Regulation of Glucocorticoid Receptor Protein and ... - Cancer Research

4 downloads 0 Views 2MB Size Report
Wayne V. Vedeckìs,2Masarrat Ali, and H. Raymond Allen. Department of Biochemistry and Molecular Biology. Louisiana State University Medical Center, New ...
[CANCER RESEARCH (SUPPL.) 49, 2295s-2302s, April 15, 1989]

Regulation of Glucocorticoid Receptor Protein and mRNA Levels1 Wayne V. Vedeckìs,2Masarrat Ali, and H. Raymond Allen Department of Biochemistry and Molecular Biology. Louisiana State University Medical Center, New Orleans, Louisiana 70112

Abstract The level of steroid receptors in target cells affects the responsiveness of the cell to the hormone. In mouse A11-211cells, it has been shown that chronic glucocorticoid treatment causes a down-regulation of glucocorticoid receptor (GR) levels (F. Svec and M. Rudis, J. Biol. Chem., 256: 5984-5987, 1981). The current study shows that chronic hormone treat ment reduces the amount of GR mRNA to about 50% of that in untreated cells. A combined treatment of the cells with an inhibitor of RNA transcription and the glucocorticoid hormone causes a more rapid de crease in steady-state GR mRNA levels than either agent alone. This suggests that glucocorticoids regulate the expression of the GR gene posttranscriptionally, perhaps via destabilization of the GR mRNA. An additional transcriptional regulation by the steroid hormone is not ruled out by this observation. It was also found that heat shocking a variety of cell types at 42°Cnot only causes an induction of heat shock proteins but also results in a dramatic decrease in the level of glucocorticoid-binding activity. GR labeled with a covalent ligand (dexamethasone 21-mesylate) was also reduced by heat shock, implying that heat shock caused an increased degradation in the GR protein itself. Finally, in vitro studies show that the GR is degraded in an ATP- and tRNA-dependent fashion in rabbit reticulocyte lysate. It therefore seems likely that the GR is degraded by the ubiquitin-dependent proteolytic pathway. Because ubiquitin is itself a heat shock protein, this may be the reason that the GR is rapidly degraded in heat-shocked cells. These studies point to possible mechanisms whereby the responsiveness of the cell to steroid hormones is altered by the regulation of the steroid receptor protein and mRNA levels.

protein itself was decreased by the hormone, rather than the receptor merely being converted into a state incapable of bind ing hormone (1). In an attempt to elucidate the molecular mechanism(s) for the hormone-mediated down-regulation of GR levels, the amount of GR mRNA has been measured in glucocorticoid-treated cells, using cloned GR gene probes. It has been found that the steady-state levels of GR mRNA are lowered to about 50% of initial levels in the rat HTC cell line (2), as well as in various rat tissues in vivo (3). To our knowledge, the level of the GR protein in hormone-treated cells has not yet been evaluated by a method independent of the hormonebinding activity. In a recent study, it has been claimed that an inhibition in GR gene transcription rate is responsible for the decreased steady-state levels of GR mRNA (4). The studies reported here analyze a number of potentially important physiological mechanisms for regulating the intracellular amounts of GR protein. Our evidence suggests a posttranscriptional mechanism for the autoregulation of GR mRNA levels in the mouse AtT-20 pituitary tumor cell line. In addition, heat shock dramatically reduces the level of GR protein in various cell lines, presumably by increasing the rate of receptor protein degradation. Finally, in vitro studies suggest that the GR may be a substrate for the ubiquitin-dependent proteolytic pathway. Materials and Methods

Introduction A wide variety of physiological processes are controlled by steroid hormones. Three minimal conditions are required for a cell to be responsive to a particular steroid: (a) the specific intracellulur receptor for the steroid hormone must be present to bind the hormone; (b) the steroid hormone level must be sufficiently high to bind to the receptor and cause receptor transformation (conversion of the receptor protein from a nonDNA-binding protein to the transformed, DNA-binding mon omer); (c) the hormone-responsive genes in the genome must be in a conformation which allows their regulation by the steroid-receptor complex. While a considerable amount of data have been obtained on the last two conditions mentioned above, a much smaller num ber of studies have been performed on the regulation of GR3

Cell Lines and Culture. Mouse AtT-20, L929, and S49 cells were cultured as described previously (5-7). Human Hep G2 cells were obtained from the American Type Culture Collection and cultured in Eagle's minimal essential medium containing 10% heat-inactivated fetal bovine serum. Chemicals. Unless noted otherwise, specialized chemicals were ob tained from Sigma Chemical Co. Salts and buffers were reagent grade from J. T. Baker. Radioactive isotopes used in these studies were as follows: [3H]TA (42.5 Ci/mmol; New England Nuclear); ["Sjmethionine (1130 Ci/mmol; ICN); |3H]DM (49.9 Ci/mmol; New England Nuclear); [a-"P]UTP (600 Ci/mmol; ICN). Proteins were labeled with I25I(100 mCi/ml; Amersham) using the chloramine-T procedure.

Analysis of GR mRNA Levels. All experiments were performed on log phase growth cells. Cells were treated with l UM TA alone, 5 ¿ig/ ml actinomycin D alone, or both agents for various times. The cells were pelleted by centrifugation. Total RNA was prepared using the guanidinium thiocyanate procedure and centrifugation (16 h) through protein levels in cells. However, it has been shown that chronic a cesium chloride cushion as described previously (7). The RNA pellet treatment of cells by glucocorticoid steroid hormones causes a was redissolved in RNA sample buffer (7). Ten ^g of total RNA were down-regulation in the hormone-binding activity (1). Once the applied to a nitrocellulose filter with a Schleicher and Schuell Minifold II apparatus. A riboprobe was generated using SP6 RNA polymerase level of the GR is lowered, a replenishment can occur upon from an ~l-kilobase pair subclone of the mouse GR gene (8), kindly removal of the hormone. Because this replenishment is blocked provided by Drs. M. Danielsen and G. M. Ringold. The processing of by cycloheximide, it was suggested that the level of the GR the filters, prehybridization, and hybridization were carried out as 1Presented at the Symposium on "Glucocorticoid Receptors: Evolution, Struc described by Northrop et al. (8). ture, Function and Abnormalities," July 14 and IS, 1988, Osaka, Japan. In Vivo Labeling of Cell Components. In the studies described below, intracellular proteins were labeled with [35S]methionine (50 ¿¿Ci/ml) in This research was supported by NIH Grant DK36086 (to W. V. V.) and by grants from Cancer Crusaders (to M. A.), Tri-Parish Golf, LA (to M. A.), and methionine-free medium for the last 30 min of incubation at either the Cancer Association of Greater New Orleans (to M. A.). H. R. A. was supported 37°Cor 42°C(heat shock). To analyze the labeled proteins, the cells by NIH Training Grant T32 CA09482. 2 To whom requests for reprints should be addressed. were washed extensively in Tris-saline (5) to remove the unincorporated 3 The abbreviations used are: GR, glucocorticoid receptor: TA, triamcinolone methionine, dissolved in Lai-umili sample buffer, sonicated to shear the acetonide; DM, dexamethasone 21-mesylate; TCA, trichloroacetic acid; SDS, DNA, and boiled for 5 min. An aliquot was then precipitated with TCA sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; hspTO, M, and the pellet was subjected to liquid scintillation counting to determine 70,000 heat shock protein; hsp90, M, 90,000 heat shock protein; DRB, 5,6dichloro-1 -/3-D-ribofuranosylbenzimidazole. the total incorporation of the label into protein. Additional aliquots, 2295s

GLUCOCORTICOID

RECEPTOR REGULATION

containing equivalent amounts of cellular material for each sample, were analyzed by SDS-PAGE (10% gels). Samples were transferred to nitrocellulose or Immobilen (Millipore) membranes using a Gelman BioTrans A semidry electroblotting apparatus. The filters were then subjected to autoradiography. Molecular weights of the protein bands were determined by comparison to a parallel lane containing prestained molecular weight markers (Sigma; Catalogue No. SDS-7B). An anti body against human hsp70 (kindly provided by Drs. D. Macejak and R. B. Luftig) and a monoclonal antibody against hsp90 (AC88, gener ously provided by Dr. D. O. Toft) were used to confirm the identity of these two proteins in the experiments shown in Fig. 3. The GR was covalently labeled in vivo by incubation of AtT-20 cells in serum-free medium with 10-15 nM [3H]DM for 30 min at 37°C. These cells were then subjected to heat shock as described below. Total cell extracts were prepared for SDS-PAGE as described above. After electrophoresis (10% gels), the gels were immersed in Enlightning (New England Nuclear) for 30 min, dried, and fluorographed using Kodak X-Omat film. The results in Fig. 7 were obtained by quantitating the band intensities using a soft laser densitometer (Biomed Instruments). Preparation of |3H]DM-labeled GR. Cytosol was prepared from AtT20 cells as described previously (5) and labeled with 10-15 nM [3H]DM. The labeled GR was prepared by passing the cytosol through a phosphocellulose column, transforming the GR by Sephadex G-25 chromatography, step-eluting the transformed GR from a DEAEcellulose column, and step elution of the GR from a DNA-cellulose column, essentially as described elsewhere (9-11). A major band with a molecular weight of ~94,000 was observed after SDS-PAGE and fluorography of this preparation. This sample was used for the in vitro GR degradation studies described below. Heat Shock Studies. Heat shock was carried out in capped 50-ml sterile, conical, plastic culture tubes. Cells at 37°Cwere pelleted by centrifugation, the medium was removed, and prewarmed medium at 37°C(control) or 42°C(heat shock) was added. The cells were sus pended and the incubation was continued for various times at the respective temperatures in water baths. After incubation, the cells were pelleted by centrifugation and washed with ice-cold Tris-saline before analysis of various parameters (hormone-binding, labeled proteins, etc.). In all cases, equivalent numbers of cells were used in comparing the control and experimental samples. In studies measuring the levels of [3H]TA-binding activity in heatshocked and control cells the following procedures were used. Cells were incubated with [3H]TA for the last 30 min of culture. The cells were then washed free of unbound hormone with cold Tris-saline. The cells were lysed in cold 1.5 mM MgCl2 and centrifuged, and the supernatant was collected. The pellet was extracted again with 1.5 mM MgClj, and the supernatant was added to the first. The crude nuclear pellet was then extracted with ethanol. The sum of the radioactivity in the supernatant and pellet fractions represents the total GR-bound [3H]TA. For the monolayer cell cultures (L929, Hep G2), the cells were washed with cold Tris-saline after labeling. The total intracellular [3H]TA was then extracted by treating the monolayer cells with l N NaOH. The extract was neutralized and counted. Nonspecific binding in these experiments (as determined by adding excess unlabeled TA) was 5-10% of the total, and this has been subtracted in the data presented. In Vitro GR Degradation. I25l-Labeled lysozyme, bovine serum al bumin, soybean trypsin inhibitor (all at 15,000-20,000 cpm/assay), and purified [3H]DM-labeled GR (10,000 cpm/assay) were used as sub strates in in vitro proteolysis assays. ATP-depleted rabbit reticulocyte lysates were generously provided by Dr. M. Rechsteiner, University of Utah (under the auspices of NIH Grant GM27159). The final 100-Ml reaction mixture contained the radioactively labeled protein substrate, 50 ill of ATP-depleted lysate, 50 mM Tris-HCI, 25 mM KCI, 10 mM NaCl, 10 mM MgCh, 1 mM dithiothreitol, and 0.1 mM disodium EDTA, all at pH 7.8. Control samples contained no further additions, while the experimental samples received 1 mM ATP plus an ATPregenerating system (13 mM phosphocreatine and 15 units of creatine kinase). The samples were incubated for the desired period of time, after which 800 ^1 of 1% cold bovine serum albumin were added followed by 10% (final concentration) TCA. The tubes were left on ice

for 30 min, and the clear supernatant was recovered after centrifugation. Aliquots of the supernatants and the pellets were both analyzed by liquid scintillation counting. Background I ( A soluble radioactivity in the original protein substrate samples (1-5% for the iodinated proteins and 0.01% for the GR) was subtracted from the reported values. To study the RNA dependence for a given substrate, the lysate was pretreated with RNase A (10 pg/ml, 30 min, 4°C)before carrying out the reaction.

Results Autoregulation of GR niRVY Levels. Chronic treatment of mouse AtT-20 pituitary tumor cells with glucocorticoids causes a decrease in the level of hormone-binding activity to about 2025% of starting levels at 24 h of treatment (1). Similar results have also been found for human HeLa cells (12), as well as for human lymphocytes (13). To investigate this phenomenon, Okret et al. (2) analyzed the levels of the GR mRNA in rat HTC cells at various times after hormone treatment. After an initial small increase in the GR mRNA levels, they found a decrease in GR mRNA to about 50% of control levels at 2448 h after dexamethasone treatment. The amounts of GR mRNA then began to increase at 72 h of treatment. Similar results were seen with intact rat tissues after in vivo treatment with dexamethasone (3). To determine if a similar effect occurred with mouse AtT-20 cells, we treated cells with TA and measured the steady-state mRNA levels. Fig. 1 shows that the GR mRNA levels decreased to 48% of control levels at 48 h of TA treatment and then increased slightly to 67% at 72 h. We did not observe any initial increase in GR mRNA levels at early times after hormone addition. A similar lowering in GR mRNA levels was also seen with the mouse L929 fibroblast cell line (data not shown). We then carried out studies to determine if this hormonemediated down-regulation of GR mRNA is transcriptional or posttranscriptional. Cells were treated with 5 /¿g/mlof actinomycin D alone, 1 ¡MTA alone, or both agents together. Actinomycin D at this level inhibited total RNA synthesis by 90% within l h of treatment (data not shown). Therefore, if the glucocorticoid hormone acts by inhibiting GR gene transcrip tion, the further addition of TA should have no effect on the rate of loss of the GR mRNA when actinomycin D is also present. On the contrary, if the hormone were having a post transcriptional effect (for example, by destabilizing the GR message), we expected to see an enhanced rate of down-regu lation when both agents were present. Fig. 2 shows that, indeed, 100

c o o

HO

c

40

o i.

20-

60

O)