The FASEB Journal • Research Communication
Competition between glucocorticoid receptor and NFB for control of the human FasL promoter Natalia Novac,*,1 Daniela Baus,* Anja Dostert,*,2 and Thorsten Heinzel†,3 *Georg-Speyer-Haus, Institute for Biomedical Research, Frankfurt, Germany; and †Institute of Biochemistry and Biophysics, Friedrich-Schiller-University, Jena, Germany Glucocorticoids mediate a variety of biological effects via binding their intracellular receptor. Ligand-bound glucocorticoid receptor (GR) translocates to the nucleus and regulates gene transcription in a DNA binding-dependent or independent manner. The predominant biological effect of glucocorticoids on peripheral T cells is immunosupression via transcriptional repression of genes induced during T cell activation. Glucocorticoids have been implicated in the inhibition of activation-induced T cell apoptosis by virtue of their down-regulation of Fas ligand (fasL) expression. It is believed that FasL, similar to other cytokines, is repressed by glucocorticoids via GR interaction with other transcription factors, interfering with their transactivation ability. Here, we show that human fasL is directly regulated by GR in a DNA bindingdependent manner. A negative GR element found at position –990 in the fasL promoter binds GR in vitro as well as in the chromatin context. This negative glucocorticoid response element overlaps with a known NFB binding site. GR down-regulates fasL promoter by competing with NFB for binding to the common response element. Thus, fasL is the first gene described whose repression by GR is mediated by sterical occlusion of NFB DNA binding. This type of repression represents an additional mechanism for the GR-NFB mutual antagonism.—Novac, N., Baus, D., Dostert, A., Heinzel, T. Competition between glucocorticoid receptor and NFB for control of the human FasL promoter. FASEB J. 20, 1074 –1081 (2006) ABSTRACT
Key Words: nGRE 䡠 AICD 䡠 inflammation 䡠 glucocorticoid receptor 䡠 cis-repression
Glucocorticoids (gcs) are steroid hormones produced by the adrenal glands on cytokine stimulation of the hypothalamic-pituitary-adrenal axis. GCs are involved in the regulation of a variety of biological processes, including cell growth and proliferation, metabolism, development, and reproduction (1, 2). Their significant immunosuppressive abilities make them the mainstay of anti-inflammatory therapy today (3, 4). The apheliotropic actions of GCs in the cell are mediated through glucocorticoid receptor (GR), a ubiquitously expressed member of the nuclear receptor superfamily. GR in its unliganded state resides in the 1074
cytoplasm as a heterooligomeric complex containing one steroid binding protein and a multisubunit nonsteroid binding complex consisting of heat shock proteins, immunophilins, and other chaperones. Binding of the lipophilic ligand to GR induces chaperone shedding, exposing the GR nuclear localization signal, which allows its translocation to the nucleus. In the nucleus GR binds to specific DNA sequences— glucocorticoid response elements (GREs), which may be positive or negative, either activating or repressing transcription. This mode of GR action is called cisregulation of gene transcription and is generally considered to be responsible for GR-mediated side effects. Another mechanism of GR action, called trans-regulation, is currently thought to be important for most of the GR anti-inflammatory functions. GR trans-regulation is based on the ability of liganded GR monomers to interact with other transcription factors such as NFB, AP-1, or STAT5, thereby repressing or potentiating their activity (5). No genes responding positively to GCs have been identified that explain the tremendous immunosuppressive activity of GR, so that GR-mediated transcriptional repression is believed to be responsible for this process. Due to the lack of a clear palindromic GRE consensus in the genes involved in inflammatory response, most recent studies have concentrated on the elucidation of GR-mediated trans-repression. Several mechanisms have been suggested to account for this process: 1) competition with other transcription factors for coactivators (squelching); 2) disruption of the interaction of a transcription factor with the basal transcription machinery; 3) recruitment of corepressors or HDACs to inactivate a promoter by chromatin remodeling; and 4) formation of inactive complexes with other transcription factors (6 –11). None of these mechanisms have been shown to be responsible for the GR-mediated cis-repression, whose mechanism and in1
Present address: Merck, Frankfurter Str. 250, Darmstadt 64293, Germany. 2 Present address: Sandi-Aventis GmbH, Potsdamer Str. 8, Berlin 10285, Germany. 3 Correspondence: Institute of Biochemistry and Biophysics, Friedrich-Schiller-University Jena, Philosophenweg 12, Jena 07743, Germany. E-mail:
[email protected] doi: 10.1096/fj.05-5457com 0892-6638/06/0020-1074 © FASEB
volvement in the anti-inflammatory action of GR remain largely uncharacterized. Fas ligand (FasL/CD95L, Apo-1L) is a member of the TNF family that induces cell death in lymphoid cells on binding to its receptor (Fas) on the surface of target cells (12–14). The most important role of FasL is associated with its inducible expression in CD4⫹ cells (T helper cells) during inflammation, where FasL is a major effector molecule in the course of activationinduced cell death (AICD). AICD is a specific form of apoptosis triggered in cytotoxic T cells during the last stages of the immune response and is required for the elimination of large number of expanded T cells in order to prevent chronic inflammation and autoimmunity. Several transcription factors are involved in fasL induction on TCR stimulation during AICD. Among those are NFB, NFAT, Egr, and IRF family members (15). Repression of fasL expression is mediated by a number of factors, including GCs. Treatment of T cell hybridomas with GCs was found to inhibit fasL upregulation and subsequent apoptosis (16). In an attempt to elucidate the mechanism of this inhibition, a study of the effect of rifampicin (a substance that in some cases acts similarly to GCs) in Jurkat cells was performed (17). As the AICD inhibition was abrogated in an I-B␣-deficient mutant, it was concluded that GR-mediated induction of I-B␣ and subsequent sequestration of NFB to the cytoplasm are responsible for fasL inhibition. Another report suggested that GCs induce expression of Bcl-2, which can inhibit fasL expression, by binding and sequestration of calcineurin, which must dephosphorylate NFAT for its translocation to the nucleus (18, 19). It was recently reported that murine fasL (mfasL) is cis-repressed by GR although the exact molecular mechanism remained unclear (20). Since neither I-B␣ nor Bcl-2 accounted for the GR-mediated inhibition of fasL we observed, we sought to determine whether the hfasL promoter is subject to direct GR DNA binding and to resolve the molecular mechanism responsible for its inhibition.
MATERIALS AND METHODS Cells and constructs Peripheral blood T lymphocytes were separated on Ficoll gradients and stimulated first with 5 g ml⫺1 of PHA (Sigma, Taufkirchen, Germany) for 3 days, then with 100 IU ml⫺1 interleukin-2 for 6 –10 days. Jurkat T cell lymphoma cell line clone J16, hepatocellular carcinoma HepG2, and human osteosarcoma U2OS cells were maintained according to the American Type Culture Collection recommendations. –1200/⫹100 and – 860/⫹100 luciferase reporter constructs (a kind gift from Dr. Min Li-Weber) are described in ref 21. Other deletion and mutant constructs were created using the QuickChange site-directed mutagenesis kit (Stratagene, San Diego, CA, USA) according to the manufacturer’s instructions. GRwt and GRdim-expressing constructs (a kind gift from Dr. A. Cato) are described in ref 22). GR COMPETES WITH NFB FOR FASL PROMOTER BINDING
Transfection and reporter assays HepG2 cells were transfected with FuGENE 6 (Roche Diagnostic GmbH, Karlsruhe, Germany) according to the manufacturer’s protocol. After 24 h the cells were treated with 100 nM Dex for another 24 h and the luciferase activity was measured and normalized against -galactosidase activity. -Galactosidase-expressing construct pTK was cotransfected as a transfection efficiency control. All the reporter experiments were performed in triplicate and repeated at least twice. HepG2 cells were used in reporter assays since their endogenous GR is inactive in reporter assays (see Fig. 2A–D, first columns) and thus suitable for analysis of the GRdim mutant. It has also been reported that in HepG2 cells the endogenous GR is subject to proteasomal degradation on Dex treatment (23). RNA isolation and RT-PCR analysis Total RNA was isolated using the RNAeasy kit (Qiagen, Hilden, Germany) and used for subsequent RT-PCT (Superscript one-step RT-PCT (Invitrogen, San Diego, CA, USA). The following primers have been used for this analysis: FasLstart: 5⬘-cag cag ccc ttc aat tac cca tat cc-3⬘; FasLstop: 5⬘-ctt ata taa gcc gaa aaa cgt c-3⬘. ABCD assay (avidin, biotin, complex, DNA) This assay is based on the immobilization of protein-DNA complexes via binding of a biotinylated oligonucleotide to a streptavidin matrix. 200 l Jurkat whole cell extract (5 g/l) was incubated with 200 l buffer H (100 mM KCl, 20 mM HEPES pH 7.8 –7.9, 20% glycerol, 1 mM DTT, 0.1% NP-40), 2 g biotinylated oligonucleotide, 10 g herring sperm DNA (competitor), and Dex for 1 h on ice. After the addition of 40 l equilibrated streptavidin agarose beads (Amersham, Arlington Heights, IL, USA), incubation was continued for 30 min at 4°C on a rotator. Beads were washed repeatedly with buffer H (containing 50 mM KCl), boiled in Laemmli sample buffer, and separated by SDS-PAGE under reducing conditions. GR was detected by Western blot (anti-GR antibody H300, Santa Cruz Technology, Santa Cruz, CA, USA). Enhanced chemiluminescence (ECL) (Amersham Pharmacia Biotech) was used for detection. In ABCD assays the GR binds to its response element independent of ligand activation due to the absence of a physical barrier between the nuclear and cytoplasmic fractions in cell lysates. Therefore, the GR-negative U2OS cell line and recombinant GR were used for the GR/NFB competition experiment. Biotinylated oligos ordered from ThermoElectron (Bremen, Germany) comprise the following sequences: FasL-990: 5⬘-aac ata gca agt ccc cat ctg tac aaa aaa aaa-3⬘; FasL990m: 5⬘-aac ata gca agt ccc cat ctt tac gaa aaa aaa-3⬘; GRE: 5⬘- aat ctc tgc tgt aca gga tgt tct agc tac t-3⬘; GREm: 5⬘- aat ctc tgc tgt aga gga tct tct agc tac t-3⬘, 2xNFB: 5⬘-tgg gga ctt tcc gct ggg gac ttt ccg c-3⬘. ChIP assay The chromatin immunoprecipitation (ChIP) protocol used in our experiments was adopted from ref 24. Jurkat cells cultivated in RPMI medium containing 10% charcoalstripped bovine calf serum were incubated with 10⫺7 M Dex for 30 min, cross-linked by PFA (1%) for 10 min at room temperature. The reaction was stopped with 0.125 M glycine. Cells were washed in PBS, lysed, and subsequently subjected to sonication and immunoprecipitation with anti-GR and 1075
anti-NFB p50 polyclonal antibodies (Santa Cruz Biotechnology). Immunoprecipitated material was collected by centrifugation and washed once each with WBI, WBII, LiCl buffer, and twice with TE buffer. After washing, 200 l digestion buffer (50 mM Tris pH8, 1 mM EDTA, 100 mM NaCl, 0.5% SDS) supplemented with proteinase K (100 g/ml) was added and the reaction was incubated at 55°C for 3 h. The material was de-crosslinked by raising the incubation temperature to 65°C for 16 h. The recovered DNA was purified by PCT purification kit (Qiagen) according to the manufacturer’s instructions. Obtained DNA (2 l) was used in a 20 l PCR reaction (using TaqDNA Polymerase (Qiagen). The following primers have been used for the amplification of the –990 nGRE containing sequence of the FasL promoter: –1160FasL: 5⬘-gtg gtc aag ttg ggc tca gtg c-3⬘; –980FasL: 5⬘-tgt aca gat ggg gac ttg cta tg-3⬘.
RESULTS Dexamethasone rapidly inhibits human FasL expression Previous studies have shown that GCs inhibit activationinduced T cell apoptosis in murine hybridomas and human and mouse primary peripheral T lymphocytes by interfering with FasL expression (16, 20), but the molecular mechanism of GR-mediated FasL repression has not been elucidated. To investigate this, we first wanted to prove that the expression of human fasL is down-regulated by GR similarly to mfasL. We applied
RT-PCR analysis of RNA isolated from Jurkat T cells treated with ionomycin and PMA (I/PMA) alone (to mimic TCR signaling) or in combination with dexamethasone (Dex). As shown in Fig. 1A, fasL activation is rapidly and transiently inhibited by Dex in Jurkat cells. GR-dependent inhibition was also observed in human primary peripheral T cells, where Dex repressed fasL expression in a dose-dependent manner (Fig. 1B). Similar to endogenous fasL, expression of the (–1200; ⫹100) FasL luciferase reporter (FasL-Luc) was downregulated by GCs. This transrepression was also dependent on the amount of GR-expressing construct cotransfected with the reporter (Fig. 1C). These experiments confirm that hfasL is repressed by GCs, and this inhibition is GR dependent. The antiglucocorticoid RU-486 is nearly as effective as Dex in inhibiting FasL reporter (Fig. 1D). To test whether fasL inhibition is caused by GRmediated induction of I-B␣ and Bcl-2, we analyzed the expression of these proteins in the presence of Dex and I/PMA (Fig. 1E). No change of the protein concentration was detected during the first 8 h of treatment. Moreover, treatment of the cells with the inhibitor of protein synthesis cycloheximide (CHX) did not relieve Dex-dependent inhibition (data not shown). These results suggest that FasL is directly inhibited by Dex and does not require synthesis of Bcl-2, IB␣, or other proteins.
Figure 1. Activation of human fasL is rapidly inhibited by GR in a dose-dependent manner. A) Jurkat T cell lymphoma cells were treated with 100 ng/ml ionomycine and 1 g/ml PMA (I/PMA) or in combination with 100 nM dexamethasone (Dex) for the times indicated. Total cellular RNA was isolated and the levels of hFasL expression were analyzed by RT-PCR. Actin RNA expression levels serve as a loading control. B) Human primary peripheral T cells were treated with 100 ng/ml ionomycine and 1 g/ml PMA (I/PMA) or in combination with 10 nM or 100 nM dexamethasone (Dex) for 3 h. After treatment total cellular RNA was isolated and the levels of hFasL expression were analyzed by RT-PCR. Actin mRNA expression levels serve as a loading control. C) A luciferase reporter construct containing the hFasL promoter (–1200/⫹100) hFasl-Luc was cotransfected with an empty vector or with 10, 50 and 100 ng of a GR-expressing construct into HepG2 cells. After 24 h the cells were treated with 100 nM Dex for another 24 h and luciferase activity was measured and normalized against -galactosidase activity (transfection efficiency control). D) A luciferase reporter construct containing the hFasL promoter (–1200/ ⫹100) hFasl-Luc was cotransfected with an empty vector or with 200 ng of a GR-expressing construct into HepG2 cells. The cells were treated 24 h post-transfection for another 24 h with ethanol (gray bars), 100 nM Dex (black bars) or 5 M RU-486 (white bars). E) Jurkat cells were treated with 100 nM Dex and/or Iono/PMA for 8 h. Cell lysates were prepared every 2 h and used to compare protein levels of I-B and Bcl-2 by Western blot. Actin protein expression serves as a loading control. 1076
Vol. 20
June 2006
The FASEB Journal
NOVAC ET AL.
GR dimer formation is required for FasL inhibition To test whether GR-DNA binding activity is required for FasL inhibition, we used a GR dimerization defective mutant (GRdim) whose DNA binding ability is severely impaired (25). We compared the ability of GRwt and GRdim to attenuate FasL transcription by transfecting these constructs into HepG2 cells, which contain nonfunctional endogenous GR. As shown in Fig. 2, GRdim failed to transactivate murine mammary tumor virus (MMTV) as well as to repress proopiomelanocortin (POMC) promoters, whose transcriptional regulation depends on GR-DNA binding activity (26, 27), whereas
both GRwt and GRdim were able to mediate the GR binding-independent transcriptional repression of collagenase-1 (Col-1) (Fig 2B). The wtGR is responsible for most of the FasL transcriptional repression reducing reporter expression down to 30% of the initial activity. The GRdim mutant, in contrast, caused only minor inhibition, suggesting that the DNA binding-dependent function of GR is required for the full inhibition of fasL. The distal part of the FasL promoter is important for the Dex-dependent inhibition A previous study demonstrated that mfasL promoter has several nGREs (20). The identified positions of nGREs, however, are not well conserved between mouse and human promoters. In silico screening of the hfasL promoter revealed nine putative nGREs (Fig. 3A). To identify the region of the hfasL promoter responsible for GR-mediated inhibition, a serial promoter deletion analysis was performed. As shown in Fig. 3B, removal of the distal part of the hfasL promoter eliminates most of the GR-mediated repression, although basal activity was reduced. This may be due to the removal of two important transcription factor binding sites (NFB and AP-1) in this construct. This deletion analysis suggests that the functional nGRE is located between –1200 and – 860 region of the hfasL promoter. GR inhibits fasL by direct binding to the negative GRE at –990 site
Figure 2. A GR dimer is required for the inhibition of the fasL expression. A) A GRE-containing MMTV-promoter construct was cotransfected with the vector control (no GR) or the construct expressing wtGR (GR) or a GR-dimerization mutant (GRdim). Transfected cells were either untreated (gray bars) or treated with 100 nM Dex (black bars). Promoter activity is shown as a percentage of luciferase expression (left panel). Percent of relative activation/inhibition is calculated from the luciferase expression data presented on the left panel (right panel). B) Same as in panel A but with collagenase 1 (AP-1 site containing promoter). C) Same as in panel A but with proopiomelanocortin (POMC) promoter. D) Same as in A but with (–1200/⫹100) FasL-Luc as a reporter. GR COMPETES WITH NFB FOR FASL PROMOTER BINDING
To determine which of the predicted nGREs are capable of GR binding, we performed ABCD analysis of the predicted oligonucleotides. This assay identified three weak GR binding sites and one nGRE in the distal promoter region (–990 relative to transcription start) that displayed the highest affinity for GR (Fig. 4A). Two-point mutations introduced into the –990 nGREcontaining oligonucleotide abolished GR DNA binding in vitro, as shown by ABCD (Fig. 4B). The same mutations introduced into the full-length FasL promoter reduced GR-mediated inhibition of reporter gene expression down to 40% (Fig. 4C, D). Mutation of the weak nGREs did not change GR-mediated repression of the reporter (data not shown). The high affinity of GR to the –990 nGRE and its importance for the inhibition of the FasL-Luc suggest it as a major nGRE site responsible for GR-mediated cis-repression of fasL. GR competes with NFB for the binding to overlapping response elements To further elucidate the mechanism of GR-mediated inhibition of the –990 nGRE, we performed a detailed analysis of this site and the surrounding sequences. This search revealed that one of the half-sites of the –990 nGRE completely overlaps with a known NFB binding site (Fig. 5A) (28). To test whether there is competition between GR and NFB for binding to the 1077
Figure 3. The distal fasL promoter is responsible for most of the Dex-dependent inhibition. A) Schematic representation of the hFasL promoter with response elements important for hFasL activation in T cells (indicated as rectangles). Black: NFAT, white: Egr, dashed to the right AP-1; dashed to the left: IRF; checkered: NFB; dotted: Myc. Predicted nGREs are indicated by asterisks. An nGRE search pattern used for the in silico screening of hFasL promoter is shown at the top of the figure. B) Serially deleted luciferase reporter constructs of the hfasL promoter (–1200/ ⫹100; – 860/⫹100, –300/⫹100), –158/⫹100) were cotransfected with the GR-expressing construct into HepG2 cells. After treatment with 100 nM Dex for 24 h (black bars) luciferase activity was measured and compared to the untreated samples (gray bars). C) Percent of the GR-mediated inhibition calculated from the reporter assay shown in panel B.
–990 nGRE, we set up an ABCD assay using lysates of the GR-negative cell line U2OS and recombinant GR. As shown in Fig. 5B, addition of increasing amounts of recombinant GR attenuated NFB binding to the –990 nGRE but neither to the mutated nGRE nor to the 2xNFB oligonucleotide representing a tandem of two NFB response elements of the I-B␣ promoter. This result suggests that GR inhibits the activation of FasL expression at the –990 nGRE via sterical occlusion of the overlapping NFB binding element. Since in the ABCD assay GR binds to its response element even in the absence of ligand (due to the lack of a physical barrier between the nuclear and cytoplasmic fractions in cell lysates), we could not trace the behavior of the endogenous GR on the –990 nGRE using this method. To test whether there is a similar competition between endogenous GR and NFB on the natural FasL pro-
moter, we employed a chromatin immunoprecipitation assay (ChIP). As shown in Fig. 5C, the ChIP analysis in Jurkat cells confirmed binding of GR to the FasL –990 nGRE, as well as the significant (threefold) reduction of NFB binding on Dex treatment.
DISCUSSION Our analysis of the hFasL promoter has shown that hfasL is negatively regulated by GCs via binding of GR to its cognate –990 nGRE site. This GR binding masks the overlapping NFB site, thereby interfering with NFB-mediated hfasL transactivation. Multiple mechanisms of the GR-NFB antagonism have been described (reviewed in ref 29). Most are believed to be DNA binding-independent based on protein-protein interac-
Figure 4. GR inhibits fasL by direct binding to the negative GRE at –990 site. A) Mapping of the fasL promoter by ABCD assay. Biotinylated oligonucleotides corresponding to the predicted nGREs were analyzed for their in vitro GR binding by ABCD assay. B) Biotinylated oligonucleotides of the –990 nGRE (wild type or mutated at two positions, nGRE990m) were used for the ABCD assay. Oligonucleotides representing the GRE of tyrosine aminotransferase (TAT) known to bind GR (GRE) and its mutated version (GREm) were used as positive and negative controls, respectively. Sequences of the –990 nGRE and –990 nGREm are indicated below the ABCD blot. Substitutions in the –990 nGREm are marked by bold letters, two half-sites of –990 nGRE are enclosed in black frames. C) hFasL-Luc, its mutated analog nGRE990m carrying two point mutations shown in (B) and the truncated FasL promoter (⫹100/⫺860) reporters were transiently cotransfected with the GRexpressing construct into HepG2 cells. Luciferase activity was measured after 24 h treatment with 100 nM Dex and compared with the untreated cells (black vs. gray bars). D) Percent of the GR-mediated inhibition calculated from the reporter assay shown in panel C. 1078
Vol. 20
June 2006
The FASEB Journal
NOVAC ET AL.
Figure 5. GR competes with NFB for binding to overlapping response elements at –990 nGRE. A) DNA sequence of the –990 nGRE and the overlapping NFB binding element. Two half-sites of the –990 nGRE and NFB binding element are enclosed in black frames. B) Biotinylated oligonucleotides of nGRE990, nGRE990m, and 2xNFkB were used in an ABCD assay with U2OS cell lysates. NFB p50 and GR binding was monitored without and with the addition of the indicated amounts of recombinant GR. C) ChIP was performed in Jurkat cells treated with the indicated substances and immunoprecipitated with either GR or NFB antibodies. Input represents PCR amplification of the unprecipitated DNA. PCR of the irrelevant gene U6 snRNA was used as a negative control.
tion between GR and NFB family members or mediated via the induction of the NFB inhibitor I-B by GR. Therefore, the mechanism of GR-mediated sterical occlusion represents a novel additional way of GRNFB antagonism, which, unlike the others, is dependent on GR DNA binding. In our analysis of the hfasL promoter, we always observed residual GR-mediated repression of hfasL even after inactivation or removal of –990 nGRE. This repression may be explained by the GR-mediated transrepression of fasL. Probably both mechanisms are involved in fasL regulation. There are many transcription factors whose transcriptional activity might be interfered by GR, like AP-1 or NFAT. Several sites for these transcription factors have been found in the fasL promoter. The GR-mediated fasL repression is also transient, and this might be in accordance with the biological significance of fasL inhibition. During the later stages of an inflammatory response, AICD is required for the elimination of excessive T cells from the inflammatory stage, thereby terminating the immune response. The anti-inflammatory action of GCs should logically support or at least not interfere with this process, thus promoting fasL activation. On the other hand, cells dying from apoptosis are rapidly cleared by macrophages, and GR-mediated inhibition of fasL during AICD might be required to slow down T cell apoptosis whenever the massive cell death overwhelms the macrophage clearance, as is the case during strong inflammation and in viral infections (30). Thus, FasL inhibition should be transient as it is not intended to completely prevent AICD but only to slow down this process. GR-mediated fasL inhibition would contribute to the anti-inflammatory action of GCs preventing the second wave of inflammation as noncleared apoptotic cells may progress to secondary necrosis, thereby triggering another wave of inflammation (31, 32). Although the GRE consensus motif consisting of two GR COMPETES WITH NFB FOR FASL PROMOTER BINDING
inverted hexameric repeats separated by three nucleotides, was identified more than 10 years ago, only a few genes containing GREs in their regulatory sequences have been described. Those include enzymes involved in gluconeogenesis, such as tyrosine aminotransferase, alanine aminotransferase, and phosphoenolpyruvate carboxykinase. Another GRE-containing promoter that has become a classical model to study GR-mediated transcriptional cis-activation is the MMTV promoter. Much less attention has been paid to the genes containing negative GREs. This might be explained in part by the nGRE consensus, being very loose compared to GRE, such as only one half-site of it appears to be conserved in nGREs. The lack of well-conserved response elements hinders the identification of nGREs by simple in silico promoter sequence analysis. In the early 1990s, several nGRE-containing promoters, including those for POMC, CRH, and prolactin, were identified using footprinting analysis. Since then only a few reports have been published on nGREs, and the exact molecular mechanism by which the expression of these genes is inhibited by GR is not fully understood. In the case of trans-repression, GR has been shown to recruit corepressors such as N-CoR and RIP140, or even directly interact with HDACs, thereby generating the repressive chromatin context. Whether GR can actively repress transcription by binding to nGRE similar to trans-repression remains a conundrum. In the case of the hfasL promoter, the GR antagonist RU-486 represses promoter activity almost as well as the GR agonist dexamethasone does. It is known that the conformation of the RU-486 bound GR does not permit the binding of coactivators, but instead favors the binding of corepressors (24). This finding indicates that DNA binding of GR rather than the recruitment of the specific cofactors may be the critical determinant of FasL promoter inhibition. The only mechanism responsible for GR cis-inhibition of nGRE-containing promoters described so far is 1079
steric occlusion. At least two nGRE-containing promoters are regulated by GR in a similar manner to hfasL. In the POMC promoter, GR binds an nGRE overlapping the Nur77 binding site, thereby preventing recruitment of this transcription factor to the promoter (33). Another example of sterical occlusion has been described for the osteocalcin promoter nGRE, which overlaps with the TATA box. GR binding to this nGRE interferes with TBP binding and the subsequent assembly of the basal transcription machinery (34, 35). The mechanism of sterical occlusion explains why these genes cannot be activated by factors whose sequences are overlapping with nGREs, but it does not explain why GR is unable to activate gene transcription as it does on positive GREs on DNA binding. One possible explanation for this could be that the nGRE consensus DNA sequence can induce the inhibitory conformation of GR. Indeed, our preliminary results using partial protease digestion show that the GR conformation on nGREs differs from that on GREs (A. Dostert, unpublished results). The precedent for this hypothesis are B sites, which can determine not only transcription factor binding selectivity but also the specificity for the second layer of cofactors (36). The nGRE site may also represent a type of composite site, which would recruit another transcription factor together with GR and that transcription factor may dictate the inactive GR conformation. GR trans-repression is believed to be responsible for the anti-inflammatory action of GCs whereas GR cisregulation has been implicated in GC-mediated metabolic side effects. According to this model, a promising novel pharmaceutical strategy could be based on dissociated GR ligands, discriminating cis, and trans-GR activities (37). However, the fact that FasL, an important regulator of inflammatory cell vitality, is cis-regulated may change our view on this issue. Understanding the role of cis- and trans- repressive mechanisms of GR in the context of inflammation should help in the future design of drugs specific to each particular type of inflammation.
5. 6.
7.
8.
9.
10.
11.
12.
13. 14. 15. 16.
17.
18.
We thank Cheng Kong Shih and Susan Goldrick for recombinant GR, Min Li-Weber and A. Cato for provided constructs, Maresa Eck and AnneMarie Schimp for their excellent technical assistance. This work was supported by a DFG grant to T.H.
19.
20.
REFERENCES 1.
Schleimer, R. P. (2000) Interactions between the hypothalamicpituitary-adrenal axis and allergic inflammation. J. Allergy Clin. Immune. 106, S270 –274 2. Umland, S. P., Schlemiel, R. P., and Johnston, S. L. (2002) Review of the molecular and cellular mechanisms of action of glucocorticoids for use in asthma. Pulm. Pharmacol. Ther. 15, 35–50 3. Reichardt, H. M. (2004) Immunomodulatory activities of glucocorticoids: insights from transgenesis and gene targeting. Curr. Pharm. Des. 10, 2797–2805 4. Buttgereit, F., Burmester, G. R., and Lipworth, B. J. (2005) Optimised glucocorticoid therapy: the sharpening of an old spear. Lancet 365, 801– 803
1080
Vol. 20
June 2006
21.
22.
23.
Dostert, A., and Heinzel, T. (2004) Negative glucocorticoid receptor response elements and their role in glucocorticoid action. Curr. Pharm. Des. 10, 2807–2816 Kamei, Y., Xu, L., Heinzel, T., Torchia, J., Kurokawa, R., Gloss, B., Lin, S. C., Heyman, R. A., Rose, D. W., Glass, C. K., and Rosenfeld, M. G. (1996) A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 85, 403– 414 De Bosscher, K., Schmitz, M. L., Vanden Berghe, W., Plaisance, S., Fiers, W., and Haegeman, G. (1997) Glucocorticoid-mediated repression of nuclear factor-kappaB-dependent transcription involves direct interference with transactivation. Proc. Natl. Acad. Sci. USA 94, 13504 –13509 Nissen, R. M., and Yamamoto, K. R. (2000) The glucocorticoid receptor inhibits Unflappable by interfering with serine-2 phosphorylation of the RNA polymerase II carboxyl-terminal domain. Genes Dev. 14, 2314 –2329 Subramaniam, N., Teeter, E., and Okra, S. (1999) Receptor interacting protein RIP140 inhibits both positive and negative gene regulation by glucocorticoids. J. Biol. Chem. 274, 18121– 18127 Ito, K., Barnes, P. J., and Adcock, I. M. (2000) Glucocorticoid receptor recruitment of his tone deacetylase 2 inhibits interleukin-1beta-induced his tone H4 acetylating on lysine’s 8 and 12. Mol. Cell. Biol. 20, 6891– 6903 Paliogianni, F., Abuja, S. S., Below, J. P., Below, J. E., and Bumpass, D. T. (1993) Novel mechanism for inhibition of human T cells by glucocorticoids. Glucocorticoids inhibit signal transduction through IL-2 receptor. J. Immune. 151, 4081– 4089 Peter, M. E., Hein, J., Egret, A., Halberd, S., Walkman, H., Maidenhair, G., and Kramer, P. H. (1995) APO-1 (CD95) -dependent and -independent antigen receptor-induced apoptosis in human T and B cell lines. Int. Immune. 7, 1873–1877 Bellgrau, D., Gold, D., Salary, H., Moore, J., Franzusoff, A., and Duke, R. C. (1995) A role for CD95 ligand in preventing graft rejection. Nature 377, 630 – 632 Griffith, T. S., Brunner, T., Fletcher, S. M., Green, D. R., and Ferguson, T. A. (1995) Fas ligand-induced apoptosis as a mechanism of immune privilege. Science 270, 1189 –1192 Kavurma, M. M., and Khachigian, L. M. (2003) Signaling and transcriptional control of Fas ligand gene expression. Cell Death Differ. 10, 36 – 44 Yang, Y., Mercep, M., Ware, C. F., and Ashwell, J. D. (1995) Fas and activation-induced Fas ligand mediate apoptosis of T cell hybridomas: inhibition of Fas ligand expression by retinoic acid and glucocorticoids. J. Exp. Med. 181, 1673–1682 Yerramasetti, R., Gollapudi, S., and Gupta, S. (2002) Rifampicin inhibits CD95-mediated apoptosis of Jurkat T cells via glucocorticoid receptors by modifying the expression of molecules regulating apoptosis. J. Clin. Immunol. 22, 37– 47 Zipp, F., Wendling, U., Beyer, M., Grieger, U., Waiczies, S., Wagenknecht, B., Haas, J., and Weller, M. (2000) Dual effect of glucocorticoids on apoptosis of human autoreactive and foreign antigen-specific T cells. J. Neuroimmunol. 110, 214 –222 Srivastava, R. K., Sasaki, C. Y., Hardwick, J. M., and Longo, D. L. (1999) Bcl-2-mediated drug resistance: inhibition of apoptosis by blocking nuclear factor of activated T lymphocytes (NFAT)induced Fas ligand transcription. J. Exp. Med. 190, 253–265 Baumann, S., Dostert, A., Novac, N., Bauer, A., Schmid, W., Fas, S. C., Krueger, A., Heinzel, T., Kirchhoff, S., Schu¨tz, G., and Krammer, P. H. (2005) Glucocorticoids inhibit activation-induced cell death (AICD) via direct DNA-dependent repression of the CD95 ligand gene by a glucocorticoid receptor dimer. Blood 106, 617– 625 Li-Weber, M., Laur, O., Hekele, A., Coy, J., Walczak, H., and Krammer, P. H. (1998) A regulatory element in the CD95 (APO-1/Fas) ligand promoter is essential for responsiveness to TCR-mediated activation. Eur. J. Immunol. 28, 2373–2383 Heck, S., Kullmann, M., Gast, A., Ponta, H., Rahmsdorf, H. J., Herrlich, P., and Cato, A. C. (1994) A distinct modulating domain in glucocorticoid receptor monomers in the repression of activity of the transcription factor AP-1. EMBO J. 13, 4087– 4095 Sengupta, S., and Wasylyk, B. (2001) Ligand-dependent interaction of the glucocorticoid receptor with p53 enhances their degradation by Hdm2. Genes Dev. 15, 2367–2380
The FASEB Journal
NOVAC ET AL.
24.
25.
26.
27.
28.
29. 30.
Metivier, R., Penot, G., Hubner, M. R., Reid, G., Brand, H., Kos, M., and Gannon, F. (2003) Estrogen receptor-alpha directs ordered, cyclical, and combinatorial recruitment of cofactors on a natural target promoter. Cell 115, 751–763 Reichardt, H. M., Kaestner, K. H., Tuckermann, J., Kretz, O., Wessely, O., Bock, R., Gass, P., Schmid, W., Herrlich, P., Angel, P., and Schu¨tz, G. (1998) DNA binding of the glucocorticoid receptor is not essential for survival. Cell 93, 531–541 Prefontaine, G. G., Lemieux, M. E., Giffin, W., Schild-Poulter, C., Pope, L., LaCasse, E., Walker, P., and Hache, R. J. (1998) Recruitment of octamer transcription factors to DNA by glucocorticoid receptor. Mol. Cell. Biol. 18, 3416 –3430 Nakai, Y., Usui, T., Tsukada, T., Takahashi, H., Fukata, J., Fukushima, M., Senoo, K., and Imura, H. (1991) Molecular mechanisms of glucocorticoid inhibition of human proopiomelanocortin gene transcription. J. Steroid Biochem. Mol. Biol. 40, 301–306 Kasibhatla, S., Brunner, T., Genestier, L., Echeverri, F., Mahboubi, A., and Green, D. R. (1998) DNA damaging agents induce expression of Fas ligand and subsequent apoptosis in T lymphocytes via the activation of NF-kappa B and AP-1. Mol. Cell 1, 543–551 Smoak, K. A., and Cidlowski, J. A. (2004) Mechanisms of glucocorticoid receptor signaling during inflammation. Mech. Ageing Dev. 125, 697–706 Wyllie, A. H. (1985) The biology of cell death in tumours. Anticancer Res. 5, 131–136
GR COMPETES WITH NFB FOR FASL PROMOTER BINDING
31. 32. 33.
34.
35.
36. 37.
Devitt, A., and Gregory, C. D. (2004) Measurement of apoptotic cell clearance in vitro. Methods Mol. Biol. 282, 207–221 Lucas, M., Stuart, L. M., Savill, J., and Lacy-Hulbert, A. (2003) Apoptotic cells and innate immune stimuli combine to regulate macrophage cytokine secretion. J. Immunol. 171, 2610 –2615 Philips, A., Maira, M., Mullick, A., Chamberland, M., Lesage, S., Hugo, P., and Drouin, J. (1997) Antagonism between Nur77 and glucocorticoid receptor for control of transcription. Mol. Cell. Biol. 17, 5952–5959 Meyer, T., Gustafsson, J. A., and Carlstedt-Duke, J. (1997) Glucocorticoid-dependent transcriptional repression of the osteocalcin gene by competitive binding at the TATA box. DNA Cell Biol. 16, 919 –927 Okabe, T., Takayanagi, R., Adachi, M., Imasaki, K., and Nawata, H. (1998) Nur77, a member of the steroid receptor superfamily, antagonizes negative feedback of ACTH synthesis and secretion by glucocorticoid in pituitary corticotrope cells. J. Endocrinol. 156, 169 –175 Leung, T. H., Hoffmann, A., and Baltimore, D. (2004) One nucleotide in a kappaB site can determine cofactor specificity for NF-kappaB dimers. Cell 118, 453– 464 Scha¨cke, H., Rehwinkel, H., and Asadullah, K. (2005) Dissociated glucocorticoid receptor ligands: compounds with an improved therapeutic index. Curr. Opin. Invest. Drugs 6, 503–507 Received for publication November 29, 2005. Accepted for publication January 31, 2006.
1081
