| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Biochemistry and Molecular Biology, Pharmacia Corp., St. Louis, Missouri 63198
Address all correspondence and requests for reprints to: Deepak S. Lala, Molecular Sciences and Technologies, Ann Arbor Laboratories, Pfizer Inc., 2800 Plymouth Road, Ann Arbor, Michigan 48105. E-mail: deepak.s.lala{at}pfizer.com.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
B and activator protein 1. Ligand-bound GR can also activate gene transcription (transactivation) via direct binding to glucocorticoid response elements. Transactivation by GR is potentiated by accessory coactivators such as steroid receptor coactivator 1 and peroxisome proliferator-activated receptor 
coactivator 1, whereas the role of these proteins in transrepression is unclear.
Here, we show that GR can recruit several coactivator receptor interacting domains in a ligand-dependent manner. All interactions require the charge clamp defined by K579/E755, while a subset also requires a second charge clamp defined by R585/D590, within the GR ligand-binding domain. A point mutation, E755A, abolished all GR-receptor interacting domain interactions and led to a decrease in GR-mediated transactivation, but did not significantly affect GR-mediated transrepression of Gal4-p65 activity. Overexpression of a GR-interacting coactivator peptide blocked transactivation but did not affect transrepression of p65 or TNF
-induced IL-6 promoter activity. Finally, the GR antagonist RU486 did not recruit coactivators to GR but maintained the ability to transrepress p65 activity.
Our data suggest that different coactivators utilize distinct contact points to interact with GR. Although GR interactions with specific coactivators are critical for transactivation, they appear to be dispensable for at least certain aspects of GR-mediated transrepression of nuclear factor-B. This is consistent with the notion that all GR- mediated repression is not intrinsically linked to activation and can be separated mechanistically.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
coactivator 1 (PGC-1) (2, 3, 4, 5). These interactions are thought to play an important role in GR-mediated transactivation.
GR can also mediate repression of genes in a DNA-dependent or -independent manner (6). The latter function is considered important in the antiinflammatory actions of GR, and glucocorticoids are very effective agents in the battle against inflammatory disorders. Their beneficial effect relies primarily on their ability to repress nuclear factor-B (NF-B) and/or activator protein 1 (AP-1)-driven proinflammatory gene expression (6, 7, 8). Interestingly, coactivators also appear to be important for the function of NF-B, a widely expressed proinflammatory transcription factor that typically exists as a heterodimer of p65 (rel A) and a p50 subunit. This complex plays a key role in the up-regulation of several genes involved in immune responses and inflammation (9). CBP was shown to directly interact with rel A (p65), and enhance its transcriptional activity (10, 11). Similar potentiating effects were observed on NF-B- and AP-1-driven transcription by SRC-1 (12, 13). These data suggested a competition model for GR-mediated repression in which limited amounts of cellular coactivators were competed away from NF-B and/or AP-1 by ligand-bound GR. In support of this model, it was shown that both GR and p65 bind to the same region of CBP, and overexpression of CBP and SRC-1 rescue GR- repressed p65-dependent transactivation (14). However, other experiments suggest a different model wherein repression of NF-B and also AP-1 by GR occurs irrespective of coactivator levels in the cell (15, 16).
To further understand the role of coactivators in GR-mediated transrepression of NF-B, we designed a molecular strategy in which we first developed GR-dependent transactivation and transrepression assays as well as GR-coactivator interaction assays. We demonstrate that receptor-interacting domains (RIDs) from many coactivators interact with GR but require different contact points within the GR-LBD for these interactions. A mutational analysis revealed that the so-called "charge clamp" comprising the amino acids lysine (K579) and glutamic acid (E755) in helices 3 and 12, respectively (17), within GR, are important for mediating interactions with all the RIDs tested. In particular, E755 is critical for GR-mediated interactions with coactivators, and a single-point mutation leads to a loss in all GR-RID interactions and to a decrease in GR-dependent transactivation. However, the same mutations have no effect on GR-dependent transrepression of p65, the transcriptionally active unit of NF-B. Furthermore, the GR antagonist RU486 that lacked the ability to recruit key coactivator peptides and failed to transactivate was also able to transrepress p65. These data suggest that specific coactivators, while critical for GR-mediated transactivation, are not necessary for certain aspects of GR-mediated transrepression of NF-B.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). A GR-dependent transrepression assay in HeLa cells was developed using the human p65 domain of NF-B fused to the Gal4-DNA-binding domain (DBD), previously shown to have high constitutive activity that could be repressed by dexamethasone, presumably due to the presence of endogenous GR (15). We also observed some repression of p65 activity in these cells in the absence of GR (
20%); however, upon titrating increasing amounts of GR to the cells, we were able to see much greater repression, with a 1:5 ratio of p65 to GR, leading to approximately 90% repression of Gal-p65 activity (Fig. 1B). A full dose-response curve of dexamethasone-mediated inhibition of p65 activity also demonstrated stronger repression in the presence of GR and gave the expected EC50 value indicating this system could be used to assess GR- dependent repression of p65 (Fig. 1C).
|
indicated that the ligand-dependent activation function 2 domain is critical for ligand-mediated activation of the receptor and recruitment of coactivators (2, 17, 19, 20). GR is also dependent on its activation function 2 domain for transactivation and for recruitment of coactivators in response to ligand (21). Glutamate and lysine residues present within helix 12 and 3, respectively, are highly conserved in LBDs of nuclear receptors and are thought to form a charge clamp that contacts the backbone atoms of specific LxxLL motifs conserved within multiple coactivators (17). For estrogen receptor and peroxisome proliferator-activated receptor-, this charge clamp has been shown to be critical, not only in ligand-dependent activation, but also in transrepression (22, 23). Although these residues are also conserved in GR, their role in mediating transrepression is not clear.
Some studies using overexpressed coactivators have suggested that these proteins are important for GR-mediated transrepression (14); however, other studies indicate that the GR can repress NF-B and AP-1 irrespective of coactivator levels within cells (15, 16). To further assess the role of the charge clamp and that of coactivators in GR-mediated NF-B transrepression, we mutated the key glutamate and lysine residues within the LBD and tested their ability to interact with coactivators in a mammalian two-hybrid system. Figure 2A shows the positions and specific mutations within the GR-LBD (left). These mutations were then tested alongside the wild-type GR for interactions with a LxxLL-containing peptide derived from the coactivator PGC-1, shown previously to interact with GR in a ligand-dependent manner (5). Either mutation leads to a block in the ability of the receptor to interact with the coactivator peptide (Fig. 2A, right). Thus, consistent with other nuclear receptors, GR also requires the K579/E755 charge clamp for interacting with coactivators.
|
, wild-type GR interacts with several coactivator peptides in the presence of dexamethasone. Interestingly, several interacting peptides contain an L immediately N-terminal (or position -1) of the LxxLL motif (indicated by L in the figure; for specific LxxLL motifs see Materials and Methods). A single mutation (E755A) within the first charge clamp completely blocked the ability of GR to interact with all peptides tested (Fig. 2C). Previous experiments performed by Darimont et al. (21) had also indicated the importance of the Leucine residue at position -1 in mediating GR interactions. To further test the importance of this residue, we mutated this common residue within the PGC-1 LxxLL peptide (L 
A) to see whether this would affect its ability to interact with GR. We also converted three amino acids at the N-terminal sequence of NR box 1 derived from GRIP1, which does not normally interact with GR, to those present in PGC-1 (QTK PSL) (Fig. 2D, top). The ability of these mutants to interact with GR was then tested in mammalian two-hybrid experiments. The L A mutation in PGC-1 completely prevented its interaction with GR, whereas the GRIP1 NR1 box mutation (QTK PSL) gained the ability to interact with GR (Fig. 2D, bottom). These data indicate that Leucine at position -1 plays an important role in mediating several coactivator interactions with GR. It should be pointed out that, clearly, other flanking amino acids are also critical in mediating specificity because some interacting peptides did not possess L-1, and one peptide with an L at position -1 did not interact with GR (data not shown).
Next, we tested the consequence of mutating the second charge clamp (R585A/D590A). Interestingly, this mutation blocked the interactions of only a few peptides while it permitted the interactions of several others (Fig. 2E). Interestingly, the subset of peptides affected by mutating the second charge clamp contain specific amino acid residues R + 2, D + 6, (indicated by R/D in the figure; for specific sequence information see Materials and Methods) that are present in the third LxxLL motif of certain coactivators (e.g. GRIP1) and were recently shown to interact specifically with the R585/D590 charge clamp (24). However, other LxxLL peptides lacking these residues (e.g. PGC-1) were not affected by the second charge clamp mutations. These data indicate that the K579/E755 charge clamp is critical for a wide range of LxxLL interactions, whereas the second R585/D690 clamp is critical for a subset of LxxLL interactions. Thus, the requirements for GR-GRIP1 vs. GR-PGC-1 interactions appear to be quite different.
Role of the K579/755E Charge Clamp in Transactivation and Transrepression
Next, we focused our attention on understanding the role of the K579/E755 charge clamp in mediating GR-dependent transactivation and transrepression. For this, we created single-point mutations within the wild-type GR LBD and tested their ability to either activate the GRE-reporter or repress p65 in our GR-dependent assays. Consistent with the notion that coactivator recruitment is critical for transactivation, mutations within the first charge clamp (E755A/K579A) led to a significant loss in ligand-dependent transactivation (Fig. 3A). Because the E755A mutant appeared to be necessary for all of the coactivator interactions in our studies, we examined its effect on transrepression of p65 in our GR-dependent transrepression assay. As shown in Fig. 3B, overexpression of E755A led to a repression curve indistinguishable from that of wild-type GR. The K579A mutation also repressed Gal-p65 in a similar manner (data not shown). We also extended our studies using other cell lines that do not contain GR, e.g. CV-1 and HT29 cells (25, 26). In both cell types the effect of the E755A mutant was similar to that of wild-type GR in mediating repression of p65 transcriptional activity (Fig. 3, C and D).
|
To further explore the dependency of coactivators in transactivation and repression, we designed a dominant negative strategy, as shown for other receptors (27), whereby overexpression of a coactivator peptide should block GR function that is coactivator dependent. Our hypothesis was that overexpression of such a peptide would block GR-dependent transactivation but, based on our mutational analysis, would not block the ability of GR to transrepress p65 activity. We therefore overexpressed the PGC-1 LxxLL and tested its ability to block transactivation in our GR-dependent transactivation assay. We also set up a modified repression assay in which the IL-6 promoter, known to be inducible by TNF and to contain NF-B response elements (9), was linked to a luciferase reporter system. This was necessary because our LxxLL-containing peptides were fused to the Gal4DBD (see Materials and Methods) and would interfere with the Gal-p65 repression system through interference with DNA binding. As predicted, overexpression of wild-type PGC-1 blocked the ability of GR to activate the GRE-dependent reporter. In contrast, and consistent with our hypothesis that GR-dependent repression of p65 is not dependent on its interactions with coactivators, PGC-1 LxxLL overexpression did not significantly affect TNF-induced IL-6 promoter-reporter activity (Fig. 3E). Although we cannot completely exclude the possibility that GR-coactivator interactions in a repression complex, but not GR-coactivator complexes in an activation complex, are resistant to disruption by the LxxLL peptides, these data are consistent with our mutational studies. An isolated NF-B response element-reporter construct, when activated by TNF, was also blocked in the presence of either wild-type GR or GRE755A, in a ligand-dependent manner (data not shown).
RU486 Does Not Elicit GR-Coactivator Peptide Interactions but Has the Ability to Transrepress
Finally, we compared the ability of dexamethasone and the GR antagonist RU486 to facilitate GR-coactivator interactions in the mammalian two-hybrid system. RU486 has also been shown previously to have some ability to transrepress and to exhibit a dissociated profile (11, 28). We therefore examined the ability of these compounds to recruit multiple LxxLL-containing peptides derived from the coactivators PGC-1, GRIP-1 SRC-1, thyroid hormone receptor-associated protein 220 (TRAP220), or other synthetic coactivator peptides: pep293, pepD2, pep14, and pep6 (27).
In contrast to dexamethasone, RU486 did not recruit any of the coactivators tested (Fig. 4A); furthermore, it antagonized the ability of dexamethasone to recruit PGC-1 (Fig. 4B). These data clearly demonstrate that RU486 causes GR to adopt a conformation that is very different from dexamethasone. Next, we tested the ability of RU486 to repress p65 activity in our transrepression assay. As shown (Fig. 4C), RU486 was as effective as dexamethasone in repressing p65 activity in this system. Together, these data support the notion that whereas transactivation requires endogenous coactivators, transrepression of nuclear p65 activity can occur independently of interactions with cellular coactivators.
|
B and whether a compound that prevents all GR-coactivator interactions will be as efficacious as dexamethasone in mediating all aspects of repression. It is clear from previous studies that the ligand- and DNA-binding domains of GR are critical for its antiinflammatory effects, although DNA binding itself is not (29, 30), but other requirements necessary for GR-dependent repression remain elusive.
Based on our studies, it appears that ligand-dependent interactions of coactivators via the K579/E755 charge clamp are not essential for, at least, certain aspects of transrepression. We also observed that although RU486 did not recruit any of the coactivators tested, it was fully capable of blocking p65 activity. It is important to note that RU486 has also been shown to inhibit p65-associated histone acetyltransferase activity (28). However, it is less effective than dexamethasone in inhibiting specific proinflammatory genes in a more physiologically relevant state; this may be due to its inability to recruit accessory proteins that contain histone deacetylating activity (28). In our hands, RU486 can also repress endogenous IL-6 protein induction; however, it appears to be less effective than dexamethasone (Ref. 27 and data not shown). Thus, it is possible that for a compound to effectively distinguish transrepression from transactivation, not only must it be nonpermissive for GR-coactivator interactions but then must be permissive for GR interactions with some corepressor proteins that recruit histone deacetylases to GR on NF-B-driven promoters. Other possibilities include recruitment of proteins that might interfere with phosphorylation of specific serine residues within the pol II carboxyl-terminal domain, in complexes containing GR-NF-B and basal transcription factors (29). It is thus possible that ligand-dependent interactions with other novel proteins, perhaps binding to other regions of GR, such as its DBD, will be important for effective transrepression. Recently, a GR-interacting protein, named GT198, was identified using the GR-DNA binding domain as bait to screen a rat pituitary cDNA library (31), but functional assays indicated it acts as a coactivator for GR-dependent transcription of the mouse mammary tumor virus promoter (31). It would be interesting to determine whether using similar approaches to screen cDNA libraries would identify GR corepressors in the presence of glucocorticoids.
Although GR can repress nuclear p65 transcriptional activity in the absence of coactivator interactions, it is likely that broader transrepression by GR is more complex involving multiple layers. For example, repression of other proinflammatory transcription factors, such as AP-1, may have other requirements; repression may also be highly dependent on the context of the promoter. Recent data suggest that GRIP-1 is a key component of transcriptional repression of the collagenase-3 promoter by GR (32, 33). Although this appears somewhat paradoxical, it is possible that some coactivator proteins exhibit dual functions and are required under specific conditions for transrepression. Thus, another hypothesis is that selective interactions of coactivators with GR may be a way to distinguish between different GR functions. This should be achievable because we have shown that the requirements of coactivator peptides containing R +2/D+6 are different compared with other peptides that lack these specific sequences yet interact strongly with GR. Another study suggests that GR can also repress NF-B via association with the catalytic subunit of protein kinase A (34). It is also important to note that certain aspects of GR-dependent repression might require activation of certain genes.
Although several mechanisms appear to be involved, it is likely that a GR ligand that exhibits a beneficial profile in vivo will result from causing a unique conformation within the receptor, leading to selective GR interactions with cofactors, either between coactivators or other proteins involved in modulating transcription.
In summary, our molecular studies indicate that the primary charge clamp within the GR LBD, with respect to coactivator interactions, is K579/E755, whereas the R585/D590 residues appear to be important for only a subset of LxxLL-containing peptides. A mutation within E755 blocked GR-mediated transactivation but permitted GR-mediated transrepression of p65; overexpression of a dominant negative coactivator peptide also blocked transactivation but not TNF-induced IL-6 promoter activity. Finally, the GR antagonist RU486 was also able to repress p65 activity, although it lacked the ability to recruit several coactivator peptides. These studies suggest that at least some aspects of GR-mediated transrepression can occur in the absence of interactions with coactivators. It is interesting to note that of all the different coactivators that interact with GR, PGC-1 was recently shown to potentiate glucocorticoid induction of phosphoenolpyruvate carboxykinase, a key liver enzyme involved in regulating gluconeogenesis (35, 36). As noted above, although GR-mediated transrepression is clearly complex, it is tempting to speculate that GR ligands that lack the ability to recruit select coactivators, such as PGC-1, will act more selectively to block certain aspects of GR-mediated side effects but still maintain some ability to repress cytokine genes in vivo. Such compounds would be useful tools to further dissect GR-mediated activities and potentially represent novel pharmaceuticals that might be effective as antiinflammatory or antidiabetic agents but with limited side effects.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Sequences of Coactivator Peptides
Mutagenesis of the GR-LBD
Oligos for mutagenesis were synthesized by Life Technologies. Site-directed mutagenesis experiments were performed using QuikChange Site-Directed Mutagenesis Kit from Stratagene. Mutations were verified by sequencing.
Cell Culture and Transfections
Human liver cells (HUH-7) were maintained at 37 C in an atmosphere of 5% CO2 in RPMI 1640 medium containing 10% fetal bovine serum (FBS). Hela cells were cultured in DMEM containing 10% FBS. Transfections were performed as follows: on d 1, Hela cells (10,000 cells per well) or HuH 7 cells (10,000 cells per well) were plated onto 96-well plates and allowed to grow overnight. HT-29 cells (American Type Culture Collection, Manassas, VA; catalog no. HTB-38) were cultured in McCoy’s 5a medium with 10% heated-inactivated FBS. Transfection was performed on the second day with a total of 0.2 µg plasmids per well by using FuGENE6 transfection reagent (Roche, Indianapolis, IN). ß-gal plasmid was cotransfected to normalize the transfection efficiency. On the third day, cells were washed with PBS and phenol red-free media containing 5% charcoal/dextran-treated FBS (HyClone, Logan, UT) and the desired concentration of specific compounds was added. Cells were washed and lysed with LucLite Plus reagents (Packard BioScience B.V., Boston, MA) the next day. Luciferase was measured using TopCount NXT microplate scintillation and luminescence counter (Packard Instruments, Meriden, CT).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: AP-1, Activator protein 1; CBP, CREB-binding protein; DBD, DNA-binding domain; FBS, fetal bovine serum; GRE, glucorticoid response element; GRIP, GR-interacting protein; LBD, ligand-binding domain; NF-B, nuclear factor-B; PGC-1, peroxisome proliferator-activated receptor coactivator 1; RID, receptor-interacting domain; SRC-1, steroid receptor coactivator 1; TRAP220, thyroid hormone receptor-associated protein 220.
Received for publication November 11, 2002. Accepted for publication September 29, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTS AND DISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
B and steroid receptor-signaling pathways. Endocr Rev 20:435–459B-dependent transcription involves direct interference with transactivation. Proc Natl Acad Sci USA 94:13504–13509B (nuclear factor-B) and glucocorticoid receptor physical interactions and antagonism. Mol Endocrinol 14:1222–1234B-mediated transactivations. J Biol Chem 273:10831–10834B signaling by CREB-binding protein and steroid receptor coactivator-1. J Biol Chem 273:29291–29294B-driven genes by disturbing the interaction of p65 with the basal transcription machinery, irrespective of coactivator levels in the cell. Proc Natl Acad Sci USA 97:3919–3924. Nature 395:137–143[CrossRef][Medline]
-dependent repression of the inducible nitric oxide synthase gene. Mol Cell Biol 20:4699–4707-subunit of the epithelial sodium channel is an aldosterone-induced transcript in mammalian collecting ducts, and this transcriptional response is mediated via distinct cis-elements in the 5'-flanking region of the gene. Mol Endocrinol 15:575–588 and ß. Mol Cell Biol 19:8226–8239B by interfering with serine-2 phosphorylation of the RNA polymerase II carboxy-terminal domain. Genes Dev 14:2314–2329B and the glucocorticoid receptor. Proc Natl Acad Sci USA 97:11893–11898NURSA Molecule Pages Link:
This article has been cited by other articles:
 |
|
 |
|
  K. Suino-Powell, Y. Xu, C. Zhang, Y.-g. Tao, W. D. Tolbert, S. S. Simons Jr., and H. E. Xu Doubling the Size of the Glucocorticoid Receptor Ligand Binding Pocket by Deacylcortivazol Mol. Cell. Biol., March 15, 2008; 28(6): 1915 - 1923. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Newton and N. S. Holden Separating Transrepression and Transactivation: A Distressing Divorce for the Glucocorticoid Receptor? Mol. Pharmacol., October 1, 2007; 72(4): 799 - 809. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Anttonen, H Parviainen, A Kyronlahti, M Bielinska, D B Wilson, O Ritvos, and M Heikinheimo GATA-4 is a granulosa cell factor employed in inhibin-{alpha} activation by the TGF-{beta} pathway. J. Mol. Endocrinol., June 1, 2006; 36(3): 557 - 568. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L.-G. Bladh, J. Liden, A. Pazirandeh, I. Rafter, K. Dahlman-Wright, S. Nilsson, and S. Okret Identification of Target Genes Involved in the Antiproliferative Effect of Glucocorticoids Reveals a Role for Nuclear Factor-{kappa}B Repression Mol. Endocrinol., March 1, 2005; 19(3): 632 - 643. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L.-G. Bladh, J. Liden, K. Dahlman-Wright, M. Reimers, S. Nilsson, and S. Okret Identification of Endogenous Glucocorticoid Repressed Genes Differentially Regulated by a Glucocorticoid Receptor Mutant Able to Separate between Nuclear Factor-{kappa}B and Activator Protein-1 Repression Mol. Pharmacol., March 1, 2005; 67(3): 815 - 826. [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 |
90%), demonstrating GR-dependent repression. C, Dose-response curve for GR-mediated transrepression in HeLa cells. Assays were performed as in panel B except the ratio of p65 to GR was set at 1:5 and increasing amounts of dexamethasone were added to generate a dose-response curve. In the presence of GR there is a greater repression of p65 than in the absence of GR. The IC50 generated in the presence of GR is approximately 500 pM.
