Factor-assisted DNA binding as a possible general mechanism for steroid receptors. Functional heterogeneity among activated receptor-steroid complexes

We previously reported that activated glucocorticoid receptor-steroid complexes from rat HTC cell cytosol exist as at least two sub-populations, one of which requires a low molecular weight (700–3000 Da) factor(s) for binding to DNA. This factor is removed by Sephadex G-50 chromatography and is found predominantly in extracts of crude HTC cell nuclei. We have now determined that factor is not limited to HTC cells since an apparently identical factor(s) was found in nuclear extracts of rat kidney and liver as well as human HeLa and MCF-7 cells. Furthermore, the DNA binding of a sub-population of human glucocorticoid receptors depends on factor. While these results were obtained with agonist (dexamethasone) bound receptors, a sub-population of HTC cell receptors covalently labeled by the antiglucocorticoid dexamethasone 21-mesylate also displayed factor-dependent DNA binding. This receptor heterogeneity was not an artifact of cell-free activation since the cell-free nuclear binding of dexamethasone mesylate labeled complexes was, as in intact cells, less than that for dexamethasone bound complexes. Earlier results suggested that the increased DNA binding with factor involved a direct interaction of receptor with factor(s). We now find that the factor-induced DNA binding is retained by amino terminal truncated (42 kDa) glucocorticoid receptors from HTC cells. Thus the ability of receptor to interact with factor(s) is encoded by the DNA and/or steroid binding domains. Two dimensional gel electrophoresis analysis of dexamethasone-mesylate labeled 98 kDa receptors revealed multiple charged isoforms for both sub-populations but no differences in the amount of the various isoforms in each sub-population. Finally, activated progesterone and estrogen receptor complexes were also found to be heterogeneous, with a similar, if not identical, small molecular weight factor(s) being required for the DNA binding of one sub-population. The observations that functional heterogeneity of receptors is not unique to glucocorticoid receptors, whether bound by an agonist or antagonist, and that the factor(s) is neither species nor tissue specific suggests that factor-assisted DNA binding may be a general mechanism for all steroid receptors.     

Interaction between 1,25-dihydroxyvitamin D3 receptors and intestinal nuclei. Binding to nuclear constituents in vitro

The molecular action of 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) is thought to involve its localization within the nucleus of target cells, a process mediated by intracellular receptors. This report probes both the association between chick intestinal 1,25(OH)2D3 receptors and purified homologous nuclei and the interaction between this receptor and nucleic acids. 1,25(OH)2D3 receptors bound to purified nuclei in a apparently saturable manner (Kd = 2.2-4.8 X 10(-10) M) under conditions of intermediate ionic strength and constant protein concentration. Nuclear binding was hormone-dependent; whereas receptor-hormone complex (Rs) binds to nuclei under the ionic conditions employed here (greater than 70%), hormone-free (R0) receptors do not bind (less than 10%). Binding was localized to the nuclear chromatin fraction and was extremely sensitive to KCl concentration both in the incubation medium and during postincubation treatment of nuclei. The interaction appeared to be temperature-independent, suggesting the lack of a classic activation event characteristic of most steroid receptors. Partial digestion of intestinal nuclei with DNase I eliminated subsequent receptor binding by greater than 95%, pointing to the involvement of DNA in the binding interaction. In turn, receptors were found to bind to both DNA and RNA, a characteristic independent of receptor aggregation, but sensitive to disruption with increasing ionic strength buffers. Elution of both Rs and R0 from DNA appeared identical (0.28 M KCl), whereas the strength of interaction with RNA was much less (0.12 M KCl). Thus, while there appeared to be a fundamental difference between R0 and Rs, such that only the binding of receptor-hormone complex to nuclei was allowed under the conditions employed here, this characteristic was not observed during DNA binding. Nevertheless, the possibility exists that the in vivo interaction between 1,25(OH)2D3 receptor and nuclei involves DNA and that this nuclear constituent may be the ultimate site of action of this unique sterol hormone.     

Nuclear interactions of wild type and variant glucocorticoid receptors

Nuclease digestion of nuclei from glucocorticoid sensitive and resistant lymphoma cell lines was used to study the nuclear compartmentalization of wild type and variant glucocorticoid receptors. In comparison with wild type, the variant line (S49 143r) had an increased capacity to translocate to the nucleus (nti), but was more completely released from nuclei by nuclease digestion. Approximately 20% of the receptor in wild type nuclei was resistant to release by DNase I digestion, while only less than 5% of the receptor from nti nuclei was retained under the same conditions. Studies with wild type nuclei show that the nuclease resistant portion of receptors was also more resistant to release by increased ionic strength.     

Hormone-free mouse glucocorticoid receptors overexpressed in Chinese hamster ovary cells are localized to the nucleus and are associated with both hsp70 and hsp90

In this work, we examine the cellular localization and protein interactions of mouse glucocorticoid receptors that have been overexpressed in Chinese hamster ovary (CHO) cells (Hirst, M. A., Northrop, J. P., Danielsen, M., and Ringold, G. M. (1990) Mol. Endocrinol. 4, 162-170). We demonstrate that wild-type unliganded mouse glucocorticoid receptor, which is expressed in CHO cells to a level approximately 10 times that of L cells, is localized entirely to the nucleus by indirect immunofluorescence with the BuGR antireceptor monoclonal antibody. Overexpressed receptors that have either no hormone binding activity or no DNA binding activity because of point mutations also localize to the nucleus, providing genetic proof that the nuclear localization cannot reflect a steroid-mediated shift of the receptor from the cytoplasm to the nucleus and that DNA binding activity is not required for nuclear localization. Like unliganded progesterone receptors, which also associate in a loosely bound "docking" complex with the nucleus, the mouse glucocorticoid receptor overexpressed in CHO cells is associated with both hsp90 and hsp70. This is in contrast to the untransformed mouse glucocorticoid receptor in L cell cytosol, which is associated with hsp90 but not hsp70. The difference in hsp70 association between cell types could reflect overexpression of the receptor in CHO cells. However, like receptors in CHO cells selected for very high levels of overexpression, receptors in CHO cells selected for an intermediate level of receptor expression that is comparable to that of L cells are also bound to hsp70. This observation argues against an explanation of hsp70 association based purely on receptor overexpression, and we speculate that association of the unliganded glucocorticoid receptor with hsp70 might be a consequence of its nuclear localization in the CHO cells. Although there are differences between the mouse receptor in CHO cells and L cells, the nuclear localization signal of the untransformed mouse receptor reacts equivalently with the AP64 antibody against NL1 in cytosols prepared from both cell types.     

Mechanisms controlling steroid receptor binding to specific DNA sequences.

Mammalian progesterone receptors activated by hormone binding in nuclei of intact cells exhibit substantially higher binding activity for specific DNA sequences than receptors bound with hormone and activated in cell-free cytosol. Differences in DNA-binding activity occur despite the fact that both activated receptor forms sediment at 4S on sucrose gradients and are apparently dissociated from the heat shock protein 90. This suggests that hormone-induced release of heat shock protein 90 from receptors is necessary, but not sufficient for maximal activation of DNA binding. This report is a review of studies from our laboratories that have examined the role of receptor interaction with other nuclear protein factor(s), and receptor dimerization in solution, as additional regulatory steps involved in the process of receptor activation and binding to specific gene sequences.     

Evidence for nuclear and DNA binding forms of the receptor-glucocorticoid complex from hepatoma tissue culture cells

Binding to DNA associated with cellulose has been used to investigate the receptor-glucorticoid complex isolated from a line of rat hepatoma tissue culture cells. The amount of activated complex that bound to DNA was approximately half that which bound to nuclei. Additional results suggest the existence of two forms of the activated glucocorticoid receptor-steroid complex in about equal amounts: one form binds only to nuclei and the other binds to DNA and nuclei. The two forms also differ in their stability, with the DNA/nuclei binding form being relatively labile. The binding of either form to the appropriate acceptor is reduced by cytosol inhibitors by the same mechanism.     

Complex human glucocorticoid receptor dim mutations define glucocorticoid induced apoptotic resistance in bone cells

A mutation in the D-loop of the second zinc finger of the DNA-binding domain of the human glucocorticoid receptor (hGR), A458T (GR(dim)), has been suggested to be essential for dimerization and DNA binding of the GR, and genetically altered GR(dim) mice survive, whereas murine GR knockout mice die. Interestingly, thymocytes isolated from the GR(dim) mice were reported to be resistant to glucocorticoid-induced apoptosis. To further evaluate the dim mutations in glucocorticoid-induced apoptosis, we stably expressed either the hGR(dim) (A458T) or the hGR(dim4) (A458T, R460D, D462C, and N454D) mutant receptors in human osteosarcoma (U-2 OS) cells that are devoid of hGR and unresponsive to glucocorticoids. We analyzed these cell lines by comparison with a stable expression hGRα U-2 OS cell line, which undergoes apoptosis after glucocorticoid treatment. Transient reporter gene assays with glucocorticoid response element-driven vectors revealed that the hGR(dim) mutation had diminished steroid responsiveness and cells carrying the hGR(dim4) mutation were unresponsive to steroid, whereas glucocorticoid-induced nuclear factor κB repression was unaffected by either mutation. Interestingly, both the hGR(dim) and hGR(dim4) receptors readily formed dimers as measured by immunoprecipitation. Examination of GR-mediated apoptosis showed that hGR(dim) cells were only partially resistant to apoptosis, whereas hGR(dim4) cells were completely resistant to glucocorticoid-induced cell death despite remaining sensitive to other apoptotic stimuli. Global gene expression analysis revealed that hGR(dim4) cells widely regulated gene expression but differentially regulated apoptotic mRNA when compared with cells expressing wild-type hGRα. These studies challenge conclusions drawn from previous studies of GR dim mutants.     

Signal transduction by steroid hormones: nuclear localization is differentially regulated in estrogen and glucocorticoid receptors.

The glucocorticoid receptor accumulates in nuclei only in the presence of bound hormone, whereas the estrogen receptor has been reported to be constitutively nuclear. To investigate this distinction, we compared the nuclear localization domains of the two receptors and the capacity of their respective hormone-binding regions to regulate nuclear localization activity. As with the glucocorticoid receptor, we showed that the human estrogen receptor contained a nuclear localization signal between the DNA-binding and hormone-binding regions (amino acids 256-303); however, in contrast to the glucocorticoid receptor, the estrogen receptor lacked a second nuclear localization domain within the hormone-binding region. Moreover, the hormone-binding domain of the unliganded estrogen receptor failed to regulate nuclear localization signals, although it efficiently regulated other receptor functions. We conclude that the two receptors employ a common mechanism for signal transduction involving a novel "inactivation" function, but that they differ in their control of nuclear localization. Thus, despite the strong relatedness of the estrogen and glucocorticoid receptors in structure and activity, certain differences in their properties could have important functional implications.     

Evidence for nuclear and DNA binding forms of the receptor-glucocorticoid complex from hepatoma tissue culture cells.

Binding to DNA associated with cellulose has been used to investigate the receptor-glucocorticoid complex isolated from a line of rat hepatoma tissue culture cells. The amount of activated complex that bound to DNA was approximately half that which bound to nuclei. Additional results suggest the existence of two forms of the activated glucocorticoid receptor-steroid complex in about equal amounts: one form binds only to nuclei and the other binds to DNA and nuclei. The two forms also differ in their stability, with the DNA/nuclei binging form being relatively labile. The binding of either form to the appropriate acceptor is reduced by cytosol inhibitors by the same mechanism.     

Genetic dissection of the signaling domain of a mammalian steroid receptor in yeast.

The mechanism of signal transduction by steroid receptor proteins is complex and not yet understood. We describe here a facile genetic strategy for dissection of the rat glucocorticoid receptor "signaling domain," a region of the protein that binds and transduces the hormonal signal. We found that the characteristics of signal transduction by the receptor expressed in yeast were similar to those of endogenous receptors in mammalian cells. Interestingly, the rank order of particular ligands differed between species with respect to receptor binding and biological efficacy. This suggests that factors in addition to the receptor alone must determine or influence ligand efficacy in vivo. To obtain a collection of receptors with distinct defects in signal transduction, we screened in yeast an extensive series of random point mutations introduced in that region in vitro. Three phenotypic classes were obtained: one group failed to bind hormone, a second displayed altered ligand specificity, and a third bound hormone but lacked regulatory activity. Our results demonstrate that analysis of glucocorticoid receptor action in yeast provides a general approach for analyzing the mechanism of signaling by the nuclear receptor family and may facilitate identification of non-receptor factors that participate in this process.     

Glucocorticoid receptors in lung. Mechanism of specific glucocorticoid uptake by fetal rabbit lung nuclei.

After exposure of fetal rabbit lungs to glucocorticoid in vivo or in vitro, the hormone binds to specific receptors localized in the cytoplasm and in the nuclei. The present studies are compatible with a mechanism by which the nuclear receptor originates from the cytoplasm and arises from a hormone-, temperature-, and ionic strength-dependent transfer of the cytoplasmic receptor into the nucleus. This conclusion is reached from the following observations. Specific binding of glucocorticoid to nuclei from lungs not previously exposed to the hormone is not observed unless the cytosol is also present. In the presence of cytosol, nuclear uptake of the hormone is very slow at 0 degrees but is highly enhanced with increasing temperature. Concomitantly with the increased nuclear uptake there is an equiivalent loss of glucocortoid-receptor complex from the cytosol, indicating that the complex is transferred to the nuclei by a temperature-dependent process. Although the nuclei do not bind the cytoplasmic complex at 0 degrees, they do so provided that the cytosol is briefly heated in the presence of hormone prior to mixing with the nuclei. Thus the cytoplasmic complex must first be activated before it can bind to nuclei..     

The subcellular distribution of glucocorticoid-receptor complexes as studied by chemical crosslinking of intact HTC cells

Treatment of intact HTC cells with glutaraldehyde results in redistribution of glucocorticoid binding sites between cytosolic and nuclear fractions. The decrease in cytosolic receptors and their accumulation at the nuclear level were found to be directly related to the glutaraldehyde concentrations employed in our procedure and inversely related to the cell density of samples. When the data from eleven separate experiments were combined, and analyzed by linear regression of cytosolic and nuclear levels of receptor complexes vs the ratios between the DNA and glutaraldehyde concentration of our samples, two lines were obtained whose intercepts on the ordinate yielded values of cytosolic and nuclear receptors corresponding to 37.5 and 62.5% of the total cellular pool, respectively. When we compared the subcellular redistribution of glucocorticoid receptor to that of the cytosolic enzyme lactate dehydrogenase upon HTC cell crosslinking with glutaraldehyde, we found that the cytosolic and nuclear levels of the enzyme were 53.2 and 46.8% of the total content, respectively. If the subcellular distribution of glucocorticoid receptor is corrected for the artefactual redistribution induced by crosslinking, using the values obtained for lactate dehydrogenase, it can be concluded that glucocorticoid receptors in HTC cells are distributed between cytosol and nuclei in a ratio which is about 2:1. Our findings lend further support to theconclusion that only a portion of glucocorticoid receptor is cytosolic in intact cells.     

The nuclear matrix is the site of glucocorticoid receptor complex action in the nucleus

Binding of highly purified glucocorticoid receptor complexes to nuclear matrix was evaluated. Extraction of purified nuclei with 2M potassium chloride and brief deoxyribonuclease digestion leaves a matrix structure containing 1% of nuclear DNA and 6-12% of nuclear proteins. The nuclear matrix retained two binding sites for receptor complexes, a high affinity, low capacity site and a low affinity, high capacity site. These sites have affinities and capacities consistent with those reported for binding of these complexes to intact nuclei. More extensive deoxyribonuclease treatment of the matrix resulted in a marked reduction of high affinity complex binding. Furthermore, the DNA binding form of the receptor complex but not the unactivated receptor complex bound to DNA fibers anchored to nuclear matrix as visualized by 18 nm gold particle receptor complexes. The data suggest that the nuclear matrix is the major site for coordinating glucocorticoid hormone action in the nucleus.     

Human progesterone receptor binding to mouse mammary tumor virus deoxyribonucleic acid: dependence on hormone and nonreceptor nuclear factor(s)

Using crude progesterone receptor preparations from T47D human breast cancer cells, we show by immunoprecipitation assay that receptor specifically and with high affinity recognizes the hormone response element (HRE) of the mouse mammary tumor virus (MMTV). The use of crude preparations minimizes alterations of receptors or loss of associated factors that may occur during purification. Specific binding was obtained at 1:1 molar ratios of receptor to DNA, and HRE sequences are recognized with an affinity at least 3 orders of magnitude greater than nonspecific DNA. We have compared the DNA-binding activities of different forms of progesterone receptors. The unliganded 8S cytosol receptor had low but detectable binding activity for MMTV DNA. Addition of hormone to cytosol produced a small but consistent 2.5-fold increase. In vitro methods of transforming cytosol receptors from an 8S to a 4S species failed to increase DNA-binding further. By contrast, 4S receptors bound by R5020 in whole cells and extracted from nuclei by salt, displayed a substantially higher (average, 11-fold) binding activity than an equal number of unliganded cytosol receptors. The dissociation constants for cytosol and nuclear receptor binding to MMTV DNA were similar (approximately 2 x 10(-9) M). Thus, nuclear receptors possess a higher capacity for binding to specific recognition sequences. These results suggest that hormone or a hormone-dependent mechanism increases the intrinsic DNA-binding activity of receptors independent of receptor transformation from 8S to 4S. Further experiments indicate that a nonreceptor activity in nuclear extracts can increase the sequence-specific DNA-binding activity of cytosol receptors. This activity is present in both T47D cells and receptor-negative MDA-231 cells. We conclude that the higher DNA-binding activity of the nuclear receptor-hormone complex is due in part to receptor interaction with other nuclear proteins or factors. Such interactions may function to maintain receptors in a disaggregated active complex or to stabilize their binding to specific DNA sites.     

Mechanisms of steroid resistance

A method is described for the rapid determination of binding capacity and subcellular distribution of radioactive dexamethasone by cultured lymphoma cells. Using it, a number of steroid-resistant lymphoma cell clones were shown to be deficient in steroid binding to specific cytoplasmic receptors (r-) in transfer of the receptor-steroid complex to nuclei (nt-), or in the reactions subsequent to nuclear localization of the complex (d-). The relative proportions of those types of variants were determined in a steady-state population. About 80% were r-, and the remainder were equally divided between nt- and d-. Quantitative steroid-binding experiments suggest that certain receptor-containing variants possess altered receptor molecules.     

The Drosophila EcR gene encodes an ecdysone receptor, a new member of the steroid receptor superfamily

The steroid hormone ecdysone triggers coordinate changes in Drosophila tissue development that result in metamorphosis. To advance our understanding of the genetic regulatory hierarchies controlling this tissue response, we have isolated and characterized a gene, EcR, for a new steroid receptor homolog and have shown that it encodes an ecdysone receptor. First, EcR protein binds active ecdysteroids and is antigenically indistinguishable from the ecdysone-binding protein previously observed in extracts of Drosophila cell lines and tissues. Second, EcR protein binds DNA with high specificity at ecdysone response elements. Third, ecdysone-responsive cultured cells express EcR, whereas ecdysone-resistant cells derived from them are deficient in EcR. Expression of EcR in such resistant cells by transfection restores their ability to respond to the hormone. As expected, EcR is nuclear and found in all ecdysone target tissues examined. Furthermore, the EcR gene is expressed at each developmental stage marked by a pulse of ecdysone.     

Similar actions of glucocorticoids and calcium on the regulation of apoptosis in S49 cells

Glucocorticoid-induced lymphocyte cell death is a programmed process which is thought to involve the calcium-dependent degradation of DNA into multiples of 180 basepairs, characteristic of internucleosomal degradation. We have used the glucocorticoid-sensitive mouse lymphoma cell line S49.1 [wild-type (wt)] and the glucocorticoid-resistant cell line S49.22r (nt-) to evaluate the role of both glucocorticoid receptors and calcium in the regulation of internucleosomal DNA degradation and expression of calcium-dependent deoxyribonuclease activity. DNA was isolated from untreated (control) and dexamethasone (dex)-treated viable cells and analyzed for internucleosomal DNA degradation by agarose gel electrophoresis, followed by ethidium bromide staining. Glucocorticoid treatment resulted in substantial internucleosomal DNA degradation in wt cells, but not in nt- cells. This effect was inhibited by coincubation of cells with dex and the glucocorticoid receptor antagonist RU486. In contrast to the glucocorticoid response, administration of either of two calcium ionophores, ionomycin or A23187, produced internucleosomal degradation of DNA in both wt and nt- cells, although the latter were less sensitive to ionophore treatment. Interestingly, A23187 treatment also resulted in a loss of cell viability in HeLa S3 cells, a cell line that does not exhibit glucocorticoid-induced apoptosis. No internucleosomal DNA degradation was detected in HeLa S3 cells killed by A23187. To determine whether similar nucleases are associated with this internucleosomal DNA degradation resulting from both glucocorticoid and calcium ionophore treatment, 0.3 M NaCl nuclear protein extracts were prepared from control and treated cells and analyzed for protein composition or nuclease activity. To assay for nuclease activity, nuclear extracts were electrophoresed in sodium dodecyl sulfate-polyacrylamide gels impregnated with [32P]DNA. Nuclease activity was detected by removal of sodium dodecyl sulfate from the gel, activation with calcium, and subsequent visualization of the loss of [32P]DNA by autoradiography. Dex treatment of wt cells resulted in the appearance of several proteins within the mol wt range of 12-18 kDa, only one of which (16-18 kDa) exhibited calcium-dependent nuclease activity. The appearance of these proteins in nuclear extracts was inhibited by coincubation of glucocorticoid-treated cells with RU 486. Glucocorticoid treatment did not result in the appearance of nuclease activity in nuclear extracts from nt- cells. Interestingly, A23187 or ionomycin treatment resulted in an increase in activity of the 16- to 18-kDa nuclease in both wt and nt- cells. These findings indicate that both glucocorticoid receptors and calcium may share common features in the regulation of apoptosis in lymphoid cells.