Triamcinolone Acetonide Regulates Glucocorticoid-receptor Levels by Decreasing the Half-life of the Activated Nuclear-receptor Form*

Glucocorticoid-receptor activation in GH1 cells re- sults from the conversion of a 10 S oligomeric cytosolic form to a 4-5 S nuclear-binding species (Raaka, B. M., and Samuels, H. H. (1983) J. Biol. Chern. 258, 417-425). In this study, we report that triamcinolone acetonide (9a-flU0r0-11@,16a,l7a,21-tetrahydroxy- pregna-1,4-diene-3,2O-dione 16,17-acetonide) elicits a time- and dose-dependent reduction of total-cell (nu-clear + cytoplasmic) receptor. The mechanism of re- ceptor regulation was studied by dense amino acid labeling of receptor using media containing ’H, “C, and “N-labeled amino acids. Total cell receptor was extracted with 0.4 M KC1 and newly synthesized dense receptor was separated from pre-existing receptor of normal density by centrifugation in gradients of 15-30% sucrose (w/v) in D,O. Receptor levels in cells grown without [‘H]triamcinolone acetonide was 260 f 19 fmo1/100 pg of DNA (16,000 molecules/cell), and, with 10 nM [‘H]triamcinolone acetonide, this decreased to 130 2 14 fmo1/100 pg of DNA after 30 h. Receptor half-life was 19 2 1.9 h in the absence

Glucocorticoid-receptor activation in GH1 cells results from the conversion of a 10 S oligomeric cytosolic form to a 4-5 S nuclear-binding species (Raaka, B. M.,  J. Biol. Chern. 258,[417][418][419][420][421][422][423][424][425]. In this study, we report that triamcinolone acetonide (9a-flU0r0-11@,16a,l7a,21-tetrahydroxypregna-1,4-diene-3,2O-dione 16,17-acetonide) elicits a time-and dose-dependent reduction of total-cell (nuclear + cytoplasmic) receptor. The mechanism of receptor regulation was studied by dense amino acid labeling of receptor using media containing 'H, "C, and "N-labeled amino acids. Total cell receptor was extracted with 0.4 M KC1 and newly synthesized dense receptor was separated from pre-existing receptor of normal density by centrifugation in gradients of 15-30% sucrose (w/v)  During the approach to steady-state conditions, 10 nM [3H]triamcinolone acetonide shortened receptor halflife almost immediately to the value in cells grown with [3H]triamcinolone acetonide for 24 h or longer. Cycloheximide did not prevent the triamcinolone acetonide-mediated decrease in receptor half-life and the shortening of receptor half-life is rapidly reversed by removal of hormone. These studies support a model of receptor regulation in which triamcinolone acetonide converts the unactivated 10 S receptor to the activated 4-5 S form which is degraded at an increased rate by the cell.
The glucocorticoid hormone-mediated conversion of receptor from a cytosolic to the nuclear-binding form has been referred to as receptor "transformation" (1) or "activation" (2). Receptor activation can be achieved by hormone with ~~~ ~ * This research was supported by Grant AM 21566 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Supported by a National Research Service Award from the National Institutes of Health. isolated cytosol preparations in uitro at 15-30 "C or without hormone by high ionic strength (3,4). The details of glucocorticoid hormone-mediated receptor activation have also been examined in uiuo in several different cell types (5-7). In GHl cells, a growth-hormone-producing rat pituitary cell line, glucocorticoids act synergistically with thyroid hormone to increase growth-hormone synthesis and mRNA accumulation (8,9). A previous study from this laboratory analyzed the influence of [3H]triamcinolone acetonide' on the receptor forms generated as a consequence of hormone-mediated receptor activation (7).
In GH1 cells, cultured without hormone and harvested in hypotonic buffer, receptor is localized to the cytosol and sediments as 10 S in low-salt gradients (7). When GH1 cells are incubated at 37 "C with [3H]TA, the 10 S cytosolic receptor decreases and a 4-5 S cytosolic receptor reciprocally increases along with the extent of nuclear translocation. Only the 4-5 S cytosolic receptor is a DNA-binding protein, suggesting that the 4-5 S and not the 10 S species represents the activated cytosolic receptor form (7). In the absence of ligand, 0.4 M KC1 also converts the 10 S receptor to the 4-5 S form in cytosol in uitro ( 7 ) . Dense amino acid labeling, followed by sedimentation of cytosol in sucrose/D20 gradients containing 0.4 M KCl, identified discrete 4-5 S forms of newly synthesized dense receptor and pre-existing receptor of normal density. In contrast, sedimentation in low-salt sucrose/DzO gradients demonstrated no discrete dense or normal 10 S species but showed as a progressive density shift indicating density heterogeneity of the 10 S form.
These studies support the view that the unactivated 10 S receptor is an oligomer (most likely a tetramer) which is in rapidly exchanging equilibrium in intact cells with a 4-5 S activated hormone-binding subunit (7). The binding of hormone to 10 S receptor appears to markedly shift the oligomersubunit equilibrium in favor of the 4-5 S component which reversibly associates with chromatin (7). This model is also supported by the observation that when hormone is removed from cells, the 4-5 S cytosolic receptor and receptor bound to nuclei rapidly regenerate the unactivated 10 S form without requirement for new receptor synthesis ( 7 ) .
Receptor half-life was about 20 h in cells cultured without hormone. 13H]TA rapidly decreases the half-life to about 10 h without affecting receptor synthesis and accounts for the decrease in the steady-state amount of receptor. Removal of hormone, which regenerates the 10 S receptor from the activated 4-5 S form, results in prolongation of the half-life to a level observed in control cells. These studies further support a reversible model for receptor activation in which t3H]TA converts the 10 S receptor to the 4-5 S form which is degraded at an increased rate by the cell.

EXPERIMENTAL PROCEDURES
Materials-[6,7-3H]TA (32 Ci/mmol) was obtained from New England Nuclear. A mixture of amino acids enriched in the dense isotopes 2H, 13C, and 15N to 98, 80, and 90 atom %, respectively, were obtained from Merck, Sharp and Dohme, Canada. Amino acids with normal density, D20, TA, protamine sulfate, and other chemicals were obtained from Sigma. Cell-culture media and serum were supplied by Gibco. The preparation of medium containing dense amino acids (dense medium) was as described previously (7), except that no serum component was added. Each 100 ml of dense medium contained 50 mg of a mixture of amino acids enriched in the dense isotopes *H, 13C, and 15N. Dense medium was sterilized by filtration through a 0.2pm filter.
Cell-culture Conditions-GH1 cells were grown in monolayer cultures at 37 "C in 75-cm2 plastic flasks as described previously (13). Before each experiment, cells were cultured for at least 24 h in Ham's F-10 medium supplemented to 10% (v/v) with hormone-depleted calf serum (14) followed by at least 24 h in serum-free Ham's F-10. At times ranging from 1-48 h prior to harvesting the cells, the medium was replaced with serum-free medium supplemented with various concentrations of [3H]TA. For density-labeling experiments, the medium was replaced with dense medium, with or without hormone, at various times prior to harvesting the cells. Rat HTC cells and Reuber H-35 hepatoma cells were maintained in Dulbecco's minimal essential medium supplemented with penicillin (50 pg/ml) and streptomycin (50 pg/ml), 15 mM Hepes, pH 7.4, pyruvate (110 milligrams/liter) and 10% fetal bovine serum (v/v). Prior to experiments, the HTC and Reuber H-35 hepatoma cells were cultured for at least 24 h in Dulbecco's minimal essential medium prepared as above, except that the serum component was 10% (v/v) hormone-depleted fetal bovine serum prepared as described (14).
Cell Harvesting and Preparation of Nuclei-The culture medium was removed and the cells were washed three times with isotonic saline at 4 "C and allowed to drain thoroughly to remove the last drops of saline. The cells were harvested with a rubber policeman at 4 "C in 0.25 ml of buffer, pH 7.4 at 25 "C, containing 20 mM Tris-HC1,5 mM sodium molybdate, 2 mM dithiothreitol, and 0.02% Triton X-100 (low-salt buffer). After mixing for 10 s at the maximum setting of a Vortex Genie, the suspension of broken cells was centrifuged at 4 "C for 10 min at 12,000 X g in an Eppendorf Model 5412 microcentrifuge. For most density-labeling studies, the broken-cell preparations were adjusted to contain 0.4 M KC1 prior to centrifugation. In some studies, the broken-cell preparations were centrifuged prior to increasing the KC1 concentration to separate the nuclear and cytoplasmic fractions. In one experiment, the nuclei were then extracted with 0.4 M KC1 and the supernatant was adjusted to 0.4 M KC1 prior to analysis. The derived supernatants were applied to sucrose gradients of various compositions as described below. The remaining pellet fraction, which contained nuclear material, was washed twice by resuspension in 1 ml of low-salt buffer. The washed nuclear pellet was incubated with 0.5 ml of absolute ethanol for 30 min at 37 "C to extract [3H]TA. Radioactivity in the ethanol extract was determined by scintillation counting at 43% efficiency. The DNA content of the nuclei was determined by the method of Burton (15). In all sedimentation experiments, parallel flasks in duplicate were incubated with [3H]TA & a 1,000-fold excess of nonradioactive steroid to assess receptor levels.
Gradient Centrifugation-In every case, discontinuous sucrose gradients consisted of 1.3 ml each of three sucrose solutions and were prepared in 4-ml polyallomer tubes. Gradients were chilled to 5 "C prior to use. For density-labeling studies, sucrose solutions were prepared in 95% DzO, 5% H20, containing 0.4 M KC], 20 mM Tris-HCI, pH 7.4 at 25 "C, 5 mM sodium molybdate, and 2 mM dithio-threitol. Gradients consisted of 30, 22.5, and 15% sucrose solutions (w/v) for sedimentation of dense and normal receptor forms (7). Samples (0.25 ml) containing [3H]TA-labeled receptor and 0.4 M KC1 were centrifuged at 58,000 rpm for 48 h at 5 "C in an SW 60 Ti rotor.
For separation of 10 S and 4-5 S cytosol receptor forms in experiments which did not involve density labeling, gradients consisted of 30, 20, and 10% sucrose (w/v) prepared in Hz0 containing 20 mM Tris-HC1, pH 7.4 at 25 "C, 5 mM sodium molybdate, and 2 mM dithiothreitol. These gradients were centrifuged at 58,000 rpm for 16 h at 4 "C in an SW 60 Ti rotor. In all cases, the cell samples were assayed for protein content by the method of Bradford (16) and equal amounts of protein were applied to each gradient. Gradients were fractionated by puncturing the bottom of the tube, and the radioactivity in each fraction was determined by scintillation counting at 43% efficiency. In each experiment, gradient studies represent duplicate cell cultures which were sedimented in parallel.
Quantitation of Receptor with Protamine Sulfate-10 S cytosolic receptor occupied with [3H]TA was quantitated by precipitation with protamine sulfate. A solution of protamine sulfate in 20 mM Tris-HCI, pH 7.4 at 25 'C, was added to cytosol so that the final concentration of protamine sulfate was 0.5 mg/ml. After 20 min at 4 "C, the precipitate was collected by centrifugation at 1,000 x g for 10 min. The precipitate was washed twice at 4 "C with 20 mM Tris-HC1, pH 7.4 at 25 "C. Bound [3H]TA was extracted from the washed precipitate by incubating with 0.5 ml of absolute ethanol for 30 min at 37 "C.
Statistical Analysis-Whenever appropriate throughout this paper, data are presented as the mean & standard error of the mean of the sample.

Quantitation of Glucocorticoid Receptor by Protamine Sulfate Precipitation and Sucrose-gradient Sedimentation-Ad-
ditional evidence that the 10 S receptor form represents unactivated receptor comes from the protamine sulfate-precipitation study illustrated in Fig. 1. Kalimi (17) reported that protamine sulfate can precipitate the unactivated but not the activated glucocorticoid receptor from rat liver. Fig. 1 shows that protamine sulfate precipitates the 10 S form and not the 4-5 S form which we have previously shown to be a DNAbinding protein (7). GH, cells, cultured in glucocorticoid-free medium for 48 h, were incubated for 1 h (Fig. 1, A  sedimented in low-salt sucrose/H20 gradients to separate 10 S and 4-5 S receptor forms (7).
In the control cytosols ( Fig. 1, A and C), both 10 S and 4-5 S receptor forms are found in the gradients. The percentage of cytosolic [3H]TA. receptor complexes in the 10 S form was 33 f 1% as shown on this and 34 other gradient profiles (not illustrated). Protamine sulfate treatment removes approximately 90% of the 10 S receptor without affecting the abundance of the slower-sedimenting form but slightly decreases its sedimentation coefficient (Fig. 1, B and D). Radioactivity corresponding to the 10 S form lost from the gradients was quantitatively recovered in the protamine sulfate precipitate. Although the nuclear-bound receptor was about 65% of totalcell receptor at both the 1-h and 24-h [3H]TA-incubation times, receptor levels were reduced at 24 h compared to the 1-h [3H]TA incubation. This is also reflected by the decreased abundance of cytosolic receptor forms at the 24-h incubation time both before and after protamine sulfate precipitation ( TA occupies greater than 90% of the receptor-binding sites.
In GH, cells exposed to hormone for 1 h, total cytosolic and nuclear binding of [3H]TA to receptor was approximately 260 fmo1/100 pg of DNA. Cells exposed to 10 nM hormone for longer times show an initial rapid decrease in the amount of nuclear and cytosolic receptor that by 6 h is about 75% of the 1-h value. By 24 h, total glucocorticoid receptor is 130 fmol/ 100 pg of DNA which is 50% of the control value. By 48 h, the amount of receptor appears to reach a new steady-state level which is typically 40-50% of the value in cells grown without hormone. Receptor reduction occurs in both the nuclear and cytosolic fractions and, at all times, 60-70% of the total receptor is localized in the nuclear compartment.

Influence of rH]Triamcinolone Acetonide ' on Receptor Abundance in HTC and Reuber H-35 Hepatoma Cells-The time-dependent depletion of glucocorticoid receptor levels by
[3H]TA has been confirmed in two related cell types. HTC and Reuber H-35 hepatoma cells have been extensively studied as model systems for glucocorticoid-hormone action (1% 22). A previous study in HTC cells demonstrated that t3H] T A converts the glucocorticoid receptor from a 10 S to a 4-5 S form and that, in the absence of hormone, the receptor in HTC cells has approximately the same half-life as in GH1 cells (7). Table I (0) and cytosolic (A) fractions were separated by centrifugation and used to calculate total receptor (0). The amount of labeled steroid bound specifically to glucocorticoid receptor was determined using protamine sulfate precipitation for the 10 S cytosolic and by ethanol extraction of nuclei as described under "Experimental Procedures." The 4-5 S receptor component was estimated based on the amount of 10 S receptor using protamine sulfate precipitation and the ratio between the 10 S and 4-5 S forms determined by gradient centrifugation. Parallel incubations containing 10 GM unlabeled TA were used to assess nonspecific binding.

TABLE I Glucocorticoid receptor depletion in HTC and Reuber H-35 hepatoma
cells Duplicate flasks of HTC and Reuber H-35 hepatoma cells previously grown for 24 h in medium with 10% hormone depleted serum were incubated for 48 h in serum-free Dulbecco's minimal essential medium. At the times indicated in the table prior to harvesting the cells, the medium was supplemented with 10 nM [3H]TA. Control cells were treated identically except that the [3H]TA was added to the culture medium 1 h prior to harvesting the cells. The nuclear and cytoplasmic glucocorticoid receptor components were quantitated as described in Fig. 2 and under "Experimental Procedures." Parallel incubations containing 10 g M unlabeled TA were used to assess nonmecific bindine.

Distribution of Pre-existing and Newly Synthesized Glucocorticoid Receptor in the Nuclear and Cytosolic Compartments
of GH1 Cells-In order to perform density-labeling studies, it was necessary to extract the maximum amount of glucocorticoid receptor from the nuclear fraction prior to sucrose/D20 centrifugation. Glucocorticoid hormone-receptor complexes can be extracted from isolated nuclei with high concentrations of KC1 (21), with pyridoxal phosphate (23,24), or by digestion with micrococcal nuclease (25). An analysis was made of the efficiency of extraction of nuclear glucocorticoid receptor using KC1 concentrations from 0.4-0.6 M, micrococcal nuclease digestion followed by extraction with high salt, and, finally, 5 mM pyridoxal phosphate alone and in combination with high salt (data not shown). In all cases, the degree of extraction of nuclear-bound receptor was approximately 55% and was independent of the 13H]TA concentration or incubation time. Therefore, extraction with 0.4 M KC1 was chosen for all subsequent studies. This also had the additional advantage of converting the oligomeric 10 S receptor to the 4-5 S form to allow for estimation of receptor half-life and synthesis rates by dense amino acid labeling (7).
Previous short-term incubation studies (3 h) (7) indicated that the nuclear and cytoplasmic forms of the glucocorticoid receptor are in a state of rapidly exchanging equilibrium. To examine this with long-term hormonal exposure and to assess whether the nuclear and cytoplasmic pools have similar turnover rates, dense amino acid labeling of nuclear and cytosolic receptor was examined. Cells were incubated with 10 nM [3H] TA for a total of 44 h and the culture medium was exchanged for dense medium containing 10 nM [3H]TA at the times indicated prior to harvesting the cells. Nuclear and cytosolic fractions were isolated and the receptor was extracted from the nuclear fraction with 0.4 M KC1. After adjusting the cytosolic fraction to 0.4 M KC1, normal and density-labeled receptors from each compartment were separated by velocity sedimentation in sucrose/D,O gradients (Fig. 3). The amounts of normal and dense receptor were determined by summing the appropriate regions of each gradient, which were corrected for the overlap of the two components (7). Cells cultured for 6 h in dense medium had approximately 30% of the receptor in the newly synthesized dense form. After 10 h in dense medium, the amount of dense receptor increased to approximately 50%. By 14 h, the nuclear receptor form is 75% dense and the cytosolic form of the receptor appears to be predom-inantly dense. The similar distribution of newly synthesized and pre-existing receptor in both cytoplasmic and nuclear compartments after various times of dense amino acid labeling suggests that receptor rapidly equilibrates between these two compartments.
In a related study, GH1 cells previously depleted of glucocorticoid hormones were grown in dense medium for 6-12 h. Exactly 1 h prior to harvesting the cells, 10 nM [3H]TA was added to the culture medium to occupy and elicit nuclear translocation of receptor. The nuclear and cytosolic fractions were separated by centrifugation, adjusted to 0.4 M KC1, and sedimented in sucrose/D,O gradients (data not shown). After each time period of density labeling, the ratio of dense to normal receptors was essentially identical to the nuclear and cytoplasmic compartments. Since the relative abundance of dense and normal receptor was the same in both compartments after short-or long-term incubations with [3H]TA, subsequent analysis of receptor half-life and synthetic rates was made with receptor extracted from broken-cell preparations with 0.4 M KCI.
Dense Amino Acid Labeling of Receptor in the Absence of Hormne-GH1 cells previously depleted of glucocorticoid hormones were cultured in medium containing dense amino acids without [3H]TA for 4,12,16, and 22 h (Fig. 4). The cells then received 10 nM [3H]TA 1 h prior to harvesting, and receptors were extracted from whole cells with 0.4 M KC1 and analyzed in sucrose/D,O gradients. As illustrated in Fig. 4A, after a 4-h incubation with dense medium receptor of normal density comprises almost 90% of total receptor. After 12 h in dense medium, the amount of normal density receptor is reduced to 62% of the total with a concomitant increase in the dense form of the receptor (Fig. 4B). Incubation with dense medium for 16 h (Fig. 4C)  density remaining as a percentage of total receptor content (not illustrated). Thus, the change in receptor half-life appears to fully account for the reduction in receptor levels, sugesting that T A had little or no effect on the receptor synthetic rate. Since the receptor levels represent steady-state values, the synthetic rate (k,) can be calculated from the degradation rate constant the 10 S to the 4-5 S form which could increase the susceptibility of the receptor to proteolytic cleavage. Alternatively, the change in half-life could be a delayed effect of the binding of hormone to receptor, such as induction of a receptorspecific protease. To distinguish between short-and longerterm effects, density labeling was used to estimate receptor half-life in cells incubated for 7 h with 10 nM [3H]TA. The amount of receptor present during nonsteady-state conditions decreases during these experiments and, therefore, receptor half-life cannot be determined by the graphic method shown in Fig. 5.

Effect of fH]Triamcinolone Acetonide on Receptor Half-life and Synthetic Rate during Steady-state Conditions-Dense
An alternative method was used in Fig. 6 where receptor half-life was estimated using the formula: R, = R+o)e-kdf (27). Rt is the amount of receptor of normal density remaining in cells at any time ( t ) after the addition of dense amino acids to cells. The amounts of receptor present at the time of the addition of dense amino acids (R+oJ was determined by summing the dense and normal receptor in Fig. 6A which received [3H]TA for 1 h. The value for the degradation constant (kd) derived from the above equation was used to calculate receptor half-life (tlh = 0.693/kd) (26, 27). Receptor half-life estimated under these conditions decreased from about 20 h in the absence (Fig. 6A) to about 9 h in the presence of hormone for 7 h (Fig. 6B). Cells cultured in parallel flasks for 30 h with t3H]TA (Fig. 6C) showed a receptor half-life of 8 h. The estimated half-life of 9 h found after a 7-h exposure to hormone is in excellent agreement with the average value of 9.5 f 0.3 h derived from 24 separate determinations in receptor-depleted cells treated for 24 h or longer with 10 nM [3H]TA. It can be calculated that for this excellent agreement to occur, [3H]TA would have to shorten receptor half-life from 20 h to 9 h within 1 h of incubation.   (Fig. 6) can be reversed by removing the hormone (Fig. 7). Cells were cultured in medium containing 3 nM [3H]TA for 1 h, a concentration which will activate approximately 75% of the receptors (7). After removing hormone from cells by exchanging the culture medium several times over a 210-min period with hormone-free medium, the cells were cultured in dense medium for 11 h with or without 3 nM (3H]TA. This washout procedure removes greater than 99% of cell-associated [3H]TA. Receptor halflife was calculated from the amount of normal-density receptor found in gradients (Fig. 7, B and C) using the total receptor (dense + normal) as R(t=o) in the gradient shown in Fig. 7A.  (Fig. 7A) to 10 h in cells exposed to 3 nM [3H]TA continuously (Fig. 7B). In cells exposed to hormone for 1 h (Fig. 7C), receptor half-life was calculated as 15 h. Although in close agreement with the control cells in Fig.  7A, the 15 h half-life (Fig. 7C) may be an underestimate of the actual value. The 1-h preincubation with 3 nM [3H]TA would be expected to reduce the receptor by about 10% prior to the 11-h incubation with dense amino acids. This would result in a slight decrease in the Rt/R(t=o) ratio and therefore yield a half-life which is slightly less than the actual value. The results of Fig. 6 along with Fig. 7 suggest that the receptor half-life decreases rapidly after exposure to hormone and that, after removal of hormone, receptor half-life approaches the value found in cells cultured without hormone. Following the removal of hormone from cells, the rate of synthesis of receptor was 9.8 fmo1/100 pg of DNA/h (Fig. 7C). This was calculated under nonsteady-state conditions using the formula, k8 = kdRt/(l - (26,27). This rate is similar to the average ks of 9.6 f 0.8 fmo1/100 pg of DNA/h found in parallel cell

h. A third set of cells (Panel A ) was incubated for 1 h in medium without glucocorticoids.
Hormone was removed from cell cultures ( B and C) by replacing the culture medium four times over a 210-min period with hormone-free medium. After this washing procedure, cells were incubated with dense medium for 11 h either in the presence ( B ) or absence ( A and C) of 3 nM [3H]TA. Cells which did not receive hormone ( A and C) during the dense amino acid-labeling period were incubated with 3 nM [3H]TA for 1 h prior to harvesting. Normal and dense receptors were extracted and quantitated as described in Fig. 4. Estimates of receptor half-life were made using the formulas indicated in the legend to Fig. 6. was determined by summing the dense and normal receptor in Panel A . The gradient profiles illustrated are fractions 1-28 of a total of 40 fractions collected from each gradient.

15
cultures grown both in the presence (Fig. 7B) and absence (Fig. 7A) of hormone.
Relationship of Receptor Occupancy and Receptor Half-life to Receptor Depletion-A dose-response experiment further confirmed that a reduction of receptor half-life fully accounts for the hormone-mediated decrease in steady-state receptor levels (Fig. 8). GH, cells were incubated for 43 h in medium containing [3H]TA concentrations which are known to occupy a specific percentage of total glucocorticoid receptor (0.2 nM, During the last 11.5 h of incubation, normal medium was exchanged for medium containing dense amino acids and the [3H]TA concentrations were maintained. One h prior to harvesting the cells, the [3H]TA concentration was adjusted to 10 nM in each cell culture and the dense and normal forms of the receptor were resolved in sucrose/D20 gradients. Degradation rate constants and half-life values were calculated using Rt as the amount of normal density receptor, t = 11.5 h, and R+o) as the steady-state amount of receptor (dense + normal) for each [3H]TA concentration. As illustrated in Fig. 8, half-maximal depletion of total receptor occurred at about 0.5 nM TA. This value is similar to the concentration required for half-maximal occupancy of receptor (Fig. 8) (7), suggesting that receptor activation is necessary for receptor depletion. Normal medium was replaced with dense medium 11.5 h before harvesting the cells and the concentration of the hormone in the medium was maintained. The concentration of [3H]TA was adjusted to 10 nM in each culture 1 h before harvesting. Normal and dense receptors in total cell 0.4 M KC1 extracts were separated and quantitated as described in Fig. 4.  for the various [3H] TA concentrations was estimated using the formulas indicated in the legend to Fig. 6. R+,,), the amount of receptor present at the time of addition of dense amino acids, was considered the total amount of dense and normal receptor from each gradient (0). The per cent occupancy (A) is derived from a previous study in which the equilibrium dissociation constant of dense and normal glucocorticoid receptor for [3H]TA in GH, cells was determined (7).

DISCUSSION
Dense amino acid labeling has been used to analyze the influence of [3H]TA on the level of glucocorticoid receptor in GHl cells. When these cells are cultured in medium without glucocorticoid, receptor half-life was about 19 h and receptor was synthesized at a rate of approximately 10 fmo1/100 pg of DNA/h. These rates establish the steady-state amount of receptor in cells cultured without glucocorticoid at about 10/ (0.693/19) = 275 fmo1/100 r g of DNA. When GH1 cells are cultured with 10 nM [3H]TA, receptor half-life is about 10 h and the average synthetic rate was also approximately 10 fmo1/100 pg of DNA/h. These rates account for a new steadystate level of receptor of 10/(0.693/10) = 144 fmo1/100 pg of DNA, which is about 50% of the value found in control cells cultured without hormone.
The amount of glucocorticoid receptor in GH1 cells begins to decrease within 3 h after addition of 10 nM TA in cells previously cultured without glucocorticoid (Fig. 2). This decrease appears to be secondary to a rapid effect of [3H]TA on shortening receptor half-life without altering receptor synthetic rates (Fig. 6). Based on the dense amino acid-labeling time of 7 h shown in Fig. 6 (Fig. 2).
In this study, estimation of receptor half-life and the influence of [3H]TA on receptor turnover was performed using serum-free conditions. In a previous study, our laboratory reported (7) a 10-h half-life for the glucocorticoid receptor in GH1 cells which was determined by dense amino acid labeling with media containing 10% hormone-depleted calf serum (v/ v). The %fold difference in receptor half-life in cells cultured without hormone in serum-free (20 h) uersus serum-containing media (10 h) has been consistently reproduced in our laboratory.
[3H]TA also elicits receptor depletion in GH, cells cultured in media supplemented with hormone-depleted serum. However, the maximal depletion is approximately 35% compared to 50% with serum-free conditions.' We do not know the reason for these differences. However, since we were primarily interested in studying receptor regulation by glucocorticoid, and serum contains a variety of undefined factors, we performed our studies under serum-free conditions.
In GHl cells, the mechanism of glucocorticoid-receptor activation and nuclear translocation is consistent with a model in which the unactivated receptor is a 10 S oligomer (tetramer) that is in rapid equilibrium with a 4-5 S subunit. Glucocorticoid agonists shift the 10 S to 4-5 S equilibrium in the direction of the 4-5 S form, which appears to be activated receptor and shows high affinity for DNA (7). Evidence to support this model comes from: 1) dense amino acid-labeling studies of the 10 S and the 4-5 S receptor forms, 2) glucocorticoid agonist-dose response conversion of the 10 S to the 4-5 S form which parallels nuclear translocation of receptor and, 3) the loss of nuclear receptor and the reappearance of 10 S cytosolic receptor after removal of hormone from intact cells without requirement for new receptor synthesis (7). Additional evidence that the activated receptor in cytosol is in equilibrium with nuclear-bound receptor comes from the dense amino acid-labeling studies illustrated in Fig. 3. In this study, cells were incubated with 10 nM [3H]TA for 44 h followed by dense amino acid labeling in the presence of [3H] T A for 6, 10, and 14 h prior to harvesting. The ratio of newly synthesized dense to pre-existing receptor of normal density in the cytosolic and nuclear compartments were virtually W. R. McIntyre and H. H. Samuels, unpublished observation. identical, indicating that nuclear-bound receptor is in constant exchange with receptor in the cytosol.
Several lines of evidence suggest that [3H]TA decreases receptor half-life as a consequence of ligand mediated conversion of the 10 S cytosolic receptor to the 4-5 S nuclearbinding form. First, glucocorticoid receptor half-life and steady-state receptor levels are inversely related to the extent of receptor occupancy and activation in the intact cell (Fig.  8). Secondly, we estimate that [3H]TA shortens receptor halflife from 20 h to 9 h within 1 h of incubation (Fig. 61, at which time receptor is fully occupied and nuclear translocation is complete (7). Finally, removal of hormone, which rapidly regenerates the 10 S unactivated receptor from the 4-5 S nuclear-binding form (7), restores receptor half-life to the value observed in cells cultured without hormone (Fig. 7). Since reduction of receptor half-life by [3H]TA occurs when protein synthesis is inhibited by cycloheximide (Table 11), receptor degradation is not secondary to hormonal induction of a receptor-specific protease but appears to reflect ligandmediated formation of the 4-5 S receptor form which is degraded at an increased rate by the cell.
Although these studies suggest that the 4-5 S receptor form is degraded more rapidly than the 10 S receptor by the cell, dense amino acid-labeling techniques cannot distinguish whether the 4-5 S receptor form is degraded more rapidly in the nuclear or cytosolic compartment. However, since the nuclear and cytosolic receptor pools are in a rapidly exchanging equilibrium (Fig. 3) (7), the ligand-mediated reduction of receptor is evident in both the cytosolic and nuclear compartments. Although [3H]TA translocates about 65% of the glucocorticoid receptor to the nucleus in GHl cells, only 55% of nuclear [3H]TA. receptor complexes are extracted with 0.4 M KCl. Pyridoxal phosphate or nuclease digestion, alone or in combination with 0.4-0.6 M KCl, did not increase the efficiency of nuclear-receptor extraction. Since all cytosolic receptors can be examined, our dense amino acid-labeling studies analyzed approximately 70% ((35% cytosol) + (0.55)(65% nuclear)) of total-cell receptor while the remainder represents unextracted nuclear receptor.
The partial extraction of nuclear [3H]TA. receptor complexes may reflect the trapping of receptor in chromatin during the extraction process (21,29) or the binding of [3H] TA. receptor complexes to a limited number of high-affinity nuclear-binding sites as has been suggested for the estrogen receptor (30). Nevertheless, the effect of [3H]TA on receptor half-life appears to influence both the 0.4 M KC1 extractable and nonextractable nuclear-receptor pools since the ratio of salt-sensitive to salt-resistant nuclear-bound glucocorticoid receptor complexes was constant in control and receptordepleted cells. In addition, the receptor half-lives derived from dense amino acid-labeling analysis (extracted nuclear + cytosol receptor) accounts for the extent of receptor depletion in the salt-resistant fraction as well.
As alternative explanation for hormone-mediated receptor depletion is reversible inactivation of the receptor-binding site rather than acceleration of the rate of receptor degradation. In this case, inactivation refers to loss of glucocorticoid binding rather than inability to interact with DNA or other nuclear components. Wheeler et al. (31) have shown that when IM-9 lymphocytes are incubated in glucose-free medium without oxygen, the glucocorticoid-receptor-binding capacity decreases approximately 2-%fold. When glucose and oxygen were reintroduced, the receptor-binding activity returned and was independent of protein synthesis. Therefore, TA-mediated receptor depletion in GHl cells could result from an effect of hormone which interferes with some metabolic events and results in inactivation of the receptor-binding site. Our results, however, do not support this possibility since random inactivation of the receptor-binding site would be expected to occur in both pre-existing normal as well as newly synthesized dense receptor. This would result in no change in the calculated half-life since the ratio of Rt/R(t=o) would remain the same as without hormone. In addition, we also find that receptor synthesis is essentially identical in the presence and absence of hormone. Since receptor synthetic rates are calculated from the amount of [3H]TA bound to receptor, identical values would not be obtained if reduction in receptor was due to loss of a fraction of the receptor-binding sites. The observation that glucocorticoid-mediated receptor activation to the DNA binding form involves a decrease in the sedimentation coefficient of the receptor (7) is not unique to GH1 cells. Using intact rat thymus cells, Holbrook et al. (32) subsequently reported that hormone-mediated receptor activation involves the conversion of a 9.2 S unactivated receptor to a 4.8 S activated form. Using cytosol from AtT-20 mouse pituitary cells, Vedeckis (33) reported that receptor activation results from the conversion of 9 S species to an activated form which sedimented at 3.2 S or 5 S, depending upon the KC1 concentrations in the gradient. Sherman et al. (34) first reported that a variety of steroid hormone cytosolic receptors can exist in a 9-10 S form as well as slower-sedimenting species. For example, receptors for glucocorticoid hormones (35-38), estrogens (35, 39), progestins (35, 40), mineralocorticoids (41), and androgens (42) have been reported to sediment as 8-10 S on low-salt gradients and 3.5-5 S on high-salt sucrose gradients. This suggests that most steroid-hormone receptors can exist as oligomeric forms and that activation to the DNA-binding form may be related to dissociation into subunits.
Recent studies by Welshons et al. (43) and by King and Greene (44) have suggested that, in the absence of ligand, the estrogen receptor is an intrinsic nuclear protein. These authors have suggested that identification of the receptor in the cytosol is an artifact which is a consequence of cell lysis. Although this has not been reported for the glucocorticoid receptor, it remains possible that the ligand-mediated 10 S to 4-5 S activation step occurs in the nucleus and is observed in the cytosol only after disrupting the cell. It should be emphasized, however, that a nuclear site of activation does not alter our conclusions concerning the regulatory effect of ligand on glucocorticoid receptor levels and half-life.
Dense amino acid labeling of the 10 S and 4-5 S glucocorticoid receptor forms is compatible with a reversible model to explain hormone-mediated receptor activation and reversion to an inactive form after hormone removal (Figs. 3 and 7) (7). To account for our observations, we proposed a model in which unactivated and activated receptor are in equilibrium and hormone acts to shift the equilibrium to the activated DNA-binding form (7). Munck and Holbrook (45) recently proposed an alternative cyclic but irreversible model of receptor activation which requires energy for regeneration of unactivated receptor. Although our current study supports the view that activated receptor is degraded more rapidly by the cell, is does not distinguish between the reversible or irreversible models for receptor activation. Mineralocorticoids (46), progesterone (47, 48), and estrogen (49) have been reported to elicit rapid decreases in total-cell homologous-receptor content. Recently, Eckert et al. (50) reported that estrogen can shorten the half-life of the estrogen receptor in MCF-7 cells. Therefore, a decrease in receptor abundance and rapid shortening of receptor half-life may be a general phenomena which occurs as a consequence of steroid-hormone-mediated receptor activation in the cell.