The authors have no conflict of interest.
1,25-Dihydroxyvitamin D3 Stimulates Cyclic Vitamin D Receptor/Retinoid X Receptor DNA-Binding, Co-activator Recruitment, and Histone Acetylation in Intact Osteoblasts†
Article first published online: 16 NOV 2004
Copyright © 2005 ASBMR
Journal of Bone and Mineral Research
Volume 20, Issue 2, pages 305–317, February 2005
How to Cite
Kim, S., Shevde, N. K. and Pike, J. W. (2005), 1,25-Dihydroxyvitamin D3 Stimulates Cyclic Vitamin D Receptor/Retinoid X Receptor DNA-Binding, Co-activator Recruitment, and Histone Acetylation in Intact Osteoblasts. J Bone Miner Res, 20: 305–317. doi: 10.1359/JBMR.041112
- Issue published online: 4 DEC 2009
- Article first published online: 16 NOV 2004
- Manuscript Accepted: 14 SEP 2004
- Manuscript Revised: 5 AUG 2004
- Manuscript Received: 14 JUN 2004
- vitamin D receptor;
- retinoid X receptor;
- vitamin D antagonist;
- chromatin immunoprecipitation
1,25(OH)2D3 induces gene expression through the VDR. We used chromatin immunoprecipitation techniques to explore this 1,25(OH)2D3-induced process on the 25-hydroxyvitamin D3-24-hydroxylase (Cyp24) and Opn gene promoters in intact osteoblasts. Our studies show that 1,25(OH)2D3-induced transactivation is a dynamic process that involves promoter-specific localization of VDR and RXR, recruitment of histone acetyltransferase complexes, and in the case of the Cyp24 gene, modification of histone 4.
Introduction: The vitamin D receptor (VDR) binds as a retinoid X receptor (RXR) heterodimer to target DNA sequences and facilitates the recruitment of protein complexes that are essential for transcriptional modulation. These complexes include an acetyltransferase component that contains members of the p160 family and p300/CBP as well as human mediator that contains D receptor interacting protein (DRIP205). The objective of this study was to investigate the kinetics of VDR/RXR binding to 25-hydroxyvitamin D3-24-hydroxylase (Cyp24) and osteopontin (Opn) target gene promoters and to explore the recruitment and subsequent activities of co-activator complexes on these target genes in intact cells.
Materials and Methods: Mouse osteoblastic MC3T3-E1 cells and mouse primary calvarial osteoblasts (MOBs) were cultured in αMEM medium supplemented with 10% FBS. Confluent cells were treated with 1,25-dihydroxyvitamin D3 ‘1,25(OH)2D3’ or the vitamin D antagonist ZK159222, and the ability of these compounds to induce localization of VDR and RXR to specific regions of Cyp24 and Opn target genes was examined using chromatin immunoprecipitation techniques. The ability of both compounds to induce the recruitment of co-activator proteins such as p160 family members, CBP and DRIP205, and to increase the level of histone acetylation on the two gene promoters in MC3T3-E1 cells was also examined.
Results: 1,25(OH)2D3 induces rapid association of the VDR and RXR with both the Cyp24 and the Opn gene promoters in both MC3T3-E1 osteoblasts and MOBs, interactions that are both rapid and cyclic in nature. 1,25(OH)2D3 treatment also induces rapid recruitment of co-regulators such as SRC-1, -2, and -3, CBP, and p300 to both promoters, recruitment that leads to acetylation of histone 4 on Cyp24 but not the Opn. DRIP205 is also recruited to the two promoters in response to hormonal stimulation, an appearance that correlates directly with entry of RNA pol II. Studies with the vitamin D antagonist ZK159222 suggest a complex mode of action of this compound in blocking 1,25(OH)2D3-induced transcription. Our studies indicate that 1,25(OH)2D3-induced transactivation in intact osteoblasts is a dynamic process that involves promoter-specific localization of VDR and RXR as well as the recruitment of a number of co-regulators essential to 1,25(OH)2D3-induced transcription.
Conclusions: We conclude that co-regulators essential for the transcriptional activity of the steroid receptor gene family are indeed critical for the actions of 1,25(OH)2D3. Selective use of co-regulators by target genes, however, may provide a mechanism for the unique and perhaps gene-selective responses observed with synthetic analogs such as ZK159222.
STEROID RECEPTORS ARE members of a large multigene family that function to regulate transcription.(1) Receptors are activated by hormonal ligands such as estrogens, androgen, and glucocorticoids that are produced in specific endocrine organs and by metabolic ligands such as prostaglandins and certain fatty acid derivatives that are produced locally in a wide variety of tissues and cells.(2) Transcriptional regulation occurs largely through direct interaction with specific DNA sequences located within functional proximity of target gene promoters, although additional mechanisms that do not require direct receptor DNA binding also have been identified.(3) The primary role of nuclear receptors is to facilitate the recruitment of large co-regulatory protein complexes that retain selective enzymatic capabilities necessary for transcriptional modification.(4) These actions are central to numerous developmental processes, cell and tissue growth and differentiation, and adult homeostasis.
The recruitment of co-modulatory complexes to individual promoters is instrumental to nuclear receptor action. Several co-activator complexes have been identified, including the chromatin remodeling complex SWI/SNF,(5) the p300/CBP co-activator complex together with receptor interacting proteins such as SRC-1, −2. and/or −3,(6) and human mediator complex TRAP/DRIP/ARC that contains the D receptor interacting protein DRIP205.(7) Each complex seems to contribute a unique enzymatic function to transcriptional activation including ATP-dependent chromatin remodeling, histone acetylation, and RNA polymerase II recruitment.(5–7) Complexes that function to repress transcription are also numerous and are comprised, in part, of members of the histone deacetyltransferases (HDAC) family that promote chromatin condensation.(8) Nuclear receptor linkage to this class of proteins occurs through silencing mediator for retinoid and thyroid hormones (SMRT),(9) nuclear receptor co-repressor (NCoR),(10) and perhaps others.(11).
Co-modulatory factors were discovered in part through yeast two-hybrid techniques using nuclear receptors as baits. Whereas the biological relevance of many of these factors in nuclear receptor action has now been determined through gene deletion experiments in mice,(12) it has been only recently that compelling evidence has emerged in intact cells for the recruitment of these protein complexes to activated gene promoters.(13–15) Using chromatin immunoprecipitation (ChIP) analysis, Shang et al.(13) observed that 17β-estradiol prompts a rapid localization of the estrogen receptor (ER) onto estrogen-responsive gene targets. This binding is cyclic and is accompanied by a unique temporal recruitment of individual co-activators such as SRC-1, p300, and CBP. Additional studies indicate that the general antagonist ICI182780 and the tissue-specific antagonist tamoxifen are unusually active in the recruitment of SMRT and NCoR and that the cell-specific actions of these unique antagonists are dependent on the relative cellular expression of co-activators and co-repressors.(16) Similar studies have also been carried out for the androgen receptor.(14, 17) More recently, Gannon and colleagues(18, 19) have reported that both unliganded and liganded ER cycle onto target genes and that this cycling correlates temporally with the timely recruitment of many co-regulators and basic elements of the transcriptional machinery. Taken together, these studies offer significant new insights into nuclear receptor action.
Vitamin D receptors (VDRs) mediate the transcriptional actions of 1,25-dihydroxyvitamin D3 [1,25(OH)2D3] in the intestine, kidney, and bone, organs whose integrated activities function to regulate calcium and phosphorus homeostasis.(20) 1,25(OH)2D3 also exerts pleiotropic effects on tissues and cells other than the above to regulate cell growth, differentiation, and function.(21) Numerous 1,25(OH)2D3 target genes have been identified over the past decade, including the osteoblastic genes osteocalcin, osteopontin (Opn), calbindin D9K, 25-hydroxyvitamin D3-24-hydroxylase (Cyp24), and Rankl, a factor essential to osteoclast formation.(22, 23) Indeed, vitamin D response elements (VDREs) that interact with the VDR have been identified in many of these promoters.(24, 25) Not surprisingly, the VDR also interacts with many of the co-regulators described above. The consequence of these interactions in intact cells is unknown, however.
In this report, we explore ligand-induced interaction of the VDR with the promoters for Cyp24 and Opn in osteoblasts. We show that this interaction is ligand-dependent, involves RXR, and like other nuclear receptors, is cyclic in nature. 1,25(OH)2D3 also stimulates the recruitment to these genes of key co-regulators as well as specific histone modifications. A vitamin D antagonist, on the other hand, seems incapable of stimulating such actions. These data support a highly dynamic interaction between the VDR and DNA before and during transcriptional activation.
MATERIALS AND METHODS
1,25(OH)2D3 was obtained from Solvay (da Weesp, The Netherlands). ZK159222 was kindly provided by Drs Andreas Steinmeyer and Ekkehard May of Schering AG (Berlin, Germany). αMEM and DMEM were purchased from Life Technologies (Grand Island, NY, USA). Oligonucleotide primers were obtained from IDT (Coralville, IA, USA). Anti-VDR (H-81), RXR (ΔN197), SRC-1 (M-341), SRC-2, SRC-3, p300, CBP, and DRIP205 antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-tetra-acetyl lysine H3 (no. 06–599) and H4 (no. 06–866) antibodies were obtained from Upstate (Charlottesville, VA, USA). Anti-VDR-9A7 has been described previously.(26) Anti-RNA polymerase II (RNA pol II) antibodies (8WG16) were obtained from COVANCE, Berkeley Antibody Company (Richmond, CA, USA). Lipofectamine Plus was obtained from Invitrogen (Carlsbad, CA, USA). α-Amanitin was purchased from Sigma Chemical (St Louis, MO, USA).
MC3T3-E1 cells were cultured in αMEM medium supplemented with 10% FBS from BioWhittaker (Walkerville, MD, USA). Primary calvarial osteoblasts from neonatal wildtype and VDR-null mice were also obtained as previously described(27) and cultured in αMEM medium supplemented with 10% FBS from BioWhittaker. All cells were plated at 80% confluence 24–72 h before ligand treatment. 1,25(OH)2D3 was added in ethanol (0.1% maximum final concentration).
Plasmids and transfection analysis
MC3T3-E1 cells were seeded into 24-well plates at a density of 5 × 104 cells/well and transfected 16–24 h later using Lipofectamine Plus.(28) The plasmids pGAl4(5x)-luc, phVDR-VP16, and pM-SRC1-NR Gal4-DBD have been previously described.(28) Individual wells received 400 ng of DNA comprised of 250 ng pGAl4(5x)-luc, 50 ng phVDR-VP16, 50 ng pM-SRC1-NR Gal4-DBD, and 50 ng pCH110-β-gal. After transfection, the cells were normally cultured in medium supplemented with 10% FBS for 24 h with vehicle, 1,25(OH)2D3, ZK159222, or a combination of 1,25(OH)2D3 and ZK159222. Cells were harvested 24 h after stimulation, and lysates were assayed for luciferase and β-galactosidase (β-gal) activities using standard methods. Luciferase units were normalized to β-gal activity.
RNA isolation and analysis
MC3T3-E1 cells were plated in 100-mm dishes in αMEM supplemented with 10% FBS at densities of 5 × 105 cells/ml and treated for up to 24 h without or with the indicated concentration of either vehicle, 1,25(OH)2D3, ZK159222, or the combination. Total RNA was isolated using the TRI reagent (MRC, Cincinnati, OH, USA). PCR analysis was carried out using total RNA that was reverse transcribed using the SuperScript First Strand Synthesis System for RT-PCR from Invitrogen (Carlsbad, CA, USA) and standard methods (see also ChIP analysis).
Western blot analysis
MC3T3-E1 cells were treated with either vehicle or 1,25(OH)2D3 for the indicated periods. Cells were washed twice with PBS and dissolved directly on the plate in 10 mM Tris-HCl, pH 7.6, 0.3 M KCl, and 1% NP-40. After centrifugation for 10 minutes, lysates were evaluated for protein content, and 60 μg was subjected to SDS-PAGE. Proteins were transferred to Immunoblot polyvinylidene difluoride membranes (BioRad, Hercules, CA, USA) and subjected to Western blot analysis using the anti-VDR monoclonal antibody 9A7 and the anti-RXR pan antibody delta N197 (Santa Cruz Biotechnology).
Chromatin immunoprecipitation was performed as described previously.(13, 28) Briefly, MC3T3-E1 cells or MOBs from wildtype or VDR-null mice were plated in αMEM supplemented with 10% FBS 72 h before the experiment and treated with or without 1,25(OH)2D3 for the times and conditions indicated. Treated cells were washed several times with PBS and subjected to a cross-linking reaction with 1% formaldehyde. Cells were extracted sequentially in 5 mM Pipes, pH 8.0, 85 mM KCl, and 0.5% NP-40, and then in 1% SDS, 10 mM EDTA, and 50 mM Tris-HCl, pH 8.1, and the chromatin pellets were sonicated to an average DNA size of 300–500 bp DNA (assessed by agarose gel electrophoresis) using a Fisher Model 100 Sonic Dismembranator at a power setting of 1. The sonicated extract was centrifuged and diluted into ChIP buffer (16.7 mM Tris-HCl, pH 8.1, 150 mM NaCl, 0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA). Immunoprecipitations were performed overnight at 4°C with the indicated antibodies and collected by centrifugation after a 1-h incubation with salmon sperm DNA- and BSA-pretreated Zysorbin (Zymed, San Francisco, CA, USA). Precipitates were washed, and cross-links were reversed during an overnight incubation in 1% SDS and 0.1 M NaHCO3 at 65°C. DNA fragments were purified using Qiagen QIAquick PCR Purification Kits (Valencia, CA, USA) and subjected to PCR using primers designed to amplify fragments of murine Cyp24 promoter (−234 to −51) or coding (+11055 to +11252) regions and murine Opn promoter (−854 to −658) or coding (+5181 to +5380) regions. Primers were also designed to amplify segments of the Cyp24 promoter from −234 to −528, −529 to −733, −734 to −933, and −934 to −1133. Analyses for each primer set were carried out at a predetermined linear range of DNA amplification. PCR products were resolved on 2% agarose gels and visualized using ethidium bromide staining. DNA acquired before precipitation was used to assess the presence of genes following the ChIP procedure and designated “Input.” PCR evaluation was performed using 10% of input DNA. Densitometric quantitation was carried out using Kodak 1D Image Analysis (software version 3.5).
1,25(OH)2D3 induces Cyp24 and Opn mRNA in mouse MC3T3-E1 osteoblasts
We explored two genes that are particularly relevant to vitamin D action in bone, the genes that encode Cyp24, a P450-containing enzyme essential to vitamin D metabolism,(29) and Opn, a cellular matrix protein involved in the recruitment and activation of bone resorbing osteoclasts.(30) A time course of induction of both of these genes by 1,25(OH)2D3 in the osteoblast-like cell line MC3T3-E1 is documented in Fig. 1. Cells were treated for periods ranging from 0 to 24 h with a previously established maximal dose of 1,25(OH)2D3 (10−8 M) and harvested for RNA isolation and RT-PCR analysis. As can be seen, basal levels of expression of Cyp24 and Opn mRNA differ substantially in the absence of ligand (Fig. 1), whereas Cyp24 mRNA was undetectable, Opn levels were significantly elevated. Despite this difference, the levels of both of these mRNAs increased significantly within 3 h in response to 1,25(OH)2D3 and continued to accumulate until the end of the time course at 24 h. The finding that Opn is significantly expressed in the absence of 1,25(OH)2D3, whereas Cyp24 is fully dependent on hormone for expression, makes these genes ideal subjects for studying the differential regulatory properties of 1,25(OH)2D3. Importantly, the action of 1,25(OH)2D3 on both Cyp24 and Opn expression is known to be transcriptional in nature and mediated in part by natural VDREs located at specific sites within each promoter.(31, 32)
Localization of VDR to the Cyp24 and Opn promoters using ChIP
Having confirmed the activation of Cyp24 and Opn gene expression by 1,25(OH)2D3 in MC3T3-E1 cells, we next examined whether the hormone could promote VDR binding to the proximal promoter regions of these genes using ChIP. MC3T3-E1 cells were seeded into 100-mm plates and cultured in steroid-free αMEM supplemented with 10% FBS. The medium was replaced with fresh serum-free or FBS-supplemented αMEM containing either ethanol vehicle (0.1%) or 1,25(OH)2D3 (10−7 M) and cultured for 12 h. Cells were harvested, and after fixation with 1% formaldehyde, lysed, sonicated to prepare chromatin fragments of 300–500 bp average length, and subjected to immunoprecipitation with antibodies to the VDR. The immunoprecipitated DNA was isolated and analyzed by PCR using oligonucleotide primers capable of detecting either gene promoter or coding region DNA. A schematic of the Cyp24 and Opn promoters and the location of the amplification primers are seen in Fig. 2A. In the absence of the ligand, small amounts of both Cyp24 and Opn promoter DNA but not coding region DNA are evident after immunoprecipitation with anti-VDR antibodies (Fig. 2B). Treatment of the cells with 1,25(OH)2D3, however, leads to a tremendous increase in the level of detection of both gene promoter DNA segments. Neither promoter nor coding segments were evident when immunoprecipitations were carried out in the absence of antibody (Fig. 2B) or when an irrelevant antibody was used (data not shown) despite the fact that input DNA was equivalent in all the samples. These findings indicate that 1,25(OH)2D3 strongly promotes the formation in intact cells of Cyp24-VDR and Opn-VDR gene complexes. The extent to which the VDR is bound in the hormone-free condition is difficult to assess. However, pretreatment with α-amanitin, a reagent known to dissociate residual ER from DNA targets,(18) had no effect on the level of promoter DNA immunoprecipitated by anti-VDR antibody (data not shown), suggesting that the presence of this DNA was not caused by selectively bound and cross-linked VDRs.
RXR localization on the Cyp24 and Opn promoters is induced by 1,25(OH)2D3
Chromatin derived from 1,25(OH)2D3-treated MC3T3-E1 cells was also subjected to immunoprecipitation with anti-RXR pan antibodies, and the precipitates were analyzed for the presence of both Cyp24 and Opn promoter DNA. The results reveal that 1,25(OH)2D3 also promotes an antibody-dependent, promoter-specific association of RXR with each of the vitamin D target genes (Fig. 2B). Preliminary experiments indicated that RXRα is the principle RXR isotype (MB Meyers and JW Pike, unpublished data, 2004). Small amounts of both Cyp24 and Opn promoter DNA were similarly evident in the precipitates derived from untreated cells. Again, however, these levels were unaffected by pretreatment with α-amanitin (data not shown). These results provide further support for RXR's role in the transcriptional activation of these two genes by 1,25(OH)2D3.
VDR and RXR co-localize on both the Cyp24 and Opn gene promoters
We also assessed whether both VDR and RXR co-localized on the same DNA fragment using a double immunoprecipitation assay. Sonicated chromatin from 1,25(OH)2D3-treated MC3T3-E1 cells was subjected to immunoprecipitation with anti-RXR or anti-VDR antibodies, and the resulting products were immunoprecipitated a second time with either IgG or antibodies to the crossover receptor partner. Both Cyp24 and Opn promoter DNA can be detected after a sequential precipitation first by VDR and then RXR antibodies as well as by RXR and then VDR antibodies (Fig. 2C). These data provide additional support for the likelihood that VDR and RXR are bound to the same DNA fragments. Although the lack of resolution of the assay prevents a definitive conclusion regarding whether VDR and RXR form heterodimers on these two promoter DNA fragments,(33) the hormone-dependent appearance of RXR on this DNA suggests a direct interaction between the two proteins.
Localization of VDR and RXR to the Cyp24 and Opn promoters in neonatal mouse osteoblastic cells derived from wildtype and VDR-null mice
1,25(OH)2D3 also induces the expression of both Cyp24 and Opn genes in neonatal MOBs.(27, 28) To confirm that this induction was similarly initiated through VDR and RXR DNA binding and to confirm the nature of the immunoprecipitated VDR and RXR signals, we treated MOBs derived from wildtype or VDR-null neonatal mice for 6 h with 1,25(OH)2D3, cross-linked and lysed the cells, and subjected the sonicated chromatin to ChIP using antibodies to both VDR and RXR. As can be seen in Fig. 3, 1,25(OH)2D3 induces the localization of both VDR and RXR to the Cyp24 and Opn genes in wildtype but not VDR-null MOB cells. These experiments confirm the ability of 1,25(OH)2D3 to promote VDR and RXR binding to both the Cyp24 and Opn genes in primary osteoblasts. They also indicate that initial RXR binding to these target genes is fully dependent on the presence of a functional VDR.
VDR and RXR binding to the Cyp24 and Opn promoters is cyclic
The positive interaction of VDR and RXR with the Cyp24 and Opn genes prompted us to examine the kinetics of receptor association. We assessed binding in MC3T3-E1 cells in 15-minute intervals over a period of 165 minutes after treatment with 1,25(OH)2D3 and also assessed longer-term exposures up to 24 h. VDR was detected on both the Cyp24 and Opn promoters at 30 minutes and peaked at ∼45-60 minutes (Figs. 4A-4C). Surprisingly, peak binding was followed by a decline in activity that reached a nadir at 90 minutes, followed by a second wave that peaked at 135 minutes. RXR binding to the Opn promoter correlated temporally with that of VDR over the time period assessed, cycling with a pattern similar to that seen for the VDR (Figs. 4A, 4B and 4D). Interestingly, however, RXR binding to the Cyp24 promoter seemed to differ substantially (Fig. 4D). In this case, RXR binding increased steadily at the earliest time-points together with the VDR, but stabilized after 60 minutes and did not seem to cycle. Although detailed time-points beyond 165 minutes were not examined, the binding of VDR and RXR to both the Cyp24 and the Opn promoters seemed to plateau after about 3 h of treatment and remained largely unchanged for periods up to 24 h (Fig. 4E and data not shown). This suggests that VDR binding may eventually becomes noncyclic on these promoters. 1,25(OH)2D3 treatment also led to a substantial increase in VDR and a modest increase in RXR protein over these time-points, as assessed by Western blot analysis (Fig. 4F). These increases in receptor protein were progressive, however, and are thus unlikely to be responsible for the cyclic receptor binding patterns that were observed after exposure to 1,25(OH)2D3.
Recruitment of the p160 co-activators
The p160 co-activators SRC-1, SRC-2 (GRIP1), and SRC-3 (pCIP) can be recruited to hormone-activated gene promoters through a direct protein-protein interaction with nuclear receptors.(4) The interaction of the VDR with many of these co-activators has been shown in vitro, although their localization to vitamin D-responsive promoters in intact cells has not been evaluated. To assess this potential recruitment, MC3T3-E1 cells treated with 1,25(OH)2D3 as above were lysed at 15-minute intervals and subjected to ChIP analysis using antibodies to SRC-1, GRIP-1, and pCIP. As can be seen, all three p160 co-activators localized to the two gene promoters in response to 1,25(OH)2D3 (Figs. 5A-5E). The patterns of recruitment for each of the p160 proteins seem to be relatively similar between the two genes, although some minor differences are evident. For example, the recruitment of SRC-2 onto the Cyp24 promoter exhibits a very rapid peak at 15 minutes, followed by two waves of activity at 60 and 120 minutes, whereas on the Opn promoter, this earliest peak at 15 minutes is absent (Fig. 5D). Some differences are also noted among each of the p160s. SRC-3 seems to be more rapidly recruited to each of the promoters, but its recruitment dissipates after 60 minutes and does not show evidence of a secondary peak during these time intervals (Fig. 5E).
Recruitment of the CBP/p300 co-activators
We also assessed the temporal recruitment of the CBP and p300 co-activators on the Cyp24 and Opn promoters. Both appear in a similar fashion in response to 1,25(OH)2D3 (Figs. 5A, 5B, 5F and 5G). The recruitment is rapid but not sustained in the first several hours, leading to a general loss of binding after 90 minutes. Interestingly, these factors, as well as SRC-2 and −3, are detected on the target genes at a time that seems to precede that of VDR and RXR. To assess whether this was a conceptual difference or simply a technical limitation because of differences in assay sensitivity between co-activator and VDR, we carried out a sequential immunoprecipitation first with antibody to p300 and then with either a control or anti-VDR antibody. The data indicate that the VDR is indeed present on the same Cyp24 and Opn promoter fragments after an initial precipitation with the p300 antibody (Fig. 5H). These results suggest that the initial recruitment of p300, as well as the other co-activators that appear early on the Cyp24 and Opn genes, is indeed mediated by the VDR.
Co-activator recruitment by VDR on the Cyp24 promoter induces histone 4 acetylation
p160/CBP complexes function to acetylate the amino terminal lysine residues of nucleosomal histones, causing chromatin remodeling events that facilitate transcriptional activation.(4) To assess whether the recruitment of these factors to the Cyp24 and Opn promoters leads to such modification, we treated MC3T3-E1 cells with 1,25(OH)2D3 for 3 h and subjected the cell lysates to ChIP analysis using anti-acetyl-lysine histone 3 (H3) and histone 4 (H4) antibodies. On the Cyp24 promoter, 1,25(OH)2D3 treatment strongly enhanced lysine acetylation on H4; H3, in contrast, was constitutively acetylated (Fig. 6A). At the Opn promoter, however, acetylation of both H3 and H4 was constitutive, and 1,25(OH)2D3 seemed unable to modify these levels. H4 acetylation in the proximal region of the Cyp24 promoter was also time-dependent (Fig. 6B). As can be seen, the modification increased as a function of time over the first 60 minutes much like that seen for the co-activators; at 3 and 6 h, the levels were even higher, suggesting a rather sustained response to 1,25(OH)2D3 (Fig. 6B). Interestingly, although coding region DNA (+11055 to +11252) was not present, DNA fragments from regions located upstream of the VDRE in the Cyp24 gene (−529 to −733, −734 to −933, and −934 to −1133) were clearly evident after immunoprecipitation with the anti-acetyl-lysine H4 antibody (Fig. 6B). The time course of modification in these regions was similar to that seen at the VDREs. These findings suggest that the recruitment of co-factors to the Cyp24 promoter leads to histone acetylation. This modification was not restricted to the proximal VDRE containing region of the gene, however, but extended upstream as well.
Recruitment of DRIP205 co-activator and RNA pol II
In a final assessment of co-activator binding, we explored the recruitment of DRIP205 and RNA pol II to the Cyp24 and Opn promoters. DRIP205 was identified previously as a nuclear receptor interacting protein, provides linkage to mammalian mediator-like complexes,(7) and is believed to both facilitate the entry of RNA pol II(34) and participate in ubiquitination.(35) DRIP205 is recruited to both the Cyp24 and the Opn promoters in a cyclic pattern that generally reflects that seen for the VDR (Figs. 7A-7C). This pattern of recruitment was also observed for RNA pol II (Figs. 7A, 7B and 7D).
ZK159222 displays gene-selective agonist and antagonist activities
Many synthetic analogs of natural hormones act as partial agonists or antagonists of transcriptional regulation. The molecular basis for these diverse and often tissue-specific activities is the ability of analogs to induce receptor conformations that are incompatible or selectively compatible with the normal transcriptional activation.(36, 37) To solidify the mechanisms observed above, we examined, in a final set of experiments, the properties of ZK159222, one of several vitamin D analogs that have been shown to antagonize the activity of 1,25(OH)2D3 [see Fig. 8A for the side-chain structure of ZK159222 compared with 1,25(OH)2D3].(38) As observed in Fig. 8B, ZK159222 does indeed antagonize 1,25(OH)2D3's ability to stimulate Cyp24 and Opn transcriptional output in MC3T3-E1 cells. Interestingly, however, whereas ZK159222 completely blocked the ability of 1,25(OH)2D3 to induce Cyp24 mRNA levels, it was only partially effective in preventing Opn upregulation, displaying a significant partial agonist activity when used alone (Fig. 8B). These observations suggest that ZK159222 is indeed an antagonist, although its actions may be gene target dependent and therefore more complex.
VDR/RXR DNA binding in response to ZK159222 is gene promoter-dependent
The above actions are likely due either to an inability of the compound to promote VDR/RXR DNA binding or to an inability to recruit essential co-activators. To test the first possibility, we treated MC3T3-E1 cells with vehicle, 1,25(OH)2D3, ZK159222, or both ligands and performed ChIP analysis on the harvested cells using antibodies to VDR or RXR. Interestingly, whereas 1,25(OH)2D3 was fully active in stimulating VDR or RXR binding to the Cyp24 promoter, ZK159222 was only weakly active on its own and blocked the activity of 1,25(OH)2D3 when added in combination (Fig. 9A). ZK159222s ability to induce VDR and RXR DNA binding to the Opn promoter, however, was virtually equivalent to that of 1,25(OH)2D3 (Fig. 9A). These surprising findings suggest that the individual characteristics of the two promoters influence the degree of VDR/RXR DNA binding after ZK159222 activation. Despite this, the data do not fully explain either the partial agonist or the antagonist activities of this vitamin D analog.
ZK159222-liganded VDR exhibits a limited capacity to recruit p160 and p300 co-activators and to induce H4 acetylation
In a final set of experiments, we determined whether ZK159222 was capable of promoting the recruitment of co-activators to the Cyp24 or Opn promoters and inducing histone acetylation. ZK159222 is much less effective than 1,25(OH)2D3 in stimulating recruitment of various members of the p160 family or p300 to the Cyp24 and Opn promoters (Fig. 9B). This weak interaction was confirmed in transfected MC3T3-E1 cells using a mammalian two hybrid system. Accordingly, ZK159222 also failed to promote an interaction between chimeric VDR-VP16 and SRC-1-Gal4-DBD in MC3T3-E1 cells and almost completely blocked the interaction induced by 1,25(OH)2D3 (Fig. 9C). As might be predicted, ZK159222 was also much less effective in stimulating H4 acetylation in the VDRE region of the Cyp24 promoter (Fig. 9D). The constitutive acetylation of H3 and H4 observed on the Opn promoter was unaffected (Fig. 9D). These findings suggest that, whereas the capacity of ZK159222 to promote VDR DNA binding varied depending on the gene target, the ability of the VDR/RXR heterodimer to recruit co-activators and to modify chromatin structure was significantly compromised.
In this study, we explored the molecular events associated with the transcriptional activation of the Cyp24 and Opn genes by 1,25(OH)2D3. We show using ChIP analysis that 1,25(OH)2D3 induces rapid localization of both VDR and RXR to the VDRE-containing promoter regions of these two genes, an association that temporally precedes that of increased Cyp24 and Opn mRNA production. The binding of VDR and RXR is cyclic in its initial phases and then approaches a steady state plateau. 1,25(OH)2D3 also induces the recruitment of several transcriptional co-regulators including those of the p160/CBP families and DRIP205. The recruitment of the chromatin-modifying acetyltransferases is associated with increased H4 acetylation on the Cyp24 but not the Opn promoter. The recruitment of DRIP205 is also cyclic and correlates directly with the entry of RNA pol II. The activities of a vitamin D antagonist ZK159222 confirm the importance of these events in 1,25(OH)2D3-induced gene activation while showing unusual properties that highlight the existence of unique differences between the two promoters. These studies reveal that transcriptional activation of Cyp24 and Opn by 1,25(OH)2D3 involves both VDR and RXR as well as several sets of transcriptional co-activators. Together with earlier studies, they suggest that gene activation by 1,25(OH)2D3 is a complex and highly dynamic process.
The results obtained by ChIP indicate that treatment of osteoblasts with 1,25(OH)2D3 induces rapid association of both VDR and RXR with VDRE-containing regions of the Opn and Cyp24 gene promoter. These findings are important because they support the idea that localization of VDR and RXR to target genes is a traditional ligand-dependent process.(39) They also reveal that the appearance of RXR on the two gene promoters is dependent on a functional VDR. Small amounts of Cyp24 and Opn promoter DNA are co-precipitated by the anti-VDR and anti-RXR antibodies in the absence of hormone, raising the question as to whether significant amounts of the two receptors are bound to Cyp24 and Opn in a ligand-independent manner as is typical of TR(9, 10) and more recently ER.(13, 18, 19) Several lines of evidence suggest this is not the case. First, α-amanitin, a reagent that causes disengagement of unliganded nuclear receptors from target DNA,(18) does not alter the level of Cyp24 or Opn promoter DNA immunoprecipitated from vehicle-treated cells. Second, vehicle-treated calvarial osteoblasts derived from both VDR-null and wildtype mice also exhibit low but detectable levels of “immunoprecipitated” Cyp24 and Opn DNA (Fig. 3 and data not shown). These immunoprecipitated DNA's increase after treatment with 1,25(OH)2D3, but only in osteoblasts derived from wildtype mice. This ligand-stimulated VDR binding on the Cyp24 promoter confirms the initial findings of Zhang et al.(40) in human breast cancer cells.
Localization of the VDR on both the Opn and Cyp24 promoters in response to 1,25(OH)2D3 is cyclic. The cycling of nuclear receptors onto target promoters is now well documented, and includes ER,(12, 18) AR,(14, 17) and more recently, VDR.(40) This phenomenon also seems typical of other transcription factors as well(41, 42) suggesting that the process may be fundamental to the mechanism of transcriptional activation. We did not examine the rate of new mRNA synthesis for the two genes at these early time-points. However, the observation that RNA pol II was also recruited to these promoters in a cyclic manner supports the idea of a cyclic production of new transcripts. Over time, VDR and RXR binding to the Cyp24 and Opn genes seem to reach a steady-state level. Whether this signifies a fading of the cycling process or is simply a loss of initial hormone-induced “cell synchrony” remains to be determined.
1,25(OH)2D3 treatment is also accompanied by an increase in VDR protein levels that peak ∼6 h later. These changes in receptor protein, which were noted earlier,(43, 44) may contribute to the mechanism of 1,25(OH)2D3 regulation, but they are clearly not responsible for the cycling process. The finding that target gene DNA becomes saturated with VDR within 3 h (Fig. 4C), whereas VDR levels continue to rise (Fig. 4D), suggests the presence of distinct pools of VDR. It is also interesting that, although RXR binding to the Opn promoter reflects that of VDR, this correlation breaks down on the Cyp24 promoter. The mechanism of this “dissociation” is unknown. The Cyp24 promoter, however, contains two VDREs(29) and an Ets-1 binding site that lies immediately adjacent to the proximal VDRE.(45) Ets-1 has been shown to influence Cyp24 transcription and response to 1,25(OH)2D3(45) and can be detected on this site using ChIP assays (S Kim and JW Pike, unpublished data, 2004). Further studies will be necessary to determine the impact of the Ets-1 factor on RXR and VDR binding at the Cyp24 promoter.
Why do receptors cycle? Cycling could be essential to the production of new mRNA, while at the same time, providing an exquisite level of gene regulation by enabling frequent monitoring of cellular levels of activating hormone. It should be noted, however, that receptor and co-factor cycling is currently based on ChIP analysis, highlighted most elegantly in recent studies by Gannon and colleagues.(18, 19) More recently, Nagaich et al.,(46) using laser-mediated cross-linking, also revealed the presence of a cycling pattern for the glucocorticoids receptor on chromatin. Features of this cycling process differ from that observed by Gannon and colleagues,(18, 19) however, not only in the cycling periodicity but also with regard to the proposed mechanism. A fundamental difference between the two studies is the rate at which cross-linking occurs, the laser-mediated method being almost instantaneous. Regardless, both findings are likely to be different from the rapid receptor and co-factors association/dissociation kinetics observed in living cells seen using fluorescence recovery after photobleaching (FRAP).(47–49) Further studies will be necessary to determine the relationships between these temporal phenomena as well as their relevance to the mechanism of nuclear receptor action.
The induction by 1,25(OH)2D3 of VDR/RXR binding to the Opn and Cyp24 promoters is accompanied by the recruitment of a number of nuclear receptor co-activators. These include the p160 co-activators SRC-1, SCR-2, and SRC-3 and both CBP and p300.(4) Why are all of these co-activators recruited to the two promoters? Perhaps each has a unique role in histone modification during the complex process of chromatin remodeling.(50) Interestingly, only small differences in the temporal pattern of recruitment are evident among the co-activators in response to 1,25(OH)2D3. Perhaps most notable is the relatively early appearance of p300 and CBP. This finding is consistent with the idea that the earliest event in hormonal activation of gene expression is the remodeling of chromatin.(51) Surprisingly, although substantial differences in the basal expression levels of Cyp24 and Opn were observed, the recruitment pattern for all the co-activators onto these two genes was generally similar. A likely consequence of co-activator recruitment was the appearance of histone modification on H4 in the Cyp24 promoter. This modification extended significantly upstream of the Cyp24 VDRE. At the Opn promoter, however, both H4 and H3 were fully acetylated and remained unaffected by 1,25(OH)2D3 treatment. Perhaps chromatin remodeling is less critical for hormonal induction of Opn gene expression. Regardless, the finding that co-activators can be identified in association with VDR/RXR complexes on vitamin D target genes provides important confirmation of their importance in 1,25(OH)2D3-induced gene regulation.
The appearance of DRIP205 seems to follow that of the p160s/CBP. This is consistent with a proposal that DRIP205 functions secondarily to chromatin remodeling, directly facilitating RNA pol II entry.(52) Indeed, the cyclic recruitment of RNA pol II tracks directly with that of DRIP205. The temporal resolution of our assays is not particularly robust, however, leading to considerable uncertainty with regard to order. An additional overarching question is the nature and mechanism whereby different protein complexes are exchanged at the level of the receptor. The recent discovery of a complex that facilitates co-repressor/co-activator exchange at specific nuclear receptor bound loci(53) suggests that a similar mechanism may be at play to mediate co-regulatory exchange as well.
In the experiments outlined here, we evaluated the activity of ZK159222, a vitamin D analog that binds the VDR with affinity ∼1/10th that of 1,25(OH)2D3.(38, 54) Previous studies in vitro indicate that ZK159222 is an antagonist that is able to promote DNA binding of VDR/RXR but prevents co-activator recruitment.(38) In our studies in osteoblasts, ZK159222 proved to be an antagonist of Cyp24 but a partial agonist of Opn. Interestingly, although ZK159222 was relatively ineffective in promoting DNA binding of VDR and RXR on the Cyp24 promoter, it was fully effective in promoting such binding on the Opn promoter. It is unclear how this occurs, but seems likely to be because of unique features of the target gene itself. In contrast, ZK159222 was uniformly unable to promote co-activator recruitment regardless of the gene target. At the Cyp24 promoter, this combination of reduced ability to promote DNA binding and inability to recruit co-activators at the Cyp24 promoter resulted in limited histone modification at H4 and antagonism. Surprisingly, the overall activities of ZK159222 did not result in antagonism at the Opn promoter, but rather a partial agonistic activity. It is possible that the basal activity of the Opn promoter alters the gene's dependency on chromatin remodeling during induction by 1,25(OH)2D3, making this gene less sensitive to the unusual receptor conformation induced by ZK159222.
In conclusion, we have shown that 1,25(OH)2D3-induced transactivation of the Opn and Cyp24 genes in intact cells involves both VDR and RXR, as well as a surprising number of co-activator complexes. The association of the receptor with DNA and the recruitment of co-activators was a highly dynamic process. It will be interesting to explore in the future features of this activation process with novel 1,25(OH)2D3 analogs that exhibit unique features of altered efficacy and tissue selectivity.
The authors thank Anjali Warrier for technical assistance and members of the Pike and Shevde laboratory for helpful discussions during this study. This work was supported by NIH Grant DK-52453 (JWP).
- 11988 The steroid and thyroid hormone receptor superfamily. Science 240: 889–895.
- 21994 Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu Rev Biochem 63: 451–486.,
- 32000 Estrogen receptor pathways to AP-1. J Steroid Biochem Mol Biol 74: 311–317., , , , , ,
- 41999 Nuclear receptor co-activators: Multiple enzymes, multiple complexes, multiple functions. J Steroid Biochem Mol Biol 69: 3–12., , , , ,
- 51999 ATP-dependent remodeling and acetylation in regulators of chromatin fluidity. Genes Dev 13: 2339–2352.,
- 62002 Molecular mechanisms and cellular biology of the steroid receptor co-activator (SRC) family in steroid receptor function. Rev Endocr Metab Disord 3: 185–192.,
- 72001 Mediator complexes and transcription. Curr Opin Cell Biol 13: 274–280.,
- 82000 Co-repressor complexes and remodeling chromatin for repression. Biochem Soc Trans 28: 379–386., ,
- 92003 The N-CoR/histone deacetylase 3 complex is required for repression by thyroid hormone receptor. Mol Cell Biol 23: 5122–5131.,
- 101995 Ligand-independent repression by the thyroid hormone receptor mediated by a nuclear receptor co-repressor. Nature 377: 397–404., , , , , , , , , ,
- 112003 Physical and functional interaction between the vitamin D receptor and hairless corepressor, two proteins required for hair cycling. J Biol Chem 278: 38665–38674., , , , , ,
- 122002 Acute disruption of select steroid receptor co-activators prevents reproductive behavior in rats and unmasks genetic adaptation in knockout mice. Mol Endocrinol 16: 1511–1523., , , ,
- 132000 Co-factor dynamics and sufficiency in estrogen receptor-regulated transcription. Cell 103: 843–852., , , ,
- 142002 Formation of the androgen receptor transcription complex. Mol Cell 9: 601–610., ,
- 152002 A co-regulatory role for the TRAP-mediator complex in androgen receptor-mediated gene expression. J Biol Chem 277: 42852–42858., , ,
- 162002 Molecular determinants for the tissue specificity of SERMs. Science 295: 2465–2468.,
- 172002 Involvement of proteasome in the dynamic assembly of the androgen receptor transcription complex. J Biol Chem 277: 48366–48371., , ,
- 182003 Cyclic, proteasome-mediated turnover of unliganded and liganded ERalpha on responsive promoters is an integral feature of estrogen signaling. Mol Cell 11: 695–707., , , , , , , ,
- 192003 Estrogen receptor-alpha directs ordered, cyclical, and combinatorial recruitment of co-factors on a natural target promoter. Cell 115: 751–763., , , , , ,
- 202003 Vitamin D: More than a “bone-a-fide” hormone. Mol Endocrinol 17: 77–91.,
- 211998 Current understanding of the molecular actions of vitamin D. Physiol Rev 78: 1193–1231., ,
- 221992 Vitamin D: New actions, new analogs, new therapeutic potential. Endocr Rev 13: 765–788.
- 232003 Osteoclast differentiation and activation. Nature 423: 337–342., ,
- 241998 The nuclear vitamin D receptor: Biological and molecular regulatory properties revealed. J Bone Miner Res 13: 325–349., , , , , , ,
- 251997 Analysis of osteocalcin expression in transgenic mice reveals a species difference in vitamin D regulation of mouse and human osteocalcin genes. J Bone Miner Res 12: 1570–1576., , , , , ,
- 261996 Transcriptional activation and dimerization functions in the human vitamin D receptor. Mol Endocrinol 10: 945–957., , ,
- 272002 A potent analog of 1α,25-dihydroxyvitamin D3 selectively induces bone formation. Proc Natl Acad Sci USA 99: 13487–13491., , , , ,
- 282003 2-methylene-19-NOR-(20S)-1,25-dihydroxyyvitamin D3 potently stimulates gene-specific DNA binding of the vitamin D receptor in osteoblasts. J Biol Chem 278: 31756–31765., , , , ,
- 292000 Regulation of the porcine 1,25-dihydroxyvitamin D3-24-hydroxylase (CYP24) by 1,25-dihydroxyvitamin D3 and parathyroid hormone in AOK-B50 cells. Arch Biochem Biophys 381: 323–327., , , ,
- 301998 Osteopontin expression and function: Role in bone remodeling. J Cell Biochem Suppl 30–31: 92–102.,
- 311990 Identification of a DNA sequence responsible for binding of the 1,25-dihydroxyvitamin D3 receptor and 1,25-dihydroxyvitamin D3 enhancement of mouse secreted phosphoprotein 1 (SPP-1 or osteopontin) gene expression. Proc Natl Acad Sci USA 87: 9995–9999., , , , ,
- 321996 Functional assessment of two vitamin D-responsive elements in the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J Biol Chem 271: 30381–30385., , , , , ,
- 331991 Vitamin D receptor interaction with specific DNA. Association as a 1,25-dihydroxyvitamin D3-modulated heterodimer. J Biol Chem 266: 23296–23305., ,
- 341999 Multimeric co-activator complexes for steroid/nuclear receptors. Trends Endocrinol Metab 10: 403–407.
- 352003 Identification of mammalian mediator subunits with similarities to yeast mediator subunits Srb5, Srb6, Med11, and Rox3. J Biol Chem 278: 15123–15127., , , , , , , , , , ,
- 362002 Definition of the molecular and cellular mechanisms underlying the tissue-selective agonist/antagonist activities of selective estrogen receptor modulators. Recent Prog Horm Res 57: 295–316., , , ,
- 371998 The structural basis of estrogen receptor/co-activator recognition and the antagonism of this interaction by tamoxifen. Cell 95: 927–937., , , , , ,
- 382001 Different molecular mechanisms of vitamin D3 receptor antagonists. Mol Pharmacol 59: 1478–1485., , , , ,
- 391997 The vitamin D receptor and its gene. In: FeldmanD, GlorieuxF, PikeJW (eds.) Vitamin D. Academic Press, San Diego, CA, USA, pp. 105–125.
- 402003 NCoA62/SKIP is a nuclear matrix-associated co-activator that may couple vitamin D receptor-mediated transcription and RNA splicing. J Biol Chem 278: 35325–35336., , , , , , ,
- 412001 Kinetics of p53 binding to promoter sites in vivo. Mol Cell Biol 21: 3375–3386., ,
- 422001 Two waves of nuclear factor kappaB recruitment to target promoters. J Exp Med 193: 1351–1359., ,
- 431987 Molecular cloning of complementary DNA encoding the avian receptor for vitamin D. Science 235: 1214–1217., , , ,
- 441993 Positive regulation of the vitamin D receptor by its cognate ligand in heterologous expression systems. Mol Endocrinol 7: 833–839., , , ,
- 452000 Regulation of rat cytochrome P450C24 (CYP24) gene expression. Evidence for functional cooperation of Ras-activated Ets transcription factors with the vitamin D receptor in 1,25-dihydroxyvitamin D3-mediated induction. J Biol Chem 275: 47–55., , , ,
- 462004 Rapid periodic binding and displacement of the glucocorticoid receptor during chromatin remodeling. Mol Cell 14: 163–174., , ,
- 472004 Dynamics of nuclear receptor movement and transcription. Biochim Biophys Acta 1677: 46–51., , , ,
- 482002 Dynamic behavior of transcription factors on a natural promoter in living cells. EMBO Rep 3: 1188–1194., , , , , ,
- 492001 Ligand-mediated assembly and real-time cellular dynamics of estrogen receptor alpha-co-activator complexes in living cells. Mol Cell Biol 21: 4404–4412., , , , , , , ,
- 502002 A co-activator code for transcription. Trends Biochem Sci 27: 165–167.,
- 512002 A unified theory of gene expression. Cell 108: 439–451.,
- 521999 Nuclear receptor co-factors as chromatin remodelers. Curr Opin Genet Dev 9: 499–504.,
- 532004 A corepressor/co-activator exchange complex required for transcriptional activation by nuclear receptors and other regulated transcription factors. Cell 116: 511–526., , , ,
- 542000 Carboxylic ester antagonists of 1alpha,25-dihydroxyvitamin D3 show cell-specific actions. Chem Biol 7: 885–894., ,