of April 8, 2018.This information is current as

NF-AT/Activator Protein-1 ElementsReceptor Binding to CompositeEnhancer Function by the Glucocorticoid Glucocorticoids Involves Inhibition ofColony-Stimulating Factor Expression by Suppression of Granulocyte-Macrophage

Z. Staynov, Tak H. Lee and Paul LavenderPhilip J. Smith, David J. Cousins, Young-Koo Jee, Dontcho

http://www.jimmunol.org/content/167/5/2502doi: 10.4049/jimmunol.167.5.2502

2001; 167:2502-2510; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/167/5/2502.full#ref-list-1

, 32 of which you can access for free at: cites 74 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2001 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Suppression of Granulocyte-Macrophage Colony-StimulatingFactor Expression by Glucocorticoids Involves Inhibition ofEnhancer Function by the Glucocorticoid Receptor Binding toComposite NF-AT/Activator Protein-1 Elements1

Philip J. Smith,* David J. Cousins,* Young-Koo Jee,*† Dontcho Z. Staynov,* Tak H. Lee,* andPaul Lavender2*

Increased expression of a number of cytokines including GM-CSF is associated with chronic inflammatory conditions such asbronchial asthma. Glucocorticoid therapy results in suppression of cytokine levels by a mechanism(s) not yet fully understood. Wehave examined regulation of GM-CSF expression by the synthetic glucocorticoid dexamethasone in human T cells. Transienttransfection assays with reporter constructs revealed that dexamethasone inhibited the function of theGM-CSF enhancer, but hadno effect on regulation of GM-CSF expression occurring through the proximal promoter. Activation of theGM-CSF enhancerinvolves cooperative interaction between the transcription factors NF-AT and AP-1. We demonstrate here that glucocorticoid-mediated inhibition of enhancer function involves glucocorticoid receptor (GR) binding to the NF-AT/AP-1 sites. These elements,which do not constitute recognizable glucocorticoid response elements, support binding of the GR, primarily as a dimer. Thisbinding correlates with the ability of dexamethasone to inhibit enhancer activity of the NF-AT/AP-1 elements, suggesting acompetition between NF-AT/AP-1 proteins and GR. The Journal of Immunology, 2001, 167: 2502–2510.

G ranulocyte-macrophage CSF is a member of a family ofsoluble glycoprotein growth factors known as the CSFs(1). It is produced by T cells, monocyte-macrophages,

mast cells, fibroblasts, and endothelial and epithelial cells and af-fects both hemopoietic precursors and differentiated macrophagesand granulocytes (2, 3). Overexpression of GM-CSF is a feature ofallergic conditions such as bronchial asthma (4), and elevated lev-els of GM-CSF are found in bronchoalveolar lavage samples andbronchial biopsies taken from asthmatic subjects (5, 6). Glucocor-ticoids are a potent therapy for a wide range of inflammatory dis-eases and are the mainstay of treatment for chronic asthma (7).However, their mechanism of action in asthma remains largelyunknown. Treatment with glucocorticoids leads to suppression ofthe cellular infiltration of asthmatic airways (8, 9) and to suppres-sion of proinflammatory cytokines including GM-CSF (10–12).

Transcriptional regulation of GM-CSF in T cells involves boththe proximal promoter region and the distal enhancer (2, 13, 14).A number of regulatory elements have been defined within theGM-CSF proximal promoter. These include the conserved lym-phokine element (CLE)30, which is the most proximal to the tran-

scription start site and is bound by the transcription factors NF-AT,AP-1, Ets1, and Elf1 (15–17). Further upstream are the CLE1 andthe CLE2/GC box, which can bind NF-�B and Sp1 (18, 19). Trans-activation of the GM-CSF promoter occurs through the synergisticaction of Ets1, NF-�B, and AP-1 (20). Additionally, we have pre-viously reported a double palindromic regulatory element withinthe GM-CSF promoter that acts as an enhancer (21).

The human GM-CSF enhancer region is a 716-bp fragment lo-cated 7 kb downstream of the IL-3 gene and 2.6 kb upstream of theGM-CSF gene (22). Transfections with a construct containing theenhancer linked to the proximal promoter demonstrated the potenteffects on GM-CSF transcription that this region confers (22). Theenhancer was also found to up-regulate IL-3 expression (22). Fur-ther investigation led to the discovery of a T cell-specific enhancer14 kb upstream of the IL-3 gene, and it is this element, rather thanthe GM-CSF enhancer, that is believed to regulate IL-3 expressionin vivo (23). Cockerill et al. (24) have recently shown that thehuman enhancer is required for correctly regulated GM-CSF ex-pression in vivo.

The GM-CSF enhancer contains four composite NF-AT/AP-1elements, three of which demonstrate cooperative binding of re-combinant NF-ATp and Fos/Jun. The fourth site binds NF-ATpand Fos/Jun independently. In the three sites that demonstrate co-operative binding function as enhancers (25), both NF-AT andAP-1 components differ in their affinity for the individual elementsas determined by band retardation assays. It has been hypothesizedthat the spacing of the NF-AT and AP-1 sites is important forcooperative binding and enhancer function. Recent data has sug-gested that transcriptional repression of GM-CSF by 1,25-dihy-droxyvitamin D3 (1,25(OH)2D3) occurs through the inhibition of

*Department of Respiratory Medicine and Allergy, Kings College London, Guy’sHospital, London, United Kingdom; and †Division of Pulmonary Medicine and Al-lergy, Dankook University Medical Centre, Chonan, Korea

Received for publication June 12, 2000. Accepted for publication June 20, 2001.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported by Grants G9536930 from the Medical Research Counciland 181 from the National Asthma Campaign.2 Address correspondence and reprint requests to Dr. Paul Lavender, Departmentof Respiratory Medicine and Allergy, Kings College London, 5th Floor ThomasGuy House, Guy’s Hospital, London, SE1 9RT U.K. E-mail address:paul.lavender@kcl.ac.uk3 Abbreviations used in this paper: CLE, conserved lymphokine element;1,25(OH)2D3, 1,25-dihydroxyvitamin D3; VDR, vitamin D3 receptor; GR, glucocor-

ticoid receptor; GRE, glucocorticoid response element; PDBu, phorbol dibutyrate;Ion, ionomycin; CAT, chloramphenicol acetyltransferase; MMTV, mouse mammarytumor virus; LTR, long terminal repeat; NR, nuclear receptor; AF-2, activation func-tion 2; TR, thyroid hormone receptor; RAR, retinoic acid receptor; HDAC, histonedeacetylase; RSV, Rolls sarcoma virus.

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

enhancer function (26). Vitamin D3 receptor (VDR) competes withNF-AT1 (NF-ATc) for binding to a composite vitamin D responseelement/NF-AT1 binding site and stabilizes the binding of a Jun-Fos heterodimer to an adjacent AP-1 site (27).

Glucocorticoids exhibit their numerous functions through bind-ing to a specific cytoplasmic receptor. Following binding of ste-roid, the glucocorticoid receptor (GR) translocates to the nucleusof the cell where it is involved in both the activation and repressionof a number of target genes (28). Positive gene regulation by glu-cocorticoids occurs by binding of the GR to glucocorticoid re-sponse elements (GREs) in the promoter regions of glucocorticoid-responsive genes (29). Unlike the mechanism for positive generegulation, glucocorticoids, acting through the GR, repress geneactivation by heterogeneous mechanisms. One mechanism is thatof protein-protein sequestration of transcription factors either di-rectly, as in the case of AP-1, or indirectly through the inductionof inhibitory proteins, as in the case of I�B suppressing NF-�Bactivity (30, 31). Other mechanisms of transcriptional repressionby the GR include binding to negative GREs, as in the case of thepro-opiomelanocortin gene (32); preventing access to nearby pos-itive regulatory elements, as in the prolactin promoter (33), orcompetition for transcriptional coactivators such as CREB bindingprotein (34). Most recently, the GR RNA cofactor steroid receptorRNA activator has been demonstrated to recruit a corepressor,SMRT/HDAC1-associated repressor protein, which itself associ-ates with histone deacetylases. This data provides direct evidencefor the capacity of ligand-bound GR to recruit corepressor com-plexes to promoters (35).

In view of the importance of elucidating the mechanism of glu-cocorticoid-mediated inhibition of transcription of cytokines asso-ciated with asthma, we have studied the repression of GM-CSF byglucocorticoids. The ability of the synthetic glucocorticoid dexa-methasone to influence the function of the known transcriptionalregulatory sequences of the GM-CSF gene has been examined. Wedemonstrate that suppression of GM-CSF expression by steroids ismediated in part through inhibition of enhancer function. Addi-tionally, we propose that this inhibition involves binding of GR tocomposite NF-AT/AP-1 sites within the enhancer.

Materials and MethodsCells and culture conditions

The human T cell lines Jurkat and HUT 78 were grown in RPMI 1640 (LifeTechnologies, Gaithersburg, MD) medium and A293 cells in DMEM, bothsupplemented with 10% FCS (Sigma, St. Louis, MO), L-glutamine (2 mM),penicillin (100 IU/ml), and streptomycin (100 �g/ml), at 37°C, 5% CO2 inhumidified air. Where indicated, cells were activated with 100 ng/ml phor-bol dibutyrate (PDBu) and 1 �g/ml ionomycin (Ion) (Calbiochem, La Jolla,CA). Dexamethasone (Sigma) was stored at a concentration of 10�2 M inethanol, then further diluted in RPMI 1640 and was added to cells to givethe relevant final concentration.

RT-PCR

Total cellular RNA was isolated by the guanidinium isothiocyanate method(36). Five micrograms of RNA per sample was reverse transcribed with anoligo(dT) primer using Pharmacia You Prime reverse transcription tubes(Pharmacia, Uppsala, Sweden) according to the manufacturers protocol.GM-CSF levels were measured by semiquantitative PCR using an estab-lished method (37). PCR conditions were an initial 95°C for 10 min, fol-lowed by cycles of 95°C for 2 min, 60°C for 10 min, 72°C for 2 min forfour cycles, followed by 18 cycles of 95°C for 2 min, 60°C for 2 min, 72°Cfor 2 min. PCR for �-actin was conducted as an internal control (18 cycles).Primer sequences (5�–3�) were as follows: GM-CSF sense, CTAAAGTTCTCTGGAGGATGTGG; antisense, TTCTACTGTTTCATTCATCTCAGC;�-actin sense, CACCACACCTTCTACAATGAGCTGC; antisense,ACAGCCTGGATAGCAACGTACATGG. PCR products were analyzed byelectrophoresis on 2.8% agarose gels run in glycine buffer (200 mM glycine,15 mM NaOH, 2 mM Na3EDTA).

Plasmids

pHGM617 contains the first 617 bp of the GM-CSF proximal promoter linkedto the chloramphenicol acetyl transferase (CAT) gene (21). pHGMB716 con-tains the GM-CSF enhancer region, �2.6 to �3.3 kb, in addition to 617 bp ofthe promoter (22). To generate pGM716E1b, the GM-CSF enhancer regionwas obtained by digesting pHGMB716 with HindIII and XhoI, which was thenligated into HindIII/XhoI-digested pE1bCAT. pGM170, pGM330, pGM420,and pGM550 contain the respective NF-AT/AP-1 sites from the GM-CSFenhancer (25). Vectors were constructed by digesting pHGMB716 with BglIIto remove the enhancer, which was then replaced by ligation of oligonucleo-tides corresponding to each of the four sites into the reporter plasmid. Con-structs containing head-to-tail dimers linked to 617 bp of the GM-CSF pro-moter were verified by sequence analysis using an Applied Biosystems 377automated sequencer (Applied Biosystems, Foster City, CA). Oligonucleotideduplexes used for reporter plasmid construction and as probes and competitorsin band retardation assays had the following sequences. The upper strand se-quence is given, with complementary single-stranded regions used for cloningshown in lower case: GM170, gatcCTGGAGTGACTCAAGCCCCTGTTTCCTACAG; GM330, gatcGCCCCATCGGAGCCCCTGAGTCAGCATGGCT; GM420, gatcCCCTGATGTCATCTTTCCATGAGAAAGATGT; GM550, gatcTCTTATTATGACTCTTGCTTTCCTCCTTTCC.

pcDNA3.Flag GR was generated by PCR amplification of full-lengthhuman GR with oligonucleotides caggatccATGGACTCCAAAGAATCATTAACTCC (sense) and caggatccTCATTGATGAAACAGAAGTTTTTTGATATTTCC (antisense), ligating the resultant cDNA down-stream of the Flag epitope in the vector pcDNA.Flag2. Expression vectorsused in cotransfections were pRSV.GR (32), pRSV.c-Fos, pRSV.c-Jun(38), and pRSV.NF-ATc (39).

Transfections

Transfections and CAT assays were conducted as previously described(21). Briefly, 4 � 106 cells were transfected with 10 �g of reporter plasmidDNA plus or minus 250 ng to 1 �g of expression plasmid DNA. Electro-poration was conducted at 300 mV, 960 �F, � �, with a Gene Pulser(Bio-Rad, Hercules, CA). Samples were activated and treated with dexa-methasone 10 min posttransfection as indicated. Duplicate samples foreach stimulus were conducted. Cells were incubated at 37°C, 5% CO2, inhumidified air for 20 h, harvested by centrifugation, and cell lysates wereassayed for CAT activity. As a control for transfection efficiency and theeffects of steroid treatment, samples were cotransfected with 2 �g ofpCMV-�-gal, and �-galactosidase assays were performed using standardprocedures (40). A293 cells were transfected by calcium phosphate pre-cipitation as previously described (41).

Immunopurification of Flag-tagged GR

Flag-tagged GR was prepared from A293 cells transfected with thecDNA3.Flag GR vector according to the protocol described (42). Cellswere lysed in IPH buffer (50 mM Tris�HCl (pH 8.0), 150 mM NaCl, 5 mMEDTA, 0.5% NP40, 1 mM PMSF), and debris was pelleted. Supernatantswere incubated with Flag agarose for 30 min at 4°C before centrifugationand repeated washing in IPH buffer. Flag protein was eluted from beadsusing acid glycine according to the manufacturer’s protocol and neutralizedin 10 mM Tris�Cl, pH 8. Eluted protein was stored at �20°C. Western blotswere conducted as previously described (41) and probed with anti-Flag(Kodak, Rochester, NY) or anti-GR antisera (Transduction Laboratories,Lexington, KY).

Band retardation assays

Recombinant human GR was obtained from Affinity BioReagents (Golden,CO) (43, 44). GR was diluted 1:10 in KED buffer (10 mM K2SO4 pH 7.6,0.1 mM EDTA, 5 mM DTT, 0.2 mM PMSF, 15 �g/ml leupeptin) beforeuse. Jurkat nuclear extract, prepared as previously described (21), and FlagGR protein, were used in band retardation assays. Oligonucleotide probeswere labeled with [�-32P]ATP with T4 polynucleotide kinase according tostandard procedures (40). Then 0.5 �l of diluted GR, or 5 �g Jurkat nuclearextract, was incubated with 7.6 fmol 32P end-labeled oligonucleotide du-plex, plus 80 ng poly(dI�dC) (500-fold excess) in binding buffer (10 mMTris�HCl (pH 7.5), 1 mM MgCl2, 0.5 mM Na3EDTA, 50 mM NaCl, 0.5mM DTT, 4% glycerol) for 15 min at room temperature. Where indicated,152 fmol (20-fold excess) unlabeled specific competitor DNA was added tothe binding reaction. Complexes were resolved on 5% polyacrylamide gelsrun in 0.3� Tris/borate/EDTA buffer (40). The GRE corresponding to�187 bp to �161 bp from the mouse mammary tumor virus (MMTV) longterminal repeat (LTR) was used as a positive control for GR binding andhas the following sequence (5�–3�): GATCGTTTATGGTTACAAACTGTTCTTAAAACA (45). GM-CSF oligonucleotides used only in EMSAs

2503The Journal of Immunology

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

were: 550(-AP-1), gatcTCTTATGCCCCGCGCAGCTTTCCTCCTTTCA;550(-NF-AT) gatcTCTTATTATGACTCTTGCGCTAGTCCTTTCA; 550.5�,gatcTCTTATTATGACTCTT; 550.3�, gatcGCTTTCCTCCTTTCA; and550.3�(-NF-AT), gatcGCTTTAAGAAT TTGA.

ResultsGM-CSF mRNA transcription is repressed by glucocorticoids

The effects of the synthetic glucocorticoid dexamethasone on GM-CSF expression from two human T cell lines were investigated.HUT 78 and Jurkat cells were stimulated with PDBu and Ion for20 h in the presence of various concentrations of dexamethasone.Dexamethasone treatment inhibited GM-CSF expression in a dose-responsive manner as determined by RT-PCR. �-actin levels wereunaffected by this treatment (Fig. 1, A and B). In both cell types,treatment with 10�6 M dexamethasone caused a �80% repressionof transcription as compared with treatment with PDBu/Ion alone.

Inhibition of GM-CSF expression by dexamethasone is mediatedby the enhancer region

To address whether the inhibitory effects of dexamethasone aremediated through the proximal promoter, HUT 78 and Jurkat cellswere transfected with the reporter construct pHGM617, whichcontains the proximal 617 bp of the GM-CSF promoter drivingexpression of the CAT gene. Cells were activated 10 min post-transfection in the presence or absence of 10�6 M dexamethasoneand were harvested after 20 h. Cell extracts analyzed for CATactivity. Dexamethasone treatment had no effect on expression ofthe reporter gene in either HUT 78 (Fig. 2A) or Jurkat cells (Fig.2B). Furthermore, overexpression of GR by cotransfection with theGR expression vector pRSV.GR in activated, dexamethasone-treated cells had no repressive effect (Fig. 2B).

Transient transfection analysis has previously shown that in ad-dition to the proximal promoter, transcriptional regulation is in-fluenced by an enhancer region located 3 kb upstream of the tran-scription start site. We tested the steroid sensitivity of the enhancerusing the reporter plasmid pHGMB716 that contains the enhancerdirectly upstream of the proximal promoter. Dexamethasone treat-ment resulted in a 38.4% inhibition and a 23.9% inhibition of CATactivity in HUT 78 and Jurkat cells, respectively (Fig. 2, A and B),suggesting that the 716-bp enhancer region is capable of mediatingsteroid responsiveness. Suppression is GR dependent, as increas-ing GR levels by cotransfection of pRSV.GR increased the level of

FIGURE 1. Dexamethasone (Dex) inhibits GM-CSF transcription inT cell lines. HUT 78 (A) and Jurkat T cells (B) were stimulated withPDBu � Ion, in the presence of the indicated concentrations of dexameth-asone, for 20 h. RNA was isolated, RT-PCR was performed, and productswere migrated on agarose gels. Band intensity was quantitated usingImageQuant software. Data are expressed as the percentage of GM-CSF:�-actin ratio in cells treated with PDBu/Ion in the absence of dexametha-sone and represent the average of two independent experiments.

FIGURE 2. Dexamethasone (Dex)-mediated inhibition of GM-CSF ex-pression involves the enhancer region. The HUT 78 (A) or Jurkat (B) T cellline were transfected with 10 �g pHGM617 or pHGMB716, plus 1 �gpRSV.GR or pRSV control (for Jurkat cells), stimulated with PDBu andIon with or without 10�6 M dexamethasone for 20 h. Cell extracts wereassayed for CAT activity. The results have been normalized to a value of100 for the activated sample and in each case are the mean of three inde-pendent experiments. Error bars indicate SE.

2504 INHIBITION OF GM-CSF BY GLUCOCORTICOIDS

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

reporter gene suppression to 42.4% (Fig. 2B). Although it is notstraightforward to draw direct comparisons between the two ex-perimental systems, the demonstration that expression of GM-CSFmRNA is almost fully repressible at a dose of 10�6 M dexameth-asone (Fig. 1, A and B) suggests that there may be other steroid-responsive regions within the GM-CSF gene. Sequences within the3� untranslated region have been shown to be involved in the reg-ulation of GM-CSF in a number of cell types (2), and there isevidence that in human fibroblasts dexamethasone may down-reg-ulate GM-CSF expression, in part, through decreasing mRNA sta-bility (46). The inhibitory effects of dexamethasone were specificto the GM-CSF enhancer, as similar levels of �-galactosidase ac-tivity were observed between the activated and steroid-treated cellswhen samples were cotransfected with the pCMV-�-gal reporterconstruct (data not shown).

Activation of the GM-CSF enhancer by NF-AT/AP-1 issuppressed by GR

To confirm that the enhancer region confers glucocorticoid respon-siveness, a plasmid was constructed in which the 716-bp enhancerdrives expression from the minimal viral promoter E1b. The con-struct had minimal activity when assayed by transient transfectionin Jurkat cells (Fig. 3A). Cotransfection of expression vectors en-coding AP-1, or NF-AT components (RSV.c-Fos � RSV.c-Jun,and RSV.NF-ATc), resulted in activation of the enhancer in adose-responsive, synergistic manner (Fig. 3A). Activation of theenhancer by AP-1/NF-ATc can be suppressed in part by dexameth-asone treatment (24.1% inhibition; Fig. 3B). Increasing the level ofGR within the cell by cotransfection with pRSV.GR resulted inincreased suppression of pGM716.E1b activity (44.5% inhibition;Fig. 3B). The level of suppression was of the same magnitude asthat observed when the enhancer was in context of the endogenousGM-CSF promoter (Fig. 2B), suggesting that the enhancer containsboth steroid-responsive and unresponsive regions and was doseresponsive to pRSV.GR levels (data not shown). Suppression wasligand dependent as cells cotransfected with pRSV.GR gave thesame level of CAT activity as those cotransfected with the controlvector pRSV (Fig. 3B). Steroid treatment had no effect on CATactivity of the parental pE1b vector (data not shown).

The GM-CSF enhancer NF-AT/AP-1 elements differ in theirability to enhance transcription and in their steroidresponsiveness

Having demonstrated that the enhancer imparts steroid responsive-ness upon an exogenous promoter, experiments were conducted todissect the responsive elements within this 716-bp fragment. Totest the hypothesis that GR might act, at least in part, by interferingwith NF-AT/AP-1 components, vectors were generated in whichthe individual 30-bp enhancer NF-AT/AP-1 elements were clonedupstream of the GM-CSF promoter as head-to-tail dimers. Trans-fection assays demonstrated the differential abilities of the indi-vidual elements to act as enhancers (Fig. 4). Element GM420 wasthe most potent enhancer, followed by GM330 and GM550. Ele-ment GM170 did not enhance the activity of the GM-CSF proxi-mal promoter. Our data are in accordance with the published ob-servations on the ability of the individual elements to act asenhancers (25). The steroid responsiveness of each of the enhancerelements was tested. Both GM550 and GM330 were found to besteroid responsive, 47.5 and 37.2% inhibition, respectively, insamples cotransfected with pRSV.GR (Fig. 4). GM170 andGM420 were both unresponsive to steroid treatment, as was thecontrol construct pHGM that contains only the proximal promoter.

GR binds to all four of the NF-AT/AP-1 elements within theGM-CSF enhancer

The observation that dexamethasone-mediated suppression ofGM-CSF occurs in part through the enhancer region led us toanalyze the mechanism of this suppression. The ability of the GRto bind to the NF-AT/AP-1 elements within the GM-CSF enhancerwas investigated using band retardation assays. None of the ele-ments contain recognizable GREs. Synthetic oligonucleotide du-plexes corresponding to the four elements were synthesized andassessed for the ability of GR to bind. The GRE from the MMTVLTR (�187 to �161 bp) was used as a positive control for GRbinding and as a specific competitor for GR binding to the en-hancer elements. Three of the four NF-AT/AP-1 elements sup-ported binding of recombinant GR (Fig. 5A), with very weak bind-ing observed to the fourth element, GM420.

To further characterize the GR complexes bound to the GM-CSF elements, we generated Flag-tagged GR protein by transfec-tion of A293 cells with a CMV.Flag GR vector, and we immuno-purified the Flag.GR protein (Fig. 5B). By a number of functional

FIGURE 3. Activation of the GM-CSF enhancer by NF-AT/AP-1 issuppressed by the GR. A, Jurkat T cells were cotransfected with 10 �gpGM716E1b and increasing amounts of each of pRSV.c-Fos, pRSV.c-Jun,and pRSV.NF-ATc and activated where indicated. The results are repre-sentative of two experiments. B, Jurkat cells were cotransfected with 10 �gof pGM716E1b, plus 500 ng each of pRSV.c-Fos, pRSV.c-Jun, andpRSV.NF-ATc, plus 1 �g pRSV.GR or pRSV. Samples were activatedwith or without dexamethasone (Dex; 10�6 M). The results have beennormalized to a value of 100 for the activated sample and are the mean offour experiments with error bars representing SE.

2505The Journal of Immunology

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

and biochemical assays this protein behaves as wild-type GR (datanot shown). We compared this material with Jurkat nuclear extractby EMSA using both consensus GRE and the GM550 sequence.Nuclear extract and Flag GR generated complexes of identical mo-bility with both oligonucleotides, and both protein sources behavedidentically in cross-competition assays (Fig. 5C). We take thesedata as strong evidence that the complex causing retardation of theGRE and GM550 oligonucleotides in the Jurkat extract is GR.

To compare the ability of each of the NF-AT/AP-1 elements tosupport GR binding, a series of band retardation assays were per-formed. The NF-AT/AP-1 elements were labeled and assessed fortheir ability to bind GR, and the unlabeled elements were used ascompetitors for GR binding (Fig. 6). Strongest competition in eachcase was provided by consensus GRE and GM550, which abol-ished GR binding to all of the elements (Fig. 6, A and B). GM550supported the strongest binding of GR, and GM420 supported theweakest (Fig. 6B). Using cumulative competition and binding data,the relative binding affinity of GR for each element was measured.Detailed analysis of the relative GR binding affinities was per-formed on a Molecular Dynamics Phosphorimager employing Im-ageQuant software (Molecular Dynamics, Sunnyvale, CA). Thisanalysis allowed the relative affinity of GR for each element to becalculated by plotting the percentage of probe bound against thecold competitor used (data not shown). This analysis confirmedthat GM550 supported the strongest binding of GR. GR had ap-proximately the same affinity for GM170 and GM330 (Fig. 6A),although lower than for GM550. GR binding to GM420 was veryweak in comparison to the other elements. The ability of GM550to support GR binding was almost as great as that of the GRE(88.3 � 3.1% the affinity of GR for GRE).

GR binds independently to both the NF-AT and AP-1 siteswithin the GM550 element

Because none of the four NFAT/AP1 elements contained consen-sus GREs, we chose GM550 to further investigate the site of GRbinding. Oligonucleotides in which either the NF-AT (550(-NF-

FIGURE 4. The four NF-AT/AP-1 elements of the GM-CSF en-hancer differ in their ability to function as enhancers and in their steroidresponsiveness. Jurkat cells were cotransfected with 10 �g of the pro-moter construct pHGM617 or the promoter plus individual enhancerelement constructs pGM170, pGM330, pGM420, or pGM550, plus 1 �gpRSV.GR or pRSV control. Samples were stimulated and treated withdexamethasone (Dex; 10�6 M), and cell extracts were assayed for CATactivity. The results have been normalized to a value of 100 for theactivated pHGM617 sample and are the mean of four experiments witherror bars representing SE.

FIGURE 5. GR binds to all four of the GM-CSF enhancer NF-AT/AP-1elements. A, Recombinant GR was incubated with labeled oligonucleotidesrepresenting the four NF-AT/AP-1 elements of the GM-CSF enhancer as de-tailed in Materials and Methods. Unlabeled GRE (20-fold excess) was used asa specific competitor. B, Immunopurified Flag.GR was electrophoresed on a10% SDS polyacrylamide gel and subjected to Western blotting with anti-GRand anti-Flag antisera. A control whole-cell extract from cells transfected withan empty expression vector is also shown. C, Immunopurified Flag GR (lanes2–4 and 9–11) and 5 �g Jurkat nuclear extract (NE) (lanes 5–7 and 12–14)were incubated with labeled GRE (lanes 1–7) and element GM550 (lanes8–14). A 20-fold excess of unlabeled GRE and GM550 were used as a specificcompetitors where indicated.

2506 INHIBITION OF GM-CSF BY GLUCOCORTICOIDS

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

AT)) or AP-1 (550(-AP-1)) site was mutated (Fig. 7A) were usedas both probes and competitors for band retardation. Fig. 7B illus-trates that both mutated oligonucleotides supported binding by GR,and both acted as efficient competitors for GR bound to GM550.These data imply that either GR is able to bind independently tothe NF-AT and AP-1 sites or that GR binds elsewhere on theoligonucleotide. To further investigate this, we synthesized oligo-nucleotides corresponding to either the 5� or 3� halves of GM550,along with a further oligonucleotide in which the NF-AT site wasmutated (Fig. 7A). Use of the two half sites as probes in bandretardation assays confirmed that each was able to support bindingby GR and that the complexes formed were efficiently competedby a molar excess of either unlabeled half site. In contrast, com-plex formation was not competed by the NF-AT mutant half site(Fig. 7C) or an unrelated oligonucleotide (Sp1) (data not shown).Furthermore, the data suggest that GR has a similar affinity forboth the AP-1 and NF-AT half sites. Taken together, these dataconfirm that GR is able to bind to the GM550 element at both

NF-AT and the nonconsensus AP-1 sites, suggesting a complexmechanism of transcriptional repression at this site.

DiscussionThe mechanism of transcriptional activation by nuclear hormonereceptors (NRs) has been the subject of intense scrutiny. SomeNRs, which are constitutively present within the nucleus, containa conserved transactivation domain known as activation function 2(AF-2) that undergoes a ligand-dependent conformational change(47, 48). A large number of cofactors have been identified thatinteract with AF-2 of NRs in a ligand-dependent manner and thatfunction in a combinatorial manner to reorganize chromatin tem-plates (49). Ligand binding leads to the dissociation of corepres-sors and the recruitment of coactivators by NRs (reviewed in Ref.50). The thyroid hormone receptor (TR) and retinoic acid receptor(RAR) are able to repress transcription both in the absence, orpresence, of their respective ligands through their interaction withspecific corepressors. Nonliganded TR or RAR interact with thenuclear receptor corepressor, ligand binding causing a reduction inaffinity for the corepressor (51, 52).

Unlike the other NRs, the GR translocates to the nucleus afterligand binding (53). Transcriptional activation by GR involves ho-modimeric binding to positive acting GREs that have a definedconsensus sequence (54). GR binding to positive GREs induces adistinct conformational change in the GR DNA-binding domain(55, 56), allowing coactivator recruitment through the carboxyl-terminal AF-2 (57, 58). In contrast, negative GREs have less con-served sequences. A common feature of negative GREs is that GRdoes not necessarily bind DNA as a homodimer. On the pro-opi-omelanocortin-negative GRE, GR has been suggested to bind as atrimer (32); furthermore, mutations that prevent the GR forming ahomodimer do not interfere with its ability to repress both thecollagenase and the proliferin genes (30, 59). In addition, GR witha point mutation in the D loop of the GR DNA-binding domain,required for dimerization, fails to bind DNA and cannot transac-tivate GRE-dependent promoters (60). Consequently, when GRacts to repress transcription, the conformational change requiredfor coactivator recruitment is not achieved and may allow the po-tential recruitment of corepressors by ligand-bound GR (61). It ispossible that repression by GR binding to DNA as a dimer mayoccur if the DNA-GR interaction does not cause the conforma-tional change required for activation. Our data suggests that inhi-bition of GM-CSF enhancer function involves binding of GRdimers. Adcock and colleagues have suggested an additionalmechanism of GR action (62), showing an ability of the receptorto recruit the corepressor histone deacetylase 2 (HDAC2) and pro-posing that the GR/deacetylase complex is also part of a CREB-binding protein-containing complex. DNA binding by GR was notinvolved in this mechanism. Therefore, these surprising data inferthat both histone acetyl transferase and deacetylase activities existwithin the same complex, raising the question of competition forspecific histone substrates. Additionally, they propose a role forGR in the stabilization of histone/DNA contacts. More recently, atranscriptional corepressor, SMRT/HDAC1-associated repressorprotein, has been isolated that interacts both with the steroid re-ceptor RNA coactivator steroid receptor RNA activator and withHDAC1 and 2, thereby providing direct evidence for recruitmentof a repressor complex to ligand-bound GR (35).

Transcriptional interference of the GR with other transcriptionfactors has been demonstrated to repress the function of both AP-1and NF-�B (30, 38, 63, 64). The importance of GR transrepressionhas been demonstrated by generation of dimerization-deficient GRmice using the point mutation in the D loop discussed above (65).

FIGURE 6. GR exhibits different binding affinities for the GM-CSF en-hancer NF-AT/AP-1 sites. A, Elements GM170 and GM330 were labeledand used as probes for recombinant GR binding. B, GR binding to GM420and GM550. Competition was provided by a 20-fold excess of unlabeledindividual enhancer NF-AT/AP-1 elements and unlabeled GRE.

2507The Journal of Immunology

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The mice are viable, unlike GR-deficient mice (66), able to tran-srepress AP-1-driven genes, but are unable to transactivate GRE-dependent genes due to the inability of their GR to bind DNA asa dimer.

In this paper, we have analyzed regions of the GM-CSF gene forglucocorticoid responsiveness and have identified a potentialmechanism for transcriptional repression of this cytokine. Thismechanism may be relevant to glucocorticoid action in asthma.Initial analysis of the GM-CSF proximal promoter, which mediatesactivation by known GR targets such as AP-1 and NF-�B (15, 18),demonstrated that this region was insensitive to steroid. The inter-action between the GR and AP-1 components is complex and doesnot always result in a negative effect on transcription. The effect ontranscription of the composite GRE contained in the proliferingene is modulated by the AP-1 components present (59). GR actssynergistically with a Jun homodimer, or represses transcription inthe presence of a Fos-Jun heterodimer. There are a number of otherexamples of GR acting either in a synergistic or repressive mannerwith AP-1 components, including regulation of dexamethasone-induced transcriptional activation of the MMTV LTR (67) andregulation of the neurotensin/neuromedin N gene (68). Regulationof the IL-2 gene occurs through a cooperative mechanism involv-ing NF-AT and AP-1 interaction at the promoter. Inhibition byglucocorticoids involves disruption of this cooperativity (69)through the GR binding to AP-1 that disrupts binding to NF-AT(70). NF-AT and AP-1 can bind to the CLE0 element in the GM-CSF promoter (15). However, NF-AT binding is very weak, andno cooperative binding with AP-1 occurs (71). Therefore, it isunlikely that glucocorticoids inhibit GM-CSF expression in thesame way that they inhibit IL-2 promoter activation. Recent data

(72) has shown that GR regulates activity of the IL-4 promoterthrough complex formation with, and inhibition of function of,NF-ATc. This illustrates the potential for multiple functional reg-ulatory events focussed at the NF-AT/AP-1 sites.

NF-�B binding to the GM-CSF promoter is another possibletarget for steroid action; however, the regulation of NF-�B activityby glucocorticoids also does not always result in repression. Dexa-methasone was found to have no effect on either synthesis of theinhibitor I�B�, or on NF-�B DNA binding ability, in two epithe-lial cell lines stimulated with IL-1� (73), or in two endothelial celllines (74). In addition, activation of NF-�B DNA binding activitywas observed following addition of dexamethasone to a PMA-stimulated monocytic cell line (75). The lack of suppression ofNF-�B activity by glucocorticoids is not confined to in vitro stud-ies. A recent investigation on the effects of the inhaled corticoste-roid fluticasone demonstrated no differences in NF-�B DNA bind-ing activity in both alveolar macrophages and in bronchial biopsiesfollowing steroid treatment (76). The same study also showed thatfluticasone caused an increase in expression of the p65 subunit ofNF-�B in the airway epithelium.

Our data indicate that repression takes place, at least in part,through the GM-CSF enhancer region and that GR binds to thecomposite NF-AT/AP-1 elements leading to repression of en-hancer function. Binding occurs despite the lack of consensusGREs within the NF-AT/AP-1 elements. This is reminiscent of theVDR-mediated inhibition of GM-CSF enhancer function. VDR in-hibits GM-CSF expression through binding to one of the NF-AT/AP-1 elements within the GM-CSF enhancer (26). Binding is to anonconsensus VDR recognition sequence and is unusual in that the

FIGURE 7. GR binds independently to both NF-AT and AP-1 halves of element GM550. Oligonucleotides mutated at either the AP-1 or NF-AT sitesof GM550 were synthesized, along with half sites and an NF-AT mutated half site (A). The full-length oligonucleotides, used either as probes or competitorsin EMSA, demonstrated the ability to be bound by GR (B). Use of half site oligonucleotides either as probes or competitors demonstrated that that eachhalf site was capable of being bound by GR and that a NF-AT mutated half site was unable to act as a competitor (C).

2508 INHIBITION OF GM-CSF BY GLUCOCORTICOIDS

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

receptor binds as a monomer and not as a heterodimer in partner-ship with the retinoid X receptor. Further analysis revealed thatbinding of the VDR monomer prevents NF-AT access to its bind-ing site and interacts with c-Jun, thereby stabilizing the affinity ofthe AP-1 interaction (27).

Following the observation that steroid treatment represses thefunction of the GM-CSF enhancer in a GR-dependent manner weconducted a more detailed analysis of this region. The four NF-AT/AP-1 elements of the enhancer display differential ability tobind NF-AT/AP-1 components, to act as enhancer elements (Fig.4), and to support GR binding (Fig. 5). None of the elements con-tain consensus GREs, and they only display partial relatedness toone another with respect to NF-AT/AP-1 site spacing and primarysequence (25). Our data suggest that the suppressive effects of GRare mediated primarily through GM550 and GM330, to which NF-AT/AP-1 bind with the weakest affinities (25). The functional abil-ity of steroid to inhibit enhancer function correlated with the abil-ity of the individual enhancer elements to support binding of GR.GR exhibited the strongest affinity for GM550 (Fig. 5), and dexa-methasone treatment inhibited the enhancer function of this ele-ment to the greatest extent (Fig. 4). In more detailed analysis, GRwas found to bind independently to both the NF-AT and AP-1halves of the element (Fig. 7). Interestingly, GM550 was the ele-ment demonstrated to bind VDR and whose function was inhibitedby 1,25(OH)2D3 treatment (26). Our data suggest that the mech-anism of repression at this site is different to that of vitamin D, dueto the additional ability of GR to bind to the AP-1 site. The factthat dexamethasone treatment did not result in a greater suppres-sion of enhancer activity is probably due to the unresponsivenessof element GM420. GR bound to this element very weakly (Figs.5 and 6), and dexamethasone had no effect on its ability to functionas an enhancer (Fig. 4). Our data and previous studies have dem-onstrated that this element is the most potent of the four enhancers,supports the strongest binding of NF-AT, and is the only one thatis contained within the essential core of the enhancer as defined bydeletion analysis (25).

Taken together, these data suggest that the GM-CSF enhanceroperates as a composite of the four individual elements. The ele-ments display differential ability to support binding of NF-AT/AP-1, GR, and possibly VDR. The ability to support GR bindingcorrelated with the steroid responsiveness of the element. We pro-pose that transcription factor competition at these sites governsactivation or repression. It is possible that GR has further repres-sive effects due to its ability to sequester both NF-AT and AP-1components, thereby destabilizing NF-AT/AP-1 cooperativity.

Band retardation analysis indicated that the GR-DNA com-plexes migrated with an identical mobility to GR-GRE, suggestingGR bound primarily as a dimer (Figs. 5C and 6). Due to the lackof homology with GRE, it is possible that the conformation ofGR-DNA is subtly different. An allosteric effect of the DNA se-quence on GR structure may allow the recruitment of a corepressorsuch as HDAC2 or, alternatively, the inability to recruit a coacti-vator. In summary, this study suggests a complex mechanism fortranscriptional repression of GM-CSF by GR brought about by theability of the receptor to compete with NF-AT and AP-1 compo-nents for binding sites within the GM-CSF enhancer.

AcknowledgmentsWe thank Drs. Peter Cockerill, Jacques Drouin, Tariq Enver, and TonyKouzarides for providing plasmids.

References1. Baldwin, G. C. 1992. The biology of granulocyte-macrophage colony-stimulating

factor: effects on hematopoietic and nonhematopoietic cells. Dev. Biol. 151:352.

2. Gasson, J. C. 1991. Molecular physiology of granulocyte-macrophage colony-stimulating factor. Blood 77:1131.

3. Nicola, N. A., D. Metcalf, G. R. Johnson, and A. W. Burgess. 1979. Separationof functionally distinct human granulocyte-macrophage colony-stimulating fac-tors. Blood 54:614.

4. Kay, A. B. 1994. T cells, cytokines and asthma. J. R. Coll. Physicians London28:325.

5. Robinson, D. S., Q. Hamid, S. Ying, A. Tsicopoulos, J. Barkans, A. M. Bentley,C. Corrigan, S. R. Durham, and A. B. Kay. 1992. Predominant TH2-like bron-choalveolar T-lymphocyte population in atopic asthma. N. Engl. J. Med. 326:298.

6. Ackerman, V., M. Marini, E. Vittori, A. Bellini, G. Vassali, and S. Mattoli. 1994.Detection of cytokines and their cell sources in bronchial biopsy specimens fromasthmatic patients: relationship to atopic status, symptoms, and level of airwayhyperresponsiveness. Chest 105:687.

7. Barnes, P. J. 1995. Inhaled glucocorticoids for asthma. N. Engl. J. Med. 332:868.8. Djukanovic, R., J. W. Wilson, K. M. Britten, S. J. Wilson, A. F. Walls,

W. R. Roche, P. H. Howarth, and S. T. Holgate. 1992. Effect of an inhaledcorticosteroid on airway inflammation and symptoms in asthma. Am. Rev. Respir.Dis. 145:669.

9. Robinson, D., Q. Hamid, S. Ying, A. Bentley, B. Assoufi, S. Durham, andA. B. Kay. 1993. Prednisolone treatment in asthma is associated with modulationof bronchoalveolar lavage cell interleukin-4, interleukin-5, and interferon-� cy-tokine gene expression. Am. Rev. Respir. Dis. 148:401.

10. John, M., S. Lim, J. Seybold, P. Jose, A. Robichaud, B. O’Connor, P. J. Barnes,and K. F. Chung. 1998. Inhaled corticosteroids increase interleukin-10 but reducemacrophage inflammatory protein-1�, granulocyte-macrophage colony-stimulat-ing factor, and interferon-� release from alveolar macrophages in asthma.Am. J. Respir. Crit. Care Med. 157:256.

11. Sousa, A. R., R. N. Poston, S. J. Lane, J. A. Nakhosteen, and T. H. Lee. 1993.Detection of GM-CSF in asthmatic bronchial epithelium and decrease by inhaledcorticosteroids. Am. Rev. Respir. Dis. 147:1557.

12. Wang, J. H., C. J. Trigg, J. L. Devalia, S. Jordan, and R. J. Davies. 1994. Effectof inhaled beclomethasone dipropionate on expression of proinflammatory cyto-kines and activated eosinophils in the bronchial epithelium of patients with mildasthma. J. Allergy Clin. Immunol. 94:1025.

13. Shannon, M. F., L. S. Coles, M. A. Vadas, and P. N. Cockerill. 1997. Signals foractivation of the GM-CSF promoter and enhancer in T cells. Crit. Rev. Immunol.17:301.

14. Cousins, D., D. Z. Staynov, and T. H. Lee. 1999. Regulation of the cytokine genecluster on chromosome 5q. In Molecular Biology of the Lung. R. A. Stockley, ed.,Birkhauser, Basel, p. 71.

15. Masuda, E. S., H. Tokumitsu, A. Tsuboi, J. Shlomai, P. Hung, K. Arai, andN. Arai. 1993. The granulocyte-macrophage colony-stimulating factor promotercis-acting element CLE0 mediates induction signals in T cells and is recognizedby factors related to AP1 and NFAT. Mol. Cell Biol. 13:7399.

16. Thomas, R. S., M. J. Tymms, A. Seth, M. F. Shannon, and I. Kola. 1995. ETS1transactivates the human GM-CSF promoter in Jurkat T cells stimulated withPMA and ionomycin. Oncogene 11:2135.

17. Wang, C. Y., A. G. Bassuk, L. H. Boise, C. B. Thompson, R. Bravo, andJ. M. Leiden. 1994. Activation of the granulocyte-macrophage colony-stimulat-ing factor promoter in T cells requires cooperative binding of Elf-1 and AP-1transcription factors. Mol. Cell Biol. 14:1153.

18. Schreck, R., and P. A. Baeuerle. 1990. NF-�B as inducible transcriptional acti-vator of the granulocyte-macrophage colony-stimulating factor gene. Mol. CellBiol. 10:1281.

19. Masuda, E. S., Y. Yamaguchi-Iwai, A. Tsuboi, P. Hung, K. Arai, and N. Arai.1994. The transcription factor Sp1 is required for induction of the murine GM-CSF promoter in T cells. Biochem. Biophys. Acta 205:1518.

20. Thomas, R. S., M. J. Tymms, L. H. McKinlay, M. F. Shannon, A. Seth, andI. Kola. 1997. ETS1, NF�B and AP1 synergistically transactivate the humanGM-CSF promoter. Oncogene 14:2845.

21. Staynov, D. Z., D. J. Cousins, and T. H. Lee. 1995. A regulatory element in thepromoter of the human granulocyte-macrophage colony-stimulating factor genethat has related sequences in other T-cell-expressed cytokine genes. Proc. Natl.Acad. Sci. USA 92:3606.

22. Cockerill, P. N., M. F. Shannon, A. G. Bert, G. R. Ryan, and M. A. Vadas. 1993.The granulocyte-macrophage colony-stimulating factor/interleukin 3 locus is reg-ulated by an inducible cyclosporin A-sensitive enhancer. Proc. Natl. Acad. Sci.USA 90:2466.

23. Duncliffe, K. N., A. G. Bert, M. A. Vadas, and P. N. Cockerill. 1997. A Tcell-specific enhancer in the interleukin-3 locus is activated cooperatively by Octand NFAT elements within a DNase I-hypersensitive site. Immunity 6:175.

24. Cockerill, P. N., A. G. Bert, D. Roberts, and M. A. Vadas. 1999. The humangranulocyte-macrophage colony-stimulating factor gene is autonomously regu-lated in vivo by an inducible tissue-specific enhancer. Proc. Natl. Acad. Sci. USA96:15097.

25. Cockerill, P. N., A. G. Bert, F. Jenkins, G. R. Ryan, M. F. Shannon, andM. A. Vadas. 1995. Human granulocyte-macrophage colony-stimulating factorenhancer function is associated with cooperative interactions between AP-1 andNFATp/c. Mol. Cell Biol. 15:2071.

26. Towers, T. L., and L. P. Freedman. 1998. Granulocyte-macrophage colony-stim-ulating factor gene transcription is directly repressed by the vitamin D3 receptor:implications for allosteric influences on nuclear receptor structure and function bya DNA element. J. Biol. Chem. 273:10338.

27. Towers, T. L., T. P. Staeva, and L. P. Freedman. 1999. A two-hit mechanism forvitamin D3-mediated transcriptional repression of the granulocyte-macrophage

2509The Journal of Immunology

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

colony-stimulating factor gene: vitamin D receptor competes for DNA bindingwith NFAT1 and stabilizes c-Jun. Mol. Cell Biol. 19:4191.

28. Evans, R. M. 1988. The steroid and thyroid hormone receptor superfamily. Sci-ence 240:889.

29. Gronemeyer, H. 1992. Control of transcription activation by steroid hormonereceptors. FASEB J. 6:2524.

30. Jonat, C., H. J. Rahmsdorf, K. K. Park, A. C. Cato, S. Gebel, H. Ponta, andP. Herrlich. 1990. Antitumor promotion and antiinflammation: down-modulationof AP-1 (Fos/Jun) activity by glucocorticoid hormone. Cell 62:1189.

31. Auphan, N., J. A. DiDonato, C. Rosette, A. Helmberg, and M. Karin. 1995.Immunosuppression by glucocorticoids: inhibition of NF-�B activity through in-duction of I�B synthesis. Science 270:286.

32. Drouin, J., Y. L. Sun, M. Chamberland, Y. Gauthier, A. De Lean, M. Nemer, andT. J. Schmidt. 1993. Novel glucocorticoid receptor complex with DNA elementof the hormone-repressed POMC gene. EMBO J. 12:145.

33. Akerblom, I. E., E. P. Slater, M. Beato, J. D. Baxter, and P. L. Mellon. 1988.Negative regulation by glucocorticoids through interference with a cAMP re-sponsive enhancer. Science 241:350.

34. Sheppard, K. A., K. M. Phelps, A. J. Williams, D. Thanos, C. K. Glass,M. G. Rosenfeld, M. E. Gerritsen, and T. Collins. 1998. Nuclear integration ofglucocorticoid receptor and nuclear factor-�B signaling by CREB-binding pro-tein and steroid receptor coactivator-1. J. Biol. Chem. 273:29291.

35. Shi, Y., M. Downes, W. Xie, H. Y. Kao, P. Ordentlich, C. C. Tsai, M. Hon, andR. M. Evans. 2001. Sharp, an inducible cofactor that integrates nuclear receptorrepression and activation. Genes Dev. 15:1140.

36. Chomczynski, P., and N. Sacchi. 1987. Single-step method of RNA isolation byacid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162:156.

37. Staynov, D. Z., and T. H. Lee. 1992. Expression of interleukin-5 and granulocyte-macrophage colony-stimulating factor in human peripheral blood mononuclearcells after activation with phorbol myristate acetate. Immunology 75:196.

38. Yang-Yen, H. F., J. C. Chambard, Y. L. Sun, T. Smeal, T. J. Schmidt, J. Drouin,and M. Karin. 1990. Transcriptional interference between c-Jun and the glucocor-ticoid receptor: mutual inhibition of DNA binding due to direct protein-proteininteraction. Cell 62:1205.

39. Northrop, J. P., S. N. Ho, L. Chen, D. J. Thomas, L. A. Timmerman, G. P. Nolan,A. Admon, and G. R. Crabtree. 1994. NF-AT components define a family oftranscription factors targeted in T-cell activation. Nature 369:497.

40. Ausubel, F., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith,and K. Struhl. 1994. Current Protocols in Molecular Biology. John Wiley andSons, New York.

41. Lavender, P., L. Vandel, A. J. Bannister, and T. Kouzarides. 1997. The HMG-boxtranscription factor HBP1 is targeted by the pocket proteins and E1A. Oncogene14:2721.

42. Brehm, A., S. J. Nielsen, E. A. Miska, D. J. McCance, J. L. Reid, A. J. Bannister,and T. Kouzarides. 1999. The E7 oncoprotein associates with Mi2 and histonedeacetylase activity to promote cell growth. EMBO J. 18:2449.

43. Srinivasan, G., and E. B. Thompson. 1990. Overexpression of full-length humanglucocorticoid receptor in Spodoptera frugiperda cells using the baculovirus ex-pression vector system. Mol. Endocrinol. 4:209.

44. Tsai, S. Y., G. Srinivasan, G. F. Allan, E. B. Thompson, B. W. O’Malley, andM. J. Tsai. 1990. Recombinant human glucocorticoid receptor induces transcrip-tion of hormone response genes in vitro. J. Biol. Chem. 265:17055.

45. Scheidereit, C., S. Geisse, H. M. Westphal, and M. Beato. 1983. The glucocor-ticoid receptor binds to defined nucleotide sequences near the promoter of mousemammary tumour virus. Nature 304:749.

46. Tobler, A., R. Meier, M. Seitz, B. Dewald, M. Baggiolini, and M. F. Fey. 1992.Glucocorticoids downregulate gene expression of GM-CSF, NAP-1/IL-8, andIL-6, but not of M-CSF in human fibroblasts. Blood 79:45.

47. Danielian, P. S., R. White, J. A. Lees, and M. G. Parker. 1992. Identification ofa conserved region required for hormone dependent transcriptional activation bysteroid hormone receptors. EMBO J. 11:1025.

48. Durand, B., M. Saunders, C. Gaudon, B. Roy, R. Losson, and P. Chambon. 1994.Activation function 2 (AF-2) of retinoic acid receptor and 9-cis retinoic acidreceptor: presence of a conserved autonomous constitutive activating domain andinfluence of the nature of the response element on AF-2 activity. EMBO J. 13:5370.

49. Pollard, K. J., and C. L. Peterson. 1998. Chromatin remodeling: a marriage be-tween two families? Bioessays 20:771.

50. Glass, C. K., and M. G. Rosenfeld. 2000. The coregulator exchange in transcrip-tional functions of nuclear receptors. Genes Dev. 14:121.

51. Horlein, A. J., A. M. Naar, T. Heinzel, J. Torchia, B. Gloss, R. Kurokawa,A. Ryan, Y. Kamei, M. Soderstrom, C. K. Glass, et al. 1995. Ligand-independentrepression by the thyroid hormone receptor mediated by a nuclear receptor co-repressor. Nature 377:397.

52. Kurokawa, R., M. Soderstrom, A. Horlein, S. Halachmi, M. Brown,M. G. Rosenfeld, and C. K. Glass. 1995. Polarity-specific activities of retinoicacid receptors determined by a co-repressor. Nature 377:451.

53. Denis, M., L. Poellinger, A. C. Wikstom, and J. A. Gustafsson. 1988. Require-ment of hormone for thermal conversion of the glucocorticoid receptor to a DNA-binding state. Nature 333:686.

54. Beato, M. 1989. Gene regulation by steroid hormones. Cell 56:335.55. Hard, T., E. Kellenbach, R. Boelens, B. A. Maler, K. Dahlman, L. P. Freedman,

J. Carlstedt-Duke, K. R. Yamamoto, J. A. Gustafsson, and R. Kaptein. 1990.Solution structure of the glucocorticoid receptor DNA-binding domain. Science249:157.

56. Luisi, B. F., W. X. Xu, Z. Otwinowski, L. P. Freedman, K. R. Yamamoto, andP. B. Sigler. 1991. Crystallographic analysis of the interaction of the glucocor-ticoid receptor with DNA. Nature 352:497.

57. Horwitz, K. B., T. A. Jackson, D. L. Bain, J. K. Richer, G. S. Takimoto, andL. Tung. 1996. Nuclear receptor coactivators and corepressors. Mol. Endocrinol.10:1167.

58. Torchia, J., C. Glass, and M. G. Rosenfeld. 1998. Co-activators and co-repressorsin the integration of transcriptional responses. Curr. Opin. Cell Biol. 10:373.

59. Diamond, M. I., J. N. Miner, S. K. Yoshinaga, and K. R. Yamamoto. 1990.Transcription factor interactions: selectors of positive or negative regulation froma single DNA element. Science 249:1266.

60. Heck, S., M. Kullmann, A. Gast, H. Ponta, H. J. Rahmsdorf, P. Herrlich, andA. C. Cato. 1994. A distinct modulating domain in glucocorticoid receptor mono-mers in the repression of activity of the transcription factor AP-1. EMBO J.13:4087.

61. Lefstin, J. A., and K. R. Yamamoto. 1998. Allosteric effects of DNA on tran-scriptional regulators. Nature 392:885.

62. Ito, K., P. J. Barnes, and I. M. Adcock. 2000. Glucocorticoid receptor recruitmentof histone deacetylase 2 inhibits interleukin-1�-induced histone H4 acetylationon lysines 8 and 12. Mol. Cell Biol. 20:6891.

63. Ray, A., and K. E. Prefontaine. 1994. Physical association and functional antag-onism between the p65 subunit of transcription factor NF-�B and the glucocor-ticoid receptor. Proc. Natl. Acad. Sci. USA 91:752.

64. Mukaida, N., M. Morita, Y. Ishikawa, N. Rice, S. Okamoto, T. Kasahara, andK. Matsushima. 1994. Novel mechanism of glucocorticoid-mediated gene repres-sion: nuclear factor-�B is target for glucocorticoid-mediated interleukin 8 generepression. J. Biol. Chem. 269:13289.

65. Reichardt, H. M., K. H. Kaestner, J. Tuckermann, O. Kretz, O. Wessely, R. Bock,P. Gass, W. Schmid, P. Herrlich, P. Angel, and G. Schutz. 1998. DNA bindingof the glucocorticoid receptor is not essential for survival. Cell 93:531.

66. Cole, T. J., J. A. Blendy, A. P. Monaghan, K. Krieglstein, W. Schmid, A. Aguzzi,G. Fantuzzi, E. Hummler, K. Unsicker, and G. Schutz. 1995. Targeted disruptionof the glucocorticoid receptor gene blocks adrenergic chromaffin cell develop-ment and severely retards lung maturation. Genes Dev. 9:1608.

67. Maroder, M., A. R. Farina, A. Vacca, M. P. Felli, D. Meco, I. Screpanti, L. Frati,and A. Gulino. 1993. Cell-specific bifunctional role of Jun oncogene familymembers on glucocorticoid receptor-dependent transcription. Mol. Endocrinol.7:570.

68. Harrison, R. J., G. P. McNeil, and P. R. Dobner. 1995. Synergistic activation ofneurotensin/neuromedin N gene expression by c-Jun and glucocorticoids: noveleffects of Fos family proteins. Mol. Endocrinol. 9:981.

69. Vacca, A., M. P. Felli, A. R. Farina, S. Martinotti, M. Maroder, I. Screpanti,D. Meco, E. Petrangeli, L. Frati, and A. Gulino. 1992. Glucocorticoid receptor-mediated suppression of the interleukin 2 gene expression through impairment ofthe cooperativity between nuclear factor of activated T cells and AP-1 enhancerelements. J. Exp. Med. 175:637.

70. Paliogianni, F., A. Raptis, S. S. Ahuja, S. M. Najjar, and D. T. Boumpas. 1993.Negative transcriptional regulation of human interleukin 2 (IL-2) gene by glu-cocorticoids through interference with nuclear transcription factors AP-1 andNF-AT. J. Clin. Invest. 91:1481.

71. Jenkins, F., P. N. Cockerill, D. Bohmann, and M. F. Shannon. 1995. Multiplesignals are required for function of the human granulocyte-macrophage colony-stimulating factor gene promoter in T cells. J. Immunol. 155:1240.

72. Chen, R., T. F. Burke, J. E. Cumberland, M. Brummet, L. A. Beck, V. Casolaro,and S. N. Georas. 2000. Glucocorticoids inhibit calcium- and calcineurin-depen-dent activation of the human IL-4 promoter. J. Immunol. 164:825.

73. Newton, R., L. A. Hart, D. A. Stevens, M. Bergmann, L. E. Donnelly,I. M. Adcock, and P. J. Barnes. 1998. Effect of dexamethasone on interleukin-1�-(IL-1�)-induced nuclear factor-�B (NF-�B) and �B-dependent transcription inepithelial cells. Eur. J. Biochem. 254:81.

74. Brostjan, C., J. Anrather, V. Csizmadia, D. Stroka, M. Soares, F. H. Bach, andH. Winkler. 1996. Glucocorticoid-mediated repression of NF�B activity in en-dothelial cells does not involve induction of I�B� synthesis. J. Biol. Chem. 271:19612.

75. Wang, Y., J. J. Zhang, W. Dai, K. Y. Lei, and J. W. Pike. 1997. Dexamethasonepotently enhances phorbol ester-induced IL-1� gene expression and nuclear fac-tor NF-�B activation. J. Immunol. 159:534.

76. Hart, L., S. Lim, I. Adcock, P. J. Barnes, and K. F. Chung. 2000. Effects ofinhaled corticosteroid therapy on expression and DNA-binding activity of nuclearfactor �B in asthma. Am. J. Respir. Crit. Care Med. 161:224.

2510 INHIBITION OF GM-CSF BY GLUCOCORTICOIDS

by guest on April 8, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from