Glucocorticoid-mediated gene suppression of rat cytokine-induced neutrophil chemoattractant CINC/gro, a member of the interleukin-8 family, through impairment of NF-kappa B activation.

The glucocorticoid dexamethasone inhibited the production of the rat cytokine-induced neutrophil chemoattractant CINC/gro, a counterpart of human melanoma growth-stimulating activity that belongs to the interleukin-8 (IL-8) family, in the normal rat kidney epithelial cell line NRK-52E stimulated with interleukin-1β (IL-1β), lipopolysaccharide, or tumor necrosis factor α. The accumulation of CINC/gro mRNA induced by these activators was also decreased comparably by dexamethasone. A nuclear run-on assay revealed that dexamethasone decreased the IL-1β-induced transcription of the CINC/gro gene. The half-life of CINC/gro mRNA transcripts did not change significantly after exposure to dexamethasone, suggesting that this glucocorticoid acts mainly at the transcriptional level. Transfection with luciferase expression vectors containing 5′-deleted and mutated CINC/gro gene sequences demonstrated that the 5′-flanking region containing the NF-κB binding site is involved in the IL-1β- and dexamethasone-induced activation and repression of the CINC/gro gene expression, respectively. Furthermore, a tandem repeat of the NF-κB sequence in the CINC/gro gene conferred the inducibility by IL-1β and suppression of luciferase activity by dexamethasone. In an electrophoretic mobility shift assay, dexamethasone diminished the IL-1β-induced formation of NF-κB complexes, which consisted of p65 and p50. Western blotting revealed that dexamethasone inhibited the IL-1β-induced translocation of p65 from the cytoplasm into the nucleus, while the nuclear level of NF-κB p50 remained almost unchanged. In addition, the degradation of IκB-α induced by IL-1β was not inhibited by dexamethasone. These results indicated that the suppression of the CINC/gro gene transcription by glucocorticoid occurs through the impairment of NF-κB activation, possibly by interference with the translocation of NF-κB p65 from the cytoplasm into the nucleus, thereby suppressing transactivation of the CINC/gro gene.

Neutrophil accumulation at sites of inflammation is induced by chemoattractants, including C5a (1) and leukotriene B 4 (2). In addition to these, we found a novel polypeptide neutrophil chemotactic factor, cytokine-induced neutrophil chemoattrac-tant (CINC), 1 which is produced by the normal rat kidney epithelial cell line NRK-52E (3,4). According to its amino acid sequence, CINC appears to be a member of the human cytokine family that includes interleukin-8 (IL-8) and three closely related gro gene products, GRO␣ , GRO␤, and GRO␥ (melanoma growth-stimulating activity ␣, ␤, and ␥, respectively), all of which have chemotactic activities for human neutrophils (5)(6)(7)(8).
Since rat CINC has closer sequence homology to GRO than to , CINC is the rat equivalent of human GRO but not of IL-8 and hereafter will be referred to as CINC/gro.
The IL-8 family acts as a functional chemoattractant for neutrophils in vivo (6, 9 -12). In our previous studies (13,14), CINC/gro is shown to contribute to neutrophil infiltration into inflammatory sites in lipopolysaccharide (LPS)-induced inflammation models in rats. Moreover, several lines of evidence have implicated IL-8 in several types of inflammation, such as synovitis (15), LPS-induced acute dermatitis (6), and reperfusion injury in the ischemic lung (16). These findings are suggestive of pathophysiological roles of CINC/gro and IL-8 in inflammatory reactions. Thus, the regulation of CINC/gro and/or IL-8 production is critically involved in the control of inflammatory reactions associated with neutrophil infiltration.
The production of CINC/gro is not constitutive, but can be induced by several inflammatory stimuli, such as interleukin-1␤ (IL-1␤), LPS, and tumor necrosis factor ␣ (TNF␣) (3,17,18), although little is known about the intracellular regulatory mechanisms that trigger CINC/gro up-regulation. The production of human IL-8 and GRO can also be induced by stimulation with inflammatory cytokines at the transcriptional level in a wide variety of cells (5,19,20). In addition, the NF-B binding site in the promoter region of human IL-8 and GRO is indispensable for gene expression in response to cytokine stimulation (10,(21)(22)(23)(24).
Glucocorticoids are used as anti-inflammatory and immunomodulatory agents in a wide variety of diseases. Their physiological effects may be accomplished largely by modulating the expression of many cytokine genes, such as IL-1 (25,26), IL-2 (27), TNF␣ (28), interferon-␤ (29), interferon-␥ (27), and monocyte chemotactic and activating factor (30). Glucocorticoids also inhibit the production of human IL-8 (31), and the mechanisms of this have been examined by several groups. Mukaida et al. reported that dexamethasone decreases human IL-8 mRNA expression at the transcriptional level in human fibrosarcoma (32) and in human glioblastoma cell lines (33). In normal human embryonic lung fibroblasts, dexamethasone decreases IL-8 gene expression by reducing the stability of its mRNA (34). However, a different mechanism may be responsible in primary cultured human airway epithelial cells (35). On the other hand, although glucocorticoids inhibit GRO production in mouse (36) and rat (18) cell lines, little is known about the molecular mechanisms by which this is achieved.
As inhibition of GRO production by glucocorticoids must also have important implications for their actions as anti-inflammatory agents, we analyzed the effects of dexamethasone on CINC/gro expression at the molecular level in the normal rat kidney epithelial cell line NRK-52E. Evidence is presented here that this steroid hormone significantly decreases the transcription rate of the CINC/gro gene without affecting the stability of its mRNA in NRK-52E cells activated with IL-1␤. We identified the NF-B binding site on the CINC/gro gene as the cis-element responsible for repression by dexamethasone as well as IL-1␤induced CINC/gro gene activation. Moreover, an electrophoretic mobility shift assay revealed that dexamethasone decreased IL-1␤-induced NF-B-binding site complexes, which were recognized using antibodies against p50 and p65. Furthermore, we found that this inhibition resulted from the prevention of NF-B p65 translocation from the cytoplasm into the nucleus after degradation of the NF-B inhibitor IB-␣ in the cytoplasm. Thus, glucocorticoids can inhibit NF-B activity by a novel mechanism involving a blockage of the cytokine-induced nuclear translocation of NF-B.
ELISA for CINC/gro-The content of CINC/gro in the culture supernatants was determined by an ELISA using a rat IL-8 assay kit (Panapharm Laboratories Co., Ltd., Kumamoto, Japan), following the supplier's instructions.
Northern and Slot Blot Analyses-NRK-52E cells grown to subconfluence in medium supplemented with 10% fetal calf serum were incubated with or without indicated concentrations of dexamethasone for 3 h. Various stimulants were added, and the cells were further incubated for the indicated periods. Total cellular RNA was extracted by the acid/guanidium isothiocyanate/phenol/chloroform method (37). For Northern blotting, 10 g of total RNA was loaded and separated on 1.5% agarose gels containing 2% formaldehyde and blotted onto nylon membranes (Hybond-N, Amersham). For slot blotting, 10 g of total RNA was denatured in 50% formamide, 7% formaldehyde, 1 ϫ SSC (1 ϫ SSC contains 0.15 M NaCl, 15 mM sodium citrate, pH 7.4) at 68°C for 15 min and directly blotted onto nylon membranes (Hybond-N) using a slot blot apparatus (Bio-Dot SF, Bio-Rad). The entire 0.9-kb CINC/gro cDNA was radiolabeled with [␣-32 P]dCTP (3000 Ci/mmol) using a Megaprime DNA labeling kit (Amersham), and filter hybridization with the probe proceeded for 16 h at 42°C in 5 ϫ SSPE buffer (1 ϫ SSPE contains 0.18 M NaCl, 10 mM sodium phosphate, pH 7.4, 1 mM EDTA) containing 0.1% SDS, 50% formamide, 5 ϫ Denhardt's solution, and 100 g/ml sonicated salmon sperm DNA. After hybridization, the mem-branes were washed twice with 2 ϫ SSPE, 0.1% SDS at room temperature for 10 min, once with 1 ϫ SSPE, 0.1% SDS at 50°C for 15 min and twice with 0.5 ϫ SSPE, 0.1% SDS at 50°C for 10 min. The membranes were exposed to Fuji imaging plates, and the radioactivity levels were determined using a bioimage analyzer (Fujix BAS 2000, Fuji Co., Ltd., Tokyo, Japan). Hybridization with the 0.77-kb fragment of chicken ␤-actin (Oncor, Inc., Gaithersburg, MD) or the 1.0-kb fragment of human glyceraldehyde-3-phosphate dehydrogenase (Clontech Laboratories, Inc., Palo Alto, CA) confirmed that equal amounts of RNA were applied to each blot.
Construction of the Luciferase Expression Vectors-The PstI-PstI fragment of the genomic CINC/gro DNA which spans nucleotides (nt) Ϫ1034 to ϩ7 from the start of the first exon (39) was subcloned into pUC118, digested with appropriate restriction endonucleases, and further subcloned into SmaI and HindIII sites in the polylinker region of the pGL2-Basic vector (Promega, Madison, WI), to place the CINC/gro promoter region upstream of the firefly luciferase gene to generate pGL2-CINC/gro (Ϫ1034). Deleted fragments of the CINC/gro gene 5Јflanking region (starting from nt Ϫ164 and Ϫ48 to ϩ6) were prepared by means of the polymerase chain reaction. Site-directed mutagenesis of the CINC/gro promoter was performed using a U.S.E. Mutagenesis kit (Pharmacia) according to the manufacturer's protocol. The adenovirus 2 major late promoter region, which was obtained by digesting the pAD␤ vector (Clontech Laboratories Inc.) with XhoI/NotI, was subcloned into pBluescript II SK(ϩ), digested with the appropriate restriction endonucleases, and further subcloned into the pGL2-Basic vector, to place the adenovirus promoter region upstream of the firefly luciferase gene. One copy of either wild-type or NF-B binding site-mutated CINC/gro promoter fragment from nt Ϫ164 to ϩ7 was inserted upstream of the adenovirus promoter in the antisense orientation with respect to the luciferase gene. Two copies of the CINC/gro NF-B binding site (5Ј-GGGAATTTCC-3Ј) or two copies of CINC/gro NF-IL6 binding site (5Ј-TGGAGCAAG-3Ј) were inserted upstream of the adenovirus promoter linked to the luciferase gene. The fidelity of the constructs was verified by nucleotide sequencing.
DNA Transfection and Luciferase Assay-To isolate long term stable transfectants from NRK-52E cells, the cells were co-transfected with 15 g of the luciferase expression vector pGL2-CINC/gro (Ϫ1034) and 5 g of pSV3neo by calcium phosphate co-precipitation as described (40). For transient expression studies, about 2.5 ϫ 10 5 NRK-52E cells were transfected with 2 g of plasmid DNA by DEAE-dextran (0.5 mg/ml) with ProFection™ mammalian transfection system DEAE-dextran (Promega), following the supplier's instructions. After 16 h, cells were incubated with or without dexamethasone and/or various stimulants for the times indicated before harvesting the transfected cells. Under our experimental conditions, the differences in transient transfection efficiencies between dishes were less than 13% (data not shown).
The luciferase activity in cell extracts was determined using either a luciferase assay system (Promega) or PicaGene™ (Toyo Ink Co., Tokyo, Japan), following the supplier's instructions. The light intensity was measured with a Lumat model LB953 luminometer (Berthold, Germany). The protein concentration was measured by the Bradford method (41) (Bio-Rad protein assay) with bovine serum albumin as a standard. To analysis stably transfected cells, the luciferase assay was performed at least three times with three different transfectants. For transiently transfected cells, results were confirmed by at least two independent transfection experiments with triplicate dishes and are expressed as means Ϯ S.E.
Preparation of Whole Cell, Nuclear, and Cytoplasmic Extracts-Whole cell extracts were prepared as described by Zimarino and Wu (42) with slight modifications. Briefly, cells were harvested, washed in phosphate-buffered saline, and lysed in whole cell extraction buffer (10 mM HEPES, pH 7.9, 20% (v/v) glycerol, 400 mM NaCl, 0.1 mM EGTA, pH 8.0, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride) for 30 min at 4°C by vigorous mixing. After centrifugation at 100,000 ϫ g at 4°C for 10 min, the supernatant was used as the whole cell extract. Nuclear and cytoplasmic extracts were prepared as described by Schreiber et al. (43) with slight modifications. Briefly, cells were harvested, washed in phosphate-buffered saline, resuspended in Buffer A (10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride), and incubated for 15 min on ice. Nonidet P-40 was added to a final concentration of 0.6%, and the mixture was vortexed vigorously for 10 s. After 1 min on ice, nuclei were pelleted by centrifugation at 2000 ϫ g for 5 min, and the supernatants were used as cytoplasmic extracts. The pelleted nuclei were extracted at 4°C for 30 min with Buffer C (20 mM HEPES (pH 7.9), 400 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride) prior to centrifugation at 15,000 ϫ g for 20 min at 4°C. The supernatants were dialyzed for 3 h in dialysis buffer (10 mM HEPES, pH 7.9, 1 mM EDTA, pH 8.0, 50 mM KCl, 20% glycerol, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride) prior to centrifugation at 15,000 ϫ g for 20 min, and the supernatants were used as nuclear extracts.
Preparation of DNA Probes for Electrophoretic Mobility Shift Assay-Sense and antisense oligomers containing wild-type NF-B binding site of CINC/gro gene were annealed, and the double-stranded oligomer was radiolabeled with T4 polynucleotide kinase and [␥-32 P]ATP, then used as a DNA probe. The coding sequence of the oligomers was 5Ј-GCTC-CGGGAATTTCCCTGGC-3Ј (from nt Ϫ67 to Ϫ48 in the 5Ј-flanking region of the CINC/gro promoter).
Electrophoretic Mobility Shift Assay-Binding reactions contained 10 mM HEPES, pH 7.9, 1 mM EDTA, pH 8.0, 50 mM KCl, 12% glycerol, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1 mg/ml bovine serum albumin, 0.015 mg/ml poly(dI-dC)⅐poly(dI-dC), and 1 g of nuclear extract prepared as describe above. The mixture was incubated for 20 min at 4°C with 1 ng of 32 P-labeled oligonucleotide probe before electrophoresis. For the competition assay, unlabeled wild-type or mutated (5Ј-GCTCCGAACATCTCAATGGC-3Ј) NF-B oligonucleotide was added to the binding reactions, and the mixture was incubated for 10 min at 4°C before adding the probe. In the supershift experiments, the antibodies were added to the binding reaction mixtures, and the mixtures were incubated for 1 h at room temperature before adding the probe. After incubation, the samples were loaded onto 5% native polyacrylamide gels with 0.25 ϫ Tris borate-EDTA electrophoresis buffer. After electrophoresis, gels were dried and autoradiographed.

RESULTS
Dexamethasone Inhibits CINC/gro Production by Rat NRK-52E Cells-Production of CINC/gro from the normal rat kidney epithelial cell line NRK-52E was examined by measuring the levels of CINC/gro protein by ELISA in the culture supernatants. As shown in Fig. 1A, CINC/gro protein production was low in unstimulated cells, but was increased markedly by treatment with IL-1␤, LPS, or TNF␣ for 5 h. The induced CINC/gro protein synthesis was reduced by 1 M dexamethasone by about 50% (Fig. 1A). Dexamethasone at concentrations higher than 1 nM significantly inhibited CINC/gro production by the cells stimulated with IL-1␤ (Fig. 1D).
Effects of Dexamethasone on CINC/gro mRNA Induction in NRK-52E Cells-We next examined the levels of CINC/gro mRNA in NRK-52E cells incubated with various stimuli and/or FIG. 1. Dexamethasone inhibits CINC/gro transcription in NRK-52E cells. A and D, effects of dexamethasone (Dex) on CINC/gro production. NRK-52E cells were cultured in 24-well plastic dishes in 0.5 ml of Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum until reaching subconfluence. The cells were then incubated with the indicated concentrations of dexamethasone for 3 h, then stimulated with the indicated stimuli for 5 h. CINC/gro contents in the culture media were determined by ELISA as described under "Experimental Procedures." The data were normalized to the IL-1␤-treated control (100%) (B). Experiments were performed three times and representative results are shown. B and E, effects of dexamethasone on CINC/gro mRNA induction. NRK-52E cells were incubated with the indicated concentrations of dexamethasone for 3 h, then stimulated with the indicated stimuli for 3 h. Slot blotting proceeded using 10 g of total RNA extracted from NRK-52E cells. The quantity of mRNA was determined using a bioimage analyzer (Fujix BAS2000). The data are presented as the -fold increase over the resting value (B) or were normalized to the IL-1␤-treated control (100%) (E) from three separate experiments. C and F, effects of dexamethasone on CINC/gro promoterdriven transcription. NRK-52E cells were permanently transfected with the CINC/gro promoter (nt Ϫ1034 to ϩ7) linked to the luciferase gene as described under "Experimental Procedures." The cells were incubated with the indicated concentrations of dexamethasone for 3 h, then stimulated with the indicated stimuli for 5 h. Thereafter, the cells were harvested to determine the luciferase activity. The data are presented as the -fold increase over resting value (C) or were normalized to IL-1␤-treated control (100%) (F) from three separate experiments. The stimulants were 100 units/ml IL-1␤, 10 g/ml LPS, 100 units/ml TNF␣, or 20 ng/ml 12-O-tetradecanoylphorbol-13-acetate (TPA). dexamethasone. Since preliminary Northern blots demonstrated a specific band for CINC/gro, ␤-actin, or glyceraldehyde-3-phosphate dehydrogenase transcripts under our hybridization/washing conditions (data not shown), we performed slot blot analysis in these experiments. Following exposure to IL-1␤, the CINC/gro mRNA transcript level increased rapidly to its maximum at 3 h and remained elevated at 8 h ( Fig. 2A). Dexamethasone (1 M) inhibited this induction by about 50%. In contrast to CINC/gro, the levels of ␤-actin and glyceraldehyde-3-phosphate dehydrogenase transcripts in the same cells did not significantly change irrespective of the presence of IL-1␤ and/or dexamethasone (Fig. 2, B and C). Dexamethasone also reduced CINC/gro mRNA levels in the cells stimulated with LPS or TNF␣ for 3 h by about 50% (Fig. 1B). Moreover, dexamethasone inhibited CINC/gro mRNA expression stimulated with IL-1␤ in a dose-dependent manner similar to that of CINC/gro protein production (Fig. 1E). These results indicated that the effects of dexamethasone were expected at least partly at the pretranslational level. An incubation with the specific glucocorticoid receptor antagonist RU486 (1 M) 2 h before dexamethasone (0.1 M) abrogated the inhibitory effect of dexamethasone on CINC/gro mRNA induction by IL-1␤ (100 units/ ml) (data not shown), suggesting that the effect of dexamethasone is mediated through the glucocorticoid receptor.
Effect of Dexamethasone on the Transcriptional Rate of the CINC/gro Gene-To examine whether dexamethasone affects CINC/gro gene expression at the transcriptional level, we performed nuclear run-on assays. As shown in Fig. 3, IL-1␤ dramatically induced the transcription of the CINC/gro gene and dexamethasone reduced this enhanced [␣-32 P]UTP incorporation into CINC/gro mRNA by about 50%, which correlated well with the degree of inhibition of CINC/gro production by dexamethasone. In contrast, the transcription rate of the control glyceraldehyde-3-phosphate dehydrogenase gene was not significantly affected (Fig. 3). This suggested that dexamethasone specifically reduced the transcriptional rate of the CINC/gro gene.
Effect of Dexamethasone on the Stability of the CINC/gro mRNA-To directly analyze whether or not dexamethasone decreases the stability of CINC/gro mRNA, we examined the half-life of CINC/gro mRNA in NRK-52E cells stimulated with IL-1␤ in the presence or absence of dexamethasone (1 M). Transcription was interrupted 3 h after stimulation by actinomycin D (10 g/ml), an inhibitor of RNA transcription. The half-life of CINC/gro mRNA of less than 1 h was not significantly changed in the presence of dexamethasone (Fig. 4). During this period, ␤-actin mRNA transcripts were relatively stable and this remained unchanged in the presence of dexamethasone (Fig. 4). This suggested that the inhibition of CINC/ gro mRNA expression by dexamethasone is not due to a decrease in its stability. Taken together, these findings indicated that dexamethasone inhibits CINC/gro production mainly at the transcriptional level.
Dexamethasone Suppressed the Activation of Luciferase Gene Linked to the CINC/gro Promoter-We examined the function of the 5Ј-flanking region of the CINC/gro gene to assess the mechanism of transcriptional regulation in activation and the suppression by several cytokines and dexamethasone, respectively. The 5Ј-upstream region (from nt Ϫ1034 to ϩ7) of the CINC/gro gene was linked to the luciferase gene and the construct was permanently transfected into NRK-52E cells. Fig. 1, C and F, shows the results of luciferase assay using one of these transformants. As shown in Fig. 1C, dexamethasone prevented the luciferase activity induced by IL-1␤, LPS, or TNF␣. Moreover, dexamethasone inhibited the luciferase activity induced by IL-1␤ in a dose-dependent manner similar to that of the CINC/gro protein synthesis and of mRNA expression (Fig. 1F). The inhibition of luciferase activity induced by IL-1␤ (100 units/ml) by dexamethasone (0.1 M) was also reversed by RU486 (1 M) (data not shown). Similar results were obtained FIG. 3. Effects of dexamethasone on the transcription of CINC/ gro gene in NRK-52E cells. NRK-52E cells were incubated with or without 1 M dexamethasone (Dex) for 3 h, then stimulated with or without 100 units/ml IL-1␤ for 1 h. Nuclei were isolated and nuclear run-on assays were performed as described under "Experimental Procedures." The quantity of the CINC/gro or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) transcript was determined using a bioimage analyzer (Fujix BAS2000). Data were normalized to IL-1␤-treated CINC/gro transcriptional rate (100) and are shown as means Ϯ S.E. of three separate experiments.
FIG. 2. Effects of dexamethasone (Dex) on CINC/gro mRNA induction in NRK-52E cells. NRK-52E cells were incubated with (filled triangles and filled circles) or without (unfilled triangles and unfilled circles) 1 M dexamethasone (Dex) for 3 h, then stimulated with (circles) or without (triangles) 100 units/ml IL-1␤ for 3 h. Slot blotting proceeded using 10 g of total RNA extracted from NRK-52E cells. The quantities of CINC/gro (A), ␤-actin (B), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (C) mRNAs were determined using a bioimage analyzer (Fujix BAS2000). The data are presented as the -fold increase over the resting value at time 0 from three separate experiments.
in the luciferase assay using another two independent stable transformants (data not shown). When the same CINC/gro promoter-luciferase construct was transiently transfected into NRK-52E cells, the results were similar (Fig. 5B).
Northern blotting analysis of total RNA from the stable transformants revealed that dexamethasone suppressed the expression of luciferase mRNA induced by IL-1␤ (data not shown), indicating that suppression of luciferase activities by dexamethasone was caused by transcriptional repression of the introduced fusion gene. These results provide evidence for the presence of sufficient information within the 5Ј-flanking sequences between nt Ϫ1034 and ϩ7 of the CINC/gro gene for transactivation, as well as for suppression by these cytokines and dexamethasone, respectively.
Dexamethasone Suppressed Transcriptional Activation through the NF-B Binding Site of the CINC/gro Gene-To delineate the sequences in the CINC/gro promoter transactivation of which by IL-1␤ is inhibited by dexamethasone, we studied the effects of both agents on the expression of tran-siently transfected luciferase expression vectors linked to deleted CINC/gro promoters (Fig. 5). IL-1␤ greatly increased the luciferase activity in cells transfected with Ϫ1034 and Ϫ164 luciferase constructs, but showed hardly any increase when cells were transfected with the Ϫ48 luciferase construct. Dexamethasone prevented the expression of Ϫ1034/ϩ7 and Ϫ164/ϩ7 luciferase constructs induced by IL-1␤, suggesting that the minimal essential elements for the CINC/gro gene expression regulation by IL-1␤ and dexamethasone are located within the 3Ј region downstream of nt Ϫ164. There are two known cis-elements, NF-IL6 (nt Ϫ119 to Ϫ111) and NF-B (nt Ϫ62 to Ϫ53) binding sites, within the region downstream of nt Ϫ164 (Fig. 5A) (39).
The regulatory roles of the NF-IL6 and NF-B binding sites were examined by site-directed mutagenesis of each element in the Ϫ164 luciferase plasmid (Fig. 5). The deletion of the NF-IL6 binding site had little effects on CINC/gro gene activation and repression by IL-1␤ and dexamethasone, respectively, indicating that this site was not essential for gene regulation by these agents. Deletion or mutation of the NF-B binding site abolished IL-1␤-induced luciferase activity, indicating the essential role of the NF-B binding site for CINC/gro gene activation by IL-1␤.
To further examine the effect of dexamethasone on the transcriptional activity through the NF-B binding site of the CINC/gro gene, two copies of the CINC/gro NF-B binding site were linked to the adenovirus 2 major late promoter, which was inserted upstream of the luciferase gene. As shown in Fig. 6, luciferase gene expression driven from the adenovirus promoter was not affected by treatment with IL-1␤ and/or dexamethasone. Dexamethasone reduced the nt Ϫ164 to ϩ7 region of the CINC/gro promoter-driven transcription induced by IL-1␤ (Fig. 6). Mutation of the NF-B target sequence of the CINC/gro gene in the same luciferase construct abolished the induction of luciferase activity by IL-1␤. Furthermore, the transcriptional activity of the NF-B motif, but not of the NF-IL6 motif, was significantly enhanced by IL-1␤, and the increased activity was inhibited by dexamethasone (Fig. 6). These findings suggested that the NF-B binding site is the target for gene repression by dexamethasone.
Dexamethasone Suppressed NF-B Complex Formation Induced by IL-1␤-To examine the nuclear factors binding to the NF-B binding site of the CINC/gro promoter, we performed an electrophoretic mobility shift assay using the CINC/gro Ϫ67/ Ϫ48 DNA fragment containing the NF-B binding site (nt Ϫ62   FIG. 4. Effects of dexamethasone on stability of CINC/gro mRNA in NRK-52E cells. NRK-52E cells were incubated with (filled symbols) or without (unfilled symbols) 1 M dexamethasone (Dex) for 3 h, then stimulated with 100 units/ml IL-1␤ for 3 h. Actinomycin D (10 g/ml) was added at time 0, and total RNA was isolated at the indicated times. Slot blotting proceeded using 10 g of total RNA extracted from NRK-52E cells. The quantities of CINC/gro (circles) and ␤-actin (triangles) mRNAs were determined using a bioimage analyzer (Fujix BAS2000). The amounts of mRNA at various times are expressed as a function of the mRNA level at time 0. The data presented are representative of two independent experiments with two cultures.

FIG. 5. Identification of the responsive elements for dexamethasonemediated transcriptional repression of CINC/gro gene in NRK-52E cells.
A, schematic structure of the 5Ј-flanking region of the CINC/gro gene. B, delineation of dexamethasone (Dex)-responsive elements in the CINC/gro promoter. Various lengths of wild-type CINC/gro promoter, or mutants carrying the internal deletions or point mutations linked to the luciferase gene, were transiently transfected into NRK-52E cells. The cells were then incubated with or without 1 M dexamethasone for 3 h then stimulated with or without 100 units/ml IL-1␤ for 5 h, and the luciferase activity was measured. The promoter activity of each test plasmid is indicated as luciferase activity relative to that of the nt Ϫ1034/ϩ7-luciferase construct in the presence of IL-1␤. Values are means Ϯ S.E. for at least three separate experiments.
to Ϫ53), as a probe. Although one DNA-protein complex was detected using nuclear extracts from unstimulated NRK-52E cells, stimulation with IL-1␤ led to the induction of a new DNA-protein complex (Fig. 7A). The complex formation was inhibited by an unlabeled corresponding oligomer but not by mutated NF-B site oligomer (Fig. 7A), indicating the specific binding of this complex. Similar results were obtained using extracts from LPS-induced NRK-52E cells (data not shown). These results demonstrated that the NF-B site of CINC/gro is indeed the target of the binding protein which is activated in response to IL-1␤ or LPS.
To examine whether the induced complex contained any of the known forms of the Rel family proteins, we performed electrophoretic mobility shift assays using antibodies against p50, p52, p65, c-Rel, and RelB. Anti-p50 and anti-p65 antibodies supershifted the complex induced by IL-1␤, whereas the other antibodies did not (Fig. 8), indicating that the induced NF-B complexes were composed of p50-p65 heterodimers. Dexamethasone inhibited the formation of the NF-B complexes induced by IL-1␤ (Fig. 7B). The radioactivity of the bands was quantified using a bioimage analyzer (Fujix BAS2000), and we found that dexamethasone inhibited the complex formation by 30 -40% (seven separate experiments), which were similar to the values of the CINC/gro production inhibited by dexamethasone. In addition, we found that the inhibited NF-B complexes were also composed of p50 and p65 heterodimers according to the results of the supershift assay (data not shown). These results indicated that dexamethasone inhibits the binding of the NF-B transcription factor to the NF-B binding site of the CINC/gro promoter, leading to the suppression of CINC/gro mRNA and protein synthesis.
Dexamethasone Suppressed the Translocation of p65 from the Cytoplasm into the Nucleus-To further examine the mechanism of dexamethasone-mediated NF-B binding inhibition, we analyzed the effect of dexamethasone on the p65 and p50 levels in NRK-52E cells by Western blotting followed by den-sitometric quantitation of the respective bands in fluorograms. A representative experiment is shown in Fig. 9. In the presence of IL-1␤, dexamethasone, or both, the amounts of p65 in whole cell extracts were measured as 1.2-, 1.1-, or 1.2-fold that found in the absence of these agents, respectively (Fig. 9C), indicating that these agents had a little, if any, effect on the total amount of p65 protein in the cells. Under the same conditions, p65 was exclusively localized in the cytoplasmic extracts from unstimulated cells (Fig. 9, A and B). IL-1␤ markedly increased the amount of p65 in the nuclear extracts by about 6-fold in parallel with the decrease of that in cytoplasmic extracts by about 1/6 ( Fig. 9, A and B), suggesting that IL-1␤ induced the translocation of p65 from the cytoplasm into the nucleus. Dexamethasone diminished the amount of translocated p65 in the nuclear extracts by 65% in IL-1␤-treated cells in parallel with an increase in level of p65 in the cytoplasmic extracts from 16 to 42% of the control resting cells (Fig. 9, A and B), suggesting that dexamethasone interferes with the translocation of p65 from the cytoplasm into the nucleus. In contrast to p65 protein, we found little, if any, detectable differences in the amounts of nuclear p50 protein in control and IL-1␤and/or dexametha-FIG. 6. Requirement of NF-B binding site for repression of transcription of CINC/gro gene by dexamethasone. One copy of the wild-type or NF-B site-mutated CINC/gro promoter from nt Ϫ164 to ϩ7 as shown in Fig. 5 was inserted upstream of the adenovirus promoter in the antisense orientation with respect to the luciferase gene. Two copies of the CINC/gro NF-B site (5Ј-GGGAATTTCC-3Ј) or two copies of the CINC/gro NF-IL6 site (5Ј-TGGAGCAAG-3Ј) were also inserted upstream of the adenovirus promoter linked to the luciferase gene. Transfection proceeded as described under "Experimental Procedures." After transfection, the cells were incubated with or without 1 M dexamethasone (Dex) for 3 h, then stimulated with or without 100 units/ml IL-1␤ for 5 h and the luciferase activity was measured. The promoter activity of each test plasmid is indicated as luciferase activity relative to that of adenovirus promoter construct in the absence of both dexamethasone and IL-1␤. These values are means Ϯ S.E. for at least three separate experiments. sone-treated cells (Fig. 9D), suggesting that the p50 subunit of NF-B constitutively existed in the nucleus and that IL-1␤ and/or dexamethasone has little effect on the amount of nuclear p50. We also measured the amount of p50 protein in whole cell extracts and found that dexamethasone rather increased the amounts by 1.5-2-fold those of control or IL-1␤treated cells, although the mechanism remains to be elucidated (Fig. 9E).
Effect of Dexamethasone on the IL-1␤-induced Degradation of IB-␣-Release of the inhibitory protein IB from the cytoplasmic NF-B/IB complex (44,45) by rapid phosphorylation and degradation of IB (46 -49) are necessary for NF-B activation. We examined the amount of IB-␣ (MAD-3) protein in cytoplasmic extracts of NRK-52E cells by Western blotting. Fig. 10A shows the time course of IB-␣ degradation in NRK-52E cells following activation by IL-1␤. In extracts from unstimulated cells, IB-␣-specific antibody detected a single ϳ40-kDa band on Western blots (Fig. 10B, lane 1). Degradation of IB-␣ was detected by IL-1␤ stimulation for 20 min, and it was almost completely disappeared at 1 h (Fig. 10A). In contrast to another studies (48,49), no transient change in the electrophoretic mobility of IB-␣ was apparent after induction with IL-1␤, even when phosphatase inhibitors were included in the lysis buffer. 2 The degradation of IB-␣ coincided with both the increase in nuclear p65 levels (Fig. 10A) and the appearance of NF-B DNA-binding activity in nuclear extracts (Fig. 7), suggesting a causal relationship between the three events.
To determine whether dexamethasone prevents the release of NF-B from IB-␣ after exposure to IL-1␤, we examined the amount of IB-␣ protein in cytoplasmic extracts prepared 1 h after exposure to IL-1␤ in the presence or absence of dexamethasone. As shown in Fig. 10B, the amount of IB-␣ in IL-1␤-activated cells was markedly reduced to 21% of that in resting cells. Dexamethasone, however, had no effect on the IL-1␤-induced degradation of IB-␣ (Fig. 10B). These results suggested that dexamethasone inhibits the activation of NF-B by interfering with a pathway after the degradation of IB-␣. DISCUSSION Glucocorticoid hormones are highly immunosuppressive and reportedly inhibit the gene expression of several cytokines, particularly those with proinflammatory actions, such as IL-1 (25,26), IL-2 (27), TNF␣ (28), interferon-␤ (29), interferon-␥ (27), IL-8 (31), and monocyte chemotactic and activating factor (30). The results of this study demonstrated the inhibitory effect of the synthetic glucocorticoid dexamethasone on CINC/ gro production in rat NRK-52E cells and its mechanism. We found that dexamethasone suppressed IL-1␤-induced CINC/ gro gene expression (Figs. 1 and 2) by inhibiting the transcriptional rate of the CINC/gro gene (Fig. 3) without affecting the stability of CINC/gro mRNA (Fig. 4). The inhibition of CINC/ gro gene expression by dexamethasone was reversed by the specific glucocorticoid receptor antagonist RU486 (data not shown), suggesting the involvement of the glucocorticoid recep- After electrophoresis, the proteins were electroblotted onto Immobilon-P SQ polyvinylidene difluoride membranes. Transblotted membranes were incubated with either anti-p65 (A-C) or anti-p50 (D and E) antibody, followed by a reaction with horseradish peroxidase-conjugated anti-rabbit IgG. Proteins detected by the primary antibody were visualized using an ECL assay kit (Amersham) and by exposure to x-ray film. tor in transcriptional repression. Furthermore, we showed that the 5Ј-flanking region of the CINC/gro gene (extending from nt Ϫ1034 through ϩ7) is sufficient to confer responsiveness to IL-1␤ and dexamethasone, since dexamethasone inhibited the IL-1␤-induced expression of the CINC/gro promoter-driven luciferase vector stably transfected into NRK-52E cells (Fig. 1, C  and F).
Functional analysis of the regulatory sequences of the CINC/ gro gene demonstrated that the minimally essential elements for the induction by IL-1␤ and repression by dexamethasone of the CINC/gro gene were present within the 3Ј region downstream of nt Ϫ164, which contains the two known cis-elements, NF-IL6 (nt Ϫ119 to Ϫ111) and NF-B (nt Ϫ62 to Ϫ53) binding sites (Fig. 5). Deletion of the NF-IL6 binding site did not abolish the inhibition of IL-1␤-induced luciferase activity by dexamethasone (Fig. 5B), suggesting that dexamethasone does not inhibit CINC/gro gene transcription through interference with the NF-IL6 binding site. Although either a deletion or mutation of the NF-B binding site abolished luciferase activity induced by IL-1␤ (Fig. 5B), a tandem repeat of the NF-B sequence in the CINC/gro gene conferred inducibility by IL-1␤ and suppression of luciferase activity by dexamethasone (Fig. 6), suggesting that this site is the element responsible for dexamethasonemediated gene repression. In addition, an electrophoretic mobility shift assay demonstrated that dexamethasone significantly diminished the IL-1␤-induced formation of NF-B complexes (Fig. 7B), which were identified immunochemically to consist of p50 and p65 (Fig. 8). Our results suggested that the NF-B binding site is responsible for CINC/gro gene repression by dexamethasone.
Western blotting of cytoplasmic and nuclear extracts from NRK-52E cells demonstrated that IL-1␤ treatment induced the translocation of p65 from the cytoplasm into the nucleus (Fig.  9, A and B). Dexamethasone diminished the amount of p65 translocated to the nuclear extracts from cells exposed to IL-1␤ (Fig. 9A). We also found that the loss of nuclear p65 was paralleled by an increase in cytoplasmic p65 (Fig. 9, A and B) without affecting the total amount of p65 in the cells (Fig. 9C), suggesting that dexamethasone interfered with its translocation from the cytoplasm into the nucleus. Under our experimental conditions for Western blotting however, we detected neither the IL-1␤-induced translocation of p50 protein into the nucleus nor the inhibition of translocation by dexamethasone (Fig. 9D). It is possible that the amount of p50 protein translocated from the cytoplasm into the nucleus was much less than that of the nuclear p50 protein constitutively present even in unstimulated cells.
We surmised that dexamethasone prevented the release of NF-B from IB after exposure to IL-1␤. Since the phosphorylation and degradation of IB is necessary for the activation of NF-B (46 -49), we examined the amount of IB protein by Western blotting. Time course experiments showed that the activation of NF-B by IL-1␤ in NRK-52E cells is correlated closely with the degradation of IB-␣, a member of the IB family (Fig. 10A). We found that dexamethasone did not prevent the IL-1␤-induced degradation of IB-␣ (Fig. 10B), suggesting that it interfered with the translocation of NF-B into the nucleus after dissociation of NF-B/IB-␣ complexes in the cytoplasm. However, we cannot exclude the possibility that dexamethasone has some effects on the synthesis and stability of other members of the IB family, including IB-␤ (50), as well as the NF-B precursors p105 (51,52) and p100 (53).
Mechanisms of glucocorticoid receptor-mediated repression of transcription have been proposed involving the physical interaction of the glucocorticoid receptor and NF-B (33,54,55). This interaction may result in the glucocorticoid-mediated blockage of nuclear NF-B binding to NF-B binding sites. We showed that glucocorticoid decreased the levels of nuclear NF-B of IL-1␤-stimulated cells (Fig. 9A) while the glucocorticoid receptor level in the nucleus was markedly increased 3 under the conditions where dexamethasone inhibited the NF-B binding to DNA in the nucleus (Fig. 7B). Thus, although we could not immunochemically detect the glucocorticoid receptor in the NF-B complexes formed in the presence of IL-1␤ and dexamethasone by means of an electrophoretic mobility shift assay, 3 it is likely that a combination of decreasing the nuclear NF-B protein level and blocking the binding of nuclear NF-B to DNA by protein-protein interaction accounts for the total repression of CINC/gro gene expression by dexamethasone. The interaction of a glucocorticoid receptor with NF-B complexes in the nucleus is now under study to determine whether the interaction also involves suppression of NF-B activation in NRK-52E cells.
The AUUUA sequence in the 3Ј-untranslated region may be involved in the rapid degradation of mRNAs for some inflammatory cytokines and proto-oncogenes (56,57). Moreover, several reports have shown that dexamethasone decreases the stability of mRNAs containing the AUUUA sequence in the 3Ј-untranslated region (25,29). A specific protein binding to RNAs containing AUUUA has been identified (58,59), and it is thought that formation of this complex may target susceptible mRNAs for rapid cytoplasmic degradation. CINC/gro mRNA contains similar AUUUA sequences in the 3Ј-untranslated region (39). However, we found that dexamethasone does not significantly decrease the half-life of entire CINC/gro mRNA, although its half-life was much shorter than that of the ␤-actin transcript which does not contain this sequence (Fig. 4). Thus, the present findings suggest that the AUUUA sequence is responsible for destabilization of CINC/gro mRNA, but not for CINC/gro gene repression by dexamethasone in NRK-52E cells.
Early studies showed that glucocorticoid inhibits the transcription of human IL-8 gene, either by binding of the glucocorticoid receptor to the glucocorticoid responsive element in the 5Ј-flanking region of the gene (32) or by interference with the binding of NF-B to DNA without inhibiting the nuclear translocation of the factor (33). We demonstrated here that dexamethasone interfered with the binding of NF-B to its ciselement on the rat CINC/gro gene that lacks a glucocorticoid responsive element in its promoter region. Unlike human IL-8, dexamethasone decreased the nuclear level of p65, perhaps by sequestering p65 in the cytoplasm after dissociation from IB-␣ in NRK-52E cells. Thus, it is likely that glucocorticoids can inhibit NF-B activity by two novel mechanisms involving blocks of cytokine-induced nuclear translocation and DNA binding to NF-B. Given the role of NF-B in the transcriptional activation of many inflammatory cytokine genes, we propose that these types of inhibition of NF-B activation represents an important mechanism for the immunosuppressive properties of glucocorticoids.