural feature of antiestrogens is a correctly positioned alkylaminoethoxy side chain. The X-ray crystallography clearly shows that the alkylaminoethoxy side chain of raloxifene causes a specific and inappropriate molecular perturbation of the LBD and that the nitrogen in the side chain must hydrogen bond with aspartate 351 in the LBD of ER. We previously identified and characterized a naturally occurring mutation in the ER from a tamoxifen-stimulated transplantable human breast tumor line. The mutation is at AA351 of LBD, where the aspartate is changed to tyrosine (Asp351Tyr). In this report, we compared and contrasted the pharmacology of raloxifene to block or induce E2-stimulated increase in TGF-alpha mRNA in stable transfectants of ER-negative human breast cancer cells with the cDNAs from wild-type, mutant-amino acid (AA) 400 ER and mutant-AA 351 ER. Our results show that the mutation at AA 351 that replaces aspartate by tyrosine specifically changes the pharmacology of raloxifene from an antiestrogen to an estrogen. By contrast, a mutation at AA 400 does not, and the antiestrogenic properties of raloxifene are retained. These data and the fact that the nitrogen in the side chain must specifically interact with aspartate 351 makes this the key to the antiestrogenic activity of raloxifene.     

Characterization of a point mutation in the hormone binding domain of the estrogen receptor from an breast tumor

It is clear that growth of the MCF-7 breast cancer cell line is stimulated by estrogen and when estrogen is removed, growth slows. We have observed a tumor derived from MCF-7 cells that grows in athymic mice in the absence of estrogen stimulation. We hypothesized that a mutation in the estrogen receptor (ER) could be responsible for this constitutive growth. Using single stranded conformational polymorphism (SSCP) and DNA sequencing analysis, we have identified an ER containing a point mutation at position 415 (gly to val) within the hormone binding domain. The functional activity of this mutant was assessed in vitro and in vivo. Using transient transfection into an ER negative breast cancer cell with an ERE luciferase reporter gene, we found that both the wild-type and mutant receptors have similar efficacy. Additionally, the estrogenic responses were blocked by antiestrogens in a concentration related manner. We also found that tumors with the mutant receptor show similar growth response in athymic mice as wild-type: stimulation with estradiol and inhibition with antiestrogens. We conclude that the point mutation at position 415 (gly to val) is not responsible for constitutive growth.     

Structural determinants of cortisol resistance in the guinea pig glucocorticoid receptor

The guinea pig exhibits resistance to glucocorticoids in vivo which results from the guinea pig glucocorticoid receptor (GR) having a lower affinity for cortisol than the human GR. Cloning of the guinea pig GR has revealed that the amino acid sequence of the ligand-binding domain (LBD) differs from the human GR at 24 residues. The present study confirms that the decreased sensitivity and binding affinity of the guinea pig GR are conferred in vitro by the LBD. Further, the substitutions in the LBD do not confer altered relative steroid sensitivity or selectivity compared with the human GR. The altered sensitivity and binding of dexamethasone are confined to the first third of the LBD, which contains 5 nonconservative substitutions in a region that is otherwise highly conserved across several species of GR. These residues, either alone or in combination, were targeted for site-directed mutagenesis in both the human and guinea pig LBD. Trans-activation studies with these mutant GR failed to exclusively implicate or exclude any of the residues in the observed resistance. Rather, the changes, with 1 exception, caused a decrease in sensitivity, suggesting that critical intramolecular interactions involving at least 4 of these residues determine the correct conformation of this region. Recent molecular modeling of the GR LBD structure suggests that although the above region is not part of the core ligand-binding pocket, it is required to maintain the conformation of the binding pocket.     

Three-dimensional models of estrogen receptor ligand binding domain complexes, based on related crystal structures and mutational and structure-activity relationship data

On the basis of the recently determined crystal structures of the ligand binding domains (LBDs) of the retinoic acid nuclear receptors (NRs), we present a three-dimensional (3D) molecular model of the human estrogen receptor alpha (hERalpha) LBD. A literature search for mutants affecting the binding properties has been performed; 45 out of 48 published mutants can be explained satisfactorily on the basis of the model. Estradiol has been docked into the binding pocket to probe its interactions with the protein. Energy minimizations and molecular dynamics calculations were performed for various ligand orientations. To evaluate their quality, the different models were scored using known structure-activity relationship (SAR) data for selected close estradiol homologues. The two best models explain largely the binding affinities of more distantly related ligands.     

Uterine estrogen receptor assayed by the controlled pore glass (CPG) method in adult rat after steroid administration

The interaction of estrogen and androgen was studied at the estrogen receptor level of adult rat uterus by the in vivo experiment. Both the free and total estrogen receptor in the cytosol and nucleus were assayed by the controlled pore glass beads method [2]. Testosterone was given by daily injection for 5 days to the rats treated with estradiol dipropionate (long-acting estradiol) 3 or 5 days before. In the estradiol dipropionate-treated rats the plasma estradiol concentration remained at an extremely high level for 7 days and then decreased to 5-fold of that of proestrus control by the 14th day. There was no significant change in the uterine estrogen receptor of the adult rat with a high estradiol concentration induced by testosterone.     

Cellular distribution and gene regulation of estrogen receptors alpha and beta in the rat pituitary gland

The pituitary gland is a heterogeneous tissue comprised of several hormone secreting and supporting cells, most of which are targeted by estrogens. Estrogen-induced changes in the pituitary are presumably mediated via the classical estrogen receptor, ER alpha. However, a novel receptor, ER beta, and pituitary-specific truncated estrogen receptor products (TERPs) were recently identified. The objectives of this study were to examine the distribution of these receptors in the rat pituitary and compare their regulation by estradiol in Sprague-Dawley and the estrogen-sensitive Fischer 344 rats. Pituitary cryosections were subjected to immunocytochemistry for specific cell types, followed by in situ hybridization for ER alpha or ER beta. ER alpha was expressed by approximately 45% of the lactotrophs and melanotrophs, 35% of the corticotrophs and folliculo-stellate cells, and 25% of the gonadotrophs. The expression of ER beta showed a similar pattern but was generally lower than ER alpha. In two cell types, melanotrophs and gonadotrophs, ER beta expression was significantly lower than ER alpha. In the second experiment, pituitary sections were immunostained for ER alpha, followed by in situ hybridization for ER beta. Only a minute population (6-10%) of either anterior or intermediate lobe cells coexpressed ER alpha and ER beta. In the next experiment, Fischer 344 and Sprague-Dawley rats were injected with oil or estradiol for 24 h. Total RNA from dissected anterior and posterior (neurointermediate) pituitaries was subjected to RT-PCR for ER alpha, ER beta, or TERPs. Interestingly, ER alpha and ER beta were unchanged by estradiol in either lobe of the pituitary. In contrast, estradiol increased pituitary TERP messenger RNA levels 4- to 7-fold. A 20-kDa TERP protein was detected by Western blots in the pituitary but not the uterus. There were no differences in the estradiol-induced expression of any of the receptors between the two strains of rats. We conclude that: 1) ER beta is expressed in all anterior and intermediate lobe cell types examined, albeit at a lower level than ER alpha; 2) no more than 10% of pituitary cells coexpress ER alpha and ER beta; and 3) estradiol markedly increases TERP messenger RNA levels but does not alter the expression of ER alpha or ER beta. We propose that estrogen receptor heterogeneity contributes to the diversity of pituitary cell responsiveness to estrogens.     

Assuring quality, information, and choice in managed care

The aim of the study was to determine whether there is an increase in responsiveness to estrogen stimulation of maternal behavior and lordosis responsiveness during pregnancy. Using separate groups of pregnancy-terminated females, we measured the initial maternal responsiveness of hysterectomized-ovariectomized (HO) females and their responsiveness to estrogen stimulation. Maternal behavior latencies were studied in females HO on the 8th, 10th, 13th, 16th, or 19th day of pregnancy (8HO-19HO) and in nonpregnant HO (NPHO) females. Groups were injected sc with estradiol benzoate (EB) in doses ranging from 0 to 200 microgram(s)/kg and tested for maternal behavior (retrieving, crouching, and licking pups). In addition, we investigated whether there is an increase during pregnancy (following HO) in lordosis responsiveness to estrogen stimulation. Lordosis behavior was studied in pregnant HO females (days 8, 16, and 22) and NPHO females given 0 to 200 microgram(s)/kg EB. There was an increase in maternal responsiveness in oil-treated HO females starting around midpregnancy. From early pregnancy on there was also an increase in maternal responsiveness to 20 microgram(s)/kg EB. In late pregnant females (16HO) there was a further increase with 50 microgram(s)/kg EB. There was no increase in lordosis responsiveness to EB stimulation during pregnancy; pregnant and nonpregnant HO females had the same EB threshold for stimulating lordosis behavior. The results of both studies were related to increases during the latter half of pregnancy in nuclear estrogen receptor concentrations in the MPOA, an area that mediates estrogen stimulation of maternal behavior, and the absence of such increases during pregnancy in the VMH, an area that mediates estrogen stimulation of lordosis behavior.     

Post-translational processing in the Golgi plays a critical role in the trafficking of the luteinizing hormone/human chorionic gonadotropin receptor to the cell surface

Point mutations in the luteinizing hormone/human chorionic gonadotropin (LH/hCG) receptor have been shown to cause constitutive activation which results in precocious puberty in affected males. We introduced one of these mutations, Asp-556 --> Gly, into the rat LH/hCG receptor and demonstrated that the mutant receptor constitutively activated adenylate cyclase in transfected 293 T cells. The cell surface expression of the mutant receptor was lower than that of the wild type receptor. Pulse-chase studies showed that the 73-kDa precursor of both the mutant and wild type receptors was synthesized at comparable efficiencies. However, post-translational processing of the mutant receptor to the mature 92-kDa form, which has N-linked complex type oligosaccharide chains, was impaired. Sensitivity of the mutant receptor to peptide-N-glycanase F and endoglycosidase H, and insensitivity to sialidase indicated that the 73-kDa species represents the high mannose form that has not yet been trafficked through the medial and trans Golgi. Additionally, although the wild type receptor was palmitoylated, the mutant receptor was not. Although the high mannose 73-kDa species is capable of binding LH/hCG, our results show that post-translational processing in the Golgi is required for the mature 92-kDa receptor to reach the cell surface.     

Effect of Cu2+ ions on the binding of estrogen to the human myometrial estrogen binding protein

The effect of cupric ions on the binding of estrogen to the human myometrial estrogen binding protein was investigated. Dissociation constants at 20 degrees C for the estrogen receptor were .081, .21, and .24 mM for estradiol, estriol, and estrone, respectively. Affinity and the number of binding sites were affected in the presence of cupric ions, indicating a mixed type inhibition of the binding of steroid to receptor. The 2nd-order association rate constants decreased in the presence of cupric ions and the effect appeared to be reinforced by an increase in temperature. No change in the sedimentation coefficient of the estrogen receptor complex occurred in the presence of cupric ions although less steroid was bound. Cupric ion concentrations above 10 mcM were needed before marked alterations in the binding characteristics were established. Since the copper level in human uterine fluid of women using a copper IUD have been shown to be in the 20 mcM range, the interaction of cupric ions with the binding of estrogens to their specific receptors must be considered as 1 factor responsible for the biochemical change in the uterus which contributes to the contraceptive action of the copper IUD.     

An improved assay for nuclear estrogen receptor in experimental and human breast cancer

We have validated a hydroxylapatite assay for measuring estrogen receptor in extracts from breast tumor nuclei. By adsorption of receptor of hydroxylapatite prior to addition of radioactive ligand and warming, receptor degradation can be avoided. Total binding sites are measured at 30 degrees by exchange, and receptor sites unoccupied by steroid are measured at 4 degrees. A single saturating dose of 5 nM tritiated estradiol (with or without a 100-fold excess of nonradioactive diethylstilbestrol to estimate nonspecific binding) yields results similar to a six-dose Scatchard plot. Following in vivo injection of estradiol into rats bearing mammary tumors, receptor translocation in the tumors can be accurately quantitated with this assay. Applying the assay to human breast cancer, we find that tumor biopsies may contain cytoplasmic receptor alone or may also have appreciable nuclear receptor. The latter may be bound to estradiol or may be found in "free" form. The finding of free receptor in the nuclei in certain cases raises the possibility that unoccupied receptor might be able to stimulate cell replication in these cases, even in the absence of estrogen.