Identification of Target Genes Involved in the

0888-8809/05/$15.00/0 Printed in U.S.A.

Molecular Endocrinology 19(3):632–643 Copyright © 2005 by The Endocrine Society doi: 10.1210/me.2004-0294

Identification of Target Genes Involved in the Antiproliferative Effect of Glucocorticoids Reveals a Role for Nuclear Factor-␬B Repression Lars-Go¨ran Bladh, Johan Lide´n, Ahmad Pazirandeh, Ingalill Rafter, Karin Dahlman-Wright, Stefan Nilsson, and Sam Okret Departments of Medical Nutrition (L.-G.B., A.P., I.R., S.O.) and Biosciences (J.L., K.D.-W.), Karolinska Institutet, Karolinska University Hospital Huddinge, Novum, SE-141 86 Huddinge, Sweden; and KaroBio AB (L.-G.B., S.N.), Novum, SE-141 57 Huddinge, Sweden Glucocorticoid hormones (GCs) exert an antiproliferative effect on most cells. However, the molecular mechanism is still largely unclear. We investigated the antiproliferative mechanism by GCs in human embryonic kidney 293 cells with stably introduced glucocorticoid receptor (GR) mutants that discriminate between cross-talk with nuclear factor-␬B (NF-␬B) and activator protein-1 signaling, transactivation and transrepression, and antiproliferative vs. non-antiproliferative responses. Using the GR mutants, we here demonstrate a correlation between repression of NF-␬B signaling and antiproliferative response. Gene expression profiling of endogenous genes in cells containing mutant GRs identified a limited number of genes

that correlated with the antiproliferative response. This included a GC-mediated up-regulation of the NF-␬B-inhibitory protein I␬B␣, in line with repression of NF-␬B signaling being important in the GCmediated antiproliferative response. Interestingly, the GC-stimulated expression of I␬B␣ was a direct effect despite the inability of the GR mutant to transactivate through a GC-responsive element. Selective expression of I␬B␣ in human embryonic kidney 293 cells resulted in a decreased percentage of cells in the S/G2/M phase and impaired cell proliferation. These results demonstrate that GCmediated inhibition of NF-␬B is an important mechanism in the antiproliferative response to GCs. (Molecular Endocrinology 19: 632–643, 2005)

G

GCs act by binding to the intracellular GC receptor (GR), which then stimulates (transactivates) or inhibits (transrepresses) transcription of specific target genes (8, 9). With regard to GC inhibition of cell proliferation, GR may either enhance the expression of genes that exert an antiproliferative effect or alternatively repress expression of genes, which stimulate cell proliferation. Stimulation of gene expression by the GR can mechanistically occur by two distinct ways. In the first case, the ligand-activated GR interacts directly with specific DNA sequences, designated glucocorticoid response elements (GREs), present in the target genes. In the second case, GR-mediated stimulation of gene expression takes place by a GR interaction with other DNA-bound transcription factors without itself contacting DNA (Refs. 8–10 and references therein). This GR interaction with other DNA-bound transcription factors is usually referred to as tethering. Tethering seems to be the most common mechanism by which GR represses gene expression. In this case, GR binds to and interferes with the transcriptional activity of other DNA-bound transcription factors, as exemplified by GR-mediated inhibition of nuclear factor-␬B (NF␬B) and activator protein 1 (AP-1) activity (8). In some cases, repression of gene expression seems to involve a direct GR interaction with so-called negative GREs present in target genes (8, 9). Irrespective of the mechanism of GR target gene regulation, binding to GREs or negative GREs, or repression of NF-␬B or AP-1

LUCOCORTICOID HORMONES (GCs) regulate a multitude of physiological processes that include metabolism, immune responses, differentiation, and proliferation. For most cell types, GCs exert an antiproliferative effect (1–6). The antiproliferative effect is of clinical relevance, because decreased cell proliferation is involved in some of the severe side effects seen after long-term GC treatment. This includes, e.g. skin atrophy associated with dermatitis treatment, impaired wound healing, and osteoporosis possibly due to inhibition of osteoblast proliferation (7). The antiproliferative effect of GCs have been known for decades, yet, the molecular mechanisms are still poorly understood. First Published Online November 4, 2004 Abbreviations: ALP, Alkaline phosphatase; AP-1, activator protein 1; CHX, cycloheximide; CMV, cytomegalovirus; DBD, DNA-binding domain; Dex, dexamethasone; Dox, doxycycline; GC, glucocorticoid hormone; GFP, green fluorescence protein; GR, GC receptor; GRE, glucocorticoid response element; HEK, human embryonic kidney; MMP1-luc, matrix metalloproteinase-1 luciferase; MMP-1, matrix metalloproteinase-1 collagenase-1; NF-␬B, nuclear factor-␬B; PI, propidium iodide; rtTA, tetracycline-regulated reverse tTA transactivator; TPA, 12-O-tetradecanoyl-phorbol-13-acetate; TR␤, thyroid hormone receptor-␤; wtGR, wild-type GR. Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community. 632

Bladh et al. • GC-Regulated Antiproliferative Target Genes

activity, the DNA-binding domain (DBD) of the GR seems to be involved (11–14). NF-␬B, a transcription factor complex, is most commonly a heterodimer between p65 (RelA) and p50 (NF-␬B1). It regulates many cellular functions such as inflammation, immune response, apoptosis, oncogenesis, and differentiation (15, 16). It has also been implicated in regulation of cell growth (17). NF-␬B is activated by a number of stimuli that include cytokines, mitogens, phorbol esters, radiation, cellular stress, and bacterial or viral products. In nonstimulated cells, NF-␬B is sequestered in the cytoplasm in an inactive form through its association with one of several NF-␬B-inhibitory molecules (I␬Bs) including I␬B␣. Activation of NF-␬B involves a stimuli-initiated degradation of I␬B␣ through a signal cascade that causes phosphorylation at serines 32 and 36 in the I␬B␣ protein. This subsequently results in the release of NF-␬B, its nuclear translocation, binding to NF-␬B response elements in target genes, and initiation of transcription. Mutation of serines 32 and 36 results in a stable (superrepressive) form of I␬B␣ that very efficiently prevents NF-␬B activation (18). AP-1 is a dimeric transcription factor complex consisting of proteins that include the Jun and Fos subgroups of transcription factors. Jun and Fos proteins may also dimerize with other transcription factors such as members of the ATF (activating transcription factor) or Maf families (19, 20). AP-1 is activated by a plethora of physiological stimuli and environmental cellular stressors and regulates a wide range of cellular processes including cell proliferation, apoptosis, transformation, and differentiation. Many signals that activate NF-␬B will activate AP-1. Furthermore, several target genes including both inflammatory cytokines and cell cycle regulators contain binding sites for both NF-␬B and AP-1 (Refs. 8 and 21 and references therein). Based on the knowledge that AP-1, as well as NF␬B, are involved in the regulation of cell proliferation and that the activity of these transcription factors are repressed by activated GR, it has been hypothesized that GR-AP-1 and/or NF-␬B cross-talk is a mechanism involved in the antiproliferative effects exerted by GCs. However, so far, this has not been experimentally demonstrated. In this report, we tested this hypothesis using GR DBD mutants that separate GR-mediated repression of NF-␬B and AP-1 signaling, respectively. We used GR mutants in which subdomains of the GR DBD had been exchanged with the corresponding region of the thyroid hormone receptor ␤ (TR␤). The basis for this approach was that previous results have indicated that different GR DBD subdomains are involved in NF-␬B and AP-1 repression, respectively (11, 12). Furthermore, the mutants are unable to transactivate via GREs, excluding genes thought to be activated through this mechanism from being involved in the GC-mediated antiproliferative effect. Using two GR DBD mutants, one able to exert an antiproliferative effect and the other not stably introduced into human

Mol Endocrinol, March 2005, 19(3):632–643 633

embryonic kidney (HEK)293 cells, we found a better correlation between the antiproliferative response and NF-␬B repression than to AP-1 repression. Gene expression profiling identified the I␬B␣ gene as a primary candidate target gene involved in the GC-mediated inhibition of cell proliferation. This was in line with the observation that the NF-␬B signaling pathway is an important target for the antiproliferative effect by GCs. This was verified by selective expression of the I␬B␣ gene.

RESULTS Generation of GR DBD Mutants Separating Cross-Talk with AP-1 and NF-␬B, Respectively To examine the role of cross-talk with the AP-1 and NF-␬B signaling pathways for the antiproliferative response of GCs, we stably introduced GR DBD mutants into Flp-In HEK293 cells. Flp-In HEK293 cells, which contain a single integrated Flp recombination target site, allow stable GR integration at a specific genomic site and, subsequently, similar GR expression in all cell clones. Previous experiments have indicated that different parts of the GR DBD are involved in transactivation vs. transrepression including crosstalk with AP-1 and NF-␬B signaling. We used chimeric GR proteins in which various parts of the DBD, i.e. the N- or C-terminal zinc finger or the linker region in between the two fingers, have been replaced with the corresponding region(s) of the TR␤ (Fig. 1). We generated pooled cell clones of Flp-In HEK293 cells expressing the different GR mutants. They all expressed similar amounts of GR protein as analyzed by Western blotting (Fig. 2A). A ligand-binding assay showed that the receptor number was approximately 80,000 GR molecules per cell (data not shown). Nontransfected

Fig. 1. Schematic Representation of the Different GR Mutants with Chimeric GR/TR␤ DBDs Used for Identifying Regions in the GR DBD Important for Repression of NF-␬B and AP-1 Activity and Involved in the Antiproliferative Response GR mutants were created in the context of the full-length GR, in which individual parts of the GR DBD (open bars) were replaced by the corresponding regions of the TR␤ DBD (gray bars). For this purpose the GR DBD was divided into three parts, the N-terminal zinc finger, the linker region, and the C-terminal zinc finger, respectively.

634 Mol Endocrinol, March 2005, 19(3):632–643

Fig. 2. Transactivation and Transrepression in HEK293 Flp-In Cells Stably Transfected with wtGR and GR Mutants A, Western blotting of GR in nontransfected (⫺) and stably transfected HEK293 Flp-In cells with wtGR (GgggG) and GR mutants. B, HEK293 Flp-In cells were transfected with the reporter gene 2xGRE-luc. Fold induction represents the ratio of the luciferase activity between nontreated (EtOH) and Dex (100 nM)-treated cells after 20 h. C, Repression of NF-␬B activity by the different GRs was tested after transfection of the stably GR expressing HEK293 Flp-In cells with the 3xNF␬B-luc reporter gene. Cells were left untreated (open bars) or treated with 5 ng/ml TPA in the absence (black bars) or presence of 100 nM Dex (gray bars) for 20 h. The relative light units (RLU) after treatment with TPA alone was set to 100. D, Repression of AP-1 activity by the different GRs was tested after transfection of the stably GR expressing HEK293 Flp-In cells with the MMP1-luc reporter gene. Cells were left


Flp-In HEK293 cells contain no, or very low levels, of endogenous GR protein (Fig. 2A). Transfection of the different clones with a 2xGREluc plasmid showed that only the wtGR (GgggG) was able to transactivate the GRE-controlled reporter gene construct (Fig. 2B). Analysis of the ability of the GR DBD mutants in the presence of dexamethasone (Dex) to repress NF-␬B activity, after transfection with the 3xNF-␬B-luc reporter gene, demonstrated that repression of NF-␬B activity in HEK293 cells relied solely on the presence of the C-terminal zinc finger of GR (Fig. 2C). Replacement of the C-terminal zinc finger of the GR by the corresponding zinc-finger from the TR␤ (GggtG) totally abolished the ability of the receptor to repress NF-␬B activity. In fact, this variant displayed a weak stimulation of NF-␬B activity. Consistent with a critical role for the second zinc finger in mediating GR repression of NF-␬B activity, most of the repressive activity remained when the C-terminal zinc finger alone was from the GR (i.e. GttgG). If the complete GR DBD was replaced by the corresponding region from TR␤ (GgggG 3 GtttG), the resulting variant displayed less than 10% of the NF-␬B repression of the wild-type protein. To examine the ability of the different GR DBD mutants to repress AP-1 activity, the GR DBD mutantcontaining Flp-In HEK293 cells were transfected with the matrix metalloproteinase-1 (MMP1) luciferase reporter gene (MMP1-luc). It has previously been shown that repression of the MMP-1 (collagenase-1) gene by GCs takes place by cross-talk between the GR and AP-1 at the AP-1 site present in the MMP-1 promoter [⫺72 to ⫺65 bp from the transcription start site (22)] through a tethering mechanism. Whereas the wtGR (GgggG) repressed the MMP1-luc activity by 90%, the GtttG, GttgG, and the GggtG mutants retained the ability to repress the activity by 40–60% (Fig. 2D). Only the GgttG receptor had completely lost its ability to repress the MMP-1 reporter gene. The GttgG and GtttG mutants were also able to repress AP-1 activity from a reporter gene controlled by five synthetic AP-1 response elements, similar to the wtGR (Fig. 2E). Furthermore, this demonstrated the NF-␬B and AP-1 discriminatory property of GtttG. Most important, the above results from the GR DBD mutants demonstrate that NF-␬B and AP-1 repression can be separated as demonstrated by the effect of the GtttG and GggtG receptors on NF-␬B and AP-1 activity, respectively.

untreated (unfilled bars), treated with 5 ng/ml TPA (black bars), or treated with 5 ng/ml TPA and 100 nM Dex (gray bars) for 20 h. The RLU after treatment with TPA alone was set to 100. E, Nontransfected or stably transfected HEK293 Flp-In cells with GgggG, GtttG, and GttgG were transiently transfected with the 5xAP1-ALP reporter and left untreated or treated as above. Alkaline phosphatase activity was measured 20 h after treatment.


Correlation between NF-␬B-, But Not AP-1 Repression, and the Antiproliferative Response to GCs in HEK293 Cells The Flp-In HEK293 cells stably transfected with the GR DBD mutants were analyzed for the effect on proliferation after treatment with 100 nM Dex. As can be seen in Fig. 3A, Dex treatment of cells containing the wtGR (GgggG) resulted in an antiproliferative re-


sponse. No effect on proliferation by Dex was seen in the nontransfected, GtttG, GggtG, or GgttG cells. However, maintaining the GR C-terminal zinc finger alone was sufficient to preserve most of the antiproliferative response after Dex administration, as demonstrated by the GttgG cells (Fig. 3A). Flow cytometric analysis of the cell cycle of GttgG HEK293 Flp-In cells showed that the inhibition of cell proliferation was associated with a reduction in the percentage of cells in the S/G2/M phase from 42.6 ⫾ 0.5% to 33.3 ⫾ 0.6% (n ⫽ 3, P ⬍ 0.001) (Fig. 3B) with a corresponding increase of cells in the G1 phase, similar to what was observed in the wtGR cells (data not shown). No indication of apoptosis was observed after Dex treatment (data not shown). No significant effect of Dex on the distribution of the percentage of cells in the S/G2/M phase in GtttG cells was observed (45.0 ⫾ 1.0% in control vs. 42.8 ⫾ 0.7% in Dex-treated cells, n ⫽ 3) (Fig. 3B). These results show that the antiproliferative effect is dependent on the GR C-terminal zinc finger. As the GttgG receptor is unable to transactivate a GRE-dependent reporter gene, the antiproliferative effect exerted by GCs in HEK293 cells is independent of direct GR interaction with a GRE. It was observed that GtttG- and GggtG-containing cells showed no Dexmediated inhibition of cell proliferation and were unable to repress NF-␬B but still repressed AP-1 activity, whereas the GttgG cells demonstrated a Dex-mediated antiproliferative response and inhibited both AP-1 and NF-␬B activity (see above). Thus, a correlation between NF-␬B repression, rather than AP-1 repression, and antiproliferative response is demonstrated. Gene Expression Profiling after Dex Administration Identified I␬B␣ as a Potential Candidate Gene Responsible for the Antiproliferative Effect of GCs in HEK293 Cells

Fig. 3. Analysis of Cell Proliferation and Cell Cycling of DexTreated HEK293 Flp-In Cells Stably Transfected with wtGR and GR Mutants A, Proliferation of nontransfected and stably transfected HEK293 Flp-In cells with wtGR (GgggG) and the GR mutants GggtG, GgttG, GttgG, and GtttG was analyzed by counting cell numbers at various time points in the absence (open symbols and dashed line) or presence of 100 nM Dex (solid symbols and solid line). B, Flow cytometric analysis of cell cycle. Shown are DNA histograms of GttgG and GtttG HEK293 Flp-In cells after treatment with control vehicle or 100 nM Dex for 24 h. (Shown is one representative experiment of three). Cells were stained with PI.

We compared GC-regulated genes in the GtttG and the GttgG cell lines, which behave similarly with regard to their inability to transactivate through GRE and ability to transrepress AP-1 activity but differ in their effect on NF-␬B repression and antiproliferative response. Using this approach, we can exclude genes activated through GR binding to GREs as well as target genes regulated by cross-talk between GR and AP-1 signaling as part of the antiproliferative response. Comparing gene expression profiles from these two cell lines will thus limit the number of candidate genes involved in the GC-mediated antiproliferative response. To identify these candidate genes in HEK293 cells after GC administration, we performed gene expression profiling using the Affymetrix Human Genome Focus Array assaying approximately 8500 verified transcripts. Statistical analysis identified 18 genes that were significantly up-regulated at least 1.7-fold and three that were repressed by at least 30% in GttgG cells after Dex administration (Table 1). In GtttG cells, six genes were induced at least 1.7-fold, and one gene was repressed

636

Mol Endocrinol, March 2005, 19(3):632–643


Table 1. GC-Regulated Genes in GttgG and GtttG Containing HEK293 Cells Mean Ratio ⫾ Dex

RefSeq Transcript ID

Probe Set ID

GttgG 1. NM_002276 2. NM_006162

201650_at 210162_s_at

5.0 4.2

0.000034 0.000016

17.2 20.8

3. NM_003068 4. NM_000104

213139_at 202437_s_at

3.9 3.6

0.0021 0.000052

5.9 15.4

5. NM_020529

201502_s_at

3.3

0.00041

9.1

6. NM_002925 7. NM_013230

204319_s_at 208650_s_at

3.0 2.9

0.00018 0.0020

11.3 5.9

8. NM_000610

212063_at

2.6

0.00081

7.6

210002_at 220234_at 203963_at 202071_at 203365_s_at

2.4 2.4 2.4 2.3 2.1

0.00030 0.00017 0.00010 0.000064 0.00088

9.9 11.4 13.2 14.6 7.4

14. NM_002923 15. NM_181671

202388_at 219155_at

2.0 2.0

0.0036 0.00070

5.1 7.9

16. NM_004199

202733_at

1.8

0.0020

6.0

17. NM_002485 18. NM_000161

202907_s_at 204224_s_at

1.7 1.7

0.00041 0.000040

19. NM_144586 20. NM_003107 21. NM_001546

212909_at 201417_at 209291_at

0.65 0.56 0.56

0.00069 0.0000059 0.0028

⫺7.9 ⫺26.6 ⫺5.4

205826_at 209183_s_at 203365_s_at 203108_at 206424_at

3.8 3.0 1.9 1.9 1.8

0.00027 0.00013 0.0000083 0.00030 0.00053

10.1 12.1 24.4 9.8 8.5

6. NM_000633

203685_at

1.7

0.00026

10.2

7. NM_006988

222162_s_at

0.60

0.000062

9. 10. 11. 12. 13.

NM_005257 NM_004056 NM_001218 NM_002999 NM_002428

GtttG 1. NM_003970 2. NM_007021 3. NM_002428 4. NM_003979 5. NM_000783

P Value

t Statistics

9.1 16.5

⫺14.7

Gene Name

Keratin 19 Nuclear factor of activated T cells, cytoplasmic, calcineurin-dependent 1 Snail homolog 2 (DROSOPHILA) Cytochrome P450, family 1, subfamily B, polypeptide 1 Nuclear factor of ␬ light polypeptide gene enhancer in B-cells inhibitor, ␣ Regulator of G-protein signaling 10 CD24 antigen (small cell lung carcinoma cluster 4 antigen) CD44 antigen (homing function and Indian blood group system) GATA-binding protein 6 Carbonic anhydrase VIII Carbonic anhydrase XII Syndecan 4 (amphiglycan, ryudocan) MMP 15 (membrane inserted) regulator of Gprotein signaling 2, 24 kDa Phosphatidylinositol transfer protein, cytoplasmic 1 Procollagen-proline, 2-oxoglutarate 4dioxygenase (proline 4-hydroxylase), ␣ polypeptide II Nijmegen breakage syndrome 1 (nibrin) GTP cyclohydrolase 1 (dopa-responsive dystonia) Hypothetical protein MGC29643 SRY (sex-determining region Y)-box 4 Inhibitor of DNA binding 4, dominant-negative helix-loop-helix protein

Myomesin (M protein) 2, 165 kDa Decidual protein induced by progesterone MMP 15 (membrane inserted) Retinoic acid induced 3 Cytochrome P450, family 26, subfamily A, polypeptide 1 B-cell CLL/lymphoma 2 A disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motif, 1

ID, Identification; RefSeq, reference sequence. Boldface highlights a gene of particular interest.

by at least 30%. One gene regulated in GttgG cells [matrix metalloproteinase 15 (membrane-inserted)] was also regulated in GtttG cells and thus does not correlate to the antiproliferative response. This suggests that at least one of the remaining 20 genes regulated in GttgG but not in GtttG cells is involved in the GC-mediated antiproliferative response observed in HEK293 cells. One of the genes strongly regulated in the GttgG cells, but not in the GtttG cells, was I␬B␣ (Table 1). This was particularly interesting, because the previous transfection experiments (see above) suggested a correlation between

NF-␬B repression and the antiproliferative response after Dex administration. I␬B␣ Expression Is Induced in GttgG cells after Dex Treatment as Assessed by Western and Northern Blot Analysis To confirm the microarray results with regard to I␬B␣ regulation, we analyzed I␬B␣ protein expression in GttgG and GtttG cells by Western immunoblotting after Dex administration (Fig. 4A). The Western blot ver-



Selective Expression of I␬B␣ in HEK293 Cells Inhibits Proliferation To assess the effect of I␬B␣ expression on HEK293 cell cycling, cells were transiently transfected with an expression vector for the superrepressive form of I␬B␣ (I␬B␣-DN) together with an expression vector for green fluorescence protein (GFP). The cell cycle of transfected cells (GFP positive) was analyzed by flow cytometry (Fig. 5A). The results demonstrated that the percentage of cells in S/G2/M phase was significantly reduced from 44.0 ⫾ 2.9% to 36.0 ⫾ 1.4% (n ⫽ 6, P ⬍

Fig. 4. Effect of Dex Treatment of GgggG, GttgG, and GtttG Cells on I␬B␣ Expression A, Western blot analysis of I␬B␣ expression after Dex (100 nM) treatment of GgggG, GttgG, and GtttG cells for various times. The same filters were also incubated with an anti-␤tubulin antibody to check for equal protein loading. Protein (20 ␮g) was added in each well. B, Northern blot analysis of I␬B␣ mRNA induction in GttgG cells treated with or without 100 nM Dex in the presence or absence of 5 ␮g/ml CHX for 6 h. The CHX was added 1 h before hormone treatment. The same filters were also hybridized with a ␤-actin probe to control for RNA loading and transfer. The lower part of the figure shows the relative ratio of I␬B␣ to ␤-actin signals as determined by phosphor imaging analysis.

ified the microarray results and showed that I␬B␣ protein levels increased approximately 3-fold in GttgG cells after Dex treatment but was unaffected in GtttG cells. An approximately 3-fold increase in I␬B␣ protein level was also seen in wtGR (GgggG) cells after Dex treatment, thus arguing against a gain of function mutation of the GttgG receptor. To investigate whether the induction of I␬B␣ mRNA was a primary effect by the GttgG receptor or a secondary effect requiring synthesis of an intermediary factor, cells were treated with Dex in the presence and absence of the protein synthesis inhibitor cycloheximide (CHX), and I␬B␣ mRNA expression was analyzed by Northern blotting. As can be seen from Fig. 4B, I␬B␣ mRNA induction in the GttgG cells did not require ongoing protein synthesis, because Dex-stimulated I␬B␣ expression was not affected by the addition of CHX.

Fig. 5. Effect of Transient Ectopic Expression in HEK293 Cells of the I␬B␣ Superrepressor (I␬B␣-DN) on Cell Cycling HEK293 cells were transiently cotransfected with the green fluorescence protein expression vector pEGFP and a vector expressing either the I␬B␣ superrepressor (CMV-I␬B␣-DN) or a control vector. Cells were fixed 70 h after transfection, permeabilized, and DNA stained with PI. A, Flow cytometric analysis of cell cycle. In the left panels, single cells are displayed on a dot plot of PI vs. GFP, which was then used to gate GFP-positive cells (R1). In the right panels, DNA histograms for the GFPpositive cells are shown. (Shown is one representative experiment of six). B, The bars show the percent cells in the S/G2/M phase after transfection with the control and the CMV-I␬B␣-DN expression vector, respectively. n ⫽ 6 obtained from two separate experiments; P ⬍ 0.001.


0.001) in I␬B␣-DN transfected cells as compared with the cells transfected with the control vector (Fig. 5B). An analogous reduction of cells in the S/G2/M phase was obtained if cells were transfected with the wild type I␬B␣ (data not shown). No indication of apoptosis was observed. Notably, I␬B␣ expression had an effect on the cell cycle pattern similar to what was observed after Dex treatment of GttgG cells (see Fig. 3B). To more directly investigate the role of I␬B␣ expression on HEK293 cell proliferation, the I␬B␣-DN cDNA was cloned into a tetracycline-inducible vector and stably transfected into HEK293 cells containing the tetracycline-regulated reverse tTA transactivator (rtTA). Clones were isolated and treated with the tetracycline analog doxycycline (Dox) and analyzed for I␬B␣-DN expression, NF-␬B repression, and cell proliferation by counting cell numbers. As can be seen from the Western blotting experiment using an antibody recognizing a FLAG epitope present in I␬B␣-DN, Dox treatment stimulated a 3-fold increase in expression of the I␬B␣-DN protein (Fig. 6A). A small basal expression of the I␬B␣-DN was seen in the absence of Dox. To investigate the contribution of Dox-induced I␬B␣-DN expression to the total amount of I␬B␣ present in the cells, samples were also analyzed by an antibody that recognizes an epitope present both in the endogenous I␬B␣ and the transfected I␬B␣-DN proteins. This showed that 24-h treatment with Dox resulted in an approximately 2-fold increase in total I␬B␣ expression (Fig. 6B). We also analyzed the capacity of Dox-induced I␬B␣-DN expression in HEK293-rtTA I␬B␣-DN cells to repress NF-␬B activity. This was tested after transfection of the cells with the NF-␬B luc reporter gene. Figure 6C shows that stimulated NF-␬B activity was reduced by 37 ⫾ 9% (P ⬍ 0.05; n ⫽ 3) after Dox treatment. Notably, Dox also repressed NF␬B-dependent reporter gene activity by 58 ⫾ 16% (P ⬍ 0.05; n ⫽ 3) in the absence of stimulation, indicating a low constitutive NF-␬B activity present in nonstimulated HEK293 cells. Addition of Dox to the HEK293-rtTA I␬B␣-DN cells resulted in a diminished cell proliferation, and after 7 d the cell number for Dox-treated cells was 50–60% less compared with the cell number of the cells treated with the solvent alone (Fig. 6D). Dox did not show any nonspecific effects on cell proliferation, because HEK293rtTA cells, which do not express the I␬B␣-DN gene, were unaffected by Dox treatment. The slightly slower proliferation of the HEK293-rtTA I␬B␣-DN-containing cells in the absence of Dox treatment, as compared with the HEK293-rtTA cells alone, could be due to the low basal expression of I␬B␣-DN in the HEK293-rtTA I␬B␣-DN cells (see Fig. 6A). In summary, these results demonstrated that increased I␬B␣ expression will decrease NF-␬B activity and is sufficient to inhibit HEK293 cell proliferation, thus suggesting that the GC-stimulated I␬B␣ expression participates in executing the antiproliferative response upon GC administration.


Fig. 6. Ectopic Expression in Stably Transfected HEK293 Tet-On Cells of the I␬B␣ Superrepressor (I␬B␣-DN) after Dox Treatment and Effect on Cell Proliferation A, I␬B␣-DN expression in stably transfected HEK293 Tet-On cells at various time points after Dox (2 ␮g/ml) treatment as assayed by Western blot analysis. The I␬B␣-DN expression was detected by an anti-FLAG antibody recognizing a FLAG epitope present in the C-terminal end of the I␬B␣-DN protein. C0 and C96 stand for analysis of plated but nontransfected cells 0 and 96 h after initiation of the experiment. B, Total I␬B␣ expression (endogenous I␬B␣ and ectopic I␬B␣-DN) in the cells after 24 h of Dox treatment as analyzed by Western blotting using an anti-I␬B␣ antibody recognizing epitopes present in both proteins. C, HEK293 rtTA I␬B␣-DN cells were transiently transfected with the 3xNF␬B-luc reporter gene and left untreated or treated with 2 ␮g/ml Dox for 40 h. TPA (to a final concentration of 5 ng/ml) was added to some of the wells 24 h before harvesting. Untreated (open bar), Dox alone (striped bar), TPA alone (black bar), or TPA ⫹ Dox (gray bar). P ⬍ 0.05 for both Dox alone vs. untreated and TPA ⫹ Dox vs. TPA alone. Figure shows one of two experiments with similar results under each experimental conditions performed in triplicate. D, Cell proliferation of nontransfected HEK293 Tet-On cells (diamonds) or HEK293 rtTA I␬B␣-DN cells (triangles) was determined by counting cells after various times after treatment with control vehicle or Dox. Open symbols and dashed lines depict cells treated with control vehicle. Solid symbols and solid lines depict cells treated with Dox (2 ␮g/ml). Similar results were obtained with several HEK293-rtTA I␬B␣-DN clones, although the degree of growth inhibition and basal growth rate varied slightly.


DISCUSSION Despite the well-recognized knowledge of an antiproliferative effect of GCs on many cell types, the molecular mechanism has remained elusive. Although alteration in expression of many genes involved in the control of the cell cycle after GC exposure has been described, very few have been demonstrated to be primary targets for the GC-GR complex and directly responsible for the antiproliferative effect. Principally, suppressed cell proliferation by GCs can be controlled by stimulated expression of genes repressing cell cycle progression, e.g. cyclin-dependent kinase inhibitors such as p21Cip1 or p57Kip2 (6, 23, 24), or suppressed expression of genes stimulating cell cycle progression, e.g. c-myc or cyclins (3). For the latter mechanism, repression may occur through GC inactivation of transcription factor complexes such as AP-1 and NF-␬B, which control expression of the genes involved in cell cycle progression. AP-1 has been demonstrated to participate in both the control of cell cycle progression and cellular transformation (19, 20). Particularly, the AP-1 family members c-jun and c-Fos are positive regulators of cell cycle progression. Similarly, NF-␬B has been implicated in the control of cell cycle progression and tumor formation by promoting proliferation and inhibiting apoptosis (17, 25). Both AP-1 and NF-␬B have been shown to be inhibited by GCs through a tethering mechanism, which involves a direct binding of the GC-GR complex to AP-1 and NF-␬B complexes without the GR interacting with DNA (8, 9). Furthermore, GCs induce I␬B␣ expression, at least in some cell types, and may contribute to the NF-␬B-repressing effect (26, 27). In this report, we describe the novel observation that GC repression of HEK293 cell proliferation involves GR-mediated inhibition of NF-␬B activity. Several results suggested that inhibition of NF-␬B (and not AP-1) activity is responsible for the antiproliferative effects of GCs on HEK293 cells. Using GR mutants unable to transactivate through GREs and with different efficacy to repress AP-1 and NF-␬B activity, a good correlation between the antiproliferative response and the ability of GCs to suppress NF-␬B activity was observed. The antiproliferative domain of the GR was localized to the C-terminal zinc finger, which was the same region shown to be required for NF-␬B repression. Our results also indicated that GR-AP-1 cross-reactivity involves the linker region of the GR DBD, because maintaining the GR linker region (GgttG 3 GggtG) sustained approximately 50% of the AP-1 repressing activity as compared with the wtGR. AP-1 repressing activity is also present in the N-terminal zinc finger of TR␤ DBD, because replacing the GR N-terminal zinc finger with the N-terminal zinc finger of TR␤ (GgttG 3 GtttG) restored 50–60% of the AP-1 repressing activity of wtGR. However, definitive proof of important GR DBDs involved in AP-1 repression will require testing additional mutants. Furthermore, our results demon-


strated that it is possible to separate GR repression of AP-1 and NF-␬B, indicating that repression of these two signaling pathways operates through separate receptor domains and mechanisms. This is in line with a report by Yamamoto and associates (28) who demonstrated that coactivators are involved in the repression of AP-1 activity; in contrast, a recent report excluded the involvement of coactivators in the repression of NF-␬B (29). Our results in HEK293 cells are also in line with reports from Jurkat, CEM-C7 T cells, and U2OS cells in which it has been demonstrated that repression rather than activation of genes via GR binding to GREs is important for the antiproliferative response (1, 4, 30). In our case, this conclusion was based on the results obtained using the GttgG cells, which lacked the ability to transactivate target genes via GREs but maintained the transrepressive and antiproliferative effects. However, the antiproliferative mechanism of GR may differ in different cell lines, because in S49 or Saos2 cells, a GR with sustained ability to transactivate via GREs is required for the antiproliferative effect (1, 31). We have also shown that GC induction of p57Kip2 is involved in the antiproliferative response of HeLa cells, and this requires GR binding to a GRE present in the p57Kip2 promoter (32). Further evidence for the involvement of NF-␬B repression in the GC-mediated antiproliferative response was obtained from our gene expression profiling studies in HEK293 cells containing two GR mutants (GttgG and GtttG) that lack the ability to transactivate GRE-controlled target genes, of which only one can repress NF-␬B activity (GttgG), whereas both can repress AP-1 activity. Only the GttgG receptor that was able to repress NF-␬B activity executed an antiproliferative response. These studies identified increased I␬B␣ expression after GC administration in the GttgG-containing HEK293 cells, but not in the GtttG-containing HEK293 cells, in line with NF-␬B repression being involved in the GC-induced antiproliferative response. Northern and Western blotting confirmed the I␬B␣ regulation identified by the microarray analysis. Further evidence for a role of NF-␬B repression in the antiproliferative response was derived from our studies using ectopically expressed I␬B␣. Transient transfection of HEK293 cells with the I␬B␣ resulted in an increased percentage of cells in the G1 phase, similar to what was seen in wtGR or GttgGcontaining HEK293 cells after Dex treatment. A G1 accumulation is in accordance with what is seen in most other cell types in which GCs result in inhibition of cell proliferation (4–6, 24). Selective expression of I␬B␣-DN using the tetracycline-inducible system in HEK293 cells also resulted in inhibited cell proliferation. However, the growth-inhibitory effect in these cells was not as pronounced as in GR-containing (wtGR or GttgG) HEK293 cells after GC administration. This could be explained by the fact that GR not only induced I␬B␣ expression in these latter cells, but also directly inhibited NF-␬B activity by tethering, as compared with the selective increase in I␬B␣ expression in


Dox-treated HEK293-rtTA I␬B␣-DN cells. An alternative explanation is that the I␬B␣ cooperates with other Dex-induced gene products in the GttgG cells, e.g. the Nijmegen breakage syndrome gene (Table 1), that play a role in the control of cell proliferation (33). An additional possibility is that GCs, via the GR, affect phosphorylation events that will result in an altered activity but not in the abundance of proteins involved in cell cycle control, because GCs have been shown to influence, for example, ERK activity (34). By comparing two GR mutants that are transactivation deficient and in which only one executes an antiproliferative response, we could restrict the number of candidate target genes involved in the antiproliferative response to 20 genes. The expression profiling data do not exclude the involvement of additional genes in the antiproliferative response, because the arrays used in the profiling studies cover only about 8,400 of the approximately 38,000 genes present in the human genome. Furthermore, genes regulated less than the applied filter for selecting regulated genes may be involved. In addition, the microarray does not record regulation at the posttranslational level. The B-cell CLL/lymphoma 2 gene (Bcl-2), that can act as a negative regulator of cell proliferation (35) does not seem to be involved in the antiproliferative response in HEK293 cells, because its induction in GtttG cells did not result in inhibition of cell growth. Interestingly, I␬B␣ gene expression was induced by the GttgG mutant despite its inability to transactivate target genes controlled by GRE. Furthermore, because I␬B␣ mRNA induction was observed in the presence of the protein synthesis inhibitor CHX, the effect must be directly mediated by the GR mutant without the involvement of newly synthesized intermediary factors. A direct effect for GR induction of I␬B␣ gene expression is in line with what previously has been described for the wtGR (26, 27). Furthermore, the GttgG mutant did not transactivate from either a palindromic or direct repeat (spaced by 4 bp) thyroid response element (Bladh, our unpublished data). This suggests that the GttgG receptor mutant induces I␬B␣ expression by a mechanism independent of DNAbinding, e.g. through a tethering mechanism. Indeed, it has been demonstrated that tethering of the GR to other transcription factors results not only in transrepression but in some cases led to enhanced transcriptional activation of target genes (10, 36, 37). Also GR DBD mutants unable to transactivate from a GRE are able, in some cases, to enhance transcriptional activation through the tethering mechanism (10). This is in agreement with recent data showing that the dimer interface of the GR DBD is dispensable for activation of some genes (38). A tethering mechanism for GR activation of the I␬B␣ gene is also indirectly supported by the fact that a GRE has not been identified in the I␬B␣ promoter. The target protein for GR interaction with the I␬B␣ gene remains to be established. Although the involvement of NF-␬B in the control of normal cell proliferation is less clear as compared with


AP-1, a few reports suggest the involvement of NF-␬B in the control of cell proliferation. NF-␬B activation has been shown to be required for cell cycling in several cell types, such as regenerating liver cells, fibroblasts, and breast cancer cells (for a review see Ref. 17). Furthermore, inhibition of NF-␬B caused impairment of cell cycle progression in glioma cells and retarded G1/S transition in HeLa cells (39, 40). This also demonstrated a small constitutive NF-␬B activity present in these proliferating nonstimulated cells. It has also been found that the level of NF-␬B activation in Jurkat cells was linked to signals that control cell cycle progression (41). As shown in Fig. 6, a small constitutive NF-␬B activity seems to be present in nonstimulated HEK293 cells, as selective I␬B␣ expression reduced basal NF-␬B activity (Fig. 6). In line with this, we detected a weak DNA-binding NF-␬B activity in proliferating nonstimulated HEK293 cells that could be competed specifically by excess unlabeled NF-␬B binding site in gel shift experiments (data not shown). All together, this suggests, as demonstrated in this report, that GC-mediated repression of NF-␬B activity is an important mechanism contributing to the antiproliferative effect of GCs. How general this mechanism is remains to be established. However, it opens up the possibility that the identification of GR ligands that will induce the GR into a conformation that discriminates between AP-1 and NF-␬B repression may generate ligands with improved therapeutic profiles.

MATERIALS AND METHODS Reagents and Chemicals Dex, Dox, CHX, geniticin, propidium iodide (PI), and 12-Otetradecanoyl-phorbol-13-acetate (TPA) were purchased from Sigma Chemical Co. (St. Louis, MO) The culture media DMEM high glucose and F12 (HAM), penicillin/streptomycin, L-glutamine were all from Life Science Technologies (Gaithersburg, MD)/Invitrogen (San Diego, CA), and zeocin, hygromycin B, and Lipofectin reagent were all purchased from Invitrogen. Fetal bovine serum was purchased from Integro b.v. (Dieren, The Netherlands), and the chemiluminescence reagents used for measurement of alkaline phosphatase and luciferase activities were purchased from PerkinElmer Life Sciences (Boston, MA) and Biotherma (Haninge, Sweden), respectively. Expression Vectors and Reporter Genes The plasmids containing the GR mutants with chimeric DBDs have been described previously (12). Briefly, the DBDs of GR and thyroid receptor ␤ (TR␤) were separated into three parts (the N-terminal zinc finger, the linker region between the two fingers, and the C-terminal zinc finger) and termed ggg or ttt, respectively; the first, second, and third lowercase letter represent the N-terminal zinc finger, linker region, and C-terminal zinc finger region of GR (g) or TR␤ (t), respectively. The wildtype (wt) GR (GgggG) is equivalent to GRnx and GtttG to GTG (Ref. 12 and references therein). The GR cDNAs (KpnI-DraI fragments) containing the complete coding sequence were subcloned into the KpnI-EcoRV sites of the pcDNA5/FRT expression vector (Invitrogen) to generate stably expressing


Flp-In HEK293 cell lines (www.invitrogen.com). Flp-In cells, which contain a single integrated Flp recombination target site, allow stable integration of cDNAs at a specific genomic site and, subsequently, similar expression in individual cell clones. The luciferase reporter plasmids 3xNF␬B, 2xGREtk-Luc, and 5xAP1-ALP (alkaline phosphatase) have previously been described (12, 42). The MMP-1 (⫺517/⫹63 collagenase-1)luc reporter plasmid was kindly provided by M. Go¨ttlicher (Munich, Germany). The superrepressive form of I␬B␣ (I␬B␣-DN, I␬B␣ with mutations at serines 32 and 36 to a glycine and alanine, respectively) has been described elsewhere (43) and was subcloned into pcDNA3 (Invitrogen) to generate cytomegalovirus (CMV)-I␬B␣-DN. The pTRE-I␬B␣-DN-FLAG plasmid is the I␬B␣-DN containing a C-terminal FLAG epitope and was subcloned into the tetracycline-inducible vector pTRE (CLONTECH/BD Biosciences, Palo Alto, CA). Cells, Cell Culturing Conditions, and Establishment of Stably Transfected Cell Lines Flp-In HEK293 (Invitrogen) or HEK293 Tet-On cells (CLONTECH) were grown at 37 C in 5% CO2 in a 1:1 mixture of high-glucose DMEM and HAM F12 medium containing 10% fetal bovine serum, and penicillin and streptomycin at 10 IU/ml and 100 ␮g/ml, respectively, and 2 mM L-glutamine. In the case of nontransfected Flp-In HEK293 cells, the medium also contained 100 ␮g/ml Zeocin. HEK293 Tet-On cells were grown in the presence of 500 ␮g/ml geniticin. Establishing Flp-In HEK293 cells stably expressing wtGR or GR mutants was performed according to the instructions of the manufacturer (Invitrogen), using the calcium phosphate precipitation method. Individual clones were selected and maintained in the presence of 100 ␮g/ml hygromycin B (no Zeocin). Individual clones established in the HEK293 Flp-In cells expressed similar amounts of GR. Pools consisting of five to 10 individual clones were used for the subsequent experiments. To establish the HEK293-rtTA I␬B␣-DN cell lines, HEK293 Tet-On cells were cotransfected as described above with 10 ␮g of the pTRE-I␬B␣-DN-FLAG plasmid and 1 ␮g of the pKSV-Hygromycin selection plasmid. Selection was performed in the presence of 100 ␮g/ml hygromycin B, and individual clones were isolated. Transient Transfections of HEK293 Cells Lipofectin reagent was used in all transfections according to the manufacturer’s instructions. Briefly, 50,000 cells were seeded in 24-well plates 20 h before transfection. Reporter plasmid (200 ng) was used per well together with 2 ng of the CMV-ALP plasmid as an internal control to correct for differences in transfection efficiency. Cells were treated 24 h post transfection for 20 h, after which the cell medium was collected. Preparation of cell extract and measurement of alkaline phosphatase and luciferase activity were performed as previously described (12, 42). Quadruplicate wells were analyzed for each experimental condition and experiments were repeated three to five times. In the case of cell cycle analysis after I␬B␣ transfection, 10-cm plates were transfected with 5 ␮g pEGFP-C2 (CLONTECH) and 15 ␮g CMV-I␬B␣-DN or pcDNA3 (Invitrogen) expression vector per plate using the Lipofectin reagent. Medium was replaced 22 h after transfection, and incubation was continued for an additional 48 h before analysis. Western Immunoblotting Whole-cell extracts were prepared from nontreated cells or cells treated for various time periods with 100 nM Dex or 2 ␮g/ml Dox as indicated in the different experiments, and


Western immunoblotting was performed as previously described (32). For detection of GR, the E-20 antibody (sc-1003) from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) was used. I␬B␣-DN in the HEK293-rtTA I␬B␣-DN cells was specifically detected by an anti-FLAG-M2 antibody (Sigma) to a FLAG epitope present in the C-terminal part of I␬B␣-DN. For detection of endogenous I␬B␣ or total (endogenous and I␬B␣-DN) I␬B␣s, an anti-I␬B␣ antibody (sc-847) from Santa Cruz was used. To check for equal loading and transfer, the filters were probed with an anti-␤-tubulin antibody (N357, Amersham Life Sciences, Arlington Heights, IL). For detection, enhanced chemiluminescence (Amersham) was used. The intensities of the bands were quantified by a luminescent image analyzer (LAS-1000, Fuji Photo Film Co., Ltd, Stamford, CT). Northern Blotting Cells were treated with or without 100 nM Dex in the presence or absence of 5 ␮g/ml CHX for 6 h. The CHX was added 1 h before hormone treatment. Total RNA was prepared by the RNeasy kit from QIAGEN (Chatsworth, CA). Northern blotting using 10 ␮g of total RNA was performed as previously described (24). [32P]UTP-labeled cRNA probes were used using CMV-I␬B␣-DN (see above) and ␤-actin cloned into pGEM3 (24) as templates. Hybridization was quantified by phosphor imager analysis using a Bas111 phosphor imager (Fuji Photo Film Co.) Cell Proliferation Cells (n ⫽ 5000) were seeded in 24-well plates and treated with or without 100 nM Dex or 2 ␮g/ml Dox for the indicated time points. Cells were trypsinized and counted in a Bu¨rker chamber. Five wells were counted per time point, and experiments were repeated two to three times. Flow Cytometric Analysis of Cell Cycle Fixation, PI staining of cells, and flow cytometric analysis of cell cycle were performed as described previously (24). For analyzing the cell cycle after GFP/I␬B␣-DN transfection, cells were first fixed with 1% formaldehyde at 4 C for 1 h, washed in PBS, and kept in 70% ethanol at 4 C overnight before staining with PI. Gated GFP-positive cells (n ⫽ 2000) were analyzed by flow cytometry. Statistical Analysis Unless otherwise stated, results are expressed as mean ⫾ SD; n is given for each experiment. For analysis of significance, two-tailed homoscedastic Student’s t test was used. Gene Expression Profiling and Identification of Regulated Genes GtttG- and GttgG-expressing cells were treated with or without 100 nM Dex for 6 h. For nontreated cells, a similar volume of solvent was added. Three separate experiments were performed on different days. Total RNA was isolated using the RNeasy kit from QIAGEN, and its quality was determined using the Nano 6000 Chip in the Bioanalyzer from Agilent Technologies, Inc. (Palo Alto, CA). Total RNA (8 ␮g) was used for target cDNA synthesis according to the Affymetrix manual (www.affymetrix.com). Hybridization of final targets to Human Gene Focus arrays representing 8500 transcripts followed by probing and scanning was performed according to the Affymetrix manual. The scanned CEL-files from Affymetrix Microarray Suite 5.0 were used for quantile normal-


ization and quantification of expression using the robust multichip analysis (44). For statistical analysis of significantly changed genes, the R software from BioConductor (www. bioconductor.org) was employed. In Quantile t plots, genes with higher experimental t scores than 5 and lower experimental t scores than ⫺5 were scored as having significantly altered expression in Dex-treated vs. untreated cells. A minimal mean ratio of 1.7-fold was used as threshold for Dexdependent induction of t score-selected genes in the two cell lines, and 0.7 was used as threshold for repressed genes.

Acknowledgments


11.

12.

13.

We thank M. Lind and R. Toftgård for kindly providing the I␬B␣-DN and pTRE-I␬B␣-DN-FLAG plasmids and M. Go¨ttlicher for the MMP-1 luciferase reporter gene. 14. Received July 22, 2004. Accepted October 26, 2004. Address all correspondence and requests for reprints to: Professor Sam Okret, Department of Medical Nutrition, Karolinska Institutet, Karolinska University Hospital Huddinge, Novum, SE-141 86 Huddinge, Sweden. E-mail: Sam.Okret@ mednut.ki.se. This work was supported by a grant from the Swedish Cancer Society (to S.O.). L.-G.B. is supported by the Foundation for Knowledge and Competence Development at Karolinska Institutet and KaroBio AB. AFFYMETRIX core facility at Novum is supported by the Wallenberg Consortium North.

15.

16. 17.

18.

19. 20.

REFERENCES 1. Rogatsky I, Trowbridge JM, Garabedian MJ 1997 Glucocorticoid receptor-mediated cell cycle arrest is achieved through distinct cell-specific transcriptional regulatory mechanisms. Mol Cell Biol 17:3181–3193 2. Smith E, Redman RA, Logg CR, Coetzee GA, Kasahara N, Frenkel B 2000 Glucocorticoids inhibit developmental stage-specific osteoblast cell cycle. Dissociation of cyclin A-cyclin-dependent kinase 2 from E2F4–p130 complexes. J Biol Chem 275:19992–20001 3. Rhee K, Reisman D, Bresnahan W, Thompson EA 1995 Glucocorticoid regulation of G1 cyclin-dependent kinase genes in lymphoid cells. Cell Growth Differ 6:691–698 4. Helmberg A, Auphan N, Caelles C, Karin M 1995 Glucocorticoid-induced apoptosis of human leukemic cells is caused by the repressive function of the glucocorticoid receptor. EMBO J 14:452–460 5. Sanchez I, Goya L, Vallerga AK, Firestone GL 1993 Glucocorticoids reversibly arrest rat hepatoma cell growth by inducing an early G1 block in cell cycle progression. Cell Growth Differ 4:215–225 6. Corroyer S, Nabeyrat E, Clement A 1997 Involvement of the cell cycle inhibitor CIP1/WAF1 in lung alveolar epithelial cell growth arrest induced by glucocorticoids. Endocrinology 138:3677–3685 7. Scha¨cke H, Docke WD, Asadullah K 2002 Mechanisms involved in the side effects of glucocorticoids. Pharmacol Ther 96:23–43 8. De Bosscher K, Vanden Berghe W, Haegeman G 2003 The interplay between the glucocorticoid receptor and nuclear factor-␬B or activator protein-1: molecular mechanisms for gene repression. Endocr Rev 24:488–522 9. Newton R 2000 Molecular mechanisms of glucocorticoid action: what is important? Thorax 55:603–613 10. Subramaniam N, Campioˆn J, Rafter I, Okret S 2003 Cross-talk between glucocorticoid and retinoic acid

21.

22.

23.

24.

25. 26.

27.

28.

29.

signals involving glucocorticoid receptor interaction with the homoeodomain protein Pbx1. Biochem J 370: 1087–1095 Heck S, Kullmann M, Gast A, Ponta H, Rahmsdorf HJ, Herrlich P, Cato AC 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 Lide´n J, Delaunay F, Rafter I, Gustafsson J-Å, Okret S 1997 A new function for the C-terminal zinc finger of the glucocorticoid receptor. Repression of RelA transactivation. J Biol Chem 272:21467–21472 Yang-Yen HF, Chambard JC, Sun YL, Smeal T, Schmidt TJ, Drouin J, Karin M 1990 Transcriptional interference between c-Jun and the glucocorticoid receptor: mutual inhibition of DNA binding due to direct protein-protein interaction. Cell 62:1205–1215 Schu¨le R, Rangarajan P, Kliewer S, Ransone LJ, Bolado J, Yang N, Verma IM, Evans RM 1990 Functional antagonism between oncoprotein c-Jun and the glucocorticoid receptor. Cell 62:1217–1226 Karin M, Cao Y, Greten FR, Li ZW 2002 NF-␬B in cancer: from innocent bystander to major culprit. Nat Rev Cancer 2:301–310 Ghosh S, Karin M 2002 Missing pieces in the NF-␬B puzzle. Cell 109(Suppl):S81–S96 Chen F, Castranova V, Shi X 2001 New insights into the role of nuclear factor-␬B in cell growth regulation. Am J Pathol 159:387–397 Brown K, Gerstberger S, Carlson L, Franzoso G, Siebenlist U 1995 Control of I␬B-␣ proteolysis by site-specific, signalinduced phosphorylation. Science 267:1485–1488 Shaulian E, Karin M 2002 AP-1 as a regulator of cell life and death. Nat Cell Biol 4:E131–E136 Mechta-Grigoriou F, Gerald D, Yaniv M 2001 The mammalian Jun proteins: redundancy and specificity. Oncogene 20:2378–2389 Ricciardi-Castagnoli P, Granucci F 2002 Opinion: interpretation of the complexity of innate immune responses by functional genomics. Nat Rev Immunol 2:881–889 Jonat C, Rahmsdorf HJ, Park KK, Cato AC, Gebel S, Ponta H, Herrlich P 1990 Antitumor promotion and antiinflammation: down-modulation of AP-1 (Fos/Jun) activity by glucocorticoid hormone. Cell 62:1189–1204 Cha HH, Cram EJ, Wang EC, Huang AJ, Kasler HG, Firestone GL 1998 Glucocorticoids stimulate p21 gene expression by targeting multiple transcriptional elements within a steroid responsive region of the p21waf1/cip1 promoter in rat hepatoma cells. J Biol Chem 273: 1998–2007 Samuelsson MK, Pazirandeh A, Davani B, Okret S 1999 p57Kip2, a glucocorticoid-induced inhibitor of cell cycle progression in HeLa cells. Mol Endocrinol 13:1811–1822 Karin M, Lin A 2002 NF-␬B at the crossroads of life and death. Nat Immunol 3:221–227 Auphan N, DiDonato JA, Rosette C, Helmberg A, Karin M 1995 Immunosuppression by glucocorticoids: inhibition of NF-␬B activity through induction of I␬B synthesis. Science 270:286–290 Scheinman RI, Cogswell PC, Lofquist AK, Baldwin Jr AS 1995 Role of transcriptional activation of I␬B ␣ in mediation of immunosuppression by glucocorticoids. Science 270:283–286 Rogatsky I, Zarember KA, Yamamoto KR 2001 Factor recruitment and TIF2/GRIP1 corepressor activity at a collagenase-3 response element that mediates regulation by phorbol esters and hormones. EMBO J 20: 6071–6083 Wu J, Li Y, Dietz J, Lala DS 2004 Repression of p65 transcriptional activation by the glucocorticoid receptor in the absence of receptor-coactivator interactions. Mol Endocrinol 18:53–62


30. Thompson EB, Nazareth LV, Thulasi R, Ashraf J, Harbour D, Johnson BH 1992 Glucocorticoids in malignant lymphoid cells: gene regulation and the minimum receptor fragment for lysis. J Steroid Biochem Mol Biol 41:273–282 31. Chapman MS, Askew DJ, Kuscuoglu U, Miesfeld RL 1996 Transcriptional control of steroid-regulated apoptosis in murine thymoma cells. Mol Endocrinol 10:967–978 32. Alheim K, Corness J, Samuelsson MK, Bladh LG, Murata T, Nilsson T, Okret S 2003 Identification of a functional glucocorticoid response element in the promoter of the cyclin-dependent kinase inhibitor p57Kip2. J Mol Endocrinol 30:359–368 33. Chiang YC, Teng SC, Su YN, Hsieh FJ, Wu KJ 2003 c-Myc directly regulates the transcription of the NBS1 gene involved in DNA double-strand break repair. J Biol Chem 278:19286–19291 34. Caelles C, Bruna A, Morales M, Gonzalez-Sancho JM, Gonzalez MV, Jimenez B, Munoz A 2002 Glucocorticoid receptor antagonism of AP-1 activity by inhibition of MAPK family. In: Cato ACB, Scha¨cke H, Asadullah K, eds. Ernst Schering Research Foundation Workshop 40. Recent advances in glucocorticoid receptor action. Berlin: Springer-Verlag; 131–152 35. Borner C 1996 Diminished cell proliferation associated with the death-protective activity of Bcl-2. J Biol Chem 271:12695–12698 36. Stocklin E, Wissler M, Gouilleux F, Groner B 1996 Functional interactions between Stat5 and the glucocorticoid receptor. Nature 383:726–728 37. Teurich S, Angel P 1995 The glucocorticoid receptor synergizes with Jun homodimers to activate AP-1-regulated promoters lacking GR binding sites. Chem Senses 20:251–255


38. Rogatsky I, Wang JC, Derynck MK, Nonaka DF, Khodabakhsh DB, Haqq CM, Darimont BD, Garabedian MJ, Yamamoto KR 2003 Target-specific utilization of transcriptional regulatory surfaces by the glucocorticoid receptor. Proc Natl Acad Sci USA 100:13845–13850 39. Kaltschmidt B, Kaltschmidt C, Hehner SP, Droge W, Schmitz ML 1999 Repression of NF-␬B impairs HeLa cell proliferation by functional interference with cell cycle checkpoint regulators. Oncogene 18:3213–3225 40. Otsuka G, Nagaya T, Saito K, Mizuno M, Yoshida J, Seo H 1999 Inhibition of nuclear factor-␬B activation confers sensitivity to tumor necrosis factor-␣ by impairment of cell cycle progression in human glioma cells. Cancer Res 59:4446–4452 41. Perkins ND, Felzien LK, Betts JC, Leung K, Beach DH, Nabel GJ 1997 Regulation of NF-␬B by cyclin-dependent kinases associated with the p300 coactivator. Science 275:523–527 42. Barkhem T, Haldosen L-A, Gustafsson J-Å, Nilsson S 2002 pS2 Gene expression in HepG2 cells: complex regulation through crosstalk between the estrogen receptor ␣, an estrogen-responsive element, and the activator protein 1 response element. Mol Pharmacol 61: 1273–1283 43. van Hogerlinden M, Rozell BL, A¨hrlund-Richter L, Toftgård R 1999 Squamous cell carcinomas and increased apoptosis in skin with inhibited Rel/nuclear factor-␬B signaling. Cancer Res 59:3299–3303 44. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP 2003 Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4:249–264

Molecular Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.