The Vitamin D Receptor, Runx2, and the Notch Signaling Pathway Cooperate in the Transcriptional Regulation of Osteopontin*

Osteopontin (OPN), a glycosylated phosphoprotein that binds calcium, is present in bone extracellular matrix and has been reported to modulate both mineralization and bone resorption. Targeted disruption in mice of the vitamin D receptor (VDR) or Runx2 results in marked inhibition of OPN expression in osteoblasts. In this study, we addressed possible cross-talk between VDR and Runx2 in regulating OPN transcription. 1,25-Dihydroxyvitamin D3 (1,25(OH)2D3) or Runx2 stimulated OPN transcription (mouse OPN promoter -777/+79) 2–3-fold. However, coexpression of Runx2 and VDR in COS-7 cells and treatment with 1,25(OH)2D3 resulted in an 8-fold induction of OPN transcription, indicating for the first time functional cooperation between Runx2 and VDR in the regulation of OPN transcription. In ROS 17/2.8 and MC3T3-E1 cells that contain endogenous Runx2, AML-1/ETO, which acts as a repressor of Runx2, significantly inhibited 1,25(OH)2D3 induction of OPN transcription, OPN mRNA, and protein expression. Both a Runx2 site (-136/-130) and the vitamin D response element (-757/-743) in the OPN promoter are needed for cooperative activation. Chromatin immunoprecipitation analyses showed that 1,25(OH)2D3 can enhance VDR and Runx2 recruitment on the OPN promoter, further indicating cooperation between these two factors in the regulation of OPN. In osteoblastic cells, Hes-1, a downstream factor of the Notch signaling pathway, was found to enhance basal and 1,25(OH)2D3-induced OPN transcription. This enhancement was inhibited by AML-1/ETO, an inhibitor of Runx2. Immunoprecipitation assays indicated that Hes-1 and Runx2 interact and that 1,25(OH)2D3 can enhance this interaction. Taken together, these findings define novel mechanisms involving the intersection of three pathways, Runx2, 1,25(OH)2D3, and Notch signaling, that play a major role in the regulation of OPN in osteoblastic cells and therefore in the process of bone remodeling.


Osteopontin (OPN), a glycosylated phosphoprotein that binds calcium, is present in bone extracellular matrix and has been reported to modulate both mineralization and bone resorption. Targeted disruption in mice of the vitamin D receptor (VDR) or Runx2 results in marked inhibition of OPN expression in osteoblasts. In this study, we addressed possible cross-talk between VDR and Runx2 in regulating OPN transcription. 1,25-Dihydroxyvitamin D 3 (1,25(OH) 2 D 3 ) or Runx2 stimulated OPN transcription (mouse OPN promoter ؊777/؉79) 2-3-fold. However, coexpression of Runx2 and VDR in COS-7 cells and treatment with 1,25(OH) 2 D 3 resulted in an 8-fold induction of OPN transcription, indicating for the first time functional cooperation between Runx2 and VDR in the regulation of OPN transcription. In ROS 17/2.8 and MC3T3-E1 cells that contain endogenous Runx2, AML-1/ETO, which acts as a repressor of Runx2, significantly inhibited 1,25(OH) 2 D 3 induction of OPN transcription, OPN mRNA, and protein expression. Both a Runx2 site (؊136/؊130
) and the vitamin D response element (؊757/؊743) in the OPN promoter are needed for cooperative activation. Chromatin immunoprecipitation analyses showed that 1,25(OH) 2 D 3 can enhance VDR and Runx2 recruitment on the OPN promoter, further indicating cooperation between these two factors in the regulation of OPN. In osteoblastic cells, Hes-1, a downstream factor of the Notch signaling pathway, was found to enhance basal and 1,25(OH) 2 D 3 -induced OPN transcription. This enhancement was inhibited by AML-1/ETO, an inhibitor of Runx2. Immunoprecipitation assays indicated that Hes-1 and Runx2 interact and that 1,25(OH) 2 D 3 can enhance this interaction. Taken together, these findings define novel mechanisms involving the intersection of three pathways, Runx2, 1,25(OH) 2 D 3 , and Notch signaling, that play a major role in the regulation of OPN in osteoblastic cells and therefore in the process of bone remodeling.
Osteopontin (OPN) 2 is a sialic acid-rich glycosylated phosphoprotein, comprising about 2% of the noncollagenous protein in bone (1,2). OPN is produced by osteoblasts when they form bone matrix (1,2). OPN is an extracellular matrix protein that contains arginine-glycine-aspartate (RGD) integrin binding motifs and promotes attachment of bone cells to the bone surface through binding to OPN receptors such as the ␣ v ␤ 3 integrin and CD44 (1)(2)(3). OPN has been suggested to be involved in the attachment of osteoclasts during bone resorption, to play a role in osteogenesis by attachment of osteoblasts when they form bone matrix, and to act to regulate crystal size during bone mineralization (2). In addition, OPN has been suggested to be a mediator of bone remodeling in response to mechanical strain (4). OPN null mice are resistant to mineral loss and bone resorption upon estrogen deprivation and have impaired activation of osteoclasts (3,(5)(6)(7). Also, vascularization and resorption of bone discs have been reported to be significantly impaired in the absence of OPN (8). Although recent studies using OPN null mice have provided new insight into the role of OPN in vivo in bone metabolism, the factors that affect the regulation of OPN are not yet clearly defined. 1,25-Dihydroxyvitamin D 3 (1,25(OH) 2 D 3 ), the active form of vitamin D, is a major calcitropic hormone involved in calcium homeostasis (9). One of its functions in bone is to regulate the synthesis of the bone calcium-binding proteins osteocalcin (OC) and OPN (9). 1,25(OH) 2 D 3 modulates the expression of these genes through transcriptional regulation. The actions of 1,25(OH) 2 D 3 are mediated through the vitamin D receptor (VDR). Liganded VDR heterodimerizes with the retinoid X receptor and interacts with a vitamin D response element (VDRE). The VDRE in the mouse OPN promoter (at Ϫ757/ Ϫ743) is a perfect direct repeat of the motif GGTTCA spaced by three nucleotides (10). Transcription proceeds through the interaction of VDR with coactivators and coregulators, including SRC-1/NcoA1, SRC-2/GRIP-1 (GR-interacting protein)/NcoA2, SRC-3/ACTR, and the multisubunit DRIP (vitamin D receptor-interacting protein) complex (11). Although a VDRE has been identified in the mouse OPN promoter (10) and VDR null mice show marked inhibition of OPN expression in osteoblasts (12), the exact mechanisms, including proteinprotein and protein-DNA interactions, involved in 1,25(OH) 2 D 3 -regulated OPN transcription are not well understood.
Runx2/Cbfa1 is a member of the runt/Cbfa family of transcription factors that was first identified as an osteoblast-specific transcription factor and a regulator of osteoblast differentiation (13,14). Runx2 Ϫ/Ϫ mice die shortly after birth and show a complete lack of mineralized bone tissue (13,14). Marked decreases in the expression of osteopontin and osteocalcin are observed in Runx2 Ϫ/Ϫ mice, indicating the regulation of these genes by Runx2 (13). Three Runx2 binding motifs have been identified in the rat OC promoter (15). In addition, Runx2 has been shown to play a key role in the 1,25(OH) 2 D 3 regulation of rat OC (15,16). However, it is not yet known whether a similar cooperation occurs between VDR and Runx2 in the regulation of OPN.
Hes-1 (Hairy and enhancer of split homologue-1), a downstream target of the Notch signaling pathway, is a helix-loop-helix transcription factor that has been reported to play a role in developmental processes, including myogenesis and neurogenesis (17). The expression of the Hes-1 gene is widely detected in embryos as well as adults (17). Hes-1 is also expressed in osteoblastic cells (18). Hes-1 is coexpressed with Runx2 in osteoblastic cells, and Runx2 and Hes-1 physically interact (19,20). In addition, studies in Drosophila indicate that runt and hairy contribute to common transcriptional regulatory events (21,22). Due to the relationship between Hes-1 and Runx2 and the suggested role of Runx2 in OPN regulation (13,14,19,20,23), we tested the possibility that Hes-1 may cooperate with Runx2 in the regulation of OPN. Our findings define, for the first time, novel mechanisms involving the intersection of Runx2, 1,25(OH) 2 D 3 and Notch signaling that are involved in the regulation of the OPN gene.
Transient Transfection and Dual Luciferase Assay-The mouse osteopontin promoter (Ϫ777/ϩ79) firefly luciferase reporter construct was kindly provided by D. Denhardt (Rutgers University, Piscataway, NJ). pCMV-Runx2 was a gift of G. Karsenty (Baylor College of Medicine, Houston, TX), and pCMV-AML-1/ETO expression vector was from S. W. Hiebert (Vanderbilt University School of Medicine, Nashville, TN). pcDNA3-Hes1 expression vector was a gift from Dr. S. Stifani (McGill University, Montreal, Canada). Cells were seeded in a 24-well culture dish 24 h prior to transfection at 70% confluence. Cells in each well were transfected using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's instructions. Empty vectors were used to keep the total DNA concentration the same. Efficiency of transfection, as assessed by green fluorescent protein cotransfection and subsequent visualization, was estimated at 60 -70%. 1,25(OH) 2 D 3 (10 Ϫ8 M) or TSA (15 nM) was added to cells 24 h post-transfection for another 24 h. Cells were washed twice with phosphate-buffered saline (PBS) and harvested by incubating with 1ϫ passive lysis buffer, supplied by the Dual-Luciferase reporter assay kit (Promega). The luciferase activity assay was performed according to the protocol of the manufacturer and normalized to values for pRL-TK-Renilla luciferase. For all transcription studies, OPN promoter activity (firefly/Renilla luciferase) is represented as -fold induction by comparison with basal levels (basal levels refer to levels of OPN promoter activity in cells transfected with vector alone and treated with vehicle).
Site-directed Mutagenesis-Mutant mouse OPN promoter (Ϫ777/ ϩ79) luciferase reporter constructs were generated by site directed mutagenesis using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). The oligonucleotides used to generate the Runx2 mutated site (shown in lowercase) were as follows: 5Ј-CCT TTT TTT TTT TTT AAg aAC AAA ACC AGA GGA GG-3Ј (top strand) and 5Ј-CCT CCT CTG GTT TTG Ttc TTA AAA AAA AAA AAA GG-3Ј (lower strand). The oligonucleotides used to generate the VDRE mutated site (shown in lowercase) were as follows: 5Ј-CAG AGC AAC AAG Gcc CAC GAG GTT CAC GTC-3Ј (top strand) and 5Ј-GAC GTG AAC CTC GTG ggC CTT GTT GCT CTG-3Ј (bottom strand).
Northern Blot Analysis-ROS17/2.8 cells or MC3T3-E1 cells, plated at 70% confluence in 100-mm tissue culture dishes, were transfected using Lipofectamine 2000 reagent, with AML-1/ETO or Hes-1 expression vector or vector alone. 24 h after transfection, cells were treated for 24 h with 1,25(OH) 2 D 3 (10 Ϫ8 M) or vehicle control. The treated cells were then harvested by trypsinization, pelleted, and washed with PBS. Total RNA was isolated by RNA-bee RNA extraction solution (Tel-Test, Friendswood, TX) and precipitated by chloroform and isopropyl alcohol. 20 g of total RNA from each sample was used for Northern blot analysis as previously described (24). 32 P-Labeled cDNA was prepared using the Random Primers DNA labeling system (Invitrogen) according to the random primer method (25). The mouse osteopontin cDNA was generated by HindIII digestion and was a gift from D. Denhardt (Rutgers University, Piscataway, NJ). The ␤-actin cDNA was purchased from Clontech. The blots were hybridized with the 32 P-labeled mouse OPN cDNA probes for 16 h at 42°C, washed, air-dried and exposed to Eastman Kodak Co. BIOMAX MR film at Ϫ80°C for 1 day. The same blots were stripped and probed with 32 P-labeled ␤-actin cDNA. Autoradiograms were analyzed by densitometric scanning using the Dual-Wavelength Flying Spot Scanner. The relative optical density obtained using the OPN probe was divided by the relative optical density obtained after probing with ␤-actin to normalize for sample variation.
OPN Western Blot Analysis-MC3T3-E1 cells, plated at 70% confluence in 100-mm tissue culture dishes, were transfected with vector alone or pCMV-AML1/ETO and treated with vehicle or 1,25(OH) 2 D 3 (10 Ϫ8 M) for 24 h and harvested by trypsinization. For Western blot analysis, 50 g of protein from total cell lysates was loaded onto a 10% SDS-polyacrylamide gel and separated by electrophoresis. Protein was transferred onto a polyvinylidene difluoride membrane (Bio-Rad). Membranes were incubated overnight at 4°C with mouse OPN polyclonal antibody (P-18; Santa Cruz Biotechnology) at a 1:1000 dilution in PBS containing 5% nonfat milk. The membrane was washed with PBS and incubated for 1 h with the corresponding secondary antibody conjugated with horseradish peroxidase. The enhanced chemiluminescent Western blotting detection system (PerkinElmer Life Sciences) was used to detect the antigen-antibody complex.
Chromatin Immunoprecipitation (ChIP) Assay-MC3T3-E1 cells were cultured in ␣-minimal essential medium supplemented with 10% FBS to 95% confluence prior to the experiment and then treated in ␣-minimal essential medium supplemented with 2% charcoal-stripped serum under the conditions and for the times indicated. Treated cells were used for the ChIP assay (26,27). Briefly, cells were first washed with PBS and subjected to a cross-link reaction with 1% formaldehyde for 15 min. The cross-link reaction was stopped by adding glycine to a final concentration of 0.125 M. Cells were washed with ice-cold PBS twice. The cells were collected by scraping and lysed sequentially in 5 mM Pipes, pH 8.0, 85 mM KCl, 0.5% Nonidet P-40 and then in 1% SDS, 10 mM EDTA, 50 mM Tris-HCl, pH 8.1, for 20 min individually. The chromatin pellets were sonicated to an average DNA size of 500 bp DNA (assessed by 1% agarose gel electrophoresis) using a Fisher model 100 sonic dismembranator at a power setting of 1. The sonicated extract was centri-

Transcriptional Regulation of Osteopontin
In re-ChIP experiments, complexes were eluted by incubation for 30 min at 37°C in 60 l of elution buffer containing 10 mM dithiothreitol. The eluted samples were diluted 50 times with ChIP dilution buffer and subjected again to the ChIP procedure with specific antibodies.
Nuclear Extracts-Cells were washed with cold PBS twice, harvested by scraping, pelleted by centrifuging at 4,000 rpm for 4 min. The pellets were washed and lysed in hypotonic buffer containing 10 mM HEPES (pH 7.4), 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM dithiothreitol, phosphatase inhibitors (0.5 mM phenylmethylsulfonyl fluoride, 1 mg/ml pepstatin A, 2 mg/ml leupeptin, 2 mg/ml aprotinin), and 1% Triton X-100. Nuclei were pelleted at 4,000 rpm for 4 min, and cytoplasmic supernatants were separated. Nuclei were resuspended in hypertonic buffer containing 0.42 mM NaCl, 0.2 mM EDTA, 25% glycerol, and the phosphatase and protease inhibitors indicated above. After a 2-h incubation at 4°C, nuclear soluble proteins were collected by centrifuging at 13,000 rpm for 10 min. Protein concentration of the supernatant was measured by the method of Bradford (28), and aliquots were stored at Ϫ80°C.
Immunoprecipitation-To examine the association of Runx2 and Hes-1 in the presence or absence of 1,25(OH) 2 D 3 coimmunoprecipitation experiments were done. Nuclear extracts were prepared as indicated above from ROS17/2.8 cells or MC3T3-E1 cells, and protein concentration was detected by the Bradford method (28). 500 g of each preparation was used for immunoprecipitation with the addition of 4 g of Hes-1 antiserum or 4 g of Runx2 antiserum in the presence or absence of 1,25(OH) 2 D 3 (10 Ϫ8 M) for 24 h at 4°C. Then 30 l of protein A-Sepharose 4 Fast Flow Beads (Amersham Biosciences) were added to each sample, and, after further incubation by rotating at 4°C for 1 h, the immunoprecipitated complex was collected by centrifuging at 3,000 rpm for 5 min. The complex was separated by 12% SDS-PAGE and probed with Runx2 antibody or Hes-1 antibody. Immunoprecipitation experiments were also done as described above using COS-7 cells transfected with VDR and treated with 1,25(OH) 2 D 3 (10 Ϫ8 M for 24 h) and cotransfected with vector alone (pCMV) or 2 g of pCMV-Runx2 to examine the association of Hes-1 with histone deacetylase-1 in the presence or absence of Runx2. For these studies, 500 g of nuclear extract was used for immunoprecipitation with the addition of 4 g of histone deacetylase-1 antiserum followed by the addition of protein A-Sepharose 4 Fast Flow Beads incubated and collected by centrifugation as described above. The complex was separated by 12% SDS-PAGE and probed with Hes-1 antibody. Statistical Analysis-Results are expressed as the mean Ϯ S.E., and significance was determined by analysis with Student's t test for twogroup comparison or analysis of variance for multiple group comparison.
Northern blot analysis was also performed to assess the effect of AML-1/ETO on endogenous 1,25(OH) 2 D 3 -induced OPN mRNA expression. Expression of AML-1/ETO in ROS17/2.8 osteoblastic cells resulted in a significant inhibition of the levels of basal and 1,25(OH) 2 D 3 -induced OPN mRNA (Fig. 3A). Note (Fig. 3A, last two bars) that although there is a 50% decrease in basal OPN mRNA, there is a 75% decrease in 1,25(OH) 2 D 3 -induced OPN mRNA. Similar results were observed using MC3T3-E1 cells (Fig. 3B). In addition, inhibition of 1,25(OH) 2 D 3 OPN protein expression was also observed in the presence of AML-1/ETO (Fig. 3C). These findings suggest that VDR and Runx2 cooperate in vivo to regulate the expression of OPN.
Both the VDRE and the Runx2 Site Are Needed for Cooperative Activation of OPN Transcription-A Runx2 site was noted in the mouse osteopontin promoter (AACCACA at Ϫ136/Ϫ130) (23). Gel shift assays using synthetic oligonucleotides corresponding to the wild type (WT) (Ϫ136/Ϫ130) or mutated (AAgaACA) Runx2 binding sequences and nuclear extracts from Runx2-transfected COS-7 cells indicated that Runx2 interacted with the WT oligonucleotides in a dose-dependent manner (not shown). No protein-DNA interaction was detected using the mutant oligonucleotide, and preincubation with cold WT oligonucleotide but not mutant oligonucleotide resulted in a dose-dependent depletion of the binding of Runx2 to the labeled probe (not shown). These electrophoretic mobility shift assays indicated, similar to previous studies (23), that Ϫ136/Ϫ130 in the mouse OPN promoter is a binding site for Runx2. To investigate the specific contribution of the VDRE and the Runx2 site to the cooperative activation of OPN transcription, mutant OPN promoter constructs were generated with either the Runx2 site (Ϫ136/Ϫ130) mutated or the VDRE (Ϫ757/Ϫ743) mutated (Fig. 4A). Mutation of the Runx2 site did not affect the induction by 1,25(OH) 2 D 3 of OPN transcription in VDR-transfected COS-7 cells (Fig. 4B, vector-transfected, vehicle-and 1,25(OH) 2 D 3 -treated) and resulted in a decreased (but not abolished) 1,25(OH) 2 D 3 response in ROS 17/2.8 cells (not shown). However, in COS-7 cells Runx2 could no longer activate OPN transcription (Fig. 4B, Runx2- 3 versus lane 1)), indicating that Runx2 acts through this site in the mouse OPN promoter Ϫ777/ϩ79 and not through additional sites (unlike the regulation of OC by Runx2) (15). Also, mutation of the Runx2 site resulted in a loss of the cooperative response (Fig. 4B, lane 4; compare with Fig. 1, lane 4). The decrease in the response to 1,25(OH) 2 D 3 in the presence of Runx2 using the OPN promoter with the mutated Runx2 site may be due to the reported binding of Runx2 to VDR (16). Runx2, in the presence of a mutated Runx2 site in the OPN promoter, may bind to VDR, and thus less VDR would be available for 1,25(OH) 2 D 3 induced transcription. Using the OPN promoter construct bearing a mutation in the VDRE, 1,25(OH) 2 D 3 was unable to activate the OPN promoter in VDR-transfected COS-7 cells (Fig. 4C, vector-transfected (Vec), 1,25(OH) 2 D 3 -treated) or in ROS 17/2.8 cells (not shown). However, transfection of COS-7 cells with Runx2 could still result in enhanced OPN transcription, and, similar to the mutation of the Runx2 site, the cooperative response was not observed (Fig. 4C). These findings suggest that the Runx2 site at Ϫ136/Ϫ130 and the VDRE are essential for cooperative effects of Runx2 and VDR in activating mouse OPN transcription.
Runx2 and VDR Interact with the OPN Promoter in Intact Osteoblastic Cells-In order to further understand mechanisms involved in activation of OPN transcription, we examined VDR and Runx2 complex formation on the OPN promoter in MC3T3-E1 cells using the ChIP assay and specific antibodies against Runx2 and VDR. The antibodies were used to precipitate sonicated chromatin cross-links from whole cell lysates after formaldehyde cross-linking of DNA to transcription factors. DNA was amplified using specific primers directed against the VDRE or the Runx2 binding region of the OPN promoter. In the PCR procedure, the number of cycles was chosen so that the amplification was conducted in the linear range of amplification efficiency. No signal was detected in the presence of IgG (Fig. 5A). The ChIP analysis showed that 1,25(OH) 2 D 3 can enhance both VDR and Runx2 recruitment to the OPN promoter (Fig. 5A). Note that transfection of MC3T3-E1 cells with AML-1/ETO resulted in decreased recruitment of Runx2 to the OPN promoter (Fig. 5B). The 1,25(OH) 2 D 3 enhancement of Runx2 as well as VDR DNA binding affinity could be one possible mechanism involved in the cooperative activation.
Hes-1 Can Potentiate the Runx2-mediated Transactivation of OPN Transcription-Hes-1, a downstream target of the Notch signaling pathway, is coexpressed with Runx2 in osteoblastic cells, and Hes-1 and Runx2 have been reported to contribute to common transcriptional regulatory events (19,20). We therefore tested the possibility that Hes-1 may be involved in 1,25(OH) 2 D 3 -and Runx2-mediated regulation of OPN transcription. In ROS17/2.8 cells and MC3T3-E1 cells, that contain endogenous Runx2, transfection of Hes-1 (0.1-1 g) resulted in an enhancement of both basal and 1,25(OH) 2 D 3 -induced OPN transcription (Fig. 6, A and B). Expression of Hes-1 also resulted in an enhancement of basal and 1,25(OH) 2 D 3 -induced OPN mRNA expression (Fig.  6C). The enhancement of the induction of OPN transcription by Hes-1 in ROS17/2.8 cells was inhibited by AML-1/ETO, a repressor of Runx2 (Fig. 7). In COS-7 cells, in the absence of transfected Runx2, expression of Hes-1 resulted in a repression of 1,25(OH) 2 D 3 -dependent induction of OPN transcription, and co-transfection of Runx2 in COS-7 cells reversed the inhibition by Hes-1 (not shown), further suggesting functional cooperation between Hes-1 and Runx2.
Since both Runx2 and Hes-1 can interact with TLE (transducin-like enhancer of split) proteins, which can recruit histone deacetylases (20,30), we asked whether inhibition of histone deacetylation may be involved in the activation by Hes-1. In ROS17/2.8 cells, TSA, a histone deacetylase inhibitor, was able to rescue the inhibition by AML-1/ETO of Hes-1-enhanced 1,25(OH) 2 D 3 -induced OPN transcription (Fig. 9A). In addition, in COS-7 cells, in the absence of transfected Runx2, inhibition of 1,25(OH) 2 D 3 -induced OPN transcription by Hes-1 was reversed in the presence of TSA (not shown). These findings suggest that Hes-1/ Runx2 binding may interfere with Runx2-TLE and Hes-1-TLE interactions, thus preventing repression, which may be mediated, at least in part, by histone deacetylation. Coimmunoprecipitation studies showed the association of Hes-1 and histone deacetylase-1 and a decrease in this association in the presence of Runx2 (Fig. 9B). Taken together, these findings show that Hes-1 can potentiate VDR-mediated OPN transcription in the presence of Runx2 and define new mechanisms and functional interactions that are involved in the regulation of OPN and may therefore affect the process of bone remodeling.

DISCUSSION
This study describes for the first time cooperative effects between Runx2, VDR, and Hes-1 in the transcriptional regulation of OPN. Functional cooperation was demonstrated between Runx2 and VDR in the regulation of OPN transcription, OPN mRNA, and protein expression. 1,25(OH) 2 D 3 was found to enhance both VDR and Runx2 recruitment on the OPN promoter in vivo, further indicating cooperation between these two factors in the regulation of OPN. Hes-1, a downstream target of the Notch signaling pathway, was found to act as an enhancer of basal and 1,25(OH) 2 D 3 -induced OPN transcription and OPN mRNA in the presence of Runx2. Coimmunoprecipitation analysis indicated that Hes-1 and Runx2 interact, and 1,25(OH) 2 D 3 enhances this interaction. We propose that these three major pathways, Runx2, 1,25(OH) 2 D 3 , and Notch signaling, intersect and play a major role in the regulation of OPN in osteoblastic cells and therefore in the process of bone remodeling.
Runx2 was found not only to up-regulate OPN basal promoter activity but also to enhance 1,25(OH) 2 D 3 -induced OPN transcription. Runx2 has been reported to be essential for osteogenic differentiation (31,32). 1,25(OH) 2 D 3 promotes osteoblastic differentiation and directly stimulates the production of OC and OPN (33,34). OPN has been reported to be present in preosteoblasts and is present in high concentrations in the osteoblast (35,36). Bone sialoprotein (BSP), another calcium-binding protein present in bone matrix that shares structural features with OPN, is expressed after OPN but earlier than OC in the development of the osteoblast phenotype (35,36). OC is the latest of the differentiation markers to be expressed. OC is abundantly expressed in mature osteoblasts (35,36). These calcium-binding proteins may function in regulating the ordered deposition of mineral (2,37). Although much work has been done concerning the regulation of OC, we are only beginning to understand mechanisms involved in the regulation of OPN and BSP. Two Runx2 sites had previously been suggested in the OPN promoter (at Ϫ136/Ϫ130) (23) and on the reverse strand at Ϫ695/Ϫ690 (14). Mutation of the Runx2 site at Ϫ136/Ϫ130 resulted in a complete block of the activation of OPN transcription by Runx2 (Fig. 4A), indicating that Runx2 can act through this single site in the OPN promoter. This is unlike the regulation of rat OC. The rat OC promoter contains two distal Runx2 sites (A and B) and a proximal Runx2 site (C). All three sites are required for maximal OC promoter activity. Mutation of the proximal site C has the least effect on basal OC promoter activity (15,16). Three Runx sites have also been noted in the BSP promoter (38). The Runx sites in the BSP promoter mediate repression of BSP (38). 1,25(OH) 2 D 3 also represses BSP expression (39). It has been suggested that the context of the multiple Runx2 motifs within a promoter may contribute to the formation of Runx2 regulatory complexes and secondary interactions that mediate either repression or activation (38). However, previous studies have also indicated, similar to our study, that multiple Runx2 sites are not always required for regulation by Runx2. For example, Drissi et al. (40) reported that, although more than one Runx2 site is present in the Runx2 promoter, a single site within the proximal promoter is sufficient to confer negative autoregulation. It is of interest that although the function of all three calcium-binding proteins, BSP, OC, and OPN, is associated with ordered deposition of mineral, OC and OPN are induced by 1,25(OH) 2 D 3 and are positively regulated by Runx2 and BSP is inhibited by 1,25(OH) 2 D 3 , and Runx mediates its  DECEMBER 9, 2005 • VOLUME 280 • NUMBER 49 repression. Thus, differential regulation by Runx2 and 1,25(OH) 2 D 3 may be needed to regulate the timing of the expression of these proteins and to accommodate their functional role at various stages of osteoblast differentiation.

Transcriptional Regulation of Osteopontin
Although Runx2 enhanced VDR-mediated OPN transcription, the Runx2 site was not required for 1,25(OH) 2 D 3 induction of OPN transcription. Mutation of the Runx2 site in the OPN promoter did not affect the 1,25(OH) 2 D 3 response in COS-7 cells (Fig. 4B) and resulted in a decreased (but not abolished) 1,25(OH) 2 D 3 response in ROS 17/2.8 cells, indicating the involvement of tissue-specific factors and a cooperative effect of VDR and Runx2 in bone cells. However, in osteoblastic cells, mutation of the Runx2 sites in the rat OC promoter blocks 1,25(OH) 2 D 3 OC transcription (15,16). The 1,25(OH) 2 D 3 regulation of rat OC requires a functional Runx2 site B, which is adjacent to the OC VDRE (15,16). In addition, both Runx2 and AP1 binding sites are required for parathyroid hormone stimulation of collagenase 3 transcription (41,42). In the collagenase 3 promoter, there is an overlapping AP1 and Runx2 site, and Runx2 has been reported to interact with c-Fos and c-Jun (43,44). It has been suggested that parathyroid hormone-dependent collagenase 3 expression involves cooperation between Runx2 and AP1 transcription factors and the composite Runx2/TRE element as well as a distal Runx2 site (43). For the regulation of OPN, the Runx2 site is not adjacent or overlapping the VDRE (VDRE Ϫ757/Ϫ743; Runx2 site Ϫ136/Ϫ130). It is possible that the Runx2 site in the OPN promoter is not critical for 1,25(OH) 2 D 3 regulation of OPN, since the Runx2 site is not adjacent or overlapping the VDRE, and that it is the   Empty vectors were used to keep the total DNA concentration the same. OPN promoter activity was expressed as firefly/Renilla luciferase activity and is represented as -fold induction (mean Ϯ S.E.; n Ն 3 observations/group) by comparison with basal levels. B, COS-7 cells were transfected with VDR and were cotransfected with vector or Runx2 and were treated with 1,25(OH) 2 D 3 (10 Ϫ8 M, 24 h). Nuclear extracts were prepared, and 500 g of nuclear protein was used for immunoprecipitation (IP) with histone deacetylase-1 antibody. Western blot was performed with Hes-1 antibody. The top panel shows the Western blot of cell extracts prior to immunoprecipitation probed with Hes-1 antibody.