| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Medical Pharmacology, Leiden/Amsterdam Centre for Drug Research and Leiden University Medical Centre, 2300 RA Leiden, The Netherlands
Address all correspondence and requests for reprints to: S. van der Laan, Division of Medical Pharmacology, Leiden/Amsterdam Centre for Drug Research and Leiden University Medical Centre, P.O. Box 9502, 2300 RA Leiden, The Netherlands. E-mail: s.van.der.laan{at}lacdr.leidenuniv.nl.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Numerous coregulator proteins interact with nuclear receptors such as the GR to regulate the transcription of target genes (6, 7). The interactions of these proteins with the DNA-bound steroid receptor form an essential mechanism in the modulation of the genomic response. Coregulators are typically divided in two different classes, coactivators and corepressors (8). The coactivator proteins such as members of the well-documented family of p160 steroid receptor coactivator (SRCs) possess intrinsic histone acetyltransferase activity (HAT) and contain distinct regions and motifs for the recruitment of other proteins with enzymatic activities, including CREB-binding protein (CBP)/p300 and coactivator-associated arginine methyltransferase 1 (CARM-1). Acetylation of histones is important in maintaining an open chromatin structure, thereby facilitating the access of the ligand-activated GR to the proximal promoter of a target gene (9). Conversely, corepressors such as nuclear corepressor (N-CoR) and silencing mediator of the retinoid and thyroid hormone receptor (SMRT) proteins have intrinsic histone deacetyltransferase activity that catalyzes the deacetylation of the chromatin. They form docking surfaces for the recruitment of additional components of corepressor complexes, resulting in a reduction of the transcriptional activity (10). The importance of coregulators for GR-mediated transcriptional regulation has been studied mainly in the context of transactivated target genes. The ratio of coactivators and corepressors expressed in the cell has been proposed to determine the nature and the magnitude of the GR-mediated transcriptional response, particularly at subsaturating levels of corticosterone (CORT) (11).
The activity of coregulators depends on their expression levels as well as posttranslational modifications of the proteins. Recently, we have mapped in the rat brain the expression of the coregulators SRC1a, SRC1e, N-CoR, and SMRT (12, 13), which are all known to interact with the ligand-activated GR in vitro (14, 15). These coregulators are abundantly expressed in brain tissue and have distinct expression patterns. In line with the proposed models of steroid action, we hypothesized that their different distribution in brain and their specific contribution to nuclear receptor activity possibly define in part the cell-specific responses elicited by glucocorticoids. The most compelling observation in this respect is that SRC1a is highly abundant in the PVN, coinciding with the site-specific glucocorticoid-dependent repression of the CRH gene (16).
A well-established model to study GR-mediated repression of the CRH gene is the AtT-20 cell line (17). The proximal promoter of this gene contains among others an inducible CRE and a negative glucocorticoid-responsive element (nGRE). To characterize the role of the coregulators in both forskolin induction and glucocorticoid-dependent repression of the CRH gene, we performed transient transfections in AtT-20 cells. We show that SRC1a (but not SRC1e) specifically enhances GR-mediated repression at physiologically relevant concentrations of the naturally occurring ligand CORT. In addition, corepressors do not affect gene repression by GR, but reduce CREB-mediated CRH expression.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Small interference RNA (siRNA) constructs
Small interference RNA (siRNA) expression vector and mismatch control directed against the consensus sequence of human, rat, and mouse GR were designed. The sequence for siRNA against mouse GR (NM_008173) was GATCCCCGAAAGCATTGCAAACCTCATTCAAGAGATGAGGTTTGCAATGCTTTCTTTTTGGAAA for the perfect match and GATCCCCGACAGCATTGCACACCTCATTCAAGAGATGAGGTGTGCAATGCTGTCTTTTTGGAAA the mismatch control. The sense and antisense oligonucleotides of 64 bp long were annealed and cloned in BglII and HindIII sites of p-super vector (The Netherlands Cancer Institute, Amsterdam, The Netherlands). Insertion of the oligonucleotides was confirmed by sequencing. The knockdown of the GR was tested by Western blot and quantitative PCR in rat PC12 cells (van Hooijdonk, L., V. Brinks, O. C. Meijer, E. R. de Kloet, T. Schouten, T. F. Dijkmans, A. C. A. Vellinga, M. S. Oitzl, C. P. Fitzsimons, and E. Vreugdenhil, in preparation) and functionally tested in AtT-20 cells on a 3x-containing GRE reporter TAT-3-luc.
Cell culture and transient transfections
AtT-20/D-16V mouse tumor cells (kindly provided by Dr. J. van der Hoek, Erasmus Medical Centre, Rotterdam, The Netherlands) were grown and maintained in DMEM containing 4.5 g/liter glucose supplemented with 0.5% penicillin/streptomycin, 10% horse serum, and 10% fetal bovine serum (Invitrogen Life Technologies, Breda, The Netherlands) in a humidified atmosphere of 5% CO2 at 37 C. A day before transfection, 0.1 x 106 cells per well were plated in a 24-well plate (Greiner Bio-One, Alphen aan den Rijn, The Netherlands). For each well, the cells were transfected using 1,6 µl Lipofectamine 2000 (Invitrogen Life Technologies) per 0.8 µg plasmid according to the manufacturer’s instructions. To induce the CRH promoter, the cells were treated with 10 µM forskolin (Calbiochem, Darmstadt, Germany), which leads to an increase of intracellular cAMP. Subsequent protein kinase A activation results in CREB phosphorylation (19). Repression of the forskolin-induced CRH promoter was performed with the synthetic glucocorticoid dexamethasone (DEX) or the naturally occurring glucocorticoid CORT cotreatment. The cells were harvested and assayed according to the luciferase kit instructions (Promega, Madison, WI) using a luminometer (LUMAT LB 9507; Berthold, Bad Wildbad, Germany).
First, we further validated the AtT-20 cells as a model for glucocorticoid-dependent repression of the CRH promoter. For the GR knockdown, 400 ng siRNA expression plasmid, 200 ng CRH reporter, and 200 ng empty vector plasmid were transfected. After transfection, the cells were allowed to recover for 48 h before treatment. For the dose-response curves of DEX and CORT, serial 10x dilutions were prepared from a 10–6 M solution. To characterize the role of the coregulators, 200 ng reporter, 400 ng expression, and 200 ng empty vector plasmid were transfected per well. For the dose-response curves of corepressors, 200 ng reporter and a varying amount of corepressor expression plasmid were transfected. The total amount of DNA for each transfection was kept constant using empty vector.
Data analysis
Two different renilla reporters were tested (pRL-TK and pRL-CMV) for normalization. Both were found responsive to forskolin and DEX treatment in AtT-20 cells (data not shown). Consequently, the data were normalized against basal activity of the CRH promoter (fold induction) or expressed as a percentage of maximal induction of the CRH promoter after forskolin stimulation (percentage of maximal induction).
Statistics
The values are expressed as the average of four parallel transfections within one experiment, and the error bars represent the SD. All transfection experiments were performed at least twice, yielding similar results. Overall statistical analysis was performed using one-way ANOVA, and statistical significance was determined with Tukey’s multiple comparison tests with P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Unless stated otherwise, all subsequent data were generated after 4 h of treatment, the time point at which the most robust DEX-induced repression of the CRH promoter activity was observed.
|
). In this condition, DEX cotreatment resulted in 90% repression, almost completely silencing the forskolin-induced stimulation of the CRH promoter. Conversely, the siRNA-mediated knockdown of the GR completely abolished the DEX-induced repression. The mismatch control did not significantly differ from the empty vector group indicating a specific knockdown of the GR. These data together indicate that the concentration of GR protein in the AtT-20 cells determines the magnitude of the DEX-induced CRH repression. The present setting forms a suitable model to assess coregulator protein function in the GR-mediated repression of the CRH promoter activity.
DEX- and CORT-induced CRH repression
The aim of our study was to address whether brain-expressed coregulator proteins play a role in the regulation of the CRH gene expression. Although it is known that the GR activity depends on the type of ligand (20), so far, the synthetic glucocorticoid DEX has been the treatment of choice in most studies on glucocorticoid-mediated CRH repression in AtT-20 cells (21, 22, 23). To approach to a maximal extent the in vivo situation, we tested whether the naturally occurring ligand of the GR in rodents, i.e. CORT, resulted in similar repression of the CRH gene. We generated a time curve of CRH promoter activity and compared the repressive effects of 10–7 M DEX with 10–7 M CORT cotreatment (Fig. 2A). Both the natural and the synthetic glucocorticoid suppressed CRH promoter activity, although not to the same extent. A maximal repression of 40 and 70% was achieved after 4–5 h with 10–7 M CORT and 10–7 M DEX cotreatment, respectively.
|
, both efficacy and potency were higher for DEX cotreatment compared with CORT cotreatment. The effective concentration of DEX [EC50(DEX)] was 0.46 ± 0.004 nM, and the maximal repression of 75% was achieved at 10–7 M. Cotreatment with CORT was less effective, and the maximal repression of 55% was observed at 10–6 M CORT cotreatment. This was significantly lower compared with the DEX-induced maximal repression. The EC50 of the CORT-induced repression [EC50(CORT)] was 52 ± 1 nM, approximately 100 times higher than for DEX cotreatment.
Differential effects of SRC1 isoforms on GR-mediated CRH repression
In the next experiments, we tested the effect of SRC1a and SRC1e overexpression on GR-mediated repression of the CRH promoter activity. Cotransfection of SRC1a resulted in a significant GR-mediated repression at 10–8 M CORT (Fig. 3B). The potency of CORT to induce repression was approximately three times higher in the presence of SRC1a [EC50(SRC1a) = 18 ± 0.3 nM] when compared with the control situation [EC50(CORT) = 52 ± 1 nM]. On the other hand, overexpression of SRC1e tended to shift the dose-response curve to the right and resulted in a two times higher EC50 [EC50(SRC1e) = 109 ± 0.3 nM] compared with the control group (Fig. 3C). These data indicate that SRC1a and SRC1e have opposing effects on the CORT-induced repression (Fig. 3D). Moreover, the maximal repression by GR in the presence of both SRC1 isoforms significantly differed from the control group. SRC1a overexpression resulted in a maximal repression of 67%, which was higher than both the maximal repression in the control group (maximal repression 55%) and in presence of SRC1e (maximal repression 45%). In summary, both the efficacy and potency of the GR-mediated repression of the CRH promoter were higher in presence of SRC1a and tended to be reduced in presence of SRC1e.
|
, B and C), suggesting that CREB-directed transcription of the CRH gene is suppressed by both corepressors (30-fold compared with a typically 100-fold as observed in the absence of overexpressed coregulators, Fig 4A). Surprisingly, N-CoR and SMRT overexpression resulted in a similar dose-response curve for transrepression compared with the control situation (Fig. 4D). A maximal repression of approximately 55% for both situations was observed at 10–6 M CORT and a highly similar potency. The effective concentration in presence of N-CoR or SMRT was EC50(N-CoR) = 47 ± 1 nM and EC50(SMRT) = 49 ± 1 nM, respectively. Although both corepressors have been shown to interact with agonist-activated GR on positively regulated genes (15, 24), they do not modulate the GR-mediated repression of the CRH promoter at the nGRE. The effects of the coactivators and the corepressors on the GR-mediated repression of CRH gene expression are summarized in Table 1.
|
|
). We found that both N-CoR and SMRT dose-dependently suppressed the CREB-mediated transcription of the CRH promoter (Fig. 5B). On the simple 5xCRE-containing reporter, both N-CoR and SMRT suppressed CREB-mediated transcription even more potently, with a maximal inhibition of the forskolin-induced stimulation after cotransfection of 100 ng expression plasmid (Fig. 5C). These data together suggest a direct interaction of CREB with both N-CoR and SMRT at multiple CRE-containing promoters. DEX cotreatment did not affect the forskolin induction of the 5xCRE-containing promoter, indicating that the GR-mediated repression of the CRH promoter is not mediated via the CRE and/or CRE-related proteins (Fig. 6) but more likely is the result of a direct interaction of the GR with the nGRE of the CRH promoter.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recently, the coactivators SRC1 and SRC2 have been shown to act also as a corepressor of ligand-activated GR at certain GREs (25, 26), whereas the corepressor SMRT has been reported to be essential for full estrogen receptor (ER) activation (27). We have shown that the SRC1a splice variant also acted as a corepressor of ligand-activated GR on the CRH promoter. Although the mechanisms of repression remain unclear, an increasing amount of literature shows that nonreceptor factors such as coregulators determine not only the magnitude but also the nature of the transcriptional outcome.
So far, coregulator recruitment by ligand-activated GR has mainly been studied in the context of positively regulated genes. In that context, a differential potentiation of the ER- and GR-mediated transcription by the SRC1 isoforms was previously observed (14, 28). We now show that the splice variants also differentially affect the GR-mediated repression of the CRH reporter. The SRC1 splice variants are highly similar. They contain an identical centrally located nuclear receptor box with a triplet of 
-helical LXXLL motifs for binding to the nuclear receptors, two activation domains, AD1 (interacts with CBP/p300) and AD2 (interacts with CARM1), and differ only in their C-terminal sequences. The C-terminal part of SRC1a contains an additional LXXLL motif, which was found to exhibit a strong interaction with the ligand-binding domain of the GR in a yeast two-hybrid system (29). Additionally, it also has been shown to possess a repression domain in the context of ER- and GR-mediated gene transactivation on a simple reporter (14, 28). Therefore, the differences in transrepression may be caused either by differential recruitment of SRC1a and SRC1e to the DNA-bound GR and/or by the SRC1a-specific C-terminal repression domain.
It is still a topic of debate whether the GR-mediated repression on the CRH promoter occurs via the CRE and/or CRE-related proteins (e.g. CBP), via cross-talk with activator protein 1 or by direct binding of the GR to a putative nGRE in the promoter region (17, 21, 23, 30). In the neuronal BE(2)C cell line, point mutation and deletion of the putative nGRE did not affect DEX-induced repression (30). However, Dorin and colleagues (21) convincingly showed that the putative nGRE in the AtT-20 cell line is necessary for GR-mediated repression of the CRH promoter. Additionally, an argument against interaction of GR with the CREB signaling pathway (on or off the DNA) is our observation that forskolin stimulation of a simple reporter containing 5xCRE was not affected by DEX cotreatment in the same cells (Fig. 6). Furthermore, taking into consideration the sequential and combinatorial assembly of protein complexes at the promoter by DNA-bound steroid receptors (31, 32), the effects of SRC1a suggests that the binding of GR to the response element of the CRH gene is necessary. We propose that the GR-mediated repression of the CRH gene expression is promoter specific and cannot be explained by squelching of rate-limiting factors such as CBP or CRE-related proteins.
Both CRH and proopiomelanocortin (POMC) genes can be directly repressed by GR and are part of the negative feedback regulation of glucocorticoids on HPA axis activity. Interestingly, the 5'-flanking region of both CRH and the POMC genes contain a putative nGRE (21, 33). Winnay et al. (34) reported that, as opposed to the situation in wild-type mice, DEX was ineffective in suppressing pituitary POMC mRNA levels in the SRC1 knockout mice. This indicates that SRC1 expression is necessary for adequate GR-mediated repression of the POMC gene. These observations suggest that the glucocorticoid repression of HPA axis activity at the level of both the hypothalamic CRH- and pituitary POMC-expressing cells is dependent on SRC1a recruitment by the GR at a nGRE.
The corepressors N-CoR and SMRT are ubiquitously expressed in brain and show moderate differences in expression in the PVN (13). Previously, Szapary et al. (11) established a model based on the observations that coactivators and corepressors have opposing effects on the dose-response curve of agonist-bound GR-regulated gene expression. The model describes that the GR dose-response curve can be modified by coactivators and corepressors in a gene- and promoter-independent fashion. Surprisingly, our data showed no effect of the corepressors N-CoR and SMRT on the GR-mediated repression of the CRH gene. We find that coactivators and corepressors did not mediate opposing effects on a negatively regulated gene, indicating a clear promoter-dependent effect of coactivators and corepressors. It is likely that the binding of the agonist-activated GR to the nGRE induces conformational changes that disfavor corepressor recruitment.
Although N-CoR and SMRT did not affect the GR-mediated repression, they both repressed CREB-mediated induction of the CRH gene. This is in line with the observation that N-CoR can directly modulate the activity of the cointegrator CBP by binding to the same complexes (35). Interestingly, although N-CoR and SMRT can associate with distinct corepressor complexes, overexpression resulted in a very similar inhibition of CREB-mediated transcription (36). We examined promoter-specific effects of the corepressors and tested overexpression of N-CoR and SMRT on CREB-mediated transcription on the CRH-luc reporter and a simple 5xCRE-containing reporter. Both corepressors dose-dependently suppressed the CREB-mediated transcription, clearly indicating that the effects are mediated through a similar mechanism (Fig. 5).
Taken together, coregulators may play different roles in the control of hypothalamic CRH expression. The brainstem catecholaminergic system projects to the PVN and releases norepinephrine that binds to the G protein-coupled 1 adrenergic receptors and activates CREB-mediated gene expression. Additionally, glutamatergic and GABAergic interneurons also regulate the activity of the PVN and in turn expression of CRH (37, 38, 39, 40, 41). The fact that N-CoR and SMRT are expressed in the PVN, and our present finding that they inhibit the CREB-mediated transcription of the CRH gene, indicate that the corepressor activity may in part determine CRH gene expression. In addition, Shepard et al. (42) recently provided evidence for the role of the CRE modulator (CREM) and inducible cAMP early repressor (ICER) in limiting the CRH surge in the context of restraint stress. Thus, repression of the CRH gene expression after stress is likely to involve a set of different mechanisms. We provide evidence that coregulator proteins have specific roles in both induction and repression of CRH gene expression.
Paradoxically, glucocorticoids repress CRH expression in the PVN but stimulate expression in the CeA (5). The current data are based on overexpression of coregulators; the exact ratio of receptor to coregulators is difficult to ascertain. However, the data provide evidence that SRC1a, highly abundant in the PVN, increases both potency and efficacy of the GR-mediated repression of forskolin-induced CRH expression. In addition, SRC1e, which was previously found to be relatively abundant in the CeA (12), significantly reduced the CORT-dependent repression (see Table 1). Moreover, SRC1e tended to increase forskolin-induced CREB-driven transcription of the CRH gene (Fig. 3C). These observations are in line with the site-specific effects of glucocorticoids on CRH gene expression previously described in vivo. In view of the chromatin-modifying properties of the coregulators, more functional studies in stably transfected AtT-20 cells and chromatin immunoprecipitation assays would provide valuable insights on their role in the regulation of CRH expression.
In conclusion, we have shown that 1) the availability and the type of ligand, 2) the expression level of the receptor in the cells, and 3) the coregulator recruited are three determinants of glucocorticoid signaling. More specifically, the coactivator SRC1a increased GR-mediated repression, whereas the corepressors N-CoR and SMRT were found to inhibit CREB-dependent induction of the CRH gene.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online November 15, 2007
Abbreviations: CeA, Central nucleus of the amygdala; CBP, CREB-binding protein; CORT, corticosterone; CRE, cAMP-responsive element; CREB, CRE-binding protein; DEX, dexamethasone; ER, estrogen receptor; GR, glucocorticoid receptor; HPA, hypothalamus-pituitary-adrenal; N-CoR, nuclear corepressor; nGRE, negative glucocorticoid-responsive element; POMC, proopiomelanocortin; PVN, paraventricular nucleus of the hypothalamus; siRNA, small interference RNA; SMRT, silencing mediator of the retinoid and thyroid hormone receptor; SRC, steroid receptor coactivator.
Received September 6, 2007.
Accepted for publication November 8, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
transcriptional activity. Mol Cell Biol 27:5933–5948 directs ordered, cyclical, and combinatorial recruitment of cofactors on a natural target promoter. Cell 115:751–763[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  S. Lachize, E. M. Apostolakis, S. van der Laan, A. M. I. Tijssen, J. Xu, E. R. de Kloet, and O. C. Meijer Steroid receptor coactivator-1 is necessary for regulation of corticotropin-releasing hormone by chronic stress and glucocorticoids PNAS, May 12, 2009; 106(19): 8038 - 8042. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Vreugdenhil, C. S. L. Verissimo, R. Mariman, J. T. Kamphorst, J. S. Barbosa, T. Zweers, D. L. Champagne, T. Schouten, O. C. Meijer, E. Ron de Kloet, et al. MicroRNA 18 and 124a Down-Regulate the Glucocorticoid Receptor: Implications for Glucocorticoid Responsiveness in the Brain Endocrinology, May 1, 2009; 150(5): 2220 - 2228. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M. Keenan, S. Alexander, C. Irvine, and J. D. Veldhuis Quantifying Nonlinear Interactions within the Hypothalamo-Pituitary-Adrenal Axis in the Conscious Horse Endocrinology, April 1, 2009; 150(4): 1941 - 1951. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
), cotreatment with 10–7 M DEX (
) or 10–7 M CORT (
) on CRH promoter activity are shown over 8 h. The average values ± SD (n = 4) are shown and represent the fold induction. B, Dose-response curve of DEX and CORT cotreatment. The luciferase activity at each steroid concentration was measured after 4 h incubation and normalized against the maximal effect of forskolin without steroid. The average values ± SD (n = 4) are shown and represent the fold repression mediated by the two different cotreatments on CRH promoter activity. DEX is more potent in suppressing CRH promoter activity.
