&-Adrenergic Receptors in Hamster Smooth Muscle Cells Are Transcriptionally Regulated by Glucocorticoids*

Steroid hormones modulate adrenergic receptor re- sponsiveness and receptor number. To investigate the regulation of the Bz-adrenergic receptor gene by glu- cocorticoids we examined the effects of the synthetic glucocorticoid agonist triamcinolone acetonide on the expression of &,-adrenergic receptors in DDTIMF-2 hamster smooth muscle cells. Glucocorticoid treatment (1 X M) produced a 2.2 f 0.4-fold (n = 8) increase in &-adrenergic receptor number (maximum between 6 and 12 h) as determined by radioligand binding and a similar increase in catecholamine-stimulated adenylate cyclase activity. Steady-state levels of &-adrener-gic receptor mRNA, analyzed by Northern blot hybrid- ization, were increased 2.4 f 0.4-fold (n = 6) within 1 h, while actin mRNA levels were unchanged through- out the experiment. These steroid-induced increases in &-adrenergic receptor mRNA returned to control lev- els by 24 h and were followed by a much slower decline in &-adrenergic receptor in plasma membranes. The rate of &-adrenergic receptor gene transcription, as-sessed by nuclear run-off transcription assays, in- creased 3.1 f 0.1-fold (n = 2) in cells treated for 30 min with 1 X M triamcinolone acetonide. These studies indicate that glucocorticoids regulate the 82-adrenergic receptor-adenylate

The &-adrenergic receptor is a member of the family of membrane receptors involved in guanine-nucleotide regulatory protein (G-protein)-mediated signal transduction. One of the principal mechanisms for the regulation of these transmembrane signaling systems is the modulation of receptor number. This may be achieved in several ways. The loss of receptors which occurs following prolonged exposure to agonist is accompanied by a reduction in responsiveness. This phenomenon is termed "down-regulation" (1). Changes in receptor number may also arise as a consequence of pathophysiological conditions. For example, it has been known for some time that hyperthyroidism or conditions leading to elevated glucocorticosteroid levels can produce symptoms * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. reminiscent of a hyperadrenergic state in the absence of detectable changes in serum catecholamine levels (2,3). This form of regulation has been called "heterologous regulation," and it refers to the situation where hormones and drugs that are not specific ligands for adrenergic receptors can nevertheless regulate adrenergic receptor responsiveness.
Studies in vivo and in vitro have shown that glucocorticoids raise /3-adrenergic receptor levels and agonist-stimulated adenylate cyclase activity (4-8). Since a major mode of steroid hormone action is the modulation of the rate of target gene transcription (9), we sought to determine whether the 8adrenergic receptor is regulated by glucocorticoids at the level of gene expression.

MATERIALS AND METHODS
Cell Culture-DDTIMF-2 cells (10) were grown in suspension culture in Dulbecco's modified Eagle's medium (GIBCO) supplemented with 20 mM Hepes,' 5% fetal calf serum, 100 units/ml penicillin, and 100 pg/ml streptomycin. Cells were placed in fresh medium at 5 X lo' cells/ml and grown in suspension for 2 days, at which time cell density was 2.2 k 0.4 X 106 cells/ml (mean * S.E., n = 8). The synthetic steroid triamcinolone acetonide (TA) (Sigma) was added to the culture from stock solutions prepared in 100% ethanol. The final concentration of ethanol in the medium did not exceed 0.1%. At the indicated times following the addition of drugs, aliquots of cells (1-5 X lo6 for ligand binding; 1-2 X 10' for RNA) were harvested by centrifugation (800 X g for 5 min). The cell pellets were flash-frozen in liquid nitrogen and stored at -80 'C until use. Cells grown in the presence of TA essentially cease dividing. Cell viability was estimated by trypan blue exclusion to be 85-90%.
Preparation of Plasma Membranes and Assays for Radioligand Binding and Adenylate Cyclase-Frozen cell pellets were thawed in buffer (75 mM Tris, pH 7.4,12.5 mM M&12,0.25 M sucrose containing 5 pg/ml each of soybean trypsin inhibitor and leupeptin). Plasma membranes were prepared as previously described (11). Protein concentrations were determined by the method of Bradford (12). Binding of the @-adrenergic receptor-specific ligand ['261]cyanopindolol (['"I] CYP) was performed as previously described (13) in the absence or presence of 1 p~ (-)-alprenolo1 to define total and nonspecific binding, respectively. Catecholamine-sensitive adenylate cyclase activity was measured in freshly prepared membranes as described previously (14,15).
Isolation of RNA-Total cellular RNA was isolated by the cesium chloride gradient method of Chirgwin (16). Briefly, cells that had been flash-frozen at the time of harvest were thawed directly into 8 ml of 4 M guanidinium isothiocyanate, 2% sodium lauryl sarcosine, 50 mM Tris (pH 7.4), 10 mM EDTA, and 0.15 M 2-mercaptoethanol. The homogenate was passed through a 20-gauge needle three times to shear the DNA, layered over 2.0 ml of 5.7 M CsCl, 25 mM sodium acetate (pH 6), 1 mM EDTA, and centrifuged at 32,000 rpm in a Sorvall TH641 rotor for 20 h.

9067
lished by comparison with an RNA Ladder and 6x174 RFIHueIII DNA fragments (Bethesda Research Laboratories) stained with methylene blue after blotting (21). Autoradiographic bands were quantitated by densitometric scanning.
Nuclear Run-off Transcription Assay-Following the addition of either 1 X M TA or ethanol vehicle, DDTIMF-2 cells were harvested by centrifugation and washed in a buffer containing 10 mM Tris, pH 7.5, 10 mM KCl, 3 mM MgC12, 3 mM dithiothreitol. Nuclei were prepared exactly as described (22) with the following exceptions. The nuclei were sedimented (25,000 rpm in a Sorvall TH641 rotor at 4 "C) through a 3.0-ml cushion of 2.0 M sucrose in the lysis buffer and were used immediately without freezing. All subsequent steps of transcript elongation in the presence of 200 pCi of [~I -~~P ] U T P (Du Pont-New England Nuclear) (22) and purification of radiolabeled RNA (23) were performed exactly as described. The labeled RNA (1 X lo7 dpm) was hybridized as previously described (24) to linearized plasmid DNAs (2 pg each) immobilized on nitrocellulose strips using a slot blot manifold (Schleicher and Schuell). pMB1.3 contains the 1.3-kb hamster &adrenergic receptor cDNA (19). The plasmid vector pGEM (Promega Biotec, Madison, WI) was used to determine nonspecific background hybridization.

RESULTS
Preliminary dose-response experiments indicated that maximal effects on the parameters measured in these studies were observed with 1 X M TA; therefore, this concentration of the hormone was used throughout these investigations. Following the addition of 1 X M TA to DDTIMF-2 cells, the &-adrenergic receptor number increased 2-fold. Fig. 1 is representative of eight independent experiments in which the p2adrenergic receptor number steadily increased to a maximum of 2.6-fold at 8 h. While there was some variation between individual experiments, the peak of induction (2.2 & 0.4-fold) typically appeared between 6 and 12 h, plateaued, and then gradually declined over the next 2 days. In untreated control cells receptor number remained constant during the 48-h period (data not shown). From saturation binding studies there was no apparent change in the dissociation constant for ['251]CYP between control cells and treated cells (data not shown), in agreement with previous reports (5, 6, 8). In cells treated with 1 X M TA for 12 h and 20 h, isoproteronolstimulated adenylate cyclase activity was elevated 38 and 56%, respectively, over control cells (data not shown). Fluo-T I ride-stimulated activity was unchanged at all time points, but a modest rise in basal activity was observed following glucocorticoid treatment, similar to earlier findings (6, 8). Therefore, values for basal activity were subtracted from isoproteronol-stimulated activity before comparison.
A very rapid rise in steady-state &-adrenergic receptor mRNA levels was observed within 1 h of exposure of DDTIMF-2 cells to glucocorticoid by Northern blot analysis (Fig. 2 A ) . The appearance of the smaller less abundant transcript at 1.8 kb may be due to alternative polyadenylation, since in both the hamster and human &adrenergic receptor cDNAs there are two polyadenylation signals in the 3'-untranslated regions, which are 453 and 450 base pairs apart, respectively. We have observed increases in the &-adrenergic receptor message as early as 15 min after the addition of the steroid, but maximal stimulation was usually found between 1 and 2 h. This increase in mRNA then slowly decayed back to the level of the untreated control by 10 h. Throughout the experimental period there was no appreciable change in actin mRNA levels (Fig. 2B). Therefore, actin mRNA was used as an internal control for minor fluctuations in total RNA applied to the gel. When exposed relative to actin, there was an overall doubling of &adrenergic receptor mRNA in the first hour, which decayed to less than 0.5-fold at later time points (Fig. 2C). The average maximal increase in P2-adrenergic receptor mRNA from several experiments was 2.4 k 0.4-fold ( n = 6).
We next utilized nuclear run-off transcription assays to determine whether this doubling of steady-state &-adrenergic receptor mRNA levels was due to stimulation of the rate of &adrenergic receptor gene transcription. In cells treated for 30 min with 1 X M TA there was a 3-fold increase in Cells were harvested at various times after the addition of 1 X lo-' M TA, and total cellular RNA was prepared. From each time point 20 pg of RNA were fractionated through a 1.2% agarose gel and blotted as described under "Materials and Methods." Blots were probed with 32P-labeled (A) &adrenergic receptor cDNA (2.5 X lo6 dpm/ml) or ( B ) actin cDNA (2.0 X IO6 dpm/ml). Final washing conditions used were 0.1 X SSC, 0.1% SDS, 55 "C. The filters were exposed to Kodak XRP film. C, autoradiographs were quantitated by densitometric scanning, and the increase in &adrenergic receptor mRNA (includes major and minor hybridizing species) relative to actin mRNA is expressed as arbitrary units. The results shown are representative of six experiments.  (-) or presence (+) of 1 X 10" M TA for 30 min were harvested for isolation of nuclei. Transcript elongation in isolated nuclei was allowed to proceed in the presence of [cY-~'P]UTP, and the 32P-labeled run-off transcripts were hybridized to plasmids (2 pglslot) bound to nitrocellulose. pMB1.3 contains the 1.3-kb Hind111 fragment of the hamster Pz-adrenergic receptor cDNA (19). pGEM is a control vector containing no insert. Autoradiograms were scanned densitometrically to determine the average increase in &adrenergic receptor transcription rate, which was 3.10 -C 0.08-fold (mean -C S.D., n = 2). nascent &adrenergic receptor transcripts (Fig. 3). Hybridization to the hamster Pz-adrenergic receptor cDNA insert in plasmid pMB1.3 (19) was specific as indicated by the lack of hybridization to the control plasmid, pGEM.

DISCUSSION
These results demonstrate that the &-adrenergic receptor in DDT1MF-2 cells is regulated by glucocorticoids. The increase in receptor density and isoproteronol-stimulated adenylate cyclase activity are essentially identical to results which have been described previously in other tissues and species (4-8). Our experiments extend these previous findings by documenting for the first time the role of regulation of &adrenergic receptor gene expression in this process.
Immediately following the addition of steroid to the cells there was a rapid rise in steady-state levels of Pz-adrenergic receptor mRNA, which preceded the appearance of new receptors in the membrane. The peak in accumulation of Pzadrenergic receptor mRNA was reached within 1-2 h. Thus, the doubling of receptors can be attributed to equivalent increases in steady-state &adrenergic receptor mRNA levels. These elevated levels of &adrenergic receptor mRNA were not maintained, however, and they quickly returned to the level of the untreated control. Similar findings of transient mRNA accumulation and transcriptional enhancement have been reported for other steroid-regulated genes. For example, Granner and colleagues (25)' have observed that expression of the rat phosphoenolpyruvate carboxykinase gene is stimulated %fold by glucocorticoids within 1 h and then subsequently declines to 2-fold and stabilizes at that lower level. Likewise, dexamethasone has been shown to produce a severalfold increase in angiotensinogen mRNA, but this induction was temporary, approaching control levels by 24 h (26). The mechanisms responsible for this type of regulation are currently unknown but may be related to the ability of glucocorticoids to down-regulate their own receptors (27, 28), a process recently reported (29) to involve decreased transcription of the glucocorticoid receptor gene.
The return of Pz-adrenergic receptor mRNA levels to control values after approximately 1 day was followed by a more gradual decline of the Pz-adrenergic receptor number in the membrane. In other experiments we have found that by approximately 4 days, &-adrenergic receptor levels had also * D. Granner, personal communication. returned to control (data not shown). The half-life of the Pzadrenergic receptor has been estimated to be -30 hours (30). At this rate, after 90 h only 10-15% of the steroid-induced receptors would be estimated to remain, which agrees well with our observations. The decrease in &adrenergic receptor mRNA below the level of the uninduced control may be more apparent than real. Our culture conditions, which include 5% fetal calf serum, probably result in some modest level of glucocorticoid stimulation in the untreated control, as has been observed in other systems (31). In cells exposed to a serum-free formulation for 2 days we have observed a significantly lower basal level of Pz-adrenergic receptor and consequently a higher level of &adrenergic receptor induction (5fold; n = 2) following hormone treatment (data not shown). Thus, the true "basal" level of &adrenergic receptor mRNA to which the system returns may be lower than that observed in "control" cells prior to the addition of the steroid. However, these conditions were not favorable for cell viability and growth; therefore, we conducted all of our experiments in serum-containing medium.
For most steroid-responsive genes studied, expression is largely controlled by changes in the rate of transcription, although significant differences in message stability have also been described (9). By nuclear run-off transcription assays we detected a 3-fold increase in the rate of transcription of the &adrenergic receptor gene in cells treated for 30 min with TA. Therefore, these results indicate that the increase in steady-state &adrenergic receptor mRNA levels is due to enhanced transcription of the &-adrenergic receptor gene.
Heterologous hormonal regulation of &-adrenergic receptor, particularly by steroid hormones, has been the subject of considerable investigation (2-8), and it has been speculated that control of &adrenergic receptor gene transcription may be the underlying mechanism. This notion was further supported by the discovery of several consensus glucocorticoid response elements in the cloned &adrenergic receptor gene (32, 33). While we do not know to which, if any of these glucocorticoid response elements, functional significance can be ascribed, this is an obvious area for further study.