Activation of Peroxisome Proliferator-activated Receptor-γ Inhibits the Runx2-mediated Transcription of Osteocalcin in Osteoblasts

Mesenchymal cells are able to differentiate into several distinct cell types, including osteoblasts and adipocytes. The commitment to a particular lineage may be regulated by specific transcription factors. Peroxisome proliferator-activated receptor-γ (PPARγ), acting in conjunction with CCAAT/enhancer-binding protein-α, has been suggested as a key regulator of adipogenic differentiation. Previous studies have shown that the activation of PPARγ in osteoblasts suppresses osteoblast differentiation and the expression of osteocalcin, an osteoblast-specific protein. However, the mechanism of this inhibition remains unclear. We investigated the effect of PPARγ activation on the expression of osteocalcin and analyzed the molecular mechanism. Mouse osteoblastic MC3T3-E1 cells expressed PPARγ, which was transcriptionally active, whereas rat osteosarcoma ROS 17/2.8 cells did not. Treatment of MC3T3-E1 osteoblasts and ROS 17/2.8 cells stably transfected with PPARγ2 with the PPARγ activator 15-deoxy-Δ12,14-prostaglandin J2 inhibited the mRNA expression of osteocalcin and Runx2, the latter of which is a key transcription factor in osteoblast differentiation. This decreased expression of osteocalcin and Runx2 was partly explained by the decreased level of Runx2 resulting from the suppressed transcription from the Runx2 promoter. However, in addition to this indirect effect, the activation of PPARγ by 15-deoxy-Δ12,14-prostaglandin J2 directly suppressed the Runx2-mediated induction of the activities of the osteocalcin promoter and the artificial promoter p6OSE2, which contains six tandem copies of osteoblast-specific element-2, the Runx2-binding promoter sequence. This inhibition was mediated by a physical interaction between PPARγ and Runx2 and the subsequent repression of the transcriptional activity at the osteoblast-specific element-2 sequence. Thus, this study demonstrates that the activation of PPARγ inhibits osteocalcin expression both by suppressing the expression of Runx2 and by interfering with the transactivation ability of Runx2. Mesenchymal cells are able to differentiate into several distinct cell types, including osteoblasts and adipocytes. The commitment to a particular lineage may be regulated by specific transcription factors. Peroxisome proliferator-activated receptor-γ (PPARγ), acting in conjunction with CCAAT/enhancer-binding protein-α, has been suggested as a key regulator of adipogenic differentiation. Previous studies have shown that the activation of PPARγ in osteoblasts suppresses osteoblast differentiation and the expression of osteocalcin, an osteoblast-specific protein. However, the mechanism of this inhibition remains unclear. We investigated the effect of PPARγ activation on the expression of osteocalcin and analyzed the molecular mechanism. Mouse osteoblastic MC3T3-E1 cells expressed PPARγ, which was transcriptionally active, whereas rat osteosarcoma ROS 17/2.8 cells did not. Treatment of MC3T3-E1 osteoblasts and ROS 17/2.8 cells stably transfected with PPARγ2 with the PPARγ activator 15-deoxy-Δ12,14-prostaglandin J2 inhibited the mRNA expression of osteocalcin and Runx2, the latter of which is a key transcription factor in osteoblast differentiation. This decreased expression of osteocalcin and Runx2 was partly explained by the decreased level of Runx2 resulting from the suppressed transcription from the Runx2 promoter. However, in addition to this indirect effect, the activation of PPARγ by 15-deoxy-Δ12,14-prostaglandin J2 directly suppressed the Runx2-mediated induction of the activities of the osteocalcin promoter and the artificial promoter p6OSE2, which contains six tandem copies of osteoblast-specific element-2, the Runx2-binding promoter sequence. This inhibition was mediated by a physical interaction between PPARγ and Runx2 and the subsequent repression of the transcriptional activity at the osteoblast-specific element-2 sequence. Thus, this study demonstrates that the activation of PPARγ inhibits osteocalcin expression both by suppressing the expression of Runx2 and by interfering with the transactivation ability of Runx2. Mesenchymal cells are able to differentiate into several distinct cell types, including osteoblasts and adipocytes (1Grigoriadis A.E. Heersche J.N. Aubin J.E. J. Cell Biol. 1988; 106: 2139-2151Google Scholar, 2Grigoriadis A.E. Heersche J.N. Aubin J.E. Dev. Biol. 1990; 142: 313-318Google Scholar, 3Yamaguchi A. Kahn A.J. Calcif. Tissue Int. 1991; 49: 221-225Google Scholar). The mechanisms directing the cells along a particular lineage and the suppression of alternative pathways are not well established, although signals derived from the extracellular environment and several key transcription factors have been identified (4Yamaguchi A. Katagiri T. Ikeda T. Wozney J.M. Rosen V. Wang E.A. Kahn A.J. Suda T. Yoshiki S. J. Cell Biol. 1991; 113: 681-687Google Scholar, 5Tapscott S.J. Weintraub H. J. Clin. Invest. 1991; 87: 1133-1138Google Scholar, 6Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Google Scholar, 7Wu Z. Bucher N.L. Farmer S.R. Mol. Cell. Biol. 1996; 16: 4128-4136Google Scholar, 8Yeh W.C. Cao Z. Classon M. McKnight S.L. Genes Dev. 1995; 9: 168-181Google Scholar).Peroxisome proliferator-activated receptors (PPARs) 1The abbreviations used are: PPARs, peroxisome proliferator-activated receptors; OSE, osteoblast-specific element; 15-dPGJ2, 15-deoxy-Δ12,14 prostaglandin J2; mRXRα, mouse retinoid X receptor-α; PPRE, PPAR-responsive element; GST, glutathione S-transferase; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum. 1The abbreviations used are: PPARs, peroxisome proliferator-activated receptors; OSE, osteoblast-specific element; 15-dPGJ2, 15-deoxy-Δ12,14 prostaglandin J2; mRXRα, mouse retinoid X receptor-α; PPRE, PPAR-responsive element; GST, glutathione S-transferase; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum. are a family of ligand-activated transcription factors that belong to the nuclear hormone receptor superfamily (9Spiegelman B.M. Flier J.S. Cell. 1996; 87: 377-389Google Scholar, 10Kliewer S.A. Forman B.M. Blumberg B. Ong E.S. Borgmeyer U. Mangelsdorf D.J. Umesono K. Evans R.M. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 7355-7359Google Scholar). PPARγ is abundantly expressed in both white and brown adipose tissue and has been known to play a critical role in the regulation of adipocyte differentiation (10Kliewer S.A. Forman B.M. Blumberg B. Ong E.S. Borgmeyer U. Mangelsdorf D.J. Umesono K. Evans R.M. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 7355-7359Google Scholar). The transfection of fibroblastic cells with PPARγ2 and its subsequent activation with ligand have been shown to be sufficient to initiate adipogenesis (11Tontonoz P. Hu E. Spiegelman B.M. Cell. 1994; 79: 1147-1156Google Scholar). Moreover, determined myoblasts with no inherent adipogenic potential can be induced to transdifferentiate into mature adipocytes by the ectopic expression of two adipogenic transcription factors, PPARγ and CCAAT/enhancer-binding protein-α (12Hu E. Tontonoz P. Spiegelman B.M. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9856-9860Google Scholar). These results suggest that a developmental switch between these highly specialized cell types can be controlled by the expression of key adipogenic transcription factors. A few studies have suggested that PPARγ also acts as a molecular switch between the osteogenic and adipogenic pathways. Lecka-Czernik et al. (13Lecka-Czernik B. Gubrij I. Moerman E.J. Kajkenova O. Lipschitz D.A. Manolagas S.C. Jilka R.L. J. Cell. Biochem. 1999; 74: 357-371Google Scholar) showed that overexpression of PPARγ2 in stromal cell lines results in the suppression of Osf2 (osteoblast-specific factor-2)/Runx2, a key transcription factor for osteoblast differentiation (6Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Google Scholar), and osteoblast-like biosynthetic activity while promoting terminal differentiation into adipocytes. Jackson and Demer (14Jackson S.M. Demer L.L. FEBS Lett. 2000; 471: 119-124Google Scholar) also reported that the treatment of MC3T3-E1 cells with high concentrations of PPARγ ligands inhibits osteoblast maturation. These studies provide insights into the mechanisms underlying aging-related osteoporosis because a decrease in the number and differentiating potential of bone marrow precursors (15Khosla S. Atkinson E.J. Riggs B.L. Melton III, L.J. J. Bone Miner. Res. 1996; 11: 857-863Google Scholar) and an alteration in the shunting of these cells between the osteoblast and adipocyte lineages (16Meunier P. Aaron J. Edouard C. Vignon G. Clin. Orthop. Relat. Res. 1971; 80: 147-154Google Scholar) were demonstrated under this condition.One of the osteoblast-specific proteins known to be suppressed by PPARγ activation is osteocalcin (13Lecka-Czernik B. Gubrij I. Moerman E.J. Kajkenova O. Lipschitz D.A. Manolagas S.C. Jilka R.L. J. Cell. Biochem. 1999; 74: 357-371Google Scholar, 14Jackson S.M. Demer L.L. FEBS Lett. 2000; 471: 119-124Google Scholar). Osteocalcin is an ∼6-kDa γ-carboxylated protein and composes up to 15% of the noncollagenous protein of mature bone (17Price P.A. Otsuka A.A. Poser J.W. Kristaponis J. Raman N. Proc. Natl. Acad. Sci. U. S. A. 1976; 73: 1447-1451Google Scholar). The expression of osteocalcin is largely restricted to the osteoblasts of bone and the odontoblasts and cementoblasts of teeth (18McKee M.D. Glimcher M.J. Nanci A. Anat. Rec. 1992; 234: 479-492Google Scholar). The transcriptional control of osteocalcin gene expression has been extensively studied, and two different types of regulation have been identified, viz. hormonal regulation and tissue-specific regulation (19Ducy P. Karsenty G. Bilezikian J.P. Raisz L.G. Rodan G.A. Principles of Bone Biology. 1st Ed. Academic Press, Inc., San Diego, CA1996: 183-195Google Scholar). Hormonal regulation is mediated by vitamin D and glucocorticoid through vitamin D-responsive element (20Demay M.B. Roth D.A. Kronenberg H.M. J. Biol. Chem. 1989; 264: 2279-2282Google Scholar, 21Lian J. Stewart C. Puchacz E. Mackowiak S. Shalhoub V. Collart D. Zambetti G. Stein G. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 1143-1147Google Scholar) and glucocorticoid-responsive element (22Morrison N.A. Shine J. Fragonas J.C. Verkest V. McMenemy M.L. Eisman J.A. Science. 1989; 246: 1158-1161Google Scholar), respectively, which are both located in the osteocalcin promoter. Recently, transcriptional and post-transcriptional stimulation by thyroid hormone has been also reported (23Gouveia C.H. Schultz J.J. Bianco A.C. Brent G.A. J. Endocrinol. 2001; 170: 667-675Google Scholar). Apart from hormone-responsive cis-acting elements, two osteoblast-specific elements, OSE1 and OSE2, have been identified in the mouse osteocalcin promoter, and these DNA sequences are known to be involved in the regulation of the tissue-specific expression of the osteocalcin gene (24Ducy P. Karsenty G. Mol. Cell. Biol. 1995; 15: 1858-1869Google Scholar). Similar sequences responsible for cell-specific regulation have also been identified in the rat osteocalcin promoter (25Tamura M. Noda M. J. Cell Biol. 1994; 126: 773-782Google Scholar, 26Towler D.A. Bennett C.D. Rodan G.A. Mol. Endocrinol. 1994; 8: 614-624Google Scholar). Of the two identified OSEs, OSE2 binds Runx2 (Cbfa1) (core-binding factor A1)/AML3/Pebp2αA, a Runt-related transcription factor that is essential for osteoblast differentiation. Runx2 is the only osteoblast-specific transactivation factor identified to date (6Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Google Scholar, 27Geoffroy V. Ducy P. Karsenty G. J. Biol. Chem. 1995; 270: 30973-30979Google Scholar, 28Merriman H.L. van Wijnen A.J. Hiebert S. Bidwell J.P. Fey E. Lian J. Stein J. Stein G.S. Biochemistry (Mosc.). 1995; 34: 13125-13132Google Scholar), and an expression vector containing Runx2 has been shown to increase osteocalcin promoter activity through OSE2 (6Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Google Scholar). Thus, it is tempting to speculate that Runx2 might be a target of the PPARγ-mediated suppression of osteocalcin expression observed in previous studies. In this study, we addressed the mechanism through which PPARγ activation inhibits osteocalcin gene expression. The activation of PPARγ by 15-deoxy-Δ12,14-prostaglandin J2 (15-dPGJ2) in osteoblasts inhibited osteocalcin expression by direct repression of osteocalcin promoter activity as well as an indirect effect through inhibition of Runx2 expression. We present evidence that PPARγ interacts with Runx2 and that this leads to the decreased binding of Runx2 to OSE2 of the osteocalcin promoter. These results may help to explain why PPARγ activation suppresses osteoblast differentiation and the expression of osteoblast-specific genes from mesenchymal precursors.EXPERIMENTAL PROCEDURESMaterials—15-dPGJ2, ciglitazone, and antisera to PPARγ were purchased from BIOMOL Research Labs Inc. (Plymouth Meeting, PA); troglitazone was from Sankyo (Tokyo, Japan); and fenofibrate was from Sigma. TRI reagent was obtained from Molecular Research Center, Inc. (Cincinnati, OH), and Western blot detection reagents and [α-32P]dCTP were from Amersham Biosciences (Buckinghamshire, UK). Random priming kits and reagents for the luciferase assay were from Promega Corp. (Madison, WI), and nitrocellulose membranes were from Schleicher & Schüll (Dassel, Germany). LipofectAMINE Plus was obtained from Invitrogen, and anti-Cbfa1 (Runx2) antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). Oligonucleotides were synthesized by Bioneer Corp. (Chungwon, Korea); and unless otherwise indicated, all other chemicals, including tissue culture medium, were from Sigma.Expression Vectors and Reporter Plasmids—An expression vector for PPARγ2, pcDNA3-PPARγ2, was constructed by isolating full-length PPARγ2 cDNA from pSV-SPORT1-PPARγ2 (a kind gift of Dr. Bruce Spiegelman, Harvard Medical School, Boston, MA) by digestion with KpnI and SnaBI and insertion into the KpnI/EcoRV sites of the pcDNA3 vector (Invitrogen). An expression vector for Runx2 (Cbfa1/Osf2), pCMV-Osf2, was obtained from Dr. Patricia Ducy (Baylor College of Medicine, Houston, TX). pCMX-mRXRα, an expression vector for mouse retinoid X receptor-α (mRXRα), was obtained from Dr. David Mangelsdorf (University of Texas, Dallas, TX). The dominant-negative PPAR expression vector was kindly provided by Dr. V. Krishna Chatterjee (University of Cambridge, Cambridge, UK). The mouse osteocalcin II (OG2) promoter-luciferase reporter construct –1.3OG2-Luc, containing a 1.3-kb segment (positions –1316 to +13), has been described previously (29Zhang R. Ducy P. Karsenty G. J. Biol. Chem. 1997; 272: 110-116Google Scholar). The p6OSE2-Luc and p6OSE2m-Luc plasmids contain six copies of the wild-type and mutant OSE2 sequences of the osteocalcin promoter, respectively, followed by a minimal promoter, which directs the expression of luciferase (24Ducy P. Karsenty G. Mol. Cell. Biol. 1995; 15: 1858-1869Google Scholar). The pCbfa1-Luc plasmid contains a 135-bp fragment of the mouse Cbfa1/Runx2 promoter from positions –89 to +46, driving the expression of luciferase, and the OSE2 sites within this Cbfa1/Runx2 promoter segment are mutated in the pCbfa1m-Luc reporter (30Ducy P. Starbuck M. Priemel M. Shen J. Pinero G. Geoffroy V. Amling M. Karsenty G. Genes Dev. 1999; 13: 1025-1036Google Scholar). All these reporter plasmids were kindly provided by Dr. Patricia Ducy. PPREx3TK-Luc, containing three copies of the acyl-CoA oxidase PPAR-responsive element (PPRE) upstream of the herpesvirus thymidine kinase promoter, has been described previously (31Kliewer S.A. Umesono K. Noonan D.J. Heyman R.A. Evans R.M. Nature. 1992; 358: 771-774Google Scholar). The GST-PPARγ2 and GST-Runx2 constructs were kindly provided by Dr. Robert Roeder (The Rockefeller University, New York, NY) and Dr. Philip Hinds (Harvard Medical School), respectively.Cell Culture—The osteogenic sarcoma cell line ROS 17/2.8 was provided by Dr. Roberto Civitelli (Washington University School of Medicine, St. Louis, MO). ROS 17/2.8 cells have been shown to express several osteoblastic features, including the production of osteocalcin and other matrix proteins (32Majeska R.J. Nair B.C. Rodan G.A. Endocrinology. 1985; 116: 170-179Google Scholar). These cells were cultured in Dulbecco's modified Eagle's medium (DMEM)/nutrient mixture F-12 containing 10% heat-inactivated fetal bovine serum (FBS; BioWhittaker, Inc., Walkersville, MD). The mouse osteoblastic MC3T3-E1 cells were derived from spontaneously immortalized calvaria cells and represent immature osteogenic cells (33Sudo H. Kodama H.A. Amagai Y. Yamamoto S. Kasai S. J. Cell Biol. 1983; 96: 191-198Google Scholar). MC3T3-E1 cells were maintained in DMEM/nutrient mixture F-12 containing 10% FBS. During osteoblast maturation studies, MC3T3-E1 cells were cultured in DMEM/nutrient mixture F-12 containing 10% FBS supplemented with 50 μg/ml ascorbic acid and 10 mm β-glycerophosphate. Either 25 μm 15-dPGJ2 or vehicle was added at confluence and with subsequent medium changes (every 3 days). Murine embryonic mesenchymal C3H10T1/2 cells (American Type Culture Collection, Manassas, VA) are pluripotent cells that retain an immature, fibroblast-like appearance under standard tissue culture conditions. C3H10T1/2 cells were grown in Eagle's basal medium containing 10% FBS. The 3T3-L1 preadipocytic cell line was a kind gift from Dr. Jae Bum Kim (Seoul National University, Seoul, Korea). 3T3-L1 cells were maintained in an immature state by culturing in DMEM supplemented with 20% FBS and 2.0 mm glutamine.Generation of Stably Transfected Cell Lines—ROS 17/2.8 cells were seeded in p100 dishes (3 × 106 cells/dish) in DMEM/nutrient mixture F-12 containing 10% (v/v) heat-inactivated FBS. After overnight recovery, the cells were transfected with either pcDNA3-PPARγ2 or pcDNA3 without insert using LipofectAMINE Plus according to the manufacturer's protocol. Forty-eight hours later, the cells were diluted 10-fold and incubated with DMEM/nutrient mixture F-12 containing 10% (v/v) FBS and 400 μg/ml G418 (Sigma). Two weeks later, drug-resistant colonies were selected and expanded, and the expression of the exogenous gene was confirmed by Northern blot analysis as described below.Northern Blotting—Total cellular RNA was isolated from cell mono-layers using TRI reagent according to the manufacturer's instructions. Samples (20 μg/lane) were separated on 1% formaldehyde-agarose gels by electrophoresis, blotted onto nylon membranes, and UV-cross-linked. The membranes were then hybridized using 32P-labeled probes made by the random-primed oligonucleotide method (Label A Gene labeling kit, Promega Corp.) in ULTRAhyb solution (Ambion Inc., Austin, TX) at 42 °C overnight and washed twice with 2× SSC and 0.1% SDS at 42 °C, followed by one high stringency wash with 0.2× SSC and 0.1% SDS at 42 °C for 15 min. The following cDNA probes were used: 1.7-kb EcoRI fragment of mouse Runx2 (6Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Google Scholar), 470-bp EcoRI-PstI fragment of mouse osteocalcin, 600-bp XbaI-HindIII fragment of PPARγ2, and 1.9-kb BamHI fragment of rat β-actin. The level of mRNA was quantitated from digitized autoradiographic images using SigmaScan (SPSS Inc., Chicago, IL).Reverse Transcription-PCR—First-strand cDNA was synthesized from 2 μg of total RNA using a reverse transcription system kit (Promega Corp.). PCR was performed using 2 μl of cDNA, 20 pmol of each primer (synthesized by Bioneer Corp.), 200 μm each dNTP, 1 mm MgCl2, and 1 unit of Taq polymerase in a 50-μl reaction volume containing 1× Taq polymerase buffer using a PerkinElmer Life Sciences GeneAmp PCR System 2400. Primers 5′-CTCTGTCTCTCTGACCTCACAG-3′ (sense) and 5′-GGAGCTGCTGTGACATCCATAC-3′ (antisense), 5′-GAGGGCACAAGTTCTATCTGGA-3′ (sense) and 5′-GGTGGTCCGCGATGATCTTC-3′ (antisense), 5′-ATGGTTGACACAGAGATGCCA-3′ (sense) and 5′-ATGCTTTATCCCCACAGAC-3′ (antisense), 5′-GGGTGAAACTCTGGGAGATT-3′ (sense) and 5′-ATGCTTTATCCCCACAGAC-3′ (antisense), and 5′-ACCACAGTCCATGCCATCAC-3′ (sense) and 5′-TACAGCAACAGGGTGGTGGA-3′ (antisense) were used to amplify osteocalcin, Runx2, PPARγ1, PPARγ2, and glyceraldehyde-3-phosphate dehydrogenase, producing bands of 359, 387, 348, 436, and 451 bp, respectively.Western Blotting—Cell lysates were prepared by treating cells with lysis buffer (150 mm NaCl, 50 mm Tris-Cl (pH 7.4), 20 mm EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, and protease inhibitors (Sigma)). Lysates were sonicated for 20 min on ice and centrifuged at 10,000 × g for 10 min to sediment particulate material. Protein concentrations of the supernatants were measured as described by Lowry et al. (34Lowry O.H. Rosebrough N.J. Farr A.L. Randell R.J. J. Biol. Chem. 1951; 193: 265-275Google Scholar). SDS-PAGE was performed under reducing conditions on 10% polyacrylamide gels, and the resolved proteins were transferred onto nitrocellulose membranes. Membranes were blocked with 0.1% Tween 20 and Tris-buffered saline containing 2% bovine serum albumin and 3% dry milk at pH 7.4 for 1 h. Polyclonal antibody against PPARγ was added, and the incubation was continued for an addition hour. After washing with 0.1% Tween 20 and Tris-buffered saline, the membranes were incubated with horseradish peroxidase-conjugated anti-mouse antibodies for 1 h. After extensive washing, bands were visualized by chemiluminescence using an ECL kit (Amersham Biosciences) according to the manufacturer's instructions.Transfections and Reporter Assays—Transient transfections were performed in triplicate, and the transfection efficiencies were monitored using pCMV-β-gal vectors (Promega Corp.) in parallel cultures. For these experiments, osteoblastic cells were plated at high density (3 × 105 cells/well) onto 12-well plates. Appropriate plasmids were transfected into each well using LipofectAMINE Plus following the manufacturer's instructions. Cell lysates (0.25 ml/well) were prepared using the Promega luciferase assay system, and reporter activity was measured using a luminometer (Lumat LB 9507, Berthold, Wildbad, Germany). All luciferase values were normalized against the β-galactosidase activities from the cotransfected pCMV-β-gal plasmid. All values and means ± S.D. are expressed as -fold induction relative to basal promoter activity.Electrophoretic Mobility Shift Assay—Nuclear extracts were prepared according to the method of Dignam et al. (35Dignam J.D. Lebovitz R.M. Roeder R.G. Nucleic Acids Res. 1983; 11: 1475-1489Google Scholar). Briefly, cells were washed with ice-cold phosphate-buffered saline and then resuspended in hypotonic lysis buffer containing 20 mm HEPES (pH 8.0), 25% glycerol, 1.5 mm MgCl2, 0.2 mm EDTA, 0.3% Triton X-100, 0.6% ammonium sulfate, 1 mm dithiothreitol, and protease inhibitors. The protein concentrations of the nuclear extracts were determined by the Bradford assay (Bio-Rad) using bovine serum albumin as a standard. In vitro translated mouse PPARγ2 and mRXRα were obtained by transcribing and translating the pcDNA3-PPARγ2 and pCMX-mRXRα expression plasmids, respectively, using the TnT T7-coupled reticulocyte lysate system (Promega Corp.). Protein concentration was measured using parallel [35S]methionine-labeled reactions.Oligonucleotide probes corresponding to the OSE2 site in the mouse osteocalcin promoter (5′-GATCCGCTGCAATCACCAACCACAGCA-3′) (24Ducy P. Karsenty G. Mol. Cell. Biol. 1995; 15: 1858-1869Google Scholar) and the optimal consensus PPRE sequence (5′-GATCAGCTACGTGACCTTTGACCTGGT-3′) (36Brun R.P. Tontonoz P. Forman B.M. Ellis R. Chen J. Evans R.M. Spiegelman B.M. Genes Dev. 1996; 10: 974-984Google Scholar) were generated using an oligonucleotide synthesizer (Bioneer Corp.). The complementary oligonucleotides were annealed and labeled with [α-32P]dCTP. The binding reaction was performed by incubating 10 μg of nuclear protein from cultured cells or in vitro translated proteins in 20 mm HEPES (pH 8.0), 25% glycerol, 1.5 mm MgCl2, 300 mg of bovine serum albumin, and 1 mg of poly(dI-dC) in a final volume of 10 μl for 10 min at 25 °C. The labeled oligonucleotide was added to the reaction mixture and allowed to incubate for an additional 20 min in ice. To prove specific binding of Runx2 to the oligonucleotide, the nuclear lysates were preincubated for 1 h at 4 °C with anti-Cbfa1 (Runx2) antibody prior to the addition of poly(dI-dC) and radiolabeled probe DNA. The samples were electrophoresed on a 4% nondenaturing polyacrylamide gel. The gel was then dried and autoradiographed.GST Pull-down Analyses—GST fusion proteins were induced in Escherichia coli BL21 for 3 h at 25 °C by the addition of isopropyl-1-thio-β-d-galactopyranoside (100 μm final concentration) to a 100-ml bacterial culture (A600 ≈ 0.5). After induction, bacteria were pelleted for 20 min at 3000 × g and resuspended in 20 ml of ice-cold binding buffer (20 mm Tris-HCl (pH 7.5), 150 mm NaCl, 0.1% Nonidet P-40, 0.1 mm phenylmethylsulfonyl fluoride, and 1 mm EDTA). Bacteria were lysed by freeze-thawing for 5 min in liquid nitrogen, followed by thawing for 10 min at 37 °C. The lysis procedure was repeated three times. The freeze-thawed bacteria were then subjected at 4 °C to three 10-s rounds of sonication, and the bacterial debris was pelleted by centrifugation at 15,000 × g for 30 min at 4 °C. Supernatants were stored frozen at –20 °C in 100-μl aliquots until needed. Free GST lysates were prepared in a similar manner from E. coli BL21 transformed with a pGEX-3 vector. Free GST and GST fusion proteins were purified on glutathione-Sepharose 4B (Amersham Biosciences) according to the manufacturer's recommendations and dialyzed against binding buffer.For GST pull-down assays, equal amounts of purified recombinant GST or GST fusion proteins were immobilized on glutathione-Sepharose beads (Amersham Biosciences) and washed four times with 1 ml of wash buffer (20 mm Tris HCl (pH 7.5), 150 mm NaCl, 1 mm EDTA, 0.1% Nonidet P-40, 1 mm NaF, 2 μg/ml aprotinin, and 0.1 mm phenylmethylsulfonyl fluoride) at 4 °C. 35S-Labeled PPARγ2 and Runx2 were synthesized in rabbit reticulocyte lysate by coupled in vitro transcription and translation (TnT T7-coupled reticulocyte lysate system), added to immobilized GST or GST fusion proteins, and incubated for 2 h. After binding, proteins bound to the beads were eluted with elution buffer (10 mm reduced glutathione, 20 mm Tris-HCl (pH 7.5), 0.1 mm phenylmethylsulfonyl fluoride, 0.1% Nonidet P-40, and 2 μg/ml aprotinin), and the samples were separated by SDS-PAGE and analyzed by autoradiography.Immunoprecipitations—MC3T3-E1 cells were lysed in 50 mm Tris-HCl (pH 7.5), 150 mm NaCl, 1% Nonidet P-40, and 0.5% sodium deoxycholate containing a mixture of protease inhibitors. Lysates were then precleared for 3 h at 4 °C with protein G-Sepharose (Roche Applied Science, Mannheim, Germany). For immunoprecipitation of endogenous Runx2 from MC3T3-E1 cells, following incubation with goat anti-Cbfa1 (Runx2) antibody or an isotype-matched control (anti-thyroglobulin antibody), rabbit anti-goat Ig secondary antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) was used prior to precipitation. PPARγ was immunoprecipitated with anti-PPARγ anti-body or an isotype-matched control (anti-hemagglutinin antibody, Santa Cruz Biotechnology). The lysates were incubated for 3 h at 4 °C prior to incubation with protein G-Sepharose. After extensive washing, the immunoprecipitates were subjected to SDS-PAGE, and the expression levels of the proteins of interest were verified by Western analyses of the cell lysates using specific antibodies.RESULTSPPARγ Is Expressed in MC3T3-E1 Cells, but Not in ROS 17/2.8 Cells—We determined the expression of PPARγ in rodent osteoblastic and mesenchymal cell lines. MC3T3-E1 cells represent immature osteoblasts derived from mouse calvarial cells, which undergo osteoblast differentiation in culture (33Sudo H. Kodama H.A. Amagai Y. Yamamoto S. Kasai S. J. Cell Biol. 1983; 96: 191-198Google Scholar). ROS 17/2.8 cells are a rat osteosarcoma cell line often used for the study of osteoblast function (32Majeska R.J. Nair B.C. Rodan G.A. Endocrinology. 1985; 116: 170-179Google Scholar). Western blot analysis revealed that PPARγ was expressed in MC3T3-E1 cells, but not in ROS 17/2.8 cells (Fig. 1A). Embryonic mesenchymal C3H10T1/2 cells also expressed PPARγ under basal conditions, and the level was increased after treatment with the PPARγ activator 15-dPGJ2. Because PPARγ exists as two isoforms (γ1 and γ2) as a result of alternative splicing, we investigated which isoform was expressed in the MC3T3-E1 cells by reverse transcription-PCR. We were able to demonstrate the expression of PPARγ1 mRNA in this cell line; however, PPARγ2 mRNA was not detectable even up to 35 cycles (data not shown). To determine whether the PPARγ expressed in the MC3T3-E1 cells was transcriptionally active, a PPRE cloned upstream of luciferase (PPREx3TK-Luc) was transiently transfected into MC3T3-E1 cells, and the cells were then treated with 15-dPGJ2. Expression of luciferase activity was significantly induced after 15-dPGJ2 treatment (Fig. 1B), suggesting that MC3T3-E1 cells express functionally active PPARγ.PPARγ Activators Inhibit Osteocalcin Gene Expression in Rodent Osteoblasts—The possibility that PPARγ activators inhibit osteocalcin expression was investigated. ROS 17/2.8 cells, which constitutively express osteocalcin (24Ducy P. Karsenty G. Mol. Cell. Biol. 1995; 15: 1858-1869Google Scholar), were stably transfected with a PPARγ2 expression construct or an empty vector. Exposure of the cells with an empty vector to 15-dPGJ2 did not alter the level of osteocalcin expression. However, ROS 17/2.8 cells stably transfected with the PPARγ2 construct showe