Nội dung

Eur. J. Biochem. 271, 4042–4052 (2004)
FEBS 2004 doi:10.1111/j.1432-1033.2004.04342.x Differential effects of RU486 reveal distinct mechanisms for glucocorticoid repression of prostaglandin E2 release Joanna E. Chivers1, Lisa M. Cambridge1, Matthew C. Catley1, Judith C. Mak1, Louise E. Donnelly1, Peter J. Barnes1 and Robert Newton2 1 Department of Thoracic Medicine, National Heart and Lung Institute, Imperial College London, Faculty of Medicine, London, UK; Department of Biological Sciences, University of Warwick, Coventry, UK 2 In A549 pulmonary cells, the dexamethasone- and budesonide-dependent repression of interleukin-1b-induced prostaglandin E2 release was mimicked by the steroid antagonist, RU486. Conversely, whereas dexamethasone and budesonide were highly effective inhibitors of interleukin-1binduced cyclooxygenase (COX)/prostaglandin E synthase (PGES) activity and COX-2 expression, RU486 (< 1 lM) was a poor inhibitor, but was able to efficiently antagonize the effects of dexamethasone and budesonide. In addition, both dexamethasone and RU486 repressed [3H]arachidonate release, which is consistent with an effect at the level of phospholipase A2 activity. By contrast, glucocorticoid response element-dependent transcription was unaffected by RU486 but induced by dexamethasone and budesonide, whilst dexamethasone- and budesonide-dependent repression of nuclear factor-jB-dependent transcription was maximally 30–40% and RU486 (< 1 lM) was without significant effect. Thus, two pharmacologically distinct mechanisms of glucocorticoid-dependent repression of prostaglandin E2 release are revealed. First, glucocorticoiddependent repression of arachidonic acid is mimicked by RU486 and, second, repression of COX/PGES is antagonized by RU486. Finally, whilst all compounds induced glucocorticoid receptor translocation, no role for glucocorticoid response element-dependent transcription is supported in these inhibitory processes and only a limited role for glucocorticoid-dependent inhibition of nuclear factor-jB in the repression of COX-2 is indicated. Synthetic glucocorticoids are potent repressors of inflammation and are a first-line therapy for inflammatory diseases [1]. However, their clinical usage is limited by immunosuppression as well as by metabolic effects, including increased gluconeogenesis, increased blood glucose, amino and fatty acid mobilization, and loss of bone [2]. In addition, endogenous glucocorticoids participate in feedback inhibition of the hypothalamo-pituitary-adrenal axis, and longterm high-dose synthetic glucocorticoid usage may cause hypothalamo-pituitary-adrenal insufficiency and glucocorticoid dependency. Glucocorticoids are believed to act primarily via the glucocorticoid receptor (GR), which is maintained as an inactive cytoplasmic complex with heat shock proteins (hsp) and immunophilins [3]. Following ligand binding and complex dissociation, the GR translocates to the nucleus where it binds glucocorticoid response elements (GREs), as a dimer, to promote the transcription of responsive genes [2]. However, the GR may also act as a monomer to inhibit key inflammatory transcription factors, such as nuclear factor-jB (NF-jB) and activator protein-1, by direct interaction, competition for cofactors or by modifying the chromatin structure to prevent the expression of inflammatory genes [1,2]. Inflammatory prostaglandins, produced by the arachidonic acid cascade, play a pathophysiological role in edema, bronchoconstriction, fever and hyperalgesia [4]. Arachidonic acid, released from cell membranes by phospholipase A2 (PLA2), is converted to prostaglandin H2 (PGH2) by cyclooxygenase enzymes (COX), and further modified by specific isomerases and reductases to produce biologically relevant prostaglandins, including prostaglandin E2 (PGE2), which is the major prostaglandin product of both airway epithelial and A549 cells [5]. In inflammation, the inducible COX, COX-2, is normally up-regulated and accounts for the elevated levels of prostaglandins [4]. Conversely, COX-2 expression is highly sensitive to glucocorticoid inhibition, suggesting that inhibition of COX-2 is critical in the repression of prostaglandins by glucocorticoids. As cytokine-induced COX-2 and PGE2 release are highly NF-jB-dependent in A549 cells [6], and treatment with dexamethasone profoundly represses PGE2 release and COX-2 expression [7], Correspondence to R. Newton, Department of Biological Sciences, University of Warwick, Coventry CV4 7AL, UK. Fax: +44 2476 523701; Tel.: +44 2476 574187; E-mail: robert.newton@imperial.ac.uk Abbreviations: COX, cyclooxygenase; CRE, cyclic AMP response element; DAPI, 4¢,6¢-diamidino-2-phenylinole dihydrochloric hydrate; EGF, epidermal growth factor; GR, glucocorticoid receptor; GRE, glucocorticoid response element; hsp, heat shock protein; IL, interleukin; NF-jB, nuclear factor-jB; PGE2, prostaglandin E2; PGES, prostaglandin E synthase; PLA2, phospholipase A2; SFM, serum-free media. (Received 13 January 2004, revised 16 August 2004, accepted 23 August 2004) Keywords: corticosteroid; cyclooxygenase; epithelial cell; glucocorticoid receptor; prostaglandin E2. 
FEBS 2004 Glucocorticoid repression: differential mechanisms (Eur. J. Biochem. 271) 4043 we have used this system to further explore the mechanisms of glucocorticoid action. Materials and methods Cell culture A549 cells were cultured to confluence, as described previously [7]. Following overnight incubation in serumfree media (SFM), drugs (dexamethasone, budesonide, ionomycin, RU486) were added 1 h before stimulation with interleukin-1b (IL-1b) (R & D Systems, Oxon, UK). Dexamethasone and budesonide (both Sigma, Poole, UK) were dissolved in Hank’s balanced salt solution (Sigma). Ionomycin and RU486 (both Sigma) were dissolved in ethanol. Final concentrations of ethanol were less than 0.1% (v/v). PGE2 release, COX/prostaglandin E synthase (PGES) activity and COX-2 expression PGE2 released into the medium was measured using a commercially available PGE2 antibody (Sigma) [5,8]. For the assay of combined COX/PGES activity, cells were rinsed with SFM prior to incubation at 37 C for 10 min in SFM supplemented with 30 lM arachidonic acid, and released PGE2 was taken as a index of COX/PGES activity [5,8]. Northern and Western blot analyses were performed as described previously [7]. Reporter cell lines and luciferase assay A549 cells containing the NF-jB-dependent reporter, 6jBtkluc, have been described previously [9]. The 1·GREdependent and 2·GRE-dependent reporters, pGL3.neo.TATA.GRE and pGL3.neo.TATA.2GRE, respectively, were based on the parent vector pGL3.neo.TATA, which contains a modified minimal b-globin promoter, as previously described [10]. This was digested at the SmaI site, upstream of the minimal promoter, and doublestranded oligonucleotides (sense strand: 5¢-GCTGTACAG GATGTTCTAG-3¢ and 5¢-GCTGTACAGGATGTTC TAGGCTGTACAGGATGTTCTAG-3¢), containing one or two copies of a consensus GRE site (underlined) [11], were inserted to produce pGL3.neo.TATA.GRE and pGL3.neo.TATA.2GRE, respectively. A 2·GRE(mut) reporter was generated as described above, but using a mutated 2·GRE oligonucleotide (sense strand 5¢-GCTcaACAGGATcaTCTAGGCTcaACAGGATcaT CTAG-3¢) (mutated bases in lower case). The cyclic AMP response element (CRE)-dependent reporter, which contains six CRE sites, was as previously described [12]. A549 cells, stably harboring the luciferase reporters, were generated as previously described [9]. Prior to experiments, confluent plates of reporter cells were incubated overnight in serum-free, G-418-free, media. Cells were subsequently harvested in 1 · reporter lysis buffer (200 ll) (Promega) 6 h after treatment for luciferase activity assay (Promega). As each well is confluent and all the cells contain the reporter construct, we find reporter activity to be highly reproducible, and normalization to a second reporter is unnecessary [9]. [3H]Arachidonic acid release As previously described [8], cells were incubated overnight in 0.5 mL of SFM supplemented with 0.125 lCi [5,6,8,9,11,12,14,15-3H]arachidonic acid (Amersham Pharmacia). Cells were washed twice prior to treatment with dexamethasone or RU486. After 1 h, supernatants were changed to fresh SFM containing 2 mgÆmL)1 fatty acid-free BSA (Sigma) plus drugs prior to stimulation. Supernatants were collected and cells washed prior to harvesting in 1% (w/v) SDS. Release of [3H]arachidonic acid, or its metabolites, was expressed as a percentage of the total incorporated. Ligand binding At 80% confluence, A549 cells cultured in T175 flasks were transferred to SFM and harvested the following day in cell dissociation solution (C-5789; Sigma). Cells (1.5–4 · 106 cells per mL) were incubated overnight at 4 C with increasing concentrations of [3H]dexamethasone, in the presence of 10 lM dexamethasone, to determine nonspecific binding. Free radioligand was removed by the rapid filtration of cells through glass-fibre filters (GF/B) presoaked in NaCl/Pi (PBS), 0.1% (v/v) polyethylenimine, using a cell harvester [M-24R Brandel, SEMAT Technical (UK) Ltd, St. Albans, Hertfordshire, UK]. Filters were combined with Filtron-X scintillant (National Diagnostics, Atlanta, GA, USA) and radioactivity was measured using a beta counter (2200CA Tri-carb Liquid Scintillation Analyser; Canberra Packard, Berks., UK). Kd and Bmax. values were determined using saturation binding isotherms and Scatchard analysis, [Bound]/[Free] vs. [Bound], where the x-intercept ¼ Bmax. and the gradient ¼ ) 1/Kd (Fig. 1A) (PRISM 3; GraphPad, San Diego, CA, USA). Relative binding affinity was assessed by incubating cells with an increasing concentration of unlabelled steroid and 4 nM [3H]dexamethasone overnight at 4C. Bound and free radioligand were separated as described above. Specific binding was calculated by subtraction of nonspecific from total binding, and Cheng–Prusoff analysis was performed to determine the Ki value: Ki ¼ IC50/{1 + ([Free Count]/Kd)}, where IC50 is the concentration that results in 50% inhibition (Table 1) [13]. Immunocytochemistry Cells grown on coverslips were transferred at 70% confluence to SFM for 24 h. After incubation with steroid for the indicated times, cells were washed with NaCl/Pi (PBS) and fixed with 4% (w/v) paraformaldehyde before successive incubations in 0.5% (v/v) Nonidet P-40 and 100 mM glycine. Coverslips were blocked in NaCl/Pi (PBS) containing 0.1% (v/v) Tween-20, 0.1% (w/v) BSA and 10% (v/v) human serum prior to incubation for 1 h in 5 lgÆmL)1 rabbit anti-human GR (PA1–511A; Affinity Bioreagents Inc., Golden, CO, USA) or rabbit isotype control (Dako, Glostrup, Denmark). After washing with NaCl/Pi (PBS) containing 0.1% (v/v) Tween and incubation with biotinylated anti-rabbit immunoglobulins (Dako) for 1 h, cells were incubated with fluorescein isothiocyanate (FITC)linked streptavidin (Dako) for 1 h. Nuclei were then stained 
FEBS 2004 4044 J. E. Chivers et al. (Eur. J. Biochem. 271) C 100 100 100 50 50 50 0 NS IL-1 0 NS IL-1 IL+Dex IL+Bud -11 -10 -9 -8 -7 -6 -5 Log [Steroid] (M) -7-6 -5 -10 -9 -8 -7 -6 -5 Log Log [Dex] (M) [RU486] (M) -10 -9 -8 -7 -6 -5 Log [RU486] (M) 100 100 50 50 50 0 D PGE2 release (% IL-1 ) 0 -10 -9 -8 -7 -6 -5 Log [RU486] (M) COX/PGES activity (% IL-1 ) -11 -10 -9 -8 -7 -6 -5 Log [Steroid] (M) NS IL-1 IL+Dex IL+Bud 0 NS IL-1 100 100 50 -7-6 -5 -10 -9 -8 -7 -6 -5 Log Log [Dex] (M) [RU486] (M) 100 50 0 NS IL-1 NS IL-1 0 NS IL-1 COX/PGES activity (% IL-1 ) 0 NS IL-1 B PGE2 release (% IL-1 ) A -7-6 -5 -10 -9 -8 -7 -6 -5 Log Log [Bud] (M) [RU486] (M) -7-6 -5 -10 -9 -8 -7 -6 -5 Log Log [Bud] (M) [RU486] (M) Fig. 1. The effect of dexamethasone, budesonide and RU486 on interleukin-1b (IL-1b)-dependent prostaglandin E2 (PGE2) release and cyclooxygenase (COX)/prostaglandin E synthase (PGES) activity. (A) A549 cells were cultured with various concentrations of dexamethasone (j), budesonide (h) or RU486 (.) for 1 h prior to stimulation with IL-1b (1 ngÆmL)1) or no stimulation (NS). (B) Cells were treated with dexamethasone (Dex) (0.1 lM) (j) or budesonide (0.1 lM) (h), in the presence of increasing concentrations of RU486, for 1 h prior to stimulation with IL-1b (1 ngÆmL)1) or no stimulation (NS). (C and D) Cells were treated with various concentrations of dexamethasone (C) or budesonide (D) in the absence (j) or presence of RU486 at 0.1 lM (h), 1.0 lM (d) or 10.0 lM (s) for 1 h prior to stimulation with IL-1b (1 ngÆmL)1) or no stimulation (NS). In all cases, PGE2 release (upper panels of A, B and C and the left panel of D) and COX/PGES activity (lower panels of A, B and C and the right panel of D) were analyzed after 24 h. Data (A and B, n ¼ 5–7; C, n ¼ 4; D, n ¼ 4) are expressed as a percentage of the response to IL-1b and plotted as mean ± SEM. The following levels of significance were established, expressed as P-values of < 0.05 (*), < 0.01 (**) and < 0.001 (***). (A) (upper panel) Budesonide at 10)8 (**), 10)7 M (***) and 10)6 M (***); dexamethasone at 10)8 M (**), 10)7 M (***), 10)6 M (***) and 10)5 M (***); and RU486 at 10)6 (*) and 10)5 (***). (A) (lower panel) Budesonide and dexamethasone at 10)7 (***), 10)6 (***) and dexamethasone at 10)5 M (***). (B) (lower panel) Budesonide + RU486 at 10)6 M and 10)5 M (both ***); and dexamethasone + RU486 at 10)6 M and 10)5 M (both ***). with 1 lM 4¢,6¢-diamidino-2-phenylinole dihydrochloric hydrate (DAPI) (Sigma) and coverslips were mounted on glass slides using Citifluor mounting fluid (Citifluor Ltd, London, UK), prior to analysis using a Leica TCS 4D confocal microscope (Leica Microsystems, Milton Keynes, UK) equipped with argon, krypton, and ultraviolet lasers. Confocal images were acquired at ·40 magnification using TCS NT software (Leica Microsystems). Statistical analysis Statistical analysis was performed using analysis of variance (ANOVA) with a Dunn’s post-test, unless specifically stated otherwise in the figure legends. Significance was taken at P-values of < 0.05 (*), < 0.01 (**) and < 0.001 (***). Results Repression of PGE2 release, COX/PGES activity and COX-2 expression As reported previously [7,14], untreated A549 cells released low levels of PGE2 (1.2 ± 0.2 ngÆmL)1) and showed low levels of combined COX/PGES activity (3.1 ± 0.6 ngÆ mL)1Æmin)1), which were both increased upon stimulation with IL-1b (1 ngÆmL)1) (22.6 ± 3.7 ngÆmL)1 and 
FEBS 2004 Glucocorticoid repression: differential mechanisms (Eur. J. Biochem. 271) 4045 Table 1. Ki and functional properties of steroid ligands in A549 cells. Cheng–Prusoff analysis was performed using 50% inhibitory concentration (IC50) values and glucocorticoid receptor (GR) number generated by saturation and competition-binding studies (see Fig. 6). Data (n ¼ 3–5) are presented as mean ± SEM. See the text for a full description of output measurements. COX, cyclooxygenase; EC50, 50% effective concentration; GRE, glucocorticoid response element; NF-jB, nuclear factor-jB; PGE2, prostaglandin E2; PGES, prostaglandin E synthase. Radioligand binding EC50 (nM) for steroid effect on various functional outputs Steroid ligands Ki (nM) PGE2 release COX/PGES activity GRE (23GRE) NF-jB (6jBtk) Dexamethasone Budesonide RU486 4.9 ± 1.3 1.2 ± 0.4 0.5 ± 0.2 1.9 ± 1.4 2.6 ± 1.6 33.1 ± 2.4 3.2 ± 1.3 7.8 ± 1.9 4995 ± 23 54.5 ± 23.6 65.3 ± 29.7 – 3.2 ± 2.2 6.6 ± 3.5 1434 ± 837 32.8 ± 2.0 ngÆmL)1Æmin)1, respectively). In each case the IL-1b-induced release of PGE2 and combined COX/PGES activity were repressed in a concentration-dependent manner to near-basal levels by dexamethasone [50% effective concentration (EC50) values of 1.9 nM and 3.2 nM, respectively) and budesonide (EC50 values of 2.6 nM and 7.8 nM, respectively) (Fig. 1A, upper and lower panels). Similarly, RU486 produced a concentration-dependent repression of IL-1b-induced PGE2 release (EC50 ¼ 33.1 nM) (Fig. 1A, upper panel), yet was considerably less effective against combined COX/PGES activity, with concentrations of less than 1 lM being without significant effect (EC50 ¼ 5 lM) (Fig. 1A, lower panel). This effect was even more apparent when RU486 was used to antagonize the responses to dexamethasone and budesonide. Thus, whereas the glucocorticoid-dependent inhibition of IL-1b-induced PGE2 release was not antagonized (Fig. 1B, upper panel), the inhibition of COX/PGES activity was effectively antagonized by RU486 (Fig. 1B, lower panel). The abilities of dexamethasone and budesonide to inhibit both PGE2 release and COX/PGES activity were further tested in the presence of various concentrations A C D B Fig. 2. Effect of dexamethasone and RU486 on cyclooxygenase-2 (COX-2) expression. (A) Cells were either not stimulated (NS) or pretreated with various concentrations of dexamethasone (Dex) or RU486 for 1 h prior to stimulation with interleukin-1b (IL-1b) (1 ngÆmL)1). Cells were harvested at 6 h for RNA, and Northern blot (NB) analysis was performed for COX-2 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Cells harvested at 24 h were subject to Western blot (WB) analysis for COX-2. (B) Cells were pretreated for 1 h with dexamethasone (0.1 lM) (Dex) in the presence of various concentrations of RU486. Cells were harvested as described in (A) for Northern and Western blot analyses. In each case, blots representative of three or more such experiments are shown. (C) Following densitometric analysis, data (n ¼ 4–6) (upper panels, Western blots; lower panels, Northern blots) from the experiments in (A) were expressed as a percentage of IL-1b, treated and plotted as mean ± SEM. (j), Dexamethasone; (.), RU486. (D) Data (n ¼ 4–5) from the experiments in (B) were plotted as described in (C). 
FEBS 2004 4046 J. E. Chivers et al. (Eur. J. Biochem. 271) To investigate the possibility of an effect of steroids upstream of COX-2, cells were loaded with [3H]arachidonic acid prior to stimulation in the presence of dexamethasone or RU486. As IL-1b alone is a poor activator of arachidonic acid release [8], cells were also treated with ionomycin or with IL-1b + ionomycin, which provides a Ca2+ stimulus, causing translocation and membrane association of cytosolic (c)PLA2 to markedly enhance cPLA2 activity [8,17,18]. IL-1b, ionomycin and IL-1b + ionomycin increased [3H]arachidonic acid release by 1.6-fold, 3.2-fold and 7.2fold, respectively (Fig. 3A). In each case, dexamethasone produced repressions of 50, 61 and 68%, whilst RU486 resulted in repressions of 58, 53 and 63%, respectively. To further characterize this inhibition, cells were treated with various concentrations of either dexamethasone or RU486 prior to stimulation with IL-1b + ionomycin. In each case, a concentration-dependent inhibition of [3H]arachidonic acid release (EC50 ¼ 18.7 ± 10.6 and 26.2 ± 11.6 nM, respectively) was observed, thereby confirming the independent inhibitory effect of RU486 acting at the level of arachidonic acid release (Fig. 3B). Transactivation and transrepression by glucocorticoids and RU486 The effect of dexamethasone and RU486 was analyzed on GRE-dependent and NF-jB-dependent transcription. From the 1·GRE reporter, pGL3.neo.TATA.GRE, GRE-dependent transcription was increased by
4.5-fold (EC50 ¼ 46.7 ± 17.7) by dexamethasone and fivefold (EC50 ¼ 53.5 ± 20.8 nM, respectively) by budesonide (Fig. 4A). Similarly the 2·GRE-driven reporter, ** ** 5 3H arachidonic acid release (% of total incorporated) 10 * * * ** Dex 0 Ru486 NS Iono IL-1 IL+Iono arachidonic acid release (% IL-1 + ionomycin) B 100 50 * ** 0 NS IL+Iono Effect of dexamethasone and RU486 on arachidonic acid release A ** *** *** ** -6 -5 3H of RU486 (Fig. 1C,D). As shown by the rightwards shift and the reduced apparent efficacy of the inhibition curves described for both dexamethasone and budesonide, the glucocorticoid-dependent repression of COX/PGES activity was clearly antagonized by increasing the concentration of RU486. However, in marked contrast, RU486 primarily resulted in an increased overall inhibition of the response curves described for dexamethasone and budesonide on PGE2 release, as shown by the progressive flattening of the respective lines (Fig. 1C,D). These data are therefore indicative of a primary inhibitory effect of RU486 on PGE2 release, but not on combined COX/PGES activity. Analysis of COX-2 mRNA and protein expression, which is responsible for the inflammatory release of PGE2 from A549 cells [5,15], often revealed basal levels of expression, as reported previously [16]. However, in each case, and as previously shown, COX-2 expression was dramatically increased by treatment with IL-1b [7,14]. Consistent with the combined COX/PGES data, the analysis of COX-2 mRNA and protein expression revealed a concentration-dependent inhibition of COX-2 expression by dexamethasone, whereas RU486 showed little effect except at high doses (Fig. 2A,B). Consistent with Fig. 1B, 0.1 lM dexamethasone almost totally repressed both mRNA and protein expression of COX-2, and this effect was efficiently antagonized by RU486 (Fig. 2B). -10 -9 -8 -7 Log [Steroid] (M) Fig. 3. Inhibition of arachidonic acid release by dexamethasone and RU486. (A) Following loading with [3H]arachidonic acid, cells were either not treated or pretreated with dexamethasone (1 lM) (Dex) or RU486 (1 lM) for 1 h. Cells were then either not stimulated (NS) or stimulated with interleukin-1b (IL-1b) (1 ngÆmL)1), ionomycin (3 lM) (Iono) or both together (IL + Iono), and the supernantants and cells were harvested after 1 h for liquid scintillation counting. Data (n ¼ 4 or 5) are shown as arachidonate release expressed as a percentage of the total incorporated ± SEM. Significance was assessed using the Student’s t-test. *P < 0.05, **P < 0.01. (B) Cells were treated as in (A) except that various concentrations of either dexamethasone (j) or RU486 (.) were added prior to the IL-1b (1 ngÆmL)1) + ionomycin (3 lM) stimulus. After harvesting, 1 h following stimulation, arachidonate release as a fraction of the total incorporated was expressed as a percentage of the IL-1b + ionomycin stimulus and plotted as mean ± SEM. Significance was assessed using analysis of variance (ANOVA) with a Dunn’s post-test. **P < 0.01, ***P < 0.001. pGL3.neo.TATA.2GRE, gave rise to over a 15-fold (EC50 ¼ 54.5) induction by dexamethasone and a 20-fold induction (EC50 ¼ 65.3 nM) by budesonide (Fig. 4B). No response was observed with reporters containing either mutated GRE elements (pGL3.neo.TATA.2GREmut) or no GRE sites (pGL3.neo.TATA) (data not shown), which confirms the specificity of these reporter systems for the presence of GRE sites. In each case, RU486 showed little or no ability to activate GRE-dependent transcription (Fig. 4A,B), but demonstrated a profound ability to antagonize both 1·GRE and 2·GRE reporter activity induced by 0.1 lM of either dexamethasone or budesonide (Fig. 4C,D). Analysis of IL-1b-induced NF-jB-dependent transcription revealed a modest 30–40% inhibition (EC50 ¼ 
FEBS 2004 Glucocorticoid repression: differential mechanisms (Eur. J. Biochem. 271) 4047 1 GRE Fold activation 5 *** *** *** 4 3 2 0 NS 1 B 20 *** 15 10 5 0 -11 -10 -9 -8 -7 -6 -5 Log [Steroid] (M) -11 -10 -9 -8 -7 -6 -5 Log [Steroid] (M) 120 100 80 60 40 0 NS Dex Bud RU486 20 100 80 60 40 20 0 -10 -9 -8 -7 -6 -5 Log [RU486] (M) 2 GRE 120 NS Dex Bud RU486 1 GRE Luciferase activity (% Dex) D Luciferase activity (% Dex) C *** *** 2 GRE NS 6 Fold activation A -10 -9 -8 -7 -6 -5 Log [RU486] (M) Fig. 4. Effect of dexamethasone, budesonide and RU486 on glucocorticoid response element (GRE)-dependent transcription. (A) 1·GRE or (B) 2·GRE A549 reporter cells were either not stimulated (NS) or treated with various concentrations of dexamethasone (j), budesonide (h) or RU486 (.). After 6 h, cells were harvested for luciferase assay. Data (n ¼ 6–10), expressed as fold induction, are plotted as means ± SEM. (C) 1·GRE and (D) 2·GRE A549 reporter cells were activated by dexamethasone (0.1 lM) (j) or budesonide (0.1 lM) (h) in the presence of various concentrations of RU486. Cells were harvested as described above, and luciferase activity, expressed as a percentage of the activity induced by dexamethasone (0.1 lM), was plotted as mean ± SEM. The effect of no stimulation (NS), or of stimulation with dexamethasone (0.1 lM) (Dex), budesonide (0.1 lM) (Bud) or RU486 (10 lM) alone, is also shown. All data are n ¼ 6–10. In (A) and (B), the indicated levels of significance apply to both budesonide and dexamethasone. In addition, the following levels of significance were established, expressed as P-values of < 0.05 (*), < 0.01 (**) and < 0.001 (***). (B) Budesonide at 10)8 M (**) and dexamethasone at 10)8 M (*). (C) Budesonide + RU486 at 10)7 M (**), 10)6 M (**) and 10)5 M (***); dexamethasone + RU486 at 10)7 M (**), 10)6 M and 10)5 M (***). (D) Budesonide + RU486 at 10)6 M (**), and 10)5 M (**); dexamethasone + RU486 at 10)7 M (*), 10)6 M (**) and 10)5 M (**). 3.2 ± 1.3 and 7.8 ± 1.9 nM) by dexamethasone and budesonide, respectively, and just over a 50% inhibition by 10 lM RU486 (Fig. 5A). RU486 was without effect at 0.1 lM and required to be present at concentrations of
100-fold higher than either dexamethasone or budesonide to achieve similar levels (30–40%) of inhibition. It is worth noting that the inhibition of NF-jB by RU486 correlates very closely with the effects observed on COX activity and COX-2 expression (Figs 1 and 2). In addition, the ability of RU486 to antagonize the repressive effects of 0.1 lM dexamethasone or budesonide was examined. In each case, a concentration-dependent antagonism was observed up to a maximum of 0.1 lM RU486 (Fig. 5B). Above this concentration, increasing levels of inhibition were observed owing to the repressive effect of RU486 acting alone (data not shown and see Fig. 5A). The expression of COX-2 may also depend on activating transcription factors (ATFs) and activator protein-1 (AP-1)like factors acting at a CRE site located in the proximal region of the COX-2 promoter [19–21]. Consistent with this, we have previously found that a CRE-driven reporter construct was unresponsive to cAMP in A549 cells, but responded to IL-1b [10]. This was not believed to reflect a general problem with this reporter, as strong cAMP-inducibility has been demonstrated in other experimental systems [12]. Consistent with these earlier findings, IL-1b was shown to induce reporter activity twofold (Fig. 5C). In each case, both dexamethasone (0.1 lM) and RU486 (10 lM) were found to produce marked repressive effects (Fig. 5C). Binding affinity of steroid ligands and effect on GR translocation Saturation binding studies using [3H]dexamethasone demonstrated one-site binding in A549 cells and revealed 16 500 ± 2700 GR/cell with an affinity of 1.36 ± 0.10 nM, which is consistent with other reports, including primary epithelial cells, indicating an affinity in the low nM range (Fig. 6A) [22–24]. Competitive binding studies were performed to examine the relative GR-binding affinity of these steroid ligands, and the following rank order of affinity was observed: RU486 > budesonide > dexamethasone (Fig. 6B). The appropriate Ki values are given in Table 1. 
FEBS 2004 4048 J. E. Chivers et al. (Eur. J. Biochem. 271) 0 NS IL-1 50 -10 -9 -8 -7 -6 NF- B 90 80 70 60 -5 Log [Steroid] (M) C 3 Fold activation 100 100 CRE * ** 2 1 0 IL-1 Dex Bud Luciferase activity (% of IL-1 ) NF- B Luciferase activity (% IL-1 ) B A -10 -9 -8 -7 Log [RU486] (M) Dex RU486 NS IL-1 Fig. 5. Transrepression by dexamethasone, budesonide and RU486. (A) 6jBtk reporter cells were either not treated or were treated with various concentrations of dexamethasone (j), budesonide (h) or RU486 (.) for 1 h, prior to stimulation with IL-1b (1 ngÆmL)1) or no stimulation (NS). After 6 h, cells were harvested for analysis in the luciferase assay. Data (n ¼ 8), expressed as percentage of the response to IL-1b stimulation, are plotted as mean ± SEM. Significance was established, expressed as P-values of < 0.05 (*), < 0.01 (**) and < 0.001 (***), for: budesonide at 10)8 M (*), 10)7 M (**) and 10)6 M (**); dexamethasone at 10)7 M (**), 10)6 M (***) and 10)5 M (***); and RU486 at 10)5 M (**). (B) 6jBtk reporter cells, were treated with dexamethasone (0.1 lM) (j) or budesonide (0.1 lM) (h) in the presence of various concentrations of RU486. Luciferase assay data (n ¼ 7–9), expressed as a percentage of the response to IL-1b, are plotted as mean ± SEM. The effect of IL-1b + dexamethasone (0.1 lM) (Dex) and IL-1b + budesonide (0.1 lM) (Bud) alone are shown. (C) CRE reporter cells were either not treated or were treated for 1 h with dexamethasone (0.1 lM) or RU486 (10 lM) prior to no stimulation (NS) or stimulation with IL-1b (1 ngÆmL)1), as indicated. Cells were harvested after 6 h for analysis in the luciferase assay, as described above. Data (n ¼ 6), expressed as fold activation, are plotted as mean ± SEM. *P < 0.05, **P < 0.01. Nuclear translocation of GR by dexamethasone and RU486 Dexamethasone induced a rapid (within 15 min) translocation of GR from the cytoplasm to the nuclear compartment, with complete translocation observed by 1 h (Fig. 7A). Similarly, and as expected, nuclear translocation of GR was also induced by budesonide (Fig. 7B). In addition, RU486 was also efficient at inducing GR nuclear translocation, indicating that binding of the antagonist can result in dissociation of the cytoplasmic hsp–GR complex (Fig. 7B). Analysis of an isotype-control antibody revealed no significant immunoreactivity, suggesting that the observed signal was GR-specific (Fig. 7C). Discussion In the above studies, dexamethasone and budesonide produced a near-total inhibition of both PGE2 and COX/ PGES activity, and acted with similar efficacies (Table 1) and potencies. However, whilst the steroid receptor antagonist, RU486, showed reversal of both COX-2 expression and COX/PGES activity, which is consistent with a GR-dependent mechanism, RU486 was incapable of antagonizing the repression of IL-1b-induced PGE2 release produced by either dexamethasone or budesonide. In fact, RU486 resulted in the progressive repression of PGE2 release at increasing concentrations. Analysis of RU486 alone on IL-1b-induced PGE2 release revealed a concentration-dependent inhibition of PGE2 release, yet showed little or no effect on COX/PGES activity or COX-2 expression until RU486 concentrations of 1 lM were reached. This clear discrepancy strongly suggests that RU486 may exert an inhibitory effect upstream of COX-2, possibly at the level of PLA2 and arachidonic acid release. This proposal was confirmed by the analysis of [3H]arachidonate release, which revealed concentration- dependent inhibition by both dexamethasone and RU486. Interestingly, the EC50 values for repression of PGE2 release, and the repression of arachidonic acid release by RU486 (33.1 and 26.2 nM, respectively), correlate closely and therefore support the suggestion of a mechanistically distinct action for RU486 at the level of arachidonic acid release. We therefore conclude that these data document the existence of at least two functionally distinct processes for the inhibition of inflammatory PGE2 release by steroids. In the first mechanism, glucocorticoids, such as dexamethasone or budesonide, inhibit the expression of COX-2, and this response is antagonized efficiently by RU486. This contrasts with a second, and pharmacologically distinct mechanism, which occurs at the level of arachidonic acid release, in which the actions of glucocorticoids are mimicked by RU486. Previous reports have also documented the inhibition of arachidonic acid release in A549 cells by dexamethasone [25]. However, these authors did not report any inhibition by RU486 (10 nM) alone [26], and showed a 50% antagonism of the dexamethasone-dependent repression when using RU486 at 10 lM [25]. In an attempt to reconcile the apparent differences between the results of these reports and those of the present study, it is noticeable that different mechanisms of stimulation were used in each of the studies, and this alone could account for any differences. Furthermore, inspection of our current data on the repression of both PGE2 release and arachidonic acid release, suggests that the effects of 10 nM RU486 could be at the margins of experimentally discernable repression (see Figs 1A and 3B). We also note that Croxtall et al. did not seemingly test higher concentrations of RU486 acting alone for an inhibitory effect on epidermal growth factor (EGF)-stimulated arachidonic acid release [25]. This therefore leaves open the possibility that the incomplete antagonism of RU486 observed on dexamethasone-dependent repression of EGF-stimulated arachidonic acid release is, in fact, 
FEBS 2004 Glucocorticoid repression: differential mechanisms (Eur. J. Biochem. 271) 4049 A 0 0 100 2.5 75 1 0 50 5.0 25 Bound/Free x 103 4 3 2 7.5 0 Specific Binding x 103 (dpm) 10 Specific Binding x102 (dpm) 10 20 30 Log [Dex] (nM) Specific binding (%) B 100 75 50 25 0 -11 -10 -9 -8 -7 -6 -5 Log [Steroid] (M) Fig. 6. Analysis of glucocorticoid receptor (GR) number and relative affinity of ligands. (A) A typical saturation–binding isotherm, showing specific GR binding using 2.4 · 106 cells and resulting Scatchard analysis (inset), where the ratio of free to bound radioligand is plotted against log [steroid] to give a straight line with a gradient equal to )1/Kd and and an x intercept that equals Bmax. (B) Competition binding curves showing relative affinity in A549 cells, where dexamethasone (j), budesonide (h), RU24858 (d) or RU486 (.) compete with 4 nM [3H]dexamethasone to bind the GR. Data are presented as mean ± SEM for n ¼ 3–5 observations. attributable to a partial agonistic effect of RU486 acting alone [25]. It is well established that glucocorticoids can repress the transcription of inflammatory genes via transcription factors such as NF-jB [1,2]. However, whilst some degree (30–40% inhibition) of glucocorticoid-dependent inhibition of NF-jB-dependent transcription was observed in response to both dexamethasone and budesonide, this effect is clearly insufficient to account for the near-complete repression of COX-2 expression or PGE2 release observed with each of these compounds. As PGE2 release and COX-2 expression in A549 cells is highly NF-jB-dependent, and this level of inhibition of NF-jB-dependent transcription correlates very well with our previous observation that the IL-1b-induced COX-2 transcription rate was inhibited by
40% by dexamethasone, we are compelled to suggest that additional mechanisms of glucocorticoid-dependent repression of COX-2 must also exist [6,7]. Similarly, whilst GREdependent transcription was robustly increased following dexamethasone and budesonide treatment, this mechanism is unlikely to account for the repression of COX-2 or COX/PGES activity, as the EC50 for this effect is greater than 10-fold more than that required for the inhibition of PGE2 release or COX/PGES activity (Table 1). Interestingly, this shift in the concentration–response curve for transactivation effects at GREs (EC50 values of 54.5 and 65.3 nM for dexamethasone and budesonide, respectively) when compared with transrepression, for example of NF-jB (EC50 values of 3.2 and 7.8 nM for dexamethasone and budesonide, respectively), has been previously reported, although the exact mechanistic explanation is currently lacking [27]. Therefore, in respect of COX-2, these data suggest that other, non-NF-jBmediated and probably non-GRE-mediated, mechanisms of dexamethasone-dependent inhibition must be in operation to account for the full repression of COX-2 and COX/PGES activities in these cells. By contrast, the inhibition of NF-jB-dependent transcription by high concentrations of RU486 correlated very closely, in terms of both apparent efficacy and potency, with the inhibition of COX/PGES activity, thereby providing further strength to the argument that additional mechanisms, other than the inhibition of NF-jB, account for the inhibition by dexamethasone. However, the basis of this inhibition by RU486 is currently unclear to us because these levels of steroid are vastly in excess of that necessary to saturate GR, as suggested by our own, and previously reported [24,28], ligand-binding studies (Fig. 6). It is possible that at these high concentrations RU486 is acting in a GR-independent manner. Notwithstanding the inhibition at high doses, it is clear that at concentrations of 1 or 0.1 lM, RU486 shows a limited or no effect on NF-jB-dependent transcription, yet is effective at inhibiting both PGE2 and arachidonic acid release, suggesting that the inhibition of NF-jB plays no role in this response. Previous studies have suggested that, relative to dexamethasone, RU486 is a poor inducer of glucocorticoid-dependent transcription [29–35]. Similarly, in the present study, RU486-induced GRE-dependent transcription from either a 1·GRE or a 2·GRE reporter was virtually absent, and this is consistent with data from primary human bronchial epithelial cells [24]. These data therefore raise the possibility that RU486 inhibits arachidonic acid release via a mechanism that is independent of transcription. Indeed, the rapid dexamethasone-dependent repression of EGF-induced release of arachidonic acid was previously shown to be actinomycin D insensitive and therefore independent of transcription [25]. In this respect, RU486 has previously been shown to mimic other nongenomic glucocorticoid responses, including the down-regulation of GR itself [36,37]. Certainly, our data indicate that RU486, can, like dexamethasone and budesonide, bind to and induce the nuclear translocation of GR. We therefore speculate that binding of ligand, including antagonists such as RU486, to GR, and complex dissociation, may be sufficient for the inhibition of arachidonic acid release and that this represents a mechanistically distinct event from the inhibition of inflammatory gene expression. In this context it is notable that various nongenomic actions of steroid hormones have been identified [38,39], which raises the possibility of ligand-dependent nongenomic anti-inflammatory functions for GR or for GR-associated 
FEBS 2004 4050 J. E. Chivers et al. (Eur. J. Biochem. 271) A B C Fig. 7. Nuclear translocation of the glucocorticoid receptor (GR). (A) Cells were either not treated or were treated with dexamethasone (1 lM) (Dex) for 15, 30 or 60 min and then probed with a fluorescein isothiocyanate (FITC)-conjugated GR immunoglobulin (green) prior to imaging by confocal microscopy. Nuclei are indicated by 4¢,6¢-diamidino-2-phenylinole dihydrochloric hydrate (DAPI) staining of chromatin (blue). (B) Cells were treated with dexamethasone (1 lM) (Dex), budesonide (1 lM) or RU486 (1 lM) for 1 h prior to probing and analysis as described above. (C) Cells were treated as in (A) and then probed with an isotype control for the GR antibody used in (A) and (B). All images are representative of three experiments. proteins present in the GR–hsp complex. Finally, we should point out that a number of effects of glucocorticoids, which are independent of the classical GR, are also reported to occur and these could help to explain our results [39]. Thus, the mineralocorticoid receptor may mediate glucocorticoid responsiveness in the brains of GR knockout mice [40]. In addition, a pharmacologically distinct pool of membrane-localized glucocorticoid receptors have been identified by various authors [39]. For example, a membrane glucocorticoid receptor has been biochemically identified in amphibians [41]. However, it is currently unclear whether this represents a version of the classical GR [42] or P-glycoprotein/multiple drug resistance gene, a member of the ATP-binding cassette (ABC) transporters [43,44], or some other receptor [45]. In this context, P-glycoprotein is of interest as it actively exports certain steroids, and blocking its function has been shown to promote glucocorticoid actions [46,47]. In conclusion, we present data which further confirm that the inhibition of NF-jB-dependent transcription cannot account for all the repressive effects of glucocorticoids on inflammatory genes such as COX-2. Furthermore, we present evidence that glucocorticoids and RU486 also inhibit the release of arachidonic acid via a process that does not involve either inhibition of NF-jB or the activation of GRE-mediated transcription and which is mechanistically distinct from the inhibition of COX-2. Taken together, these data indicate the existence of pharmacologically distinct processes that are collectively responsible for the repression of inflammatory PGE2 release by glucocorticoids. Acknowledgements J.E.C. and M.C.C. were collaborative students with the BBSRC and the MRC, respectively, and both were supported by Aventis Pharmaceuticals. References 1. Barnes, P.J. (1999) Therapeutic strategies for allergic diseases. Nature 402 (Suppl.), B31–B38. 2. Newton, R. (2000) Molecular mechanisms of glucocorticoid action: what is important? Thorax 55, 603–613. 3. Schaaf, M.J. & Cidlowski, J.A. (2002) Molecular mechanisms of glucocorticoid action and resistance. J. Steroid Biochem. Mol. Biol. 83, 37–48. 4. Funk, C.D. (2001) Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 294, 1871–1875. 5. Mitchell, J.A., Belvisi, M.G., Akarasereenont, P., Robbins, R.A., Kwon, O.J., Croxtall, J., Barnes, P.J. & Vane, J.R. (1994) Induction of cyclo-oxygenase-2 by cytokines in human pulmonary epithelial cells: regulation by dexamethasone. Br. J. Pharmacol. 113, 1008–1014. 6. Catley, M.C., Chivers, J.E., Cambridge, L.M., Holden, N., Slater, D.M., Staples, K.J., Bergmann, M.W., Loser, P., Barnes, P.J. & Newton, R. (2003) IL-1beta-dependent activation of NF-kappaB mediates PGE2 release via the expression of 
FEBS 2004 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. Glucocorticoid repression: differential mechanisms (Eur. J. Biochem. 271) 4051 cyclooxygenase-2 and microsomal prostaglandin E synthase. FEBS Lett. 547, 75–79. Newton, R., Seybold, J., Kuitert, L.M.E., Bergmann, M. & Barnes, P.J. (1998) Repression of cyclooxygenase-2 and prostaglandin E2 release by dexamethasone occurs by transcriptional and post-transcriptional mechanisms involving loss of polyadenylated mRNA. J. Biol. Chem. 273, 32312–32321. Newton, R., Cambridge, L., Hart, L.A., Stevens, D.A., Lindsay, M.A. & Barnes, P.J. (2000) The MAP kinase inhibitors, PD098059, UO126 and SB203580, inhibit IL-1beta-dependent PGE(2) release via mechanistically distinct processes. Br. J. Pharmacol. 130, 1353–1361. Newton, R., Hart, L.A., Stevens, D.A., Bergmann, M., Donnelly, L.E., Adcock, I.M. & Barnes, P.J. (1998) Effect of dexamethasone on interleukin-1beta (IL-1beta) induced nuclear factor-kappaB (NF-kappaB) and kappaB-dependent transcription in epithelial cells. Eur. J. Biochem. 254, 81–89. Catley, M.C., Cambridge, L.M., Nasuhara, Y., Ito, K., Chivers, J.E., Beaton, A., Holden, N.S., Bergmann, M.W., Barnes, P.J. & Newton, R. (2004) Inhibitors of protein kinase C (PKC) prevent activated transcription: role of events downstream of NF-kappaB DNA binding. J. Biol. Chem. 279, 18457–18466. Strahle, U., Schmid, W. & Schutz, G. (1988) Synergistic action of the glucocorticoid receptor with transcription factors. EMBO J. 7, 3389–3395. Meja, K.K., Catley, M.C., Cambridge, L.M., Barnes, P.J., Lum, H., Newton, R. & Giembycz, M.A. (2004) Adenovirus-mediated delivery and expression of a cAMP-dependent protein kinase inhibitor gene to BEAS-2B epithelial cells abolishes the anti-inflammatory effects of rolipram, salbutamol, and prostaglandin E2: a comparison with H-89. J. Pharmacol. Exp. Ther. 309, 833–844. Cheng, Y. & Prusoff, W.H. (1973) Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol. 22, 3099–3108. Newton, R., Kuitert, L.M., Bergmann, M., Adcock, I.M. & Barnes, P.J. (1997) Evidence for involvement of NF-jB in the transcriptional control of COX-2 gene expression by IL-1b. Biochem. Biophys. Res. Commun. 237, 28–32. Newton, R., Eddleston, J., Haddad, E., Hawisa, S., Mak, J., Lim, S., Fox, A.J., Donnelly, L.E. & Chung, K.F. (2002) Regulation of kinin receptors in airway epithelial cells by inflammatory cytokines and dexamethasone. Eur. J. Pharmacol. 441, 193–202. Watkins, D.N., Peroni, D.J., Lenzo, J.C., Knight, D.A., Garlepp, M.J. & Thompson, P.J. (1999) Expression and localization of COX-2 in human airways and cultured airway epithelial cells. Eur. Respir. J. 13, 999–1007. Schievella, A.R., Regier, M.K., Smith, W.L. & Lin, L.L. (1995) Calcium-mediated translocation of cytosolic phospholipase A2 to the nuclear envelope and endoplasmic reticulum. J. Biol. Chem. 270, 30749–30754 Leslie, C.C. (1997) Properties and regulation of cytosolic phospholipase A2. J. Biol. Chem. 272, 16709–16712 Xie, W., Fletcher, B.S., Andersen, R.D. & Herschman, H.R. (1994) v-src induction of the TIS10/PGS2 prostaglandin synthase gene is mediated by an ATF/CRE transcription response element. Mol. Cell Biol. 14, 6531–6539. Inoue, H., Yokoyama, C., Hara, S., Tone, Y. & Tanabe, T. (1995) Transcriptional regulation of human prostaglandin-endoperoxide synthase-2 gene by lipopolysaccharide and phorbol ester in vascular endothelial cells. Involvement of both nuclear factor for interleukin-6 expression site and cAMP response element. J. Biol. Chem. 270, 24965–24971. Xie, W. & Herschman, H.R. (1996) Transcriptional regulation of prostaglandin synthase 2 gene expression by platelet-derived growth factor and serum. J. Biol. Chem. 271, 31742–31748. 22. Hammer, S., Spika, I., Sippl, W., Jessen, G., Kleuser, B., Holtje, H.D. & Schafer-Korting, M. (2003) Glucocorticoid receptor interactions with glucocorticoids: evaluation by molecular modeling and functional analysis of glucocorticoid receptor mutants. Steroids 68, 329–339. 23. Robin-Jagerschmidt, C., Wurtz, J.M., Guillot, B., Gofflo, D., Benhamou, B., Vergezac, A., Ossart, C., Moras, D. & Philibert, D. (2000) Residues in the ligand binding domain that confer progestin or glucocorticoid specificity and modulate the receptor transactivation capacity. Mol. Endocrinol. 14, 1028–1037. 24. LeVan, T.D., Babin, E.A., Yamamura, H.I. & Bloom, J.W. (1999) Pharmacological characterization of glucocorticoid receptors in primary human bronchial epithelial cells. Biochem. Pharmacol. 57, 1003–1009. 25. Croxtall, J.D., Choudhury, Q. & Flower, R.J. (2000) Glucocorticoids act within minutes to inhibit recruitment of signalling factors to activated EGF receptors through a receptor-dependent, transcription-independent mechanism. Br. J. Pharmacol. 130, 289–298. 26. Croxtall, J.D., Paul-Clark, M. & van Hal, P.T. (2003) Differential modulation of glucocorticoid action by FK506 in A549 cells. Biochem. J. 376, 285–290. 27. Jonat, C., Rahmsdorf, H.J., Park, K.K., Cato, A.C., Gebel, S., Ponta, H. & Herrlich, P. (1990) Antitumor promotion and antiinflammation: down-modulation of AP-1 (Fos/Jun) activity by glucocorticoid hormone. Cell 62, 1189–1204. 28. Gagne, D., Pons, M. & Philibert, D. (1985) RU 38486: a potent antiglucocorticoid in vitro and in vivo. J. Steroid Biochem. 23, 247–251. 29. Vayssiere, B.M., Dupont, S., Choquart, A., Petit, F., Garcia, T., Marchandeau, C., Gronemeyer, H. & Resche Rigon, M. (1997) Synthetic glucocorticoids that dissociate transactivation and AP-1 transrepression exhibit antiinflammatory activity in vivo. Mol. Endocrinol. 11, 1245–1255. 30. Heck, S., Bender, K., Kullman, M., Gottlicher, M., Herrlich, P. & Cato, A.C.B. (1997) IkBa-independent downregulation of NF-kB activity by glucocorticoid receptor. EMBO J. 16, 4698– 4707. 31. Wissink, S., van-Heerde, E.C., vand-der, B.B. & van-der-Saag, P.T. (1998) A dual mechanism mediates repression of NF-kappaB activity by glucocorticoids. Mol. Endocrinol. 12, 355–363. 32. Liden, J., Delaunay, F., Rafter, I., Gustafsson, J. & Okret, S. (1997) A new function for the C-terminal zinc finger of the glucocorticoid receptor. Repression of RelA transactivation. J. Biol. Chem. 272, 21467–21472. 33. Fryer, C.J., Kinyamu, H.K., Rogatsky, I., Garabedian, M.J. & Archer, T.K. (2000) Selective activation of the glucocorticoid receptor by steroid antagonists in human breast cancer and osteosarcoma cells. J. Biol. Chem. 275, 17771–17777. 34. Rogatsky, I., Hittelman, A.B., Pearce, D. & Garabedian, M.J. (1999) Distinct glucocorticoid receptor transcriptional regulatory surfaces mediate the cytotoxic and cytostatic effects of glucocorticoids. Mol. Cell Biol. 19, 5036–5049. 35. Pariante, C.M., Pearce, B.D., Pisell, T.L., Su, C. & Miller, A.H. (2001) The steroid receptor antagonists RU40555 and RU486 activate glucocorticoid receptor translocation and are not excreted by the steroid hormones transporter in L929 cells. J. Endocrinol. 169, 309–320. 36. Hoeck, W., Rusconi, S. & Groner, B. (1989) Down-regulation and phosphorylation of glucocorticoid receptors in cultured cells. Investigations with a monospecific antiserum against a bacterially expressed receptor fragment. J. Biol. Chem. 264, 14396– 14402. 37. Burnstein, K.L., Jewell, C.M., Sar, M. & Cidlowski, J.A. (1994) Intragenic sequences of the human glucocorticoid receptor complementary DNA mediate hormone-inducible receptor messenger