Glucocorticoids increase interleukin-6–dependent gene induction by interfering with the expression of the suppressor of cytokine signaling 3 feedback inhibitor

Authors


  • Potential conflict of interest: Nothing to report.

  • Supported by the Deutsche Forschungsgemeinschaft (DFG) within the frame of the SFB542 TP-B4 to F.S. and J.G.B. and TP-C14 to C.T.

  • These authors contributed equally to this work.

Abstract

Glucocorticoids are known to be potent regulators of inflammation and have been used pharmacologically against inflammatory, immune, and lymphoproliferative diseases for more than 50 years. Due to their possible and well-documented side effects, it is crucial to understand the molecular mechanisms and targets of glucocorticoid action in detail. Several modes of action have been discussed; nevertheless, none of them fully explain all the functions of glucocorticoids. Therefore, we analyzed the cross-talk between glucocorticoids and interleukin-6 (IL-6) in the liver. IL-6 exerts pro-inflammatory as well as anti-inflammatory properties and is a main inducer of the acute-phase response. The balance between the proinflammatory and anti-inflammatory activities of IL-6 is tightly regulated by suppressor of cytokine signaling 3 (SOCS3), a well-known feedback inhibitor of IL-6 signaling. Here, it is demonstrated that glucocorticoids enhance IL-6–dependent γ-fibrinogen expression. Studying of the underlying mechanism revealed prolonged activation of signal transducer and activator of transcription 3 (STAT3) caused by down-regulation of SOCS3 protein expression. Consequently, in SOCS3-deficient cells glucocorticoids do not affect IL-6–induced signal transduction. Moreover, in hepatocytes lacking the SOCS3 recruiting motif within gp130, IL-6–dependent γ-fibrinogen expression is not influenced by glucocorticoid treatment. Conclusion: Glucocorticoids interfere with IL-6–induced expression of the feedback inhibitor SOCS3, thereby leading to enhanced expression of acute-phase genes in hepatocytes. This mechanism contributes to the explanation of how glucocorticoids affect inflammation and acute-phase gene induction. (HEPATOLOGY 2012;55:256–266)

Glucocorticoids are metabolic hormones that possess the ability to suppress immune responses. This property is taken advantage of and widely used as a therapeutic approach to treat inflammatory diseases that are caused by an overactive immune system. Glucocorticoids act through induction of apoptosis in thymocytes, peripheral T cells, and blood monocytes. Moreover, they modulate the activity of pro-inflammatory (e.g., interleukin-1 [IL-1], IL-8, inducible NO synthase) and anti-inflammatory proteins (e.g., IL-1 receptor antagonist, inhibitor of κB) (reviewed in Barnes1).

Glucocorticoids exert their effects in conjunction with nuclear receptors (glucocorticoid receptors) and following translocation to the nucleus bind to glucocorticoid response elements (GREs) in the promoters of glucocorticoid-responsive genes. However, the expression of genes that lack GREs within their promoters has also been described to be enhanced by glucocorticoids. Often, these promoters harbor response elements for transcription factors of the STAT family, and the glucocorticoid receptor acts as a coactivator protein.2-6 In contrast, the interaction of STAT proteins with glucocorticoid receptors bound to GREs has been shown to suppress glucocorticoid-dependent promoter activation.7 Thus, the interaction between the glucocorticoid receptor and transcription factors can be both enhancing and repressing.

In addition to gene-specific effects, glucocorticoids can also regulate signal transduction in a more general way. Ivashkiv's group demonstrated that glucocorticoids suppress the expression of STAT1 messenger RNA (mRNA) and thus contribute to the down-regulation of the proinflammatory activities of IFN-γ.8 Furthermore, it has been shown that glucocorticoids induce the expression of the mitogen-activated protein kinase phosphatase 1 (MKP1). MKP1 has been shown to dephosphorylate MAP kinase p38, thereby reducing the mRNA stabilizing ability of this MAP kinase, leading to the deactivation of this potent signaling cascade.8 In short, glucocorticoids possess a broad variety and range of anti-inflammatory actions that are still not fully understood or explained.

IL-6 exerts pro-inflammatory as well as anti-inflammatory activities and is the major mediator of the acute-phase response in the liver.10 It has been shown that the activation of STAT proteins is crucial for the induction of acute-phase proteins (APPs) in hepatocytes such as the protease inhibitors α1-antichymotrypsin (α1-ACT), α2-macroglobulin (α2M), tissue inhibitor of metalloproteinases (TIMP)1 in response to inflammation, γ-fibrinogen (FGG) for structural components in blood clotting, and serum-amyloid A protein (SAA) for the recruitment and activation of leukocytes.11 Binding of IL-6 to its receptor complex results in phosphorylation of STAT proteins by Janus kinases,12, 13 and subsequently leads to the expression of IL-6 target genes. IL-6–induced signal transduction must be a tightly controlled process because prolonged activation of IL-6 has been linked to chronic diseases and tumor growth. Therefore, one of the immediate downstream target genes of IL-6 is suppressor of cytokine signaling 3 (SOCS3), which is involved in the negative regulation of cytokines that signal through the JAK/STAT pathway, in so doing effectively turning off signal transduction.14-16 SOCS3 mRNA14, 17 and SOCS3 protein18 are rapidly induced by IL-6. Furthermore. the half-lives of SOCS3 mRNA17, 19, 20 and SOCS3 protein are also tightly regulated.19-21

Baumann et al. observed that glucocorticoids increase the expression of IL-6–induced APPs.22, 23 Although it has been demonstrated that certain genes are activated directly at the promoter by STAT3/glucocorticoid receptor dimers,6 we propose an alternative mode of action. Our data show that glucocorticoids reduce the expression of SOCS3 and thereby lead to an enhanced IL-6–induced acute-phase response.

Abbreviations

α1-ACT, α1-antichymotrypsin; AP1, activator protein 1; APP, acute-phase protein; APR, acute phase response; α2M, α2-macroglobulin; COX, cyclooxygenase; Dex, dexamethasone; Epo, erythropoietin; FBS, fetal bovine serum; FGG, γ-fibrinogen; GC, glucocorticoid; gp130, glycoprotein 130; GRE, glucocorticoid response element; IFN, interferon; IL, interleukin; MAPK, mitogen-activated protein kinase; MKP, MAPK-phosphatase; MCP-1, monocyte chemoattractant protein 1; MEF, murine embryonic fibroblasts; NF-κB, nuclear factor κB; SAA, serum-amyloid A protein; SHP2, SH2-domain containing phosphatase 2; SOCS, suppressor of cytokine signaling; STAT, signal transducer and activator of transcription; TIMP, tissue inhibitor of metalloproteinases; WT, wild type.

Materials and Methods

Isolation and Cultivation of Cells.

Isolation and culture of mouse hepatocytes were performed according to standard procedures. Eight- to 10-week-old wild-type (WT) and gp130RasΔhepa mice, maintained on a mixed background, were used.24 Gp130RasΔhepa mice express a mutated gp130 in which the cytoplasmic tyrosine that binds SOCS3 has been replaced by phenylalanine. Animals received care according to the criteria prepared by the National Academy of Sciences (National Institutes of Health publication 86-23, revised 1985). Primary mouse hepatocytes were isolated by collagenase perfusion. Mouse liver was perfused retrograde with Krebs Ringer buffer followed by perfusion of the liver with collagenase solution. Isolated hepatocytes were washed thoroughly with phosphate-buffered saline (PBS).

After determination of viability, 7 × 105 cells were plated on collagen-coated 6-cm dishes in William's Medium E supplemented with 10% fetal bovine serum (FBS), 100 nM dexamethasone, 2 mM L-glutamine, and 1% penicillin/streptomycin solution. Four hours later, the medium was changed to William's Medium E containing 100 nM dexamethasone, 2 mM L-glutamine, and 100 mg/mL penicillin/streptomycin solution. Before the experiments were started, cells were starved for 3 hours in William's Medium E supplemented with 2 mM L-glutamine and 100 mg/mL penicillin/streptomycin without FBS or phenol red.

For in vivo experiments, mice were treated intraperitoneally with dexamethasone (4 mg/kg) and/or IL-6 (100 μg/kg) as indicated. Serum samples were collected by bleeding. SAA was measured by enzyme-linked immunosorbent assay (ELISA; Dunn Labortechnik, Asbach, Germany) according to the manufacturer's instructions. For RNA analysis, mice were sacrificed, livers were explanted, tissue samples were immediately frozen, and RNA was isolated using TRIZOL and the RNeasy Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions.

HepG2 cells (DSMZ, Braunschweig, Germany) were grown in Dulbecco's modified Eagle medium (DMEM)+F12 supplemented with 10% FBS, streptomycin (100 mg/mL), and penicillin (100 mg/mL) at 37°C in a water-saturated atmosphere containing 5% CO2. WT and SOCS3-deficient murine embryonic fibroblasts (MEFs) were grown in DMEM supplemented with 10% FBS, streptomycin (100 mg/mL), and penicillin (100 mg/mL) at 37°C in a water-saturated atmosphere containing 10% CO2. Before stimulation cells were starved overnight in medium without FBS and phenol red.

Stimulation with IL-6 was performed with 20 ng/mL and if mentioned with 1 μg/mL sIL-6R ± 1 μM dexamethasone or 1 μM RU-468. HepG2 cells were stimulated with human IL-6, whereas primary murine hepatocytes and mice were stimulated with murine IL-6.

Quantitative RT PCR.

Total RNA was isolated using the RNeasy Kit (Qiagen) according to the manufacturer's instructions. Then 1 μg of RNA was reverse transcribed into complementary DNA (cDNA) with Omniscript (Qiagen) using random hexameric primers according to the manufacturer's instructions. TaqMan gene expression assays for human FGG (Hs00241037_m1), human HPRT: (Hs99999909_m1), murine FGG (Mm00513575_m1), murine SAA (Mm00656927_g1), and murine HPRT (Mm01545399_m1) were obtained from Applied Biosystems (Carlsbad, CA), and polymerase chain reaction (PCR) was performed using qPCR Mastermix plus (Eurogentec, Cologne, Germany). The PCR reaction was done in a final volume of 10 μL containing 2 μL cDNA and 1 μL TaqMan gene expression assay solution. After denaturing for 15 minutes at 94°C, amplification was performed in 40 cycles (15 seconds at 94°C, 60 seconds at 60°C) on a Rotorgene (Qiagen). The gene of interest and the housekeeping gene were amplified in duplicates. The quantification of gene expression was calculated using the Pfaffl method.25

Reporter Gene Assay.

HepG2 cells were grown in six-well plates. Transient transfection of 0.3 μg luciferase reporter construct (either FGG, SOCS3, or α2M) and 1 μg of β-galactosidase expression vector (pCR3lacZ, Pharmacia) per plate was performed in OptiMEM using Lipofectamine 2000 transfection reagent according to the manufacturer's description (Life Technology, Carlsbad CA, USA). When indicated, 1 μg of EpoR/gp130-chimeric receptors containing the extracellular part of the Epo receptor and the transmembrane and cytoplasmic part of gp130 as described26 were cotransfected. EpoR/gp130-chimeric construct (pRcCMV-EG(YYYYYY) conforms to the WT cytoplasmic part of gp130. Construct pRcCMV-EG(FYFFFF) contains tyrosine to phenylalanin substitutions at positions 683, 767, 814, 905, and 915 within the cytoplasmic part. Construct pRcCMV-EG (YFYYYY) contains a tyrosine to phenylalanine substitution at position 759.

Five hours after transfection, cells were starved in DMEM+F12 without phenol red overnight. Cells were stimulated as indicated in the figures. Cell lysis and luciferase assays were carried out using the luciferase kit (Promega, Mannheim, Germany) according to the manufacturer's instructions. Luciferase values were normalized to transfection efficiency monitored by β-galactosidase expression.

Western Blot

For the isolation of cellular proteins, cells from a 10-cm dish were lysed in 300 μL RIPA lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% NP-40, 15% glycerol supplemented with 10 μg/mL each of aprotinin, leupeptin, and pepstatin as well as 0.8 μM Pefabloc (Roche, Mannheim, Germany), 1 mM NaF, and 1 mM Na3VO4). The protein concentration of the lysates was determined using Protein Assay (Bio-Rad, Munich, Germany). Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene fluoride membrane. Antigens were detected by incubation with specific primary antibodies (1:1,000) followed by incubation with horseradish peroxidase–coupled secondary antibodies (1:7,500) (DAKO, Hamburg, Germany). Detection occurred by means of an enhanced chemoluminescence kit (Millipore Corp., Darmstadt, Germany). The primary antibodies and their suppliers are as follows: pSTAT3-Y705 (New England Biolabs, Frankfurt am Main, Germany); STAT3 (C20), SOCS3 (M20), and GAPDH (Santa Cruz Biotechnology, Santa Cruz, CA); LAMIN A/C (BD Bioscience, Franklin Lakes, NJ); and HSP70 (Stressgen Biotechnologies, San Diego, CA).

Flow Cytometry.

HepG2 cells were harvested in ice-cold PBS-EDTA and resuspended in PBS (supplemented with 5% FBS, 0.1% NaN3) containing antibody raised against gp80 (B-R6, Hölzel Diagnostika GmbH, Cologne, Germany) or gp130 (BR-3, Hölzel Diagnostika GmbH). After 30 minutes, cells were washed and subsequently stained with R-phycoerythrin–conjugated secondary antibody for 30 minutes. Cells were analyzed with a FACSCalibur (BD, Heidelberg, Germany).

Laser Scanning Microscopy.

HepG2 cells were transfected with 1 μg STAT3–cyan fluorescent protein (CFP) expression vector per 10-cm dish. One day after transfection cells were seeded on poly-L-lysine–coated coverslips and starved overnight in DMEM+F12 without phenol red. Cells were stimulated as indicated and fixed with 4% PFA. For detection of the endogenous glucocorticoid receptor, cells were permeabilized in 90% methanol for 15 minutes and stained with an antibody specific for the glucocorticoid receptor (glucocorticoid receptor [M20]; Santa Cruz Biotechnology) and a Cy-2–coupled secondary antibody. Analysis of cells was performed on a laser scanning confocal microscope (Zeiss LSM510, Jena, Germany).

Results

Glucocorticoids Enhance IL-6–Induced Acute-Phase Gene Expression in Liver Cells.

To demonstrate the effect of glucocorticoids on IL-6–dependent acute-phase gene expression, primary murine hepatocytes were isolated from healthy untreated mice and stimulated with IL-6 in the absence or presence of dexamethasone. Levels of endogenous FGG mRNA were quantified by real-time PCR (Fig. 1A). The results demonstrate that IL-6–induced FGG expression was further enhanced in the presence of dexamethasone whereas dexamethasone alone had no effect on FGG expression. To test whether these observations are also valid in human hepatoma cells, HepG2 cells were stimulated with IL-6 and dexamethasone. As observed in primary hepatocytes, IL-6–induced FGG mRNA was also elevated in the presence of glucocorticoids in HepG2 cells (Fig. 1B). Similar results were obtained for the induction of α1-ACT mRNA (Supporting Fig. 1). These results are indicative of cross-talk between glucocorticoid and IL-6 signaling in primary hepatocytes and human hepatoma cells.

Figure 1.

IL-6–induced FGG mRNA expression is enhanced by glucocorticoids. (A) Primary murine hepatocytes and (B) HepG2 cells were stimulated with IL-6 ± dexamethasone for 4 hours. FGG mRNA expression was analyzed using quantitative real-time PCR. Maximal induction of FGG mRNA was set at 100%. Data are given as mean of three independent experiments ± SD. Student t test for paired values: *, P ≤ 0.05; **, P ≤ 0.01.

Glucocorticoids Prolong IL-6–Induced STAT3 Activation.

It is well known that IL-6 is a strong activator of the Jak/STAT pathway and the acute-phase response. However, we were interested in determining whether the expression of FGG is STAT3 dependent. Therefore, an FGG promoter-luciferase reporter assay was employed. The reporter construct was transfected along with expression vectors for chimeric EpoR/gp130 fusion proteins as described in Materials and Methods. Chimeric receptors in which the tyrosine residues of the STAT3 recruiting motifs within gp130 have been replaced by phenylalanine allow one to analyze signaling without stimulating endogenous WT receptors. This experimental setup has been characterized in detail and is a widely used and accepted method.27, 28 Following transfection, HepG2 cells were stimulated with Epo and the FGG promoter activity was determined by measuring luciferase activity. Stimulation of cells expressing the mutated receptor did not lead to promoter activation whereas cells expressing wild-type chimeric receptors strongly activated the FGG promoter (Fig. 2A), demonstrating a crucial role for STAT3 in FGG promoter activation.

Figure 2.

Glucocorticoids enhance prolonged IL-6–induced STAT3 phosphorylation. (A) HepG2 cells were transfected with an FGG reporter construct, a β-galactosidase expression vector and chimeric EpoR/gp130 constructs. Five hours after transfection, cells were starved and subsequently stimulated with Epo (7 U/mL) for 4 hours. Maximal induction of the FGG promoter was set at 100%. Data are given as mean of three independent experiments ± SD. Student t test for paired values: ***, P ≤ 0.001. (B) Primary murine hepatocytes and (C) HepG2 cells were stimulated with IL-6 either in the absence or presence of dexamethasone for the indicated times. STAT3 phosphorylation and total STAT3 protein levels were evaluted by western blotting.

Based on these results, we investigated whether glucocorticoids affect STAT3 phosphorylation. Primary murine hepatocytes were stimulated for up to 5 hours with IL-6 in the presence or absence of dexamethasone. As shown in Fig. 2B, STAT3 is transiently activated in response to IL-6. However, dexamethasone-treated cells exhibited prolonged IL-6–dependent STAT3 activation, although STAT3 protein expression was not affected. For proof of principle, the same kinetics of IL-6–induced STAT3 activation were monitored in HepG2 cells. Once again, dexamethasone led to enhanced and prolonged STAT3 activation after IL-6 stimulation (Fig. 2C; see Supporting Fig. 2 for densitometric analysis). These results demonstrate that glucocorticoids positively affect STAT3 activation.

Glucocorticoids Inhibit IL-6–Induced SOCS3 Protein Expression

STAT3 activity in response to IL-6 is on the one hand positively regulated by Janus kinases12, 13 and on the other hand negatively regulated by SOCS3, a known feedback inhibitor for Janus kinases.14-16 To test the hypothesis that glucocorticoids modulate SOCS3 expression leading to sustained STAT3 activity, IL-6–induced SOCS3 protein expression in the absence and presence of dexamethasone in primary murine hepatocytes was examined. Figure 3A shows that glucocorticoids counteract IL-6–induced SOCS3 protein expression. Again, these results could be confirmed in HepG2 cells (Fig. 3B).

Figure 3.

Glucocorticoids dampen IL-6–induced SOCS3 protein expression. (A) Primary murine hepatocytes and (B) HepG2 cells were preincubated with dexamethasone for up to 30 minutes and then stimulated with IL-6 for 1 hour. SOCS3 and HSP70 protein expression were analyzed by western blotting.

Glucocorticoids Do Not Control SOCS3 Gene Induction.

Dexamethasone causes a reduction in the amount of SOCS3 protein following IL-6 stimulation (Fig. 3).

Next, the underlying mechanism involved in this process had to be elucidated. IL-1β induces internalization and degradation of the IL-6 receptor component gp13029; therefore, we tested whether dexamethasone treatment also affected SOCS3 expression by internalization of gp80 or gp130. In contrast to IL-1β, glucocorticoids do not repress cell surface expression of any of the receptor subunits of the IL-6 receptor complex in the presence of IL-6 (Fig. 4A,B). Thus, SOCS3 induction is not reduced due to repressed cell surface expression of the IL-6 receptor components.

Figure 4.

Glucocorticoids do not inhibit SOCS3 induction. HepG2 cells were stimulated with IL-6 ± dexamethasone for the indicated times. Expression of gp80 (A) and gp130 (B) was analyzed by FACS. Data are given as mean of three independent experiments ± SD. Student t test for paired values: *, P ≤ 0.05. (C) HepG2 cells were stimulated with IL-6 ± dexamethasone for the indicated times. Expression and phosphorylation state of STAT3 were analyzed by western blotting. (D) HepG2 cells were transfected with a STAT3-CFP expression vector. One day after transfection, cells were seeded on poly-L-lysine–coated coverslips, starved, and subsequently stimulated with IL-6 ± dexamethasone for the indicated times and fixed with PFA. Nuclear translocation of STAT3 was analyzed by laser scanning microscopy. Lower panels: HepG2 cells were stimulated with dexamethasone for 30 minutes and then permeabilized with methanol. Endogenous glucocorticoid receptor was stained with a specific glucocorticoid receptor antibody and detected with a Cy-2–coupled secondary antibody by laser scanning microscopy. (E) HepG2 cells were transfected with a SOCS3 reporter plasmid and a β-galactosidase expression vector. Five hours after transfection, cells were starved and subsequently stimulated with IL-6 ± dexamethasone for 4 hours. Maximal induction of SOCS3 promoter was set at 100%. Data are given as mean of three independent experiments ± SD. Student t test for paired values: ***, P ≤ 0.001. (F) HepG2 cells were treated with IL-6 ± dexamethasone for 4 hours. The expression of SOCS3 mRNA was analyzed using quantitative real-time PCR. Maximal induction of SOCS3 mRNA was set at 100%. Data are given as mean of three independent experiments ± SD. Student t test for paired values: **, P ≤ 0.01. (G) HepG2 cells were then pretreated with dexamethasone for 30 minutes and subsequently stimulated with IL-6 for 30 minutes. After termination of transcription by addition of actinomycin D (4 μM), the amount of SOCS3 mRNA was measured by quantitative real-time PCR. Maximal induction of SOCS3 mRNA was set at 100%. Data are given as mean of three independent experiments ± SD.

SOCS3 is very rapidly induced by IL-6 in a strong STAT3-dependent manner. Hence, it is possible to affect SOCS3 induction by affecting early STAT3 phosphorylation. IL-6–induced STAT3 activation in the presence of dexamethasone was revisited by focusing only on early time points. HepG2 cells were pretreated with dexamethasone for 30 minutes and subsequently stimulated with IL-6 for up to 60 minutes. STAT3 phosphorylation was not altered in the presence of dexamethasone (Fig. 4C). Thus, these results demonstrate that glucocorticoids do not down-regulate SOCS3 expression by counteracting early STAT3 activation.

Another possible mechanism for glucocorticoids to reduce SOCS3 expression is by inhibiting the nuclear translocation of STAT3. To test this hypothesis the cellular distribution of CFP-tagged STAT3 in response to IL-6 and dexamethasone was monitored by laser scanning microscopy. As expected, treatment with IL-6 induced the nuclear translocation of STAT3-CFP; however, this translocation was not influenced by the presence of dexamethasone. Instead, dexamethasone induced nuclear translocation of the glucocorticoid receptor (Fig. 4D). These results could be further confirmed with biochemical analyses (Supporting Fig. 3) and indicate that glucocorticoids do not affect the nuclear translocation of STAT3 to prevent SOCS3 expression.

Alternatively, glucocorticoids could act on SOCS3 promoter activity, directly or by affecting promoter-bound transcription factors. Therefore, we investigated whether dexamethasone has any influence on the activity of the SOCS3 promoter. To test this possibility a SOCS3 promoter luciferase reporter construct was transfected into HepG2 cells, and IL-6–initiated promoter activity was determined in the presence or absence of dexamethasone. Figure 4E demonstrates that glucocorticoids do not affect SOCS3 promoter activity. Furthermore, dexamethasone does not reduce total SOCS3 mRNA expression (Fig. 4F). The results from these data lead us to conclude that the effects of glucocorticoids on promoters of IL-6 target genes are strongly gene specific and the SOCS3 promoter is not a target of glucocorticoid action, in contrast to the α2M promoter, which has been demonstrated to be synergistically activated by IL-6 and glucocorticoids.6

Glucocorticoids Do Not Affect SOCS3 mRNA Stability.

Because activation of the SOCS3 promoter is not affected by dexamethasone, we tested whether SOCS3 mRNA turnover is accelerated by glucocorticoids. Therefore, the decay of IL-6–induced SOCS3 mRNA in the presence or absence of dexamethasone was compared. HepG2 cells were stimulated with IL-6 to induce SOCS3 mRNA expression. Transcription was terminated by addition of actinomycin D and elimination of IL-6. SOCS3 mRNA was quantified by real-time PCR. Figure 4G shows that SOCS3 mRNA is characterized by a half-life of about 30 minutes, which is not altered by the presence of dexamethasone. Accordingly, glucocorticoids do not accelerate SOCS3 mRNA turnover to reduce the levels of SOCS3 protein expression.

The investigation was furthered by examining whether glucocorticoids could initiate SOCS3 protein degradation. The decay of SOCS3 protein in the presence or absence of dexamethasone was compared but no altered decay was observed in the presence of dexamethasone (Supporting Fig. 4).

Glucocorticoids Interfere with FGG mRNA Expression in a Transcriptional Manner.

Because dexamethasone activates the transcriptional activity of the glucocorticoid receptor, we tested whether the regulation of SOCS3 expression is affected by the transcriptional activity of the glucocorticoid/glucocorticoid receptor complex. IL-6–induced FGG mRNA expression in the presence of the glucocorticoid receptor antagonist RU-486 was monitored. RU-486 blocks the effect of dexamethasone on acute-phase expression (Fig. 5), indicating that transcriptional activity of the glucocorticoid receptor is mandatory for the regulation of FGG expression by glucocorticoids.

Figure 5.

Transcriptional activity of the glucocorticoid receptor is essential for the regulation of IL-6–induced SOCS3 expression. Primary murine hepatocytes were treated with IL-6 ± dexamethasone or RU-486 and analyzed as described in Fig. 1.

Glucocorticoids Act on IL-6–Induced Gene Expression Through Reduction of SOCS3.

The reduction of SOCS3 expression by dexamethasone may explain how glucocorticoids enhance acute-phase gene expression in the liver. To prove this hypothesis, the impact of dexamethasone in two cellular systems that are resistant to the inhibitory function of SOCS3 was tested. These approaches consisted of analyzing the effects in (1) SOCS3-deficient MEFs and (2) hepatocytes isolated from gp130Y757F mice.

Because SOCS3 deficiency is lethal, we worked with MEFs derived from SOCS3 knockout embryos; first it was determined that this approach represented a good experimental system, as shown in Supporting Fig. 5. Next, we tested whether dexamethasone affects the late IL-6–induced STAT3 activation in wild-type MEFs and in SOCS3-deficient MEFs. In WT MEFs, treatment with dexamethasone enhanced the late STAT3 activation in accordance with the results achieved in primary hepatocytes and HepG2 cells (Fig. 6A). In contrast, treatment of SOCS3-deficient MEFs did not alter the kinetics of IL-6–induced STAT3 phosphorylation (Fig. 6B), highlighting the crucial role of SOCS3 in glucocorticoid-initiated augmented STAT3 activation.

Figure 6.

The presence of SOCS3 is crucial for glucocorticoids to act on STAT3 phosphorylation. (A) WT MEF and (B) SOCS3−/− MEF cells were stimulated with IL-6 and soluble IL-6R ± dexamethasone for the indicated times. Total and phosphorylated STAT3 were analyzed by western blotting.

Phosphorylation of STAT3 was increased in both the absence and presence of dexamethasone in SOCS3-deficient MEFs compared with WT MEFs due to the absence of SOCS3-dependent feedback inhibition. However, IL-6–induced STAT3 activation can be further enhanced in SOCS3 knockout cells with higher doses of IL-6, indicating that signaling amplitude is not saturated (data not shown).

In a second approach, hepatocytes isolated from mice that express a mutant form of the IL-6 coreceptor gp130 (Y757F) (which lack the SOCS3 recruiting motif and are thus resistant to SOCS3-dependent feedback inhibition) were analyzed.30 Similar to what was observed in WT primary hepatocytes (Fig. 3A), SOCS3 expression was also lessened in primary hepatocytes from mice expressing the mutated receptor (Fig. 7A). Nevertheless, the reduced SOCS3 protein expression did not result in enhanced IL-6–induced STAT3 activation as shown for WT hepatocytes (compare Fig. 7B with Fig. 2B).

Figure 7.

The presence of the SOCS3-recruitment site in gp130 is crucial for glucocorticoids to affect the IL-6–induced STAT3 phosphorylation and FGG mRNA expression. Primary murine hepatocytes from mice expressing gp130 (Y757F) were starved and subsequently stimulated with IL-6 ± dexamethasone for the indicated times. (A) SOCS3 and HSP70 as well as (B) total and phosphorylated STAT3 were analyzed by western blotting. (C) Expression of FGG mRNA was analyzed by quantitative real-time PCR. Maximal induction of FGG mRNA was set at 100%. Data are given as mean of three independent experiments ± SD. Student t test for paired values: ***, P ≤ 0.001.

The study clearly demonstrates that glucocorticoids enhance IL-6–induced STAT3 activation by repressing the SOCS3-mediated negative feedback. If this theory is correct, then APP expression should not be enhanced in mice bearing the mutated IL-6 coreceptor gp130. To verify this conclusion experimentally, hepatocytes from mutant gp130Y757F mice were stimulated with IL-6 in either the absence or presence of dexamethasone, and FGG mRNA expression was analyzed. As expected, FGG mRNA expression levels were not affected by dexamethasone in SOCS3-resistant hepatocytes (Fig. 7C). Similarly, the effect of dexamethasone on IL-6–induced acute-phase protein expression was also tested in vivo. Figure 8A and B confirms that dexamethasone increases IL-6–induced hepatic SAA mRNA expression and serum SAA protein of treated mice. In contrast, dexamethasone did not affect SAA expression in mice lacking the SOCS3-recruiting tyrosine 757 within gp130 (Fig. 8C,D). Treatment of the mice with dexamethasone had no striking effect on APP expression in comparison with IL-6 alone (Supporting Fig. 6).

Figure 8.

IL-6–induced hepatic gene expression is enhanced by glucocorticoids in vivo. (A) RNA was isolated from whole liver extracts from wt mice after the indicated treatment. Expression of SAA mRNA was analyzed using quantitative real-time PCR. Data are given as mean of n = 3-4 animals per group ± SD. (B) Serum SAA concentration was measured by ELISA. (C) SAA mRNA expression from mice expressing a mutant form of gp130 (Y757F) was measured by qRT-PCR as described in (A). (D) Serum SAA concentration was measured by ELISA as in (B). Data are given as mean of n = 3-4 animals per group ± SD. Student t test for unpaired values: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

In summary, the influence of dexamethasone on IL-6–induced acute-phase gene expression in the liver depends on SOCS3 expression and the SOCS3 recruiting motif within the gp130 receptor.

Discussion

A stimulating effect of glucocorticoids on IL-6–induced acute-phase gene expression was described by Baumann's group in the late 1980s.22, 23 First analyses of the underlying mechanism revealed a promoter-specific action of glucocorticoids downstream of STAT3 activation.4 Lerner et al. described in detail how glucocorticoids act on the α2M promoter.6 The authors presented compelling evidence showing that the glucocorticoid receptor assembles with STAT3 to form an enhanceosome at this promoter. The present study confirms enhanced activity of the α2M promoter in the presence of dexamethasone in HepG2 cells (Supporting Fig. 7A). The reduction of SOCS3 shown here should affect all STAT3-dependent genes, therefore also influencing the activity of the α2M promoter. Consequently, the glucocorticoid effect was reduced in the presence of a receptor mutant lacking the SOCS3 recruiting motif (Supporting Fig. 7B). Thus, our data extend the results of Lerner et al.6 by presenting a more general spectrum of glucocorticoid activity.

Long-term treatment (18 hours) with dexamethasone increases FGG expression in HepG2 cells stimulated with a supramaximal concentration of IL-6 by up-regulating IL-6Rα,31, 32 indicating that IL-6Rα is the limiting factor. In contrast, the study presented here analyses short-term effects of dexamethasone on STAT3 activation and early acute-phase gene expression in the presence of lower IL-6 concentrations. Here, no change in IL-6R expression could be detected.

In addition, Paul et al. have shown an inhibition of IL-6 and growth hormone-dependent SOCS3 mRNA expression in rat hepatocytes by glucocorticoids.33 As there are no glucocorticoid binding motifs in the rat SOCS3 promoter, they suggested that glucocorticoid receptors interact with other transcription factors to inhibit the expression of SOCS3 mRNA. This is interesting because, although we observe a reduction in both human and mouse SOCS3 protein expression, neither SOCS3 promoter activation nor total SOCS3 mRNA or its degradation was influenced by dexamethasone, indicating a species-dependent mechanism. However, the fact that two distinct mechanisms lead to comparable outcomes in different species emphasises the importance of SOCS3 down-regulation by glucocorticoids.

In summary, this study demonstrates that glucocorticoids enhance IL-6–induced APP expression by reducing SOCS3 protein levels in a transcriptional manner. The effect of glucocorticoids on other branches of IL-6 signaling pathways remains to be examined. Yasukawa et al. have shown that IL-6 causes a proinflammatory effect in WT macrophages, whereas macrophages lacking SOCS3 exert an IL-10–like anti-inflammatory response following IL-6 stimulation.34 In contrast, the expression of acute-phase proteins in hepatocytes leads to a proinflammatory environment that is further enhanced in the absence of the negative feedback regulator SOCS3. Because glucocorticoids are widely prescribed for the treatment of inflammatory diseases, our data encourage further studies to decipher the role of glucocorticoid action in the body.

Acknowledgements

We thank B. Hilger and S. Fiedler (Roche, Mannheim, Germany) for the generous gift of recombinant erythropoietin (Epo), W. Doppler (Innsbruck) for the supply of GCR-cDNA, Maria Luisa Tenchini (Milan) for the supply of FGG reporter constructs, and Sibille Sauer-Lehnen and Carmen G. Tag for technical assistance.

Ancillary