![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Previous Article | Next Article 
Molecular and Cellular Biology, June 2008, p. 3817-3829, Vol. 28, No. 11
0270-7306/08/$08.00+0 doi:10.1128/MCB.01909-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Alberto Zambrano,
Ana I. Castillo, and
Ana Aranda*
Instituto de Investigaciones Biomédicas, Consejo Superior de Investigaciones Cieníficas and Universidad Autónoma de Madrid, Madrid, Spain
Received 23 October 2007/ Returned for modification 11 December 2007/ Accepted 14 March 2008
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Steroid hormone receptors are the prototype for a second model. Unbound steroid hormone receptors do not interact effectively with the corepressors and therefore do not have silencing activity. Binding of an agonist causes coactivator recruitment and activation. However, strong interactions with corepressors both in vivo and in vitro were observed with receptor-bound antagonists (36), although a weaker association with an agonist-bound androgen receptor has been recently described (46). Therefore, antagonists can convert steroid receptors into transcriptional silencers. Structural studies have shown that there is a shift in the position of the C-terminal helix (H12) of the ligand binding domain in the antagonist-bound estrogen receptor that not only disrupts coactivator interaction but might also expose the surface for corepressor binding. This helix, which contains the core ligand-dependent transcriptional activating domain (AF-2), sterically prevents corepressor binding to many nuclear receptors. For instance, the retinoid X receptor (RXR) has, at most, a weak silencing activity, but deletion of this region allows strong corepressor interaction and transforms RXR into a potent repressor (48).
Many nuclear receptors require heterodimerization with RXR for high-affinity DNA binding, and this receptor appears to play an active role in corepressor interaction (49). It has been proposed that sequence-specific interactions between H12 of RXR and a hydrophobic cleft of unliganded TR formed by residues in H3, H4, and H5 reposition H12 and unmask the corepressor interaction surface of RXR, allowing the unoccupied TR/RXR heterodimer to repress transcription. It is likely that RAR behaves similarly, but other RXR-heterodimerizing receptors would not present this interaction with RXR H12, thereby providing an explanation of variations in their repression function (48).
In this work we show that the binding of an agonist to the vitamin D receptor (VDR) induces association of the corepressors SMRT and NCoR to DNA-bound VDR/RXR heterodimers, thus demonstrating the existence of a novel type of corepressor recruitment. ChIP assays show that corepressors are recruited in a ligand-dependent manner to the 1,25-dihydroxy-vitamin D3 (VD3) target gene cyp24, and knockdown of corepressors by means of small interfering RNA (siRNA) enhances VD3-dependent transcriptional activity. Deletion of RXR H12 strongly enhances the ligand-dependent corepressor association, and the heterodimer of VDR with the truncated RXR shows a markedly reduced VD3-dependent transcriptional response. In contrast to that found with other heterodimers, enhanced binding of the corepressor requires an agonist-bound VDR and an intact VDR H12. Furthermore, although occupancy of VDR is responsible for corepressor recruitment, mutation of residues in H3, H4, and H5 of RXR (and not VDR) abolishes VD3-dependent corepressor recruitment and restores the transcriptional response to VD3. These results show that ligand binding to VDR triggers an allosteric change in the partner receptor that exposes its corepressor binding surface. Since VDR/RXR can recruit both coactivators and corepressors in an agonist-dependent manner, transcriptional activation by VD3 may be modulated specifically in different target cells by the relative levels of these coregulators.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, VDR
H12, K246A, E420Q, and RXRH12 were described earlier (3, 18). Other mutants were obtained by PCR with PfuTurbo DNA polymerase (Stratagene). All mutations were confirmed by direct sequencing. The luciferase reporter 4 x VDRE contains a multimerized DR3 (AGGTCAtgaAGGACA). In the cyp24 reporter plasmid (a gift from S. Kato), the –367 to +1 fragment of the human cyp24 promoter (which contains two vitamin D-responsive elements [VDREs] at –293/273 and –172/143) is fused to luciferase. Gal4-VDR (a gift of R. M. Evans) contains the yeast Gal4 DNA binding domain fused to the receptor ligand binding domain (118 to 427). The reporter UAS plasmid contains Gal4 DNA binding sites upstream of the E1B-TATA element fused to luciferase. The VP16 activation domain alone or fused to the C terminus of SMRT was obtained from A. Baniahmad. The constructs glutathione S-transferase (GST)-SMRT, GST-NCoR, and GST-ACTR code for the nuclear receptor-interacting domains of these proteins and have been described previously (3). The His-tagged nuclear receptor-interacting domain of TIF-2 has also been described (9). The GST-SMRT M1 (V2142A/I2143A), M2 (I2345A/I2346A), and M1/2, as well as VP16-SMRT M2 and M1/2, mutants were constructed by site-directed mutagenesis. Transfections. Cells, grown in 24-well plates, were transfected with 40 ng of reporter and 10 ng of pRL-TK-Renilla (Promega) as a normalization control using calcium phosphate. After transfection, cells were plated in a medium containing hormone-stripped serum, and after one night, incubation treatments were started. When appropriate, the reporter was cotransfected with expression vectors for RXR (12.5 ng), wild-type or mutant VDR (12.5 ng), or TIF-2 or SMRT (10 ng) or with an equivalent amount of empty vectors. Unless otherwise stated, luciferase activity was determined after 36 h of treatment with 3 nM VD3 or VD3 analogs and/or 1 µM 9-cis-RA. Experiments were performed with triplicate cultures, and each experiment was repeated at least three times. Data are represented as means ± standard deviations. Statistical P values were calculated using the one-tailed Student t test. Statistically significant differences (P < 0.05) are shown with an asterisk in the figures. siRNAs for NCoR, SMRT, and a nontarget control were purchased from Dharmacon (catalog no. M-003518-01, M-020145-01, and D-001210-01). siRNA transfections were performed using 33 nM of each siRNA and Lipofectamine 2000 (Invitrogen), as recommended by the manufacturer. The efficiency of knockdown was determined by Western blotting.
RNA extraction and quantification. Total RNA was extracted using Tri Reagent (Sigma), and mRNA levels were analyzed by quantitative reverse transcription-PCR as previously described (34), using the following primers: 5'-GGCAACAGTTCTGGGTGAAT-3' (forward) and 5'-TATTTGCGGACAATCCAACA-3' (reverse) for cyp24, 5'-TGGCGTTGTGAAGACCATGA-3' (forward) and 5'-ATAACTGTTCCACCTTGCCCCT-3' (reverse) for c-fos, 5'-ACCCTAGTTCTACCTCAGGC' (forward) and 5'-AAGATCTACTCCCCCATCAT-3' (reverse) for p21, 5'-GATGAAAGAAAGTCGCCTCG-3' (forward) and 5'-TGCAGTTTCTGCTGGACATC-3' (reverse) for cyp3A, 5'-TTGCAGTGATTTGCTTTTGC-3' (forward) and 5'-GTCATGGCTTTCGTTGGACT-3' (reverse) for osteopontin, 5'-GCACAGCCCAGAGGGTATAA-3' (forward) and 5'-CGCCTGGGTCTCTTCACTAC-3' (reverse) for osteocalcin, and 5'-GCGGAAGGGTACAGCCAAT-3' (forward) and 5'-GCAGCCGGCGCAAA-3' (reverse) for L19. Values obtained were corrected by GAPDH (glyceraldehyde-3-phosphate dehydrogenase) expression determined with primers 5'-ACAGTCCATGCCATCACTGCC-3' (forward) and 5'-CTAGCTGACCTCCTTGACCTG-3' (reverse).
Gel retardation assays. Assays were performed as described previously (3), with the DR3 oligonucleotide 5'-AGCTCAGGTCAAGGAGGTCAG-3' and 2.5 µl of in vitro-translated receptors (TNT Quick; Promega), in the presence and absence of 1 µg of GST-fused corepressors and 150, 250, or 400 ng of GST-ACTR or His-tagged TIF-2. VD3 and VD3 analogs (1 µM) or 9-cis-RA (10 µM) was present in the assays as indicated.
GST pull-down assays. Pull-down assays were performed with 5 µl of in vitro-translated [35S]methionine-labeled VDR or TR (TNT-T7 quick transcription/translation system; Promega) and 1 µg of GST-SMRT, GST-NCoR, or GST alone as previously described (18). When indicated, 1 µM VD3 or tri-iodothyronine (T3) was included in the binding reaction.
ChIP assays.
Cells growing in p150 dishes were washed twice in serum-free medium and treated for 2.5 h with 2.5 µM -amanitin. Cells were then washed and treated with 3 nM VD3 alone or in the presence of 1 µM 9-cis-RA. At the indicated time points, cells were fixed with 1% formaldehyde for 15 min at 37°C. The chromatin immunoprecipitation (ChIP) assay kit from Upstate (catalog no. 17-295) was used. Sonication was performed using a Bioruptor UCD-200TM (Diagenode) following the manufacturer's directions. For each immunoprecipitation, 2.5 x 106 to 3.0 x 106 cells were used with 2 µg of the following antibodies: anti-acetylated histone 4 (catalog no. 06-598; Upstate), anti-HDAC3 (Ab2379; Abcam), anti-SMRT (catalog no. PA1-842; Affinity BioReagents), anti-NCoR (sc-8994; Santa Cruz Biotechnology), anti-VDR (sc-9164x; Santa Cruz Biotechnology), anti-CBP (sc-369; Santa Cruz Biotechnology), anti-TIF-2 (AB3906; Chemicon); anti-SRC-1 (sc-8995; Santa Cruz Biotechnology), and normal rabbit serum immunoglobulins (sc-2027; Santa Cruz Biotechnology). DNAs were subjected to 35 cycles of PCR with primers 5'-TGCCTTCCTGGGGGTTATCTC-3' (forward) and 5'-CGTTTCCTCCTGTCCCTCTC-3' (reverse), which amplify a 
300-bp segment (–263/+22) of the cyp24 promoter; 5'-CCTCCTTTGCACAAGGTAGT-3' (forward) and 5'-AATGCACGTAAAGCGGCAAC-3' (reverse), which amplify an upstream region (–882/–676) of the gene; and 5'-TACTAGCGGTTTTACGGGCG-3' (forward) and 5'-TCGAACAGGAGGAGCAGAGAGCGA-3' (reverse), which amplify the GAPDH promoter (–460/–318). For sequential ChIP (re-ChIP), VDR immunoprecipitates were eluted in 10 mM dithiothreitol, diluted 20x in re-ChIP buffer (1% Triton X-100, 2 mM EDTA, 150 mM NaCl, and 20 mM Tris-HCl, pH 8.0), and immunoprecipitated with SMRT antibody.
Western blotting. Twenty micrograms of total protein was separated by 5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and probed with specific primary antibodies and horseradish-coupled anti-rabbit or anti-mouse immunoglobulin G (sc-2030 and sc-2005; Santa Cruz Biotechnology) as secondary antibodies. The primary antibodies used were anti-SMRT (catalog no. PA1-842; Affinity BioReagents) and anti-NCoR (catalog no. 06-892; Upstate) at 2 µg/ml, and anti-β-catenin (1/2,000) (catalog no. 610184; BD Biosciences) as a loading control.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
(PPAR) in solution but not on a peroxisome proliferator response element, a context that reflects their in vivo function more accurately (47). Therefore, we next examined recruitment of SMRT to the VDR/RXR heterodimer in gel retardation assays. Although in solution VDR is capable of binding SMRT to some extent, corepressor binding to the VDR/RXR heterodimer on a DR3 VDRE was not observed (Fig. 1B). This is more consistent with the finding that unliganded VDR does not repress transcription on this site. Additionally, a weak but detectable supershifted band was found upon incubation with VD3. This is not the case with the RXR ligand, 9-cis-RA, which did not cause corepressor recruitment. Since deletion of H12 in RXR allows strong corepressor interaction with this receptor and increases corepressor binding to PPAR/RXR (47, 48), we also analyzed binding of SMRT to the heterodimer of VDR with the H12-truncated RXR (RXRH12). Discernible SMRT binding to the mutant heterodimer was obtained in the absence of ligand (Fig. 1B, lane 7), and this binding was markedly enhanced in the presence of VD3 (lane 9). Incubation with 9-cis-RA had an opposite effect, causing SMRT release (lane 8). In contrast to steroid receptors in which antagonist ligands induce corepressor recruitment, ZK159222, a VDR ligand with very low agonistic activity, possessed a strongly reduced ability to recruit SMRT compared with VD3. This suggests an agonist-dependent binding of the corepressor to VDR/RXR. However, both the rexinoid LG100268 and the RXR-selective antagonist LG101208 were able to cause reverse corepressor binding to the VD3-bound heterodimer, thus showing that an agonist conformation of RXR is not required for this effect (Fig. 1C). The VDR agonist-dependent recruitment of corepressors was not restricted to SMRT, since the heterodimer also bound NCoR in a VD3-dependent manner (Fig. 1D). Although a DR3 is a common VDRE, VDR/RXR can also bind to elements with other configurations. SMRT binding in response to VD3 was also found using DR4 and IP-9 response elements (not illustrated). The carboxyl terminus of SMRT possesses two receptor-interacting domains (ID1 and ID2). These domains, also named CoRNR boxes, each contain the motif I/LXXI/VI (16, 28, 31). Since the receptor-interacting domains differ in their affinity for specific receptors (10, 42), we also analyzed the effect of individual and combined mutations in these domains. As shown in Fig. 1E, mutation of any of the receptor-interacting domains abolished VD3-dependent SMRT recruitment to the heterodimer.
Binding of SMRT to VDR/RXR in mammalian one-hybrid assays.
To analyze binding of SMRT to the wild-type and mutant VDR/RXR heterodimer in 293-T cells (Fig. 2A), transactivation of the VDRE reporter by VD3 was examined in cells transfected with the receptors and the VP16 activation domain alone or fused to the C terminus of SMRT containing the receptor-interacting domain (VP16-SMRT). Whereas expression of VP16 alone did not increase transactivation by VD3 in cells expressing VDR/RXR, this response was enhanced significantly by VP16-SMRT, confirming the ligand-dependent interaction of the corepressor with the wild-type heterodimer observed in the in vitro assays. Also, in support of the results obtained in the supershift assays, deletion of RXR H12 caused an increase of SMRT binding to the heterodimer. That the increase in transactivation is indeed a measure of SMRT interaction with the heterodimer was demonstrated by the finding that mutation of the ID2 motif of the corepressor was sufficient to block the increased response to VD3 that was mediated by both native and H12-deleted heterodimers (Fig. 2B). Similar results were obtained when both ID1 and ID2 were mutated (data not shown). VD3-dependent interaction of VDR/RXR with NCoR in these assays was also observed, and deletion of RXR H12 increased binding to this corepressor as well (Fig. 2C). Of interest, the increased association with corepressors caused a marked reduction of the response to VD3 mediated by VDR/RXRH12.
|
H12, the slope of the regression line was higher with the mutant than with the wild-type receptors, though these mediated a stronger transactivation by the VDR ligands. In gel retardation assays, it was also observed that only the full VDR agonists ZK161422 and ZK157202 caused strong SMRT association with the heterodimer (not illustrated).
In contrast with the results obtained in the assays performed with the receptor heterodimer and the VDRE, Gal-SMRT did not affect VD3-dependent transactivation of a UAS-containing plasmid by VP16-VDR. This result agrees with the lack of a VD3-dependent effect on SMRT binding obtained in pull-down assays and suggests an important role of RXR in this interaction. Furthermore, when similar assays were performed in the presence of RXR or RXRH12, VD3 was still unable to induce association with VP16-SMRT (data not shown). This demonstrates not only that both partner receptors are required for corepressor recruitment but also that their DNA binding domains need to be attached to an appropriate DNA response element to allow ligand-dependent SMRT recruitment.
VD3 causes recruitment of both coactivators and corepressors. Despite causing recruitment of corepressors to VDR/RXR, VD3 has a net positive effect on transcription, which indicates that coactivator recruitment should be predominant. Thus, we next examined the effect of the p160 coactivators TIF-2 and ACTR on SMRT binding. As can be observed in Fig. 3A, the mobility of the ternary complex of the receptor heterodimer with p160 coactivators was slower than that obtained with SMRT, and the coactivators displaced binding of the corepressor. These results suggest that the coactivator/corepressor ratio should determine VD3-mediated transcriptional activity. If this was the case, high coactivator levels should increase the response to VD3, whereas this response should be reduced in cells expressing high corepressor levels. This was, in fact, observed in cells expressing VDR/RXR or the heterodimer of VDR with RXR lacking H12, in which overexpression of TIF-2 was able to increase transactivation by VD3, and transfection with a SMRT expression vector had the opposite effect (P < 0.05) (Fig. 3B).
|
H12 is related to increased corepressor binding, incubation with 9-cis-RA, which blocks this binding, should have a stimulatory effect. Indeed, transactivation by the truncated heterodimer was increased significantly when the vitamin was combined with the RXR ligand (Fig. 3C). In contrast with the native heterodimer, 9-cis-RA does not cause coactivator recruitment to VDR/RXRH12 (34). Therefore, transactivation of the truncated heterodimer by the RXR ligand cannot be mediated by recruitment of coactivators and should instead be secondary to corepressor release. Increased association with SMRT by deletion of RXR H12 also produced a statistically significant reduction of the ability of VD3 to induce transcription from an endogenous gene. As shown in Fig. 3D, transcripts from the cyp24 gene were strongly induced by VD3 in 293-T cells upon expression of VDR/RXR, and this response was markedly blunted when VDR/RXRH12 was expressed. In addition, 9-cis-RA cooperated with VD3 to induce cyp24 mRNA levels, and the reduced response from VDR/RXRH12 was significantly recovered. Role of endogenous corepressors in VD3-dependent transcription. The results described above suggested that increased corepressor levels and/or recruitment diminishes VD3-dependent transcription. siRNA was used to directly address whether recruitment of endogenous SMRT and NCoR by VD3 has a physiological inhibitory effect on the transcriptional activity of VDR/RXR. For this purpose, 293-T cells were cotransfected with the heterodimer and control, SMRT, or NCoR siRNA. Corepressor levels were reduced by the corresponding siRNAs (Fig. 4E), and as shown in Fig. 4A, transactivation by VD3 was enhanced by the SMRT and NCoR siRNAs, in comparison to that found with the cells transfected with the control siRNA. As shown in Fig. 4B, siRNAs for both corepressors also increased ligand-dependent transcripts of the target gene cyp24, although in the case of the endogenous VD3-responsive gene, down-regulation of NCoR had a much stronger effect. To analyze whether knockdown of this corepressor can also activate other putative VD3 target genes (26), endogenous transcripts for p21, c-fos, cyp3A, osteopontin, and osteocalcin were also determined. Interestingly, induction of these genes by VD3 was negligible in 293-T cells transfected with a control siRNA, but a significant induction was found in cells expressing the corepressor siRNA after 4 h and 12 h of incubation with VD3 (Fig. 4D). The corepressor siRNA did not have a general effect of transcription, since no changes in the transcripts for the nontarget gene L19 were found. In addition, the influence of corepressor knockdown on cyp24 mRNA levels was also assessed in 293-T cells that were not transfected with the receptors (Fig. 4C). The low levels of endogenous receptors mediated only a weak induction of cyp24 transcripts by VD3, but this response, as well as cooperation with 9-cis-RA, was significantly enhanced in cells transfected with NCoR siRNA.
|
|
|
ChIP with anti-VDR antibody revealed that the receptor binds the target promoter in a VD3-dependent manner and with similar kinetics as SMRT. Furthermore, re-ChIP of the VDR-immunoprecipitated chromatin with SMRT antibody shows that these proteins form a complex on the cyp24 promoter (Fig. 6D). Also, in agreement with the finding that VD3 causes an increase in acetylated histone H4, ligand-dependent association of the histone acetyltransferase CBP to the promoter was observed (Fig. 6D). These results again support the view that both corepressors and coactivators are recruited to the target promoter in an agonist-dependent manner.
Furthermore, in cells transfected with SMRT or NCoR siRNAs (Fig. 6E), incubation with VD3 caused a stronger increase in acetylated histone H4, as well as a concomitant reduction in HDAC3 recruitment to the cyp24 promoter but not to a VD3-unresponsive promoter (Fig. 6F). This again supports the finding that endogenous corepressors can function to attenuate transcription by VD3-liganded native VDR/RXR heterodimers.
Role of RXR and VDR H12 in SMRT recruitment.
Since recruitment of SMRT to the DNA-bound heterodimers appears to be dependent upon the binding of an agonist to VDR, we next analyzed the effect of mutations in the AF-2 domain of this receptor on in vitro association with the corepressor. The conformational change elicited in the receptors by agonist binding involves repositioning of H12. This helix folds back against the ligand binding pocket and together with residues in H3, H4, and H5 generates a hydrophobic cleft responsible for coactivator interaction. A conserved glutamic acid residue in H12 (E420 in VDR) and an invariable lysine residue in H3 (K246 in VDR), contact directly with the coactivator and form a charge clamp that stabilizes binding (for a review, see references 5, 23, and 29). We have previously shown that deletion of H12 (VDRH12) or mutations E420Q and K246A abolish VD3-dependent binding of coactivators and transcriptional activity (18, 34). As shown in Fig. 7A, the H12 point mutant also exhibited a distinct reduction of SMRT association in response to VD3. In contrast, the K246A mutant showed an increased corepressor binding that was noticeable even in the presence of 9-cis-RA. Therefore, the H12 E420 residue of VDR has an important role in SMRT recruitment, but the charge clamp required for coactivator binding appears to be dispensable for corepressor binding.
|
H12 was not observed with different point mutations of RXR in this helix (Fig. 7B).
The pattern of SMRT recruitment by VDR/RXR was very different from that obtained with the heterodimer of RXR with TR. Binding of SMRT to unoccupied native TR/RXR was weak (Fig. 7C, left panel), since this heterodimer does not recruit corepressors with the same potency as TR homodimers (3, 44), and the corepressor was released by T3. In addition, deletion of RXR H12 also caused a marked enhancement of SMRT binding by TR/RXR (Fig. 7C, right panel), but this increase was ligand independent, and occupancy of TR by T3 caused some release of SMRT that was potentiated by 9-cis-RA. Furthermore, mutation of TR H12 in the glutamic acid residue equivalent to E420 in VDR did not reduce corepressor binding to TR/RXRH12 but, rather, significantly inhibited SMRT dissociation by both ligands.
SMRT binds to RXR in the heterodimer.
Association of corepressors with TR requires the conserved residues A, H, and T in the so-called CoR box located in H1 (15). Mutation of the homologous amino acids (AEV) in RXR lacking H12 abrogates interaction with this receptor (48). However, the AEV mutant was able to interact normally with SMRT in the context of the VDR heterodimer, showing that this region of RXR is dispensable for corepressor association in response to VD3 (Fig. 8A). CoR box residues are not exposed in the ligand binding domain surface, and the corepressor does not interact directly with this TR region but docks to a hydrophobic groove formed by H3, H4, and H5 that overlaps with the coactivator-interacting domain (16, 22, 28, 31). We then tested the effect of mutations in the putative corepressor-binding surface of RXR on VD3-dependent SMRT association. Mutations V280R in H3, V298R in H4, and W305K in H5 (equivalent to those that abrogate NCoR binding to TR), abolished recruitment of SMRT to the VDR/RXRH12 heterodimer in response to VD3 (Fig. 8A), showing that the corepressor binds to this moiety of the heterodimer.
|
The mutant heterodimers in the RXR corepressor binding surface also failed to interact in cells with the corepressor, as proven by the finding that VP16-SMRT was unable to enhance the VD3 response mediated by the H3, H4, and H5 mutants in VDR/RXRH12 (Fig. 8C). In contrast, and confirming the data obtained in the mobility shift assays, interaction of the AEV mutant with SMRT was preserved in the cellular assays. On the other hand, mutation in H3, H4, and H5 of VDR totally abolished responses to VD3 (not illustrated), since these mutations also block coactivator recruitment and transcriptional activity (34). Since mutations in the corepressor binding surface of RXRH12 blocked VD3-dependent interaction of the heterodimer with SMRT, it was expected that these mutations could restore the response to VD3 mediated by the heterodimer of VDR with RXR lacking H12. Indeed, as shown in Fig. 8C, mutations V280R, V298R, and W305K in the context of RXRH12 enhanced transactivation by VD3, reaching the levels obtained with the native receptors. These results demonstrate that the reduced VD3-dependent transcription by the truncated heterodimer was indeed due to increased corepressor binding. Confirming the importance of corepressor binding, the AEV mutant that interacted normally with VP16-SMRT still presented a reduced transcriptional activity.
The ability of the heterodimers of VDR with RXR H3, H4, and H5 mutants to mediate VD3 and 9-cis-RA induction of endogenous cyp24 transcripts was also examined. As illustrated in Fig. 9A, these mutations also reversed significantly the reduced transcriptional response to VD3 obtained with VDR/RXRH12. In addition, these mutants, which are defective both in coactivator recruitment upon incubation with 9-cis-RA and in corepressor recruitment upon incubation with VD3, did not mediate transcription by 9-cis-RA, and cooperation of both ligands was not observed. This contrasts with the results obtained with RXRH12 and the AEV mutant, which bind corepressors and in which the transcriptional effect of VD3 was further enhanced by 9-cis-RA. This difference in behavior must be a consequence of corepressor release in the presence of the RXR ligand, a behavior that cannot occur after mutation of the RXR corepressor binding surface. On the other hand, mutations in H3, H4, and H5 on the wild-type RXR also abolished transcriptional responses to 9-cis-RA (Fig. 9B). However, in contrast to H12 deletion, mutation of the corepressor binding surface did not repress but, rather, increased the response to VD3. This increase matches abolishment of corepressor recruitment by VD3 and is in accordance with the results obtained after corepressor knockdown, indicating again that corepressor recruitment by VD3 plays a physiological negative role in VD3-mediated responses.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Most importantly, as observed with the cyp24 promoter in ChIP assays, SMRT and NCoR are recruited to a target gene in a VD3-dependent manner. Since VDR is predominantly cytoplasmic and translocates to the nucleus upon ligand binding (1), association of corepressors with the unliganded receptor would not be detected by ChIP assays. However, our data show that corepressors are recruited together with the liganded receptors and that they form a complex on the promoter. Recruitment of HDAC3, which forms a tight complex with SMRT and NCoR (12, 20), is also found, showing that the corepressor complexes are recruited to the target gene in response to VD3. Binding of corepressors and the deacetylase to the VDRE-containing promoter contrasts to results obtained with genes containing binding sites for other receptor heterodimers such as TR/RXR or PPAR/RXR, in which HDAC3 as well as SMRT and NCoR are recruited in the absence of ligand and released upon ligand binding (11, 19). This mirror image ties in with the contrasting model of corepressor recruitment by VDR/RXR and those heterodimers.
Interestingly, it has been recently demonstrated (8, 25) that NCoR corepressor complex components are recruited to the 1 (OH)ase promoter after incubation with VD3. This is a key enzyme in VD3 biosynthesis, which possesses a negative VDRE to which the receptors are tethered through interaction with a basic helix-loop-helix activator (25). In light of our present results, this recruitment would not be specific for the transrepressed gene but, rather, would reflect the intrinsic property of VDR/RXR to associate with corepressors in the presence of VD3.
Role of H12 of RXR and VDR in corepressor recruitment by VDR/RXR. H12 of RXR plays an inhibitory role in VD3-dependent corepressor recruitment by VDR/RXR, since deletion of this helix enhances the ability of the heterodimer to bind SMRT and NCoR, both in vitro and in cells. The increased association with corepressors correlates with the strongly reduced transcriptional activity of VD3 through this heterodimer. Unlike that found with other unoccupied heterodimers in which deletion of RXR H12 increases corepressor recruitment, which is reversed upon ligand binding (35, 48), heterodimerization with VDR masks the ability of truncated RXR to bind SMRT, and strong binding of the corepressor occurs only upon VDR occupancy.
In contrast to the inhibitory role of RXR H12 for corepressor binding, the integrity of VDR H12 is required for efficient binding of corepressors to the heterodimer in response to VD3. In agreement with the fact that association with the corepressors correlates with the agonistic activity of VDR ligands, deletion or mutation of this helix blocks VD3-dependent recruitment of coactivators (18, 34) and also inhibits corepressor binding. This demonstrates that an agonist conformation of VDR, which implies repositioning of H12, is required for association of the corepressors to the heterodimer. Not surprisingly, the structural requirements for binding of coactivators and corepressors appear to be different. This is indicated by the dispensability of the charge clamp between H3 and H12 for corepressor association, though the clamp is crucial for coactivator recruitment. Again the role of H12 of VDR in corepressor recruitment is different from that of TR or RAR. In the case of the latter receptors, mutation or deletion of H12, rather than inhibition of corepressor binding, potentiates association of the heterodimers with the corepressors and reduces the ligand ability to dismiss these proteins (4, 21).
Corepressors associate with RXR in response to ligand binding to VDR.
Mutation analysis of TR, later confirmed in a crystallographic study with PPAR (22, 28, 31), demonstrates that corepressors dock into the hydrophobic groove that also belongs to the coactivator binding site (43), predicting that binding of coactivators and binding of corepressors are mutually exclusive. Since corepressor recruitment by VDR/RXR is obtained in response to VD3 and was dependent upon the integrity of the VDR H12, it intuitively looks as if corepressors would bind to this receptor as a consequence of the repositioning of this helix to the agonist position. However, mutation of the putative corepressor binding surface of this receptor, formed by the hydrophobic groove created by H3, H4, and H5 (16, 28, 31), has little, if any, effect on recruitment. In contrast, mutation of the same surface in RXRAF-2 abolishes corepressor recruitment, demonstrating that they are bound to this receptor. This indicates the existence of an important allosteric communication between both members of the heterodimer by which the conformational change elicited by ligand binding to VDR is transmitted to its partner. This is a mechanism reminiscent of the "phantom ligand" effect observed with liver X receptor/RXR (41), in which the activation domain of liver X receptor is required for stimulation by the RXR ligand. Crystals for VDR/RXR are presently unavailable, and the structural basis for this change is presently unknown. However, it is clear that H12 of both receptors plays a role in these interactions, and the crystal structures of other receptor dimers and heterodimers have demonstrated docking of H12 of one receptor within its partner (40). Our results suggest that the shift in position of VDR H12 upon ligand binding unmasks the corepressor binding surface of RXR. This domain in TR would have an inverse function, namely, to reposition H12 in RXR to mask this surface and to release the corepressor (48). Of interest, in contrast to that observed with RXR homodimers, mutation of the RXR CoR box does not inhibit VD3-dependent corepressor recruitment, showing that in the heterodimeric context this region is dispensable.
The RXR ligand blocks corepressor recruitment by the VDR ligand.
Unlike binding of VD3 to VDR, binding of 9-cis-RA to RXR reverses SMRT and NCoR association with VDR/RXR. Not only the natural ligand, but also an antagonist, can block VD3-dependent corepressor recruitment, indicating that occupancy of the RXR ligand binding pocket, but not acquisition of an agonist configuration, is required for inhibiting corepressor binding. Surprisingly, the RXR ligand is able to induce this effect in the absence of H12, demonstrating that reorganization of other receptor domains appears to be sufficient for this phenomenon. Corepressor release has a clear functional effect, since 9-cis-RA still has the ability to cooperate with VD3 to induce transcription through VDR/RXRAF-2, despite the fact that this heterodimer is unable to recruit coactivators in response to the RXR ligand (34). These transcriptional actions of 9-cis-RA are lost in mutant receptors in the corepressor binding surface, demonstrating that the stimulatory effect of this ligand is indeed due to corepressor release.
NCoR and SMRT have a physiological inhibitory role in VD3-dependent transcription.
Our results demonstrate that VD3 causes recruitment of the receptors and the corepressors to the cyp24 promoter, and that also induces the recruitment of coactivators and a net increase in the amount of acetylated H4 histone bound to the promoter. That binding of an agonist ligand to VDR causes recruitment of both coactivators and corepressors to VDR/RXR heterodimers suggests that corepressors could compete with coactivators and function in a negative feedback loop to attenuate VD3-induced transactivation. Therefore, the relative expression of SMRT and NCoR versus coactivators may regulate distinct VD3 responses in different target tissues. The inhibitory role of endogenous SMRT and NCoR in transcription by native VDR/RXR heterodimers was shown by siRNA experiments that demonstrate that corepressor down-regulation further enhances the VD3-dependent increase of acetylated histone H4 bound to the cyp24 promoter and transcriptional activation of this target gene. Corepressors also inhibit the response of other VD3 target genes that show little, if any, activation in cells expressing normal corepressor levels but display a significant ligand-dependent response upon corepressor knockdown. Furthermore, SMRT and NCoR knockdowns appear to show a differential effect on VD3 responses depending on the promoter and cell type. In addition, it has been recently shown that SMRT (but not NCoR) depletion reduces transcriptional activity of the agonist-bound estrogen receptor , although not that of other receptors (32), indicating that corepressors may play diverse biological roles in nuclear receptor-dependent transcription.
That both SMRT and NCoR as well as coactivators could be recruited to VDR/RXR upon VD3 binding is perhaps counterintuitive. However, the existence of corepressors such as RIP140 or LCoR, which recognize many agonist-bound receptors through LXXLL motifs and play an important role in nuclear receptor functions in vivo, is well documented (7, 37).
Agonist-dependent corepressor recruitment may have additional roles. It can also function in ligand-dependent target gene repression and may act as part of a cycle of cofactors recruited to target promoters by ligand-bound receptors. Receptors and coregulators rapidly cycle on and off target promoters (24). This cycling is accompanied by waves of histone acetylation and deacetylation. Recruitment of corepressors may contribute to promoter clearance and deacetylation.
In summary, our results are compatible with a model (Fig. 10B), in which binding of VD3 to VDR/RXR heterodimers would cause a conformational change in VDR that includes repositioning of H12 to an agonist position, as well as an allosteric change in RXR that would unmask the corepressor binding surface in this receptor. Binding of RXR ligand (either agonist or antagonist) would again conceal this surface, preventing association of the corepressor. In contrast with the requirement of VDR H12 for corepressor recruitment, RXR H12 plays an inhibitory effect, since deletion of this helix causes a large increase in VD3-dependent corepressor recruitment to the heterodimer. Furthermore, binding of 9-cis-RA to RXR lacking H12 is also able to prevent corepressor interaction, demonstrating that ligand-induced conformational changes in other RXR domains are sufficient to abolish association. Future studies will hopefully solve the structural requirements responsible for corepressor association with VDR/RXR heterodimers.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by grant BFU2004 3465 from Ministerio de Educación y Ciencia, grant RD06/0020/0036 from the Fondo de Investigaciones Sanitarias, and European grant CRESCENDO (FP-018652).
|
FOOTNOTES |
|---|
Published ahead of print on 24 March 2008.
These authors contributed equally to this work.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Castillo, A. I., R. Sanchez-Martinez, A. M. Jimenez-Lara, A. Steinmeyer, U. Zugel, and A. Aranda. 2006. Characterization of vitamin D receptor ligands with cell-specific and "dissociated" activity. Mol. Endocrinol. 20:3093-3104.
3. Castillo, A. I., R. Sanchez-Martinez, J. L. Moreno, O. A. Martinez-Iglesias, D. Palacios, and A. Aranda. 2004. A permissive retinoid X receptor/thyroid hormone receptor heterodimer allows stimulation of prolactin gene transcription by thyroid hormone and 9-cis-retinoic acid. Mol. Cell. Biol. 24:502-513.
4. Chen, J. D., and R. M. Evans. 1995. A transcriptional co-repressor that interacts with nuclear hormone receptors. Nature 377:454-457.[CrossRef][Medline]
5. Darimont, B. D., R. L. Wagner, J. W. Apriletti, M. R. Stallcup, P. J. Kushner, J. D. Baxter, R. J. Fletterick, and K. R. Yamamoto. 1998. Structure and specificity of nuclear receptor-coactivator interactions. Genes Dev. 12:3343-3356.
6. Dwivedi, P. P., G. E. Muscat, P. J. Bailey, J. L. Omdahl, and B. K. May. 1998. Repression of basal transcription by vitamin D receptor: evidence for interaction of unliganded vitamin D receptor with two receptor interaction domains in RIP13delta1. J. Mol. Endocrinol. 20:327-335.[Abstract]
7. Fernandes, I., Y. Bastien, T. Wai, K. Nygard, R. Lin, O. Cormier, H. S. Lee, F. Eng, N. R. Bertos, N. Pelletier, S. Mader, V. K. Han, X. J. Yang, and J. H. White. 2003. Ligand-dependent nuclear receptor corepressor LCoR functions by histone deacetylase-dependent and -independent mechanisms. Mol. Cell 11:139-150.[CrossRef][Medline]
8. Fujiki, R., M. S. Kim, Y. Sasaki, K. Yoshimura, H. Kitagawa, and S. Kato. 2005. Ligand-induced transrepression by VDR through association of WSTF with acetylated histones. EMBO J. 24:3881-3894.[CrossRef][Medline]
9. Germain, P., J. Iyer, C. Zechel, and H. Gronemeyer. 2002. Co-regulator recruitment and the mechanism of retinoic acid receptor synergy. Nature 415:187-192.[CrossRef][Medline]
10. Ghosh, J. C., X. Yang, A. Zhang, M. H. Lambert, H. Li, H. E. Xu, and J. D. Chen. 2002. Interactions that determine the assembly of a retinoid X receptor/corepressor complex. Proc. Natl. Acad. Sci. USA 99:5842-5847.
11. Guan, H. P., T. Ishizuka, P. C. Chui, M. Lehrke, and M. A. Lazar. 2005. Corepressors selectively control the transcriptional activity of PPARgamma in adipocytes. Genes Dev. 19:453-461.
12. Guenther, M. G., W. S. Lane, W. Fischle, E. Verdin, M. A. Lazar, and R. Shiekhattar. 2000. A core SMRT corepressor complex containing HDAC3 and TBL1, a WD40-repeat protein linked to deafness. Genes Dev. 14:1048-1057.
13. Heinzel, T., R. M. Lavinsky, T. M. Mullen, M. Soderstrom, C. D. Laherty, J. Torchia, W. M. Yang, G. Brard, S. D. Ngo, J. R. Davie, E. Seto, R. N. Eisenman, D. W. Rose, C. K. Glass, and M. G. Rosenfeld. 1997. A complex containing N-CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 387:43-48.[CrossRef][Medline]
14. Herdick, M., and C. Carlberg. 2000. Agonist-triggered modulation of the activated and silent state of the vitamin D(3) receptor by interaction with co-repressors and co-activators. J. Mol. Biol. 304:793-801.[CrossRef][Medline]
15. Horlein, A. J., A. M. Naar, T. Heinzel, J. Torchia, B. Gloss, R. Kurokawa, A. Ryan, Y. Kamei, M. Soderstrom, C. K. Glass, et al. 1995. Ligand-independent repression by the thyroid hormone receptor mediated by a nuclear receptor co-repressor. Nature 377:397-404.[CrossRef][Medline]
16. Hu, X., and M. A. Lazar. 1999. The CoRNR motif controls the recruitment of corepressors by nuclear hormone receptors. Nature 402:93-96.[CrossRef][Medline]
17. Huang, E. Y., J. Zhang, E. A. Miska, M. G. Guenther, T. Kouzarides, and M. A. Lazar. 2000. Nuclear receptor corepressors partner with class II histone deacetylases in a Sin3-independent repression pathway. Genes Dev. 14:45-54.
18. Jimenez-Lara, A. M., and A. Aranda. 1999. Lysine 246 of the vitamin D receptor is crucial for ligand-dependent interaction with coactivators and transcriptional activity. J. Biol. Chem. 274:13503-13510.
19. Li, J., Q. Lin, W. Wang, P. Wade, and J. Wong. 2002. Specific targeting and constitutive association of histone deacetylase complexes during transcriptional repression. Genes Dev. 16:687-692.
20. Li, J., J. Wang, J. Wang, Z. Nawaz, J. M. Liu, J. Qin, and J. Wong. 2000. Both corepressor proteins SMRT and N-CoR exist in large protein complexes containing HDAC3. EMBO J. 19:4342-4350.[CrossRef][Medline]
21. Lin, B. C., S. H. Hong, S. Krig, S. M. Yoh, and M. L. Privalsky. 1997. A conformational switch in nuclear hormone receptors is involved in coupling hormone binding to corepressor release. Mol. Cell. Biol. 17:6131-6138.[Abstract]
22. Marimuthu, A., W. Feng, T. Tagami, H. Nguyen, J. L. Jameson, R. J. Fletterick, J. D. Baxter, and B. L. West. 2002. TR surfaces and conformations required to bind nuclear receptor corepressor. Mol. Endocrinol. 16:271-286.
23. McInerney, E. M., D. W. Rose, S. E. Flynn, S. Westin, T. M. Mullen, A. Krones, J. Inostroza, J. Torchia, R. T. Nolte, N. Assa-Munt, M. V. Milburn, C. K. Glass, and M. G. Rosenfeld. 1998. Determinants of coactivator LXXLL motif specificity in nuclear receptor transcriptional activation. Genes Dev. 12:3357-3368.
24. Metivier, R., G. Penot, M. R. Hubner, G. Reid, H. Brand, M. Kos, and F. Gannon. 2003. Estrogen receptor-alpha directs ordered, cyclical, and combinatorial recruitment of cofactors on a natural target promoter. Cell 115:751-763.[CrossRef][Medline]
25. Murayama, A., M. S. Kim, J. Yanagisawa, K. Takeyama, and S. Kato. 2004. Transrepression by a liganded nuclear receptor via a bHLH activator through co-regulator switching. EMBO J. 23:1598-1608.[CrossRef][Medline]
26. Nagpal, S., S. Na, and R. Rathnachalam. 2005. Noncalcemic actions of vitamin D receptor ligands. Endocrinol. Rev. 26:662-687.
27. Nagy, L., H. Y. Kao, D. Chakravarti, R. J. Lin, C. A. Hassig, D. E. Ayer, S. L. Schreiber, and R. M. Evans. 1997. Nuclear receptor repression mediated by a complex containing SMRT, mSin3A, and histone deacetylase. Cell 89:373-380.[CrossRef][Medline]
28. Nagy, L., H. Y. Kao, J. D. Love, C. Li, E. Banayo, J. T. Gooch, V. Krishna, K. Chatterjee, R. M. Evans, and J. W. Schwabe. 1999. Mechanism of corepressor binding and release from nuclear hormone receptors. Genes Dev. 13:3209-3216.
29. Nolte, R. T., G. B. Wisely, S. Westin, J. E. Cobb, M. H. Lambert, R. Kurokawa, M. G. Rosenfeld, T. M. Willson, C. K. Glass, and M. V. Milburn. 1998. Ligand binding and co-activator assembly of the peroxisome proliferator-activated receptor-gamma. Nature 395:137-143.[CrossRef][Medline]
30. Palmer, H. G., J. M. Gonzalez-Sancho, J. Espada, M. T. Berciano, I. Puig, J. Baulida, M. Quintanilla, A. Cano, A. G. de Herreros, M. Lafarga, and A. Munoz. 2001. Vitamin D(3) promotes the differentiation of colon carcinoma cells by the induction of E-cadherin and the inhibition of beta-catenin signaling. J. Cell Biol. 154:369-387.
31. Perissi, V., L. M. Staszewski, E. M. McInerney, R. Kurokawa, A. Krones, D. W. Rose, M. H. Lambert, M. V. Milburn, C. K. Glass, and M. G. Rosenfeld. 1999. Molecular determinants of nuclear receptor-corepressor interaction. Genes Dev. 13:3198-3208.
32. Peterson, T. J., S. Karmakar, M. C. Pace, T. Gao, and C. L. Smith. 2007. The silencing mediator of retinoic acid and thyroid hormone receptor (SMRT) corepressor is required for full estrogen receptor alpha transcriptional activity. Mol. Cell. Biol. 27:5933-5948.
33. Pissios, P., I. Tzameli, P. Kushner, and D. D. Moore. 2000. Dynamic stabilization of nuclear receptor ligand binding domains by hormone or corepressor binding. Mol. Cell 6:245-253.[CrossRef][Medline]
34. Sanchez-Martinez, R., A. I. Castillo, A. Steinmeyer, and A. Aranda. 2006. The retinoid X receptor ligand restores defective signalling by the vitamin D receptor. EMBO Rep. 7:1030-1034.[CrossRef][Medline]
35. Schulman, I. G., H. Juguilon, and R. M. Evans. 1996. Activation and repression by nuclear hormone receptors: hormone modulates an equilibrium between active and repressive states. Mol. Cell. Biol. 16:3807-3813.[Abstract]
36. Smith, C. L., and B. W. O'Malley. 2004. Coregulator function: a key to understanding tissue specificity of selective receptor modulators. Endocrinol. Rev. 25:45-71.
37. Steel, J. H., R. White, and M. G. Parker. 2005. Role of the RIP140 corepressor in ovulation and adipose biology. J. Endocrinol. 185:1-9.
38. Tagami, T., W. H. Lutz, R. Kumar, and J. L. Jameson. 1998. The interaction of the vitamin D receptor with nuclear receptor corepressors and coactivators. Biochem. Biophys. Res. Commun. 253:358-363.[CrossRef][Medline]
39. Wen, Y. D., V. Perissi, L. M. Staszewski, W. M. Yang, A. Krones, C. K. Glass, M. G. Rosenfeld, and E. Seto. 2000. The histone deacetylase-3 complex contains nuclear receptor corepressors. Proc. Natl. Acad. Sci. USA 97:7202-7207.
40. Westin, S., R. Kurokawa, R. T. Nolte, G. B. Wisely, E. M. McInerney, D. W. Rose, M. V. Milburn, M. G. Rosenfeld, and C. K. Glass. 1998. Interactions controlling the assembly of nuclear-receptor heterodimers and co-activators. Nature 395:199-202.[CrossRef][Medline]
41. Willy, P. J., and D. J. Mangelsdorf. 1997. Unique requirements for retinoid-dependent transcriptional activation by the orphan receptor LXR. Genes Dev. 11:289-298.
42. Wong, C. W., and M. L. Privalsky. 1998. Transcriptional silencing is defin
, ZK161422; 
, VD3; 
, ZK168289; 
, ZK168492; 
, ZK191732; 
, ZK159222; 
, ZK136607. Error bars indicate standard deviations.
