Glucocorticoid receptors: ATP-dependent cycling and hormone-dependent hyperphosphorylation

The dependence of hormone binding to glucocorticoid receptors (GRs) on cellular ATP levels led us to propose that GRs normally traverse an ATP-dependent cycle, possibly involving receptor phosphorylation, and that without ATP they accumulate in a form that cannot bind hormone. We identified such a form, the null receptor, in ATP-depleted cells. GRs are basally phosphorylated, and become hyperphosphorylated after treatment with hormone (but not RU486). In mouse receptors we have identified 7 phosphorylated sites, all in the N-terminal domain. Most are on serines and lie within a transactivation region. The time-course of hormone-induced hyperphosphorylation indicates that the primary substrates for hyperphosphorylation are the activated receptors; unliganded and hormone-liganded nonactivated receptors become hyperphosphorylated more slowly. After dissociation of hormone, most receptors appear to be recycled and reutilized in hyperphosphorylated form. From these and related observations, we have concluded that the postulated ATP-dependent cycle can be accounted for by hormone-induced or spontaneous dissociation of receptor-Hsp90 complexes, followed by reassociation of unliganded receptors with Hsp90 via an ATP-dependent reaction like that demonstrated in cell-free systems. Other steroid hormone receptors might traverse a similar cycle. Four of the 7 phosphorylated sites in the N-terminal domain are in consensus sequences for p34^cdc2 kinases important in cell cycle regulation. This observation, along with the known cell cycle-dependence of sensitivity to glucocorticoids and other evidence, point to a role for receptor phosphorylation in controlling responses to glucocorticoids through the cell cycle.     

Phosphorylation of transfected wild type and mutated progesterone receptors

An expression vector encoding wild type or mutated forms of the rabbit progesterone receptor was transfected into COS-7 cells and phosphorylation was studied by incubation with 32Pi followed by specific immunoprecipitation. The features of phosphorylation of the wild type receptor were identical to those previously observed in uterine cells: there was a basal level of phosphorylation which was increased approximately 7-fold by incubation with the hormone. The hyperphosphorylated receptor had decreased electrophoretic mobility ("upshift"). These experiments thus showed that the presence of the receptor specific kinase is not restricted to the target cells. Cleavage of the receptor by hydroxylamine and cyanogen bromide, and use of receptor mutants deleted in the N-terminal region, showed the absence of any detectable phosphorylation downstream from amino acid 520 (thus in the DNA and steroid binding domains). The majority of the phosphorylation sites were localized between amino acids 166 and 520. This localization was similar for basal and hormone-induced phosphorylation. DNA binding and hormone-induced hyperphosphorylation were not directly related, since deletion of the first zinc finger provided a hyperphosphorylated receptor. We showed that the constitutive receptor (totally deleted in the steroid binding region) exhibited only a low basal level of phosphorylation, and antagonist RU 486-receptor complexes were found to be hyperphosphorylated, leading us to conclude that the active form of the receptor was not the hyperphosphorylated one. Moreover receptor down regulation and hormone-induced receptor hyperphosphorylation were two independent phenomena. Basal phosphorylation was observed for both cytoplasmic and nuclear mutants, whereas nuclear localization was necessary but not sufficient for hyperphosphorylation. Finally, the second finger region and the hormone binding domain, which are necessary for receptor hyperphosphorylation, may be involved in the hormonally induced increased affinity of the receptor toward its kinase.     

A dynamic model of glucocorticoid receptor phosphorylation and cycling in intact cells

Glucocorticoid receptors have been proposed to undergo an ATP-dependent recycling process in intact cells, and a functional role for receptor phosphorylation has been suggested. To further investigate this possibility we have examined the phosphate content of the steroid-binding protein of all glucocorticoid receptor forms which have been isolated from WEHI-7 mouse thymoma cells. By labeling of intact cells with 32Pi for 18-20 h in the absence of hormone, covalent binding of [3H]dexamethasone 21-mesylate, immunopurification and SDS-PAGE analysis, the steroid binding protein was found to contain, on average, 2-3 phosphates as phosphoserine. One third of the phosphates were associated with proteolytic fragments encompassing the C-terminal steroid-binding domain. The central DNA-binding domain was not phosphorylated, leaving the other two thirds of the phosphates localized in the N-terminal domain. The phosphate content of various receptor forms from cells incubated with 32Pi and [35S]methionine was compared using 35S to normalize for quantity of protein. In ATP-depleted cells a non-steroid-binding form of the receptor (the "null" receptor) is found tightly bound to the nucleus, even without steroid. The phosphate content of null receptors was two thirds that of cytosolic receptors from normal cells, suggesting phosphorylation-dependent cycling in the absence of hormone. Addition of glucocorticoid agonists, but not antagonist, to 32P- and 35S-labeled cells increased the phosphate content of the cytosolic steroid-binding protein up to 170%, indicating an average increase in the phosphates from about 3 to 5. After 30 min of hormone treatment the phosphate content of the steroid-binding protein of cytosolic activated (DNA-binding) and nonactivated receptors, and that of nuclear receptors extractable with high salt concentrations and/or DNase I digestion, was the same. No change in the phosphate content of the 90-kDa heat shock protein associated with unliganded and nonactivated receptors was detected following association of the free protein with the receptor and following hormone binding of the receptor. Analysis of the unextractable nuclear receptors indicated that they contained less phosphate (60% of that of cytosolic receptors), similarly to null receptors, indicating that dephosphorylation is associated with the unextractable nuclear fraction. The rate of hormone-dependent phosphorylation appeared to be much faster than the rate of dephosphorylation in the presence of hormone, the latter determined by a chase of the 32P label with unlabeled phosphate. Our results show that phosphorylation and dephosphorylation are involved in the mechanism of action of glucocorticoid receptors.(ABSTRACT TRUNCATED AT 400 WORDS)     

Compound profiling using a panel of steroid hormone receptor cell-based assays

A major focus in the current discovery of drugs targeting nuclear receptors (NRs) is identifying drugs with reduced side effects by improving selectivity, not only from other receptors but also by selective modulation of the NR of interest. Cellular assays not only provide valuable information on functional activity, potency, and selectivity but also are ideally suited for differentiating partial agonists and antagonists. The ability to partially activate a receptor is believed to be closely tied to the ability to selectively modulate the NR, resulting in expression of a subset of the normally regulated genes. To this end, the authors have built a complete panel of cell-based steroid hormone receptor assays for the androgen receptor, estrogen receptor alpha, estrogen receptor beta, glucocorticoid receptor, mineralocorticoid receptor, and progesterone receptor by stably engineering a Gal4 DNA-binding domain/nuclear receptor ligand-binding domain fusion protein into an upstream activation sequence beta-lactamase reporter cell line. Each assay was validated with known agonists and antagonists for correct pharmacology and high-throughput compatibility. To demonstrate the utility of these assays, the authors profiled 35 pharmacologically relevant compounds in a dose-response format against the panel in both agonist and antagonist modes. The results demonstrated that selective estrogen receptor modulators can be identified and differentiated, as well as mixed and partial agonists and antagonists easily detected in the appropriate assays. Importantly, a comparison of the chimeric assays with full-length reporter gene assay data from the literature shows a good degree of correlation in terms of selectivity and pharmacology of important ligands. Taken together, these steroid hormone receptor assays provide good selectivity, sensitivity, and appropriate pharmacology for high-throughput screening and selectivity profiling of modulators of steroid hormone receptors.     

Molecular interactions of steroid hormone receptor with its enhancer element: evidence for receptor dimer formation

A steroid hormone responsive element (GRE/PRE), sufficient to confer glucocorticoid and progesterone inducibility when linked to a reporter gene, was used in band-shift assays to examine its molecular interactions with steroid hormone receptors. Both progesterone and glucocorticoid receptors bound directly and specifically to the GRE/PRE. The purine contact sites for both form A and form B chicken progesterone receptor, as well as those for rat glucocorticoid receptor, are identical. A peptide fragment produced in bacteria that primarily contain the DNA binding domain of the glucocorticoid receptor binds first to the TGTTCT half-site of the GRE/PRE, and a second molecule binds subsequently to the TGTACA (half-site) of the GRE/PRE in a cooperative manner. Utilizing the peptide fragment and the protein A-linked fragment, we demonstrated that the receptor interacts with its cognate enhancer as a dimer.     

Differential hormone-dependent phosphorylation of progesterone receptor A and B forms revealed by a phosphoserine site-specific monoclonal antibody

Human progesterone receptor (PR) is phosphorylated on multiple serine residues (at least seven sites) in a manner that involves distinct groups of sites coordinately regulated by hormone and different kinases. Progress on defining a functional role for PR phosphorylation has been hampered both by the complexity of phosphorylation and the lack of simple, nonradioactive methods to detect the influence of ligands and other signaling pathways on specific PR phosphorylation sites in vivo. Toward this end, we have produced monoclonal antibodies (MAbs) that recognize specific phosphorylation sites within human PR including a basal site at Ser 190 (MAb P190) and a hormone-induced site at Ser 294 (MAb P294). Biochemical experiments showed the differential reactivity of the P190 and P294 MAbs for phosphorylated and unphosphorylated forms of PR. Both MAbs recognize specific phosphorylated forms of PR under different experimental conditions including denatured PR protein by Western blots and PR in its native conformation in solution or complexed to specific target DNA. As detected by Western blot of T47D cells treated with hormone for different times, hormone-dependent down-regulation of total PR and the Ser 190 phosphorylation site occurred in parallel, whereas the Ser 294 phosphorylation site was down-regulated more rapidly. This difference in kinetics suggests that the Ser 294 site is more labile than basal sites and is acted upon by distinct phosphatases. A strong preferential hormone-dependent phosphorylation of Ser 294 was observed on PR-B as compared with the amino-terminal truncated A form of PR. This was unexpected because Ser 294 and flanking sequences are identical on both proteins, suggesting that a distinct conformation of the N-terminal domain of PR-A inhibits phosphorylation of this site. That Ser 294 lies within an inhibitory domain that mediates the unique repressive functions of PR-A raises the possibility that differential phosphorylation of Ser 294 is involved in the distinct functional properties of PR-A and PR-B.     

Hormone-dependent androgen receptor phosphorylation is accompanied by receptor transformation in human lymph node carcinoma of the prostate cells

Phosphorylation of the androgen receptor was investigated in the absence of hormone as well as during and after transformation of the receptor to the tight nuclear binding form. Human prostate tumor cells (LNCaP) were labeled for 4 h with [32P]orthophosphate in the presence or absence of steroid. Subsequently, androgen receptors were immunoprecipitated either from total cell lysates or from nuclear extracts using a specific monoclonal antibody. The immunoprecipitated receptor preparations were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, Western blotting, using a polyclonal antiserum, and autoradiography. It was observed that the androgen receptor is already phosphorylated in the absence of hormone, but undergoes a hormone-induced additional phosphorylation. After administration of 10 nM R1881, a 1.8-fold increase in phosphorylation over nonstimulated control cells was reached. Moreover, the amount of nuclear extractable androgen receptor was increased; the acquisition of tight nuclear binding capacity was accompanied by hormone-induced receptor phosphorylation.     

Communication between the N and C Termini Is Required for Copper-stimulated Ser/Thr Phosphorylation of Cu(I)-ATPase (ATP7B)

The Wilson disease protein ATP7B exhibits copper-dependent trafficking. In high copper, ATP7B exits the trans-Golgi network and moves to the apical domain of hepatocytes where it facilitates elimination of excess copper through the bile. Copper levels also affect ATP7B phosphorylation. ATP7B is basally phosphorylated in low copper and becomes more phosphorylated (“hyperphosphorylated”) in elevated copper. The functional significance of hyperphosphorylation remains unclear. We showed that hyperphosphorylation occurs even when ATP7B is restricted to the trans-Golgi network. We performed comprehensive phosphoproteomics of ATP7B in low versus high copper, which revealed that 24 Ser/Thr residues in ATP7B could be phosphorylated, and only four of these were copper-responsive. Most of the phosphorylated sites were found in the N- and C-terminal cytoplasmic domains. Using truncation and mutagenesis, we showed that inactivation or elimination of all six N-terminal metal binding domains did not block copper-dependent, reversible, apical trafficking but did block hyperphosphorylation in hepatic cells. We showed that nine of 15 Ser/Thr residues in the C-terminal domain were phosphorylated. Inactivation of 13 C-terminal phosphorylation sites reduced basal phosphorylation and eliminated hyperphosphorylation, suggesting that copper binding at the N terminus propagates to the ATP7B C-terminal region. C-terminal mutants with either inactivating or phosphomimetic substitutions showed little effect upon copper-stimulated trafficking, indicating that trafficking does not depend on phosphorylation at these sites. Thus, our studies revealed that copper-dependent conformational changes in the N-terminal region lead to hyperphosphorylation at C-terminal sites, which seem not to affect trafficking and may instead fine-tune copper sequestration.     

Human progesterone receptor transformation and nuclear down-regulation are independent of phosphorylation

We have studied the phosphorylation of progesterone receptors (PR) in T47Dco human breast cancer cells using a monoclonal antibody directed against human PR called AB-52. This antibody recognizes both the A- (Mr approximately 94,000) and B- (Mr approximately 120,000) hormone binding proteins of PR, and was used to immunoprecipitate phosphorylated receptors isolated from cells incubated in vivo with [32P]orthophosphate. The specific activity, or phosphorylation levels, relative to protein levels was quantified by combined immunoblotting and autoradiography followed by densitometry. We find that immunopurified untransformed hormone-free receptors, which have a characteristic triplet B, singlet A structure, are phosphoproteins with similar levels of phosphate incorporation in all protein bands. If PR are first transformed to the nuclear binding form by treatment of cells with progesterone, and then labeled with [32P]orthophosphate, the receptor proteins are additionally phosphorylated. These chromatin-bound hormone occupied receptors incorporate five to 10 times more labeled phosphate per total receptor protein than do PR from untreated cells during the same [32P]incubation time. The second round of phosphorylation may also account for mobility shifts of transformed A- and B-receptors observed in sodium dodecyl sulfate-polyacrylamide gels. Both untransformed and transformed species of A- and B-receptors are phosphorylated only on serine residues, and neither the extent of phosphorylation, nor the phosphoamino acids, are affected by treatment of the cells with epidermal growth factor or insulin. We previously reported that after hormone binding and transformation of receptors to the tight chromatin binding state, PR undergo processing, or nuclear down-regulation. AB-52 was used to compare PR protein and phosphorylation levels when cells were treated for 0.5-48 h with progesterone or the synthetic progestin R5020. Both agonists lead to hyperphosphorylation of nuclear PR before phosphorylation levels decrease, in parallel with the drop in protein levels as receptors down-regulate. Treatment of cells with RU 486, an antiprogestin, leads to PR transformation as determined by immunoblotting, but subsequent down-regulation does not occur. After transformation, chromatin-bound RU 486-occupied receptors become intensely phosphorylated however, with specific activities 15 times greater than those of untransformed PR. Since these receptors are phosphorylated but not processed, the hormone-induced nuclear phosphorylation of PR is unlikely to be a signal for receptor processing.(ABSTRACT TRUNCATED AT 400 WORDS)     

Different regions of the estrogen receptor are required for synergistic action with the glucocorticoid and progesterone receptors.

Estrogen and progesterone or estrogen and glucocorticoid receptors functionally cooperate in gene activation if their cognate binding sites are close to one another. These interactions have been described as synergism of action of the steroid receptors. The mechanism by which synergism is achieved is not clear, although protein-protein interaction of the receptors is one of the favorite models. In transfection experiments with receptor expression vectors and a reporter gene containing estrogen and progesterone-glucocorticoid receptor binding sites, we have examined the effects that different portions of the various receptors have on synergism. N-terminal domains of the chicken progesterone and human glucocorticoid receptors, when deleted, abolished the synergistic action of these receptors with the estrogen receptor. Deletion of the carboxy-terminal amino acids 341 to 595 of the estrogen receptor produced a mutant receptor that could not trans-activate on its own. This mutant receptor did not affect the action of the glucocorticoid receptor but functioned synergistically with the progesterone receptor. We therefore conclude that the synergistic action of the receptors for estrogen and progesterone is mechanistically different from the synergistic action of the receptors for estrogen and glucocorticoid.     

Modulation of glucocorticoid receptor phosphorylation and transcriptional activity by a C-terminal-associated protein phosphatase

The glucocorticoid receptor (GR) is phosphorylated at three major sites on its N terminus (S203, S211, and S226), and phosphorylation modulates GR-regulatory functions in vivo. We examined the phosphorylation site interdependence, the contribution of the receptor C-terminal ligand-binding domain, and the participation of protein phosphatases in GR N-terminal phosphorylation and gene expression. We found that GR phosphorylation at S203 was greater when S226 was not phosphorylated and vice versa, indicative of intersite dependency. We also observed that a GR derivative lacking the ligand-binding domain, which no longer binds the heat shock protein 90 (Hsp90) complex, exhibits increased GR phosphorylation at all three sites as compared with the full-length receptor. A GR mutation (F602S) that produces a receptor less dependent on Hsp90 for function as well as treatment with the Hsp90 inhibitor geldanamycin also increased basal GR phosphorylation at a subset of sites. Pharmacological inhibition of serine/threonine protein phosphatases increased GR basal phosphorylation. Likewise, a reduction in protein phosphatase 5 protein levels enhanced GR phosphorylation at a subset of sites and selectively reduced the induction of endogenous GR target genes. Together, our findings suggest that GR undergoes a phosphorylation/dephosphorylation cycle that maintains steady-state receptor phosphorylation at a low basal level in the absence of ligand. Our findings also suggest that the ligand-dependent increase in GR phosphorylation results, in part, from the dissociation of a ligand-binding domain-linked protein phosphatase(s), and that changes in the intracellular concentration of protein phosphatase 5 differentially affect GR target gene expression.     

Retinoic acid receptor belongs to a subclass of nuclear receptors that do not form "docking" complexes with hsp90.

We have recently reported that, in contrast to the glucocorticoid receptor, the thyroid hormone receptor does not bind to hsp90 when the receptor is translated in rabbit reticulocyte lysate [Dalman, F. C., Koenig, R. J., Perdew, G. H., Massa, E., & Pratt, W. B. (1990) J. Biol. Chem. 265, 3615-3618]. All of the steroid receptors that are known to bind hsp90 are recovered in the cytosolic fraction when hormone-free cells are ruptured in hypotonic buffer. In contrast, unliganded thyroid hormone receptors and retinoic acid receptors are tightly associated with nuclear components. In this paper, we translated the human estrogen receptor and the human retinoic acid receptor in reticulocyte lysate and then immunoadsorbed the [35S]methionine-labeled translation products with the 8D3 monoclonal antibody against hsp90. The estrogen receptor is bound to hsp90, as indicated by coimmunoadsorption, but the retinoic acid receptor is not. Translation and immunoadsorption of chimeric proteins containing the DNA binding domain of one receptor and the N-terminal and COOH-terminal segments of the other show that the DNA binding finger region of the estrogen receptor is neither necessary nor sufficient for hsp90 binding. These observations suggest that there are two classes within the steroid receptor family. In one class (e.g., glucocorticoid, mineralocorticoid, sex hormone, and dioxin receptors), the receptors bind to hsp90 and remain in some kind of inactive "docking" mode until hormone-triggered release of hsp90 occurs. In the retinoic acid/thyroid hormone class, the unligated receptors do not bind to hsp90, and the receptors appear to proceed directly to their high-affinity nuclear acceptor sites without entering the "docking" state.