Agonist-induced Destabilization of &Adrenergic Receptor mRNA ATTENUATION OF GLUCOCORTICOID-INDUCED UP-REGULATION OF @-ADRENERGIC RECEPTORS*

down-regulated ag- onists and up-regulated glucocorticoids. interaction between these two opposing regulatory path- investigated at the levels of and in quantified dem-onstrated in the second phase of steroid induction, isoproterenol was to reduce levels altering the rate of transcription. Thus, by promoting destabilization of message, agonist can maintain counterregulation of the glucocorticoid response. These studies provide a more detailed picture of the dynamic regulation of the expression of a G-protein-linked receptor in response to counterregulation by agonist and steroid. Via the interplay of enhanced transcription (steroid effect) and altered mRNA stability (ag-onist effect) the steady-state level of receptor message and protein can be modulated dynamically by two opposing forces.

&Adrenergic receptor expression and receptor mRNA levels are down-regulated by &adrenergic agonists and up-regulated by glucocorticoids. The interaction between these two opposing regulatory pathways was investigated at the levels of receptor and receptor mRNA in DDTl MF-2 hamster vas deferens cells. Dexamethasone blunted a marked decrease in receptor expression induced by isoproterenol alone, as made visible by indirect immunofluorescence using antireceptor antibodies. Receptor mRNA levels were quantified by DNA-excess solution hybridization. Dexamethasone stimulated a sharp increase in receptor mRNA at 4 h following the addition of steroid in either the absence or the presence of isoproterenol. B y 12 h, dexamethasone treatment resulted in a new steadystate level of receptor mRNA double that observed in untreated cells. Isoproterenol blunted the dexamethasone effect observed at 12 h. Cells treated with isoproterenol and dexamethasone in combination displayed a new steady-state level only 30% greater than untreated cells. Measured by nuclear run-on assays, transcription rates of the receptor gene were unaffected in cells challenged with isoproterenol alone. Dexamethasone, in contrast, stimulated a 4-fold increase in b2adrenergic receptor gene transcription. Isoproterenol and dexamethasone in combination promoted a transcription rate comparable to dexamethasone alone. The half-life of receptor mRNA in untreated and dexamethasone-treated cells was 12 h. In contrast, &adrenergic receptor mRNA half-life declined to 5 h in cells that were treated with isoproterenol in the presence or absence of dexamethasone. Agonist-promoted destabilization and steroid-induced transcription provide mechanisms for the interplay of two opposing pathways controlling receptor mRNA levels.
Glucocorticoids and @-adrenergic agonists play important roles in the regulation of the hormone-sensitive adenylate cyclase . The @-adrenergic receptor is a prominent locus for permissive hormone effects (Davies and Lefkowitz, 1984). Glucocorticoids stimulate a 2-3-fold increase in receptor expression and sensitivity to stimulation * This work was supported in part by United States Public Health Services Grants DK25410 and DK30111 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 "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
In addition to regulation by permissive hormones, the 0adrenergic receptor is regulated by agonist stimulation. Treatment of cells with P-adrenergic agonists leads to an "uncoupling" of the receptor from the stimulatory G-protein,' G.
(the G-protein that mediates hormonal stimulation of adenylate cyclase), and a loss of radioligand binding (Sibley and Lefkowitz, 1985). Chronic stimulation of DDT, MF-2 cells with agonist promotes down-regulation of &adrenergic receptor and receptor mRNA levels. @Adrenergic agonists promote a down-regulation of receptor mRNA, establishing a new steady-state level of message at 18 h of challenge which is 40-50% lower than that observed in untreated cells (Hadcock and Malbon, 1988a). @-Adrenergic agonists promote bronchodilation and represent important therapeutic agents for the treatment of asthma. Chronic use of P-adrenergic agonists, however, leads to adaptation, manifest by a reduced ability of P-adrenergic agonists to maintain bronchodilation (Nelson, 1986). 0-Adrenergic receptors in lung are down-regulated by chronic stimulation with @-agonists. This agonist-promoted downregulation can be reversed by treatment with glucocorticoids (Davies and Lefkowitz, 1984). Exposure to glucocorticoids, in vivo or in vitro, restores the responsiveness of isoproterenolstimulated CAMP accumulation in leukocytes desensitized by chronic incubation with P-adrenergic agonists (Lee and Reed, 1977;Logson et al., 1972). The reversal of agonist-induced down-regulation of the P-adrenergic receptor by glucocorticoids has been well characterized in DDT, MF-2 cells at the level of radioligand binding. Scarpace et al. (1985) have demonstrated that glucocorticoid treatment of cells previously exposed to agonist "resensitizes" the cells to stimulation by P-agonists and promotes recovery of @-adrenergic receptors to levels greater than that observed in control cells. The recovery of receptors is associated with a complete restoration of agonist-stimulated adenylate cyclase.
lized to quantify p-adrenergic receptor mRNA levels in DDTl MF-2 and other cell lines Malbon, 1988a, 1988b;George et al., 1988;Malbon and Hadcock, 1988). The DDT, MF-2 cell line has been well characterized at the biochemical and molecular levels with regard to steroid and catecholamine actions on the regulation of the hormonesensitive adenylate cyclase (Norris and Kohler, 1977;Norris et al., 1987;Scarpace et al., 1985;Malbon, 1988a, 1988b;Collins et al., 1988). A more complete understanding of the interaction between agents with opposing actions on (3adrenergic receptor expression was sought in the present work. At the level of mRNA, DNA-excess solution hybridization and nuclear run-on transcription assays were employed. At the level of receptor protein, radioligand binding and indirect immunofluorescence were used to study the expression of p-adrenergic receptors.  Malbon, 1988a).

Materials
Cell Culture-DDT1 MF-2 cells were grown to confluence as monolayers in Dulbecco's modified Eagle's medium supplemented with 5% fetal bovine serum, 3.75 g/liter sodium bicarbonate, streptomycin (60 mg/liter), and penicillin (60 mg/liter), as described by Scarpace et al. (1985). Stock solutions of steroid were prepared in ethanol vehicle and diluted into this same cell culture medium. Untreated (control) cells received vehicle in medium only.
DNA-excess Solution Hybridization Probe Synthesis-The solution hybridization probe (PAR 170) was constructed as described (Hadcock and Malbon, 1988a;Hadcock et al., 1989). Briefly, the single-stranded probe corresponding to nucleotides 12-182 of the hamster P-adrenergic receptor cDNA coding region was subcloned into bacteriophage M13mp18 and used to transform Escherichia coli JM101. The probe, PAR 170, was prepared as described by Williams et al. (1985) as modified by Bahouth et al. (1988). The probe was uniformly radiolabeled with [ c~-~' P ]~C T P (specific activity of 222 Ci/mmol) by primer extension followed by digestion with SmaI. The probe was gel isolated and subjected to hydroxylapatite chromatography. Single-stranded probe was isolated from double-stranded DNA by a step gradient of NaP04 (20-220 mM) and desalted using a Centricon 30 (Amicon) cartridge.
DNA-excess Solution Hybridization Assays-Uniformly radiolabeled probe (1.5 fmol (100 pg)/sample) was incubated with known amounts of template DNA (used as a standard) with total cellular RNA or alone for 60 h at 68 "C in 20 mM Hepes, pH 7.0,0.3 M NaC1, 1 mM EDTA, and 100 pg/ml denatured salmon sperm DNA. S1 endonuclease (150 units/ml), denatured salmon sperm DNA (50 pg/ ml), and s1 endonuclease buffer (0.28 M NaC1,4.5 mM ZnS04, 50 mM sodium acetate, pH 4.5) were then added to each sample and the mixture incubated for 90 min at 42 "C. The samples were treated with trichloroacetic acid (7.5%) and precipitated on ice for 10 min. The S1 endonuclease-resistant hybrids were collected by vacuum filtration on Whatman GF/C filters. The extent of internal resistance of the probe to S1 endonuclease treatment, typically less than I%, was determined for each assay. Data from DNA-excess solution hybridization assays were calculated as described earlier (Hadcock and Malbon, 1988a;Williams et al., 1985) and expressed as am01 of @-adrenergic receptor mRNA/pg of total cellular RNA.
Receptor mRNA Stability-The half-life of P-adrenergic receptor mRNA was determined as described (Rodgers et al., 1985). Cells were harvested from 0 to 12 h after the addition of actinomycin D (5 pg/ ml). Total cellular RNA was extracted at each of the time points, and receptor mRNA levels were quantified by DNA-excess solution hybridization. The concentration of actinomycin D employed in these studies (5 pg/ml) is sufficient to inhibit transcription in these cells by 99% for at least 16 h, as evaluated by ['Hluridine incorporation (Hadcock and Malbon, 1988b).
Nuclear Run-on Transcription Assays-Nuclei were isolated from DDT, MF-2 cells as described (Greenberg and Ziff, 1984). Cells (1 X 10') were harvested, washed two times with cold phosphate-buffered saline, and resuspended in a lysis buffer composed of 10 mM Tris-HC1, pH 7.4, 10 mM NaCl, 3 mM MgCI,, and 0.5% Nonidet P-40. Nuclei were collected by centrifugation at 500 X g for 10 min. The resulting pellets were washed and resuspended in glycerol buffer (50 mM Tris-HCI, pH 8.3,40% glycerol, 5 mM MgC12, 0.1 mM EDTA) for storage and maintained at -70 "C. Nascent transcripts were detected as described by Greenberg and Ziff (1984). Nuclei (200 pl) were added to 200 p1 of a reaction buffer composed of 10 mM Tris-HCI, pH 8.0, 5 mM MgCl,, 0.3 M KCI, 5 mM dithiothreitol, unlabeled GTP, ATP, CTP, and 10 pl of [w3'P]UTP (Du Pont-New England Nuclear, 800 Ci/mmol). Newly transcribed labeled RNA was extracted (Greenberg and Ziff, 1984) and then incubated for 36 h at 65 "C with plasmid DNAs immobilized on Nytran (Schleicher & Schuell). After hybridization, each sample was washed two times with 2 X SSC (1 X SSC is 15 mM sodium citrate, pH 7.0, 0.15 M NaCl) for 60 min at 65 "C. The samples were then treated with RNase A for 30 min at 37 "C followed by a wash with 2 X SSC at 37 "C for 60 min. The filters were dried and subjected to autoradiography for 72 h with an intensifying screen.
Relative changes in transcription were assessed by scanning densitometry of the autoradiogram. Radioligand Binding-@-Adrenergic receptors were measured in postnuclear supernatant (160,000 X g) membrane fractions by equilibrium radioligand-binding analysis with the antagonist ligand (-) [1251]iodocyanopindolol. Nonspecific binding was measured by competition with 10 p M isoproterenol and represented -10% of the total binding (Scarpace et al., 1985).
Indirect Immunofluorescence of &Adrenergic Receptors-Indirect immunofluorescence of P-adrenergic receptors was performed with intact DDT, MF-2 cells grown on glass chamber slides. The cells were fixed with paraformaldehyde and probed with either the antireceptor antiserum (George et al., 1988;Hadcock and Malbon, 1988b) or preimmune serum, at 1:200 dilutions. The fixed cells were then stained by rhodamine-conjugated goat anti-rabbit IgG (Boehringer Mannheim) diluted 1:lOOO. Phase-contrast and epifluorescence microscopy were performed by a Zeiss Axiophot system and photographed with hypersensitized Kodak 2415 film (George et al., 1988).

RESULTS
The steady-state levels of p-adrenergic receptor mRNA in control untreated cells were found to be 0.63 f 0.05 amol padrenergic receptor mRNA/pg of total cellular RNA (amol/ pg of RNA). When cells were treated with dexamethasone alone, fl-adrenergic receptor mRNA levels increased to 1.7 f 0.2 amol/pg of RNA by 4 h (Fig. LA). By 12 h, a new steadystate level of 1.3 f 0.15 ( n = 5) amol/pg of RNA was achieved. This new steady-state remained essentially constant for the next 24 h. Challenging cells with dexamethasone and a pagonist isoproterenol similarly produced a sharp increase in the steady-state level of receptor mRNA at 4 h, rising from 0.63 to 1.53 f 0.17 ( n = 3) amollpg of RNA. In contrast to the sustained elevation of receptor mRNA which occurred at later times (>4 h) in response to steroid alone, receptor mRNA levels were not sustained in cells treated concurrently with dexamethasone and isoproterenol. By 12 h, p-adrenergic receptor mRNA declined from 1.3 to 0.85 f 0.12 ( n = 3) amol/pg of RNA. Challenge with isoproterenol alone induced a progressive decline in p-adrenergic receptor mRNA levels. After a lag of 4 h, receptor mRNA of isoproterenol-treated cells declined to 50-60% of their original levels (Fig. 1B). Dexamethasone induced a sharp increase in receptor mRNA levels in control cells (Fig. L A ) as well as in isoproterenoltreated cells (Fig. 1B) in which agonist had promoted a substantial down-regulation of receptor mRNA. By 4 h, steroid increased mRNA levels in isoproterenol-treated cells from 0.38 f 0.09 ( n = 3) to 0.9 2 0.12 (n = 3) amol/pg of RNA (Fig. 1B). This new steady state in receptor mRNA induced by dexamethasone remained elevated over the next 12 h.
Was the attenuated dexamethasone response in isoproterenol-treated cells a result reduced sensitivity of the cells to  (isoldex, A-A).
B, DDTl MF-2 cells were incubated for the indicated times with vehicle (untreated, W), 10 p~ isoproterenol (W), and 10 p~ isoproterenol followed by concurrent addition of 500 nM dexamethasone at 24 h (A-A). Cells were harvested, RNA was extracted, and DNA-excess solution hybridization was performed as described under "Experimental Procedures." The results displayed are the means (2S.E.) of three to five separate determinations. Each determination was performed in duplicate.
the steroid? The dose dependence of the steroid-induced increase of receptor mRNA was examined in control cells and cells challenged with agonist (10 p M isoproterenol for 24 h). In both control and isoproterenol-treated cells alike, the maximal response in receptor mRNA occurred a t -250 nM dexamethasone. Half-maximal stimulation by steroid was achieved a t equivalent concentrations of steroid in control and isoproterenol-treated cells (Fig. 2).
Isoproterenol and dexamethasone regulation of 0-adrenergic receptor mRNA levels was explored at the level of transcription by nuclear run-on assays. Dexamethasone (4 h, 500 nM) enhanced the relative rates of transcription of the p2adrenergic receptor gene by 4-fold (Fig. 3). Isoproterenol, in contrast, failed to alter the relative rate of transcription of the &adrenergic receptor gene. When cells were challenged with dexamethasone and isoproterenol simultaneously, transcription rates of the &adrenergic receptor gene were enhanced to levels essentially equivalent to those of cells treated with steroid alone.
One possible mechanism by which the steady-state receptor mRNA levels might be down-regulated by agonist was evalu- Dexamethasone but not isoproterenol modulates the rate of transcription of the &adrenergic receptor gene: analysis by nuclear run-on assays. DDT, MF-2 cells were treated with vehicle (basal), 1 SM dexamethasone (dex) alone, 10 p M isoproterenol (iso) alone, or dexamethasone and isoproterenol in combination (&x/ iso) for 4 h. The cells were harvested, and nuclei were isolated as described under "Experimental Procedures." Transcription elongation was allowed to continue in the presence of [a-"PJUTP and unlabeled nucleotides. After elongation, radiolabeled RNA was hybridized either to plasmid harboring the @-adrenergic receptor cDNA (BAR) or to the plasmid lacking the receptor cDNA insert (pUC).  were analyzed by scanning densitometry to quantify the relative changes in the rate of transcription. The left panel is a representative autoradiogram. The right panel is the mean (+ the range) of values derived from two independent experiments. ated by investigating the stability of receptor mRNA. Cells were treated with vehicle, isoproterenol (10 pM, 24 h) alone, dexamethasone alone (500 nM, 12 h), or isoproterenol and dexamethasone in combination for 24 h and then concurrently with actinomycin D for 12 h (Fig. 4). Cells were then harvested at 0, 4, 8, and 12 h following the addition of actinomycin D. Using the approach of Rodgers et al. (1985), we estimated the half-life of receptor mRNA. Receptor mRNA displayed a halflife of -12 h in control cells. Interestingly, receptor mRNA half-life was observed to decline from 12 to -5 h in cells challenged with @-agonist. Receptor mRNA half-life was -5 h in cells treated with isoproterenol and dexamethasone in combination. Dexamethasone, in contrast to @-agonist, failed to alter the half-life of receptor mRNA (Fig. 4).
Receptor expression at the protein level in response to steroid and 0-agonist was probed by two independent methods, radioligand binding and indirect immunofluorescence or simultaneously with isoproterenol and 500 nM dexamethasone (dex + iso,

A-A),
then concurrently with actinomycin D (5 pg/ml). Cells were harvested at 0, 4, 8, and 12 h after the addition of actinomycin D. p-Adrenergic receptor mRNA levels (amol/pg of total cellular RNA) of cells were as follows: control, 0.60; isoproterenol-treated, 0.39; dexamethasone-treated, 1.23; and dexamethasone + isoproterenol-treated, 0.76. The data are mean values of three separate experiments except the dexamethasone and isoproterenol in combination, which is the average of two separate experiments. Each separate determination of mRNA was performed in duplicate.

TABLE I Radioligand binding to DDT, MF-2 cell membranes
DDT, MF-2 cells were treated with each of the indicated or vehicle (control) for 48 h. Membranes were prepared from 1,000 X g postnuclear supernatants collected by centrifugation at 160,000 X g for 120 min. Radioligand binding was then performed on membranes  (Table I). ISOproterenol reduced @-adrenergic by half and, in addition, attenuated the glucocorticoid-induced increase in receptor expression. Indirect immunofluorescence revealed a similar set of observations (Fig. 5 ) . Challenging cells with @-agonist reduced the epifluorescence, suggesting a loss in receptor expression (compare Fig. 5, A and C ) . Glucocorticoids, in contrast, were shown to increase epifluorescence (Fig. 5E). These data agree well with those observed by radioligand binding. Isoproterenol blunted the glucocorticoid-induced increase in receptor expression when cells were challenged simultaneously with both steroid and 0-agonist (Fig. 5G).

DISCUSSION
Glucocorticoids "up-regulate'' the expression of P-adrenergic receptors (Hadcock and Malbon, 1988b;Collins et al., 1988), whereas agonists "down-regulate" the expression of these G-protein-linked receptors (Hadcock and Malbon, 1988a). The goal of the present work was to explore the counterregulation of agonist and glucocorticoids at the levels of both receptor protein and mRNA. Chronic treatment of asthmatic conditions with @-agonists promotes an adaptive response that compromises the beneficial aspects of this therapy over an extended period (Nelson, 1986). Glucocorticoids have been shown to be effective in treating this agonistinduced adaptation (Davies and Lefkowitz, 1984). These observations prompted us to explore the underlying molecular basis for the counterregulatory effects of glucocorticoids and agonists on receptor expression.
Radioligand binding and in situ indirect immunofluorescence provided a compelling picture of the interplay between glucocorticoids and @-agonist stimulation at the level of receptor expression. As observed previously (Hadcock and Malbon, 1988b), glucocorticoids increased @-adrenergic receptor levels in DDTl MF-2 cells. In uiuo, glucocorticoids promote the expression of @-adrenergic receptors in lung and other tissues (Davies and Lefkowitz, 1984;Sharma et al., 1989). Chronic stimulation by agonist promoted a loss of radioligand binding as well as a reduction in the epifluorescence signal of cells stained with antireceptor antibodies. Interestingly, the punctiform patterns of staining were not grossly altered by either up-or down-regulation. When cells were exposed to both agents simultaneously, a clear reduction in epifluorescence signal and radioligand binding was evident when compared with the unopposed action of glucocorticoid alone. The epifluorescence and binding data do suggest, however, that glucocorticoids rescue, to a limited extent, the cells from the agonist-induced down-regulation of receptors. With respect to expression of receptor, neither glucocorticoid-induced upregulation or agonist-promoted down-regulation predominates. Agonists remain capable of down-regulating receptor expression even in the presence of glucocorticoids. Glucocorticoids are capable of up-regulating receptor levels even in cells chronically stimulated by agonist.
Analysis at the level of receptor mRNA provided both a basis for the counterregulatory effects of agonist and glucocorticoids as well as a molecular explanation for the reduction in mRNA promoted by chronic stimulation by agonist. A sharp transient peak of receptor mRNA was induced by glucocorticoid alone and by glucocorticoid in combination with @-agonist. Thus, the early phase of the glucocorticoid response appears to be largely unaffected by chronic stimulation by agonist. The second phase, establishing a new steady-state level of receptor mRNA in response to glucocor- ticoid, in contrast, was sensitive to chronic stimulation with agonist. Isoproterenol stimulation reduced the steady-state level of receptor mRNA in response to steroid back to nearly control levels. Regulation of the expression of receptor mRNA and protein remains sensitive to both of these two opposing forces.
Analysis of the relative rate of transcription by nuclear runon assays revealed two important features about the counterregulation by agonist and glucocorticoids. First, agonist treatment does not appear to alter the rate of transcription of the @-adrenergic receptor gene. The presence of CAMP-response elements in the gene and the ability of @-agonists to increase intracellular CAMP prompted the suggestion that these elements may be involved in the negative control of transcription (Hadcock and Malbon, 1988a). This would have been a novel mode of regulation as CAMP-response elements have been shown to enhance rather than suppress transcription rates of target genes (Roesler et al., 1988). Our data clearly demonstrate that agonist stimulation does not suppress transcription of the receptor gene. Second, isoproterenol treatment did not alter the ability of glucocorticoids to enhance transcription of the receptor gene. These observations prompted us to evaluate the stability of receptor mRNA under the influence of agonist and steroid.
The half-life of the @-adrenergic receptor mRNA was ex-amined in control cells and cells chronically stimulated with steroid. As reported previously (Hadcock and Malbon, 1988b), the half-life for receptor in control cells was -12 h. Glucocorticoids did not alter the half-life of receptor message. Several consensus sequences for glucocorticoid-response elements exist in both the hamster and human @-adrenergic receptor genes (Kobilka et al., 1987;Emorine et al., 1987). These potential glucocorticoid-response elements have been identified in the 5'-noncoding, coding, and 3'-noncoding regions of the genes. Evidence has been presented that the glucocorticoid-response element(s) in the 5"noncoding domain of the @-adrenergic receptor gene are obligate for glucocorticoid responsiveness . In sum, these studies suggest that it is primarily the enhanced rate of transcription at a glucocorticoid-response element in the 5'noncoding portion of the gene which is responsible for the steroid-induced up-regulation of receptor mRNA. Message stability appears to play no major role in the glucocorticoid effect. Finally, the present work reveals a molecular explanation for agonist-promoted down-regulation of receptor mRNA, message destabilization. The half-life of receptor mRNA declined from -12 h in control cells to -5 h under the influence of stimulation with @-agonist. The half-life of receptor mRNA in glucocorticoid-stimulated cells was similarly reduced to -5 h following stimulation by @-agonist. Study of the time course of receptor mRNA levels in glucocorticoid-treated cells demonstrated that in the second phase of steroid induction, isoproterenol was able to reduce mRNA levels rapidly without altering the rate of transcription. Thus, by promoting destabilization of message, agonist can maintain counterregulation of the glucocorticoid response. These studies provide a more detailed picture of the dynamic regulation of the expression of a G-protein-linked receptor in response to counterregulation by agonist and steroid. Via the interplay of enhanced transcription (steroid effect) and altered mRNA stability (agonist effect) the steady-state level of receptor message and protein can be modulated dynamically by two opposing forces.