Portal de Periódicos da CAPES

E3 Ubiquitin Ligase Smurf1 Mediates Core-binding Factor α1/Runx2 Degradation and Plays A Specific Role in Osteoblast Differentiation

2003; Elsevier BV; Volume: 278; Issue: 30 Linguagem: Inglês

10.1074/jbc.m304132200

1083-351X

Ming Zhao, Mei Qiao, Babatunde O. Oyajobi, Gregory R. Mundy, Di Chen,

Bone health and treatments

Osteoblast differentiation and bone formation is stimulated by bone morphogenetic protein (BMP)-2 and its downstream signaling molecules Smad1 and -5 and the osteoblast-specific transcription factor core-binding factor α1 (Cbfa1). Proteolytic degradation of Smad1 and Cbfa1 is proteasome-dependent, and intracellular concentrations of Smad1 and Cbfa1 are enhanced by inhibition of the 26 S proteasome. Smad1 degradation is mediated by the E3 ubiquitin ligase Smurf1 (Smad ubiquitin regulatory factor 1), but the specific E3 ligase responsible for Cbfa1 degradation has not been identified. Because Cbfa1 interacts with Smad1, whose degradation is mediated by Smurf1, we examined the effect of Smurf1 on Cbfa1 degradation in osteoblast precursor cells. Smurf1 interacts directly with Cbfa1 and mediates Cbfa1 degradation in a ubiquitin- and proteasome-dependent manner. Because Smurf1 controls the intracellular concentrations of several key molecules in the bone formation cascade, we examined the effect of a mutant form of Smurf1 in osteoblasts and found that expression of mutant Smurf1 markedly enhanced osteoblas tdifferentiation. Smurf1 therefore appears to be an important regulatory factor in osteoblast differentiation and a potential molecular target for identification of bone anabolic agents. Osteoblast differentiation and bone formation is stimulated by bone morphogenetic protein (BMP)-2 and its downstream signaling molecules Smad1 and -5 and the osteoblast-specific transcription factor core-binding factor α1 (Cbfa1). Proteolytic degradation of Smad1 and Cbfa1 is proteasome-dependent, and intracellular concentrations of Smad1 and Cbfa1 are enhanced by inhibition of the 26 S proteasome. Smad1 degradation is mediated by the E3 ubiquitin ligase Smurf1 (Smad ubiquitin regulatory factor 1), but the specific E3 ligase responsible for Cbfa1 degradation has not been identified. Because Cbfa1 interacts with Smad1, whose degradation is mediated by Smurf1, we examined the effect of Smurf1 on Cbfa1 degradation in osteoblast precursor cells. Smurf1 interacts directly with Cbfa1 and mediates Cbfa1 degradation in a ubiquitin- and proteasome-dependent manner. Because Smurf1 controls the intracellular concentrations of several key molecules in the bone formation cascade, we examined the effect of a mutant form of Smurf1 in osteoblasts and found that expression of mutant Smurf1 markedly enhanced osteoblas tdifferentiation. Smurf1 therefore appears to be an important regulatory factor in osteoblast differentiation and a potential molecular target for identification of bone anabolic agents. Bone formation is regulated by a number of growth regulatory factors, key signal transduction molecules, and critical transcription factors. Important among these are the bone morphogenetic proteins (BMPs), 1The abbreviations used are: BMP, bone morphogenetic protein; Cbfa1, Core-binding factor α1; Smad, Sma/Mother against decapentaplegic; Smurf1, Smad ubiquitin regulatory factor 1; PS1, proteasome inhibitor 1; SBE, Smad1 binding element; caBMPR-IB, constitutively active type IB BMP receptor; OC, osteocalcin; ALP, alkaline phosphatase; E1, ubiquitin-activating enzyme; E2, ubiquitin-conjugating enzyme; E3, ubiquitin-protein isopeptide ligase; HA, hemagglutinin. which promote normal appositional bone growth and induce ectopic bone formation experimentally (1Wozney J.M. Mol. Reprod. Dev. 1992; 32: 160-167Crossref PubMed Scopus (613) Google Scholar). BMP-2 and -4 act through a complex serine-threonine kinase receptor mechanism to activate the signal transduction molecules Smad1 and -5, which in turn activate downstream target genes with other transcription factors. The mechanism of action and regulation of BMP signaling molecules during osteoblast differentiation are not fully defined. Cbfa1 2Other names for Cbfa1/Runx2 are PEBP2αA, AML-3, and Osf-2. is a transcription factor that belongs to the runt-domain gene family and plays an essential role in osteoblast differentiation, bone development, and postnatal bone formation (2Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Abstract Full Text Full Text PDF PubMed Scopus (3647) Google Scholar, 3Komori T. Yagi H. Nomura S. Yamaguchi A. Sasaki K. Deguchi K. Shimizu Y. Bronson R.T. Gao Y. Inada M. Sato M. Okamoto R. Kitamura R. Yoshiki S. Kishimoto T. Cell. 1997; 89: 755-764Abstract Full Text Full Text PDF PubMed Scopus (3653) Google Scholar, 4Otto F. Thornell A.P. Crompton T. Denze A. Gilmour K.C. Rosewell I.R. Stamp G.W.H. Beddington R.S.P. Mundlos S. Olsen B.R. Selby P. Owen M. Cell. 1997; 89: 765-771Abstract Full Text Full Text PDF PubMed Scopus (2412) Google Scholar, 5Ducy P. Starbuck M. Priemel M. Shen J. Pinero G. Geoffroy V. Amling M. Karsenty M. Gene Dev. 1999; 13: 1025-1036Crossref PubMed Scopus (709) Google Scholar). Its expression and activity are regulated by many bone-derived growth factors, including BMPs (2Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Abstract Full Text Full Text PDF PubMed Scopus (3647) Google Scholar, 6Chen D. Ji X. Harris M.A. Feng J.-Q. Karsenty G. Celeste A.J. Rosen V. Mundy G.R. Harris S.E. J. Cell Biol. 1998; 142: 295-305Crossref PubMed Scopus (341) Google Scholar). BMPs may stimulate osteoblast differentiation by activating Cbfa1 through Smad1, as evidenced by reports that Smad1, the downstream effector of BMP signaling, directly interacts with Cbfa1 (7Hanai J.I. Chen L.F. Kanno T. Ohtani-Fujita N. Kim W.Y. Guo W.-H. Imamura T. Ishidou Y. Fukuchi M. Shi M.J. Stavnezer J. Kawabata M. Miyazono K. Ito Y. J. Biol. Chem. 1999; 274: 31577-31582Abstract Full Text Full Text PDF PubMed Scopus (410) Google Scholar), but precise signaling mechanisms remain unclear. The activities of signaling proteins and transcription factors are regulated at both the transcriptional and post-translational levels. Recent reports (8Zhu H. Kavsak P. Abdollah S. Wrana J. Thomsen G.H.A. Nature. 1999; 400: 687-693Crossref PubMed Scopus (688) Google Scholar, 9Zhang Y. Chang C. Gehling D.J. Hemmati-Brivanlou A. Derynck R. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 974-979Crossref PubMed Scopus (427) Google Scholar, 10Tintut Y. Parhami F. Le V. Karsenty G. Demer L.L. J. Biol. Chem. 1999; 274: 28875-28879Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar) demonstrate that both Smad1 and Cbfa1 undergo ubiquitin-proteasome-mediated degradation. Protein ubiquitination involves a cascade of enzymatic reactions catalyzed by the E1 ubiquitin-activating enzyme, the E2 ubiquitin-conjugating enzymes, and the E3 ubiquitin ligases (11Hochstrasser M. Annu. Rev. Genet. 1996; 30: 405-439Crossref PubMed Scopus (1459) Google Scholar), which play a crucial role in defining substrate specificity and subsequent protein degradation by the 26 S proteasomes. Smurf1 is a member of the Hect family of E3 ubiquitin ligases and has been found to interact with the BMP-activated Smad1 and -5, thereby triggering their ubiquitination and degradation (8Zhu H. Kavsak P. Abdollah S. Wrana J. Thomsen G.H.A. Nature. 1999; 400: 687-693Crossref PubMed Scopus (688) Google Scholar). Hect domain proteins represent a major subclass of E3 ligases and contain a conserved cysteine located at the carboxylterminal of the Hect domain that is capable of forming a thioester bond with ubiquitin (8Zhu H. Kavsak P. Abdollah S. Wrana J. Thomsen G.H.A. Nature. 1999; 400: 687-693Crossref PubMed Scopus (688) Google Scholar, 12Huibregtse J.M. Scheffner M. Beaudenon S. Howley P.M. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 2563-2567Crossref PubMed Scopus (703) Google Scholar). Another motif often found in the Hect family of E3 ligase is the WW domain, which contains two highly conserved tryptophans and a conserved proline in an ∼30-amino acid region (8Zhu H. Kavsak P. Abdollah S. Wrana J. Thomsen G.H.A. Nature. 1999; 400: 687-693Crossref PubMed Scopus (688) Google Scholar, 13Rotin D. Curr. Top. Microbiol. Immunol. 1998; 228: 115-133PubMed Google Scholar). The WW domains have a preference for binding to small proline-rich sequences, PPXY motifs, and different WW domains possess different substrate specificity. Although Cbfa1 has been reported to be degraded through the ubiquitin-proteasome pathway (10Tintut Y. Parhami F. Le V. Karsenty G. Demer L.L. J. Biol. Chem. 1999; 274: 28875-28879Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar), the specific E3 ubiquitin ligase for Cbfa1 has not been identified. In the present studies, we examined the relationship between Smad1, Cbfa1, and Smurf1 in osteoblast precursor cells. Smad1 interacts directly with Cbfa1 in these cells, and both Smad1 and Cbfa1 are required for BMP-2 signaling. The intracellular concentrations of both Smad1 and Cbfa1 are dependent on osteoblast proteasomal function. Smurf1 mediates Cbfa1 degradation in a ubiquitin proteasome-dependent manner and therefore is a powerful regulator of osteoblast differentiation. Expression Plasmids—Myc-tagged Smad1 expression plasmid was obtained from Dr. R. Nishimura (Department of Biochemistry, Osaka University, Osaka, Japan), and FLAG-tagged Smurf1 and mutant Smurf1 (C710A) cDNA plasmids in SK vector were obtained from Dr. G. Thomsen and cloned into pcDNA3 expression vector. Cbfa1 cDNA was amplified by RT-PCR based on mouse Cbfa1 cDNA sequences (GenBank™ accession number AF010284) using RNA extracted from 2T3 osteoblast precursor cells and then cloned into p3×FLAG-CMV vector (Sigma). The 6×OSE2-OC-pGL3 reporter construct was generated by PCR as described by Ducy et al. (2Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Abstract Full Text Full Text PDF PubMed Scopus (3647) Google Scholar). The constitutively active type IB BMP receptor (caBMPR-IB) expression plasmid was obtained from Dr. J. Wrana (Program in Molecular Biology and Cancer, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Canada). The HA-ubiquitin expression plasmid was obtained from Dr. X. Cao (Department of Pathology, University of Alabama at Birmingham, Alabama). Cell Culture and Transfections—C2C12 cells were cultured in Dulbecco's modified Eagle's medium, and 2T3 osteoblast precursor cells were cultured in α-minimal essential medium supplemented with 8% fetal calf serum. The cDNA expression plasmids were transiently transfected into these cells using LipofectAMINE Plus reagents (Invitrogen) in a 10-cm-diameter culture dish for immunoprecipitation assay, 6-well culture plates for Western blotting analysis, and 24-well plates for the luciferase assay. 5, 1, and 0.2 μg of expression plasmid were used for these transfections. After transfection (48 h), cells were lysed with lysis buffer. In proteasome inhibitor experiments, transfected cells were incubated with the proteasome inhibitors for 2, 6, or 24 h, the proteasome inhibitor was removed, and the cells were then cultured an additional 24 h. Immunoprecipitation and Immunoblotting—Expression plasmids were transfected into C2C12 cells separately or together for immunoprecipitation. In Western analysis, the total amounts of transfected plasmids in each group were equalized by addition of empty vector cDNA. 48 h after transfection, cells were washed three times with phosphate-buffered saline and solubilized in lysis buffer (150 mm NaCl, 1% Triton X-100, 0.5% doc, and 50 mm Tris buffer, pH 7.5). For Western blotting, 0.1% SDS was included in the lysis buffer. The protease inhibitors aprotinin (10 μg/ml), leupeptin (10 μg/ml), and phenylmethylsulfonyl fluoride (1 mm) were added to the lysis buffer. Cell lysates were centrifuged for 10 min at 4 °C at 10,000 × g and incubated with anti-Myc antibody for 4 h at 4 °C, followed by immunoprecipitation with protein G-agarose (Roche Applied Science) at 4 °C overnight. Immunoprecipitates were washed with lysis buffer five times, added to 1× reducing buffer containing 0.5 m β-mercaptoethanol, and boiled for 3 min. The immunoprecipitation and Western blotting samples were separated by SDS-PAGE, transferred to nitrocellulose membrane, immunoblotted with anti-FLAG or anti-Myc antibody, and visualized with horseradish peroxidase-coupled anti-mouse IgG antibody (Amersham Biosciences) with an enhancement by ECL detection kits (Amersham Biosciences). A monoclonal antibody against human Smad1 and a polyclonal antibody against human Smurf1 were purchased from Santa Cruz Biotechnology, Santa Cruz, CA. A polyclonal antibody against human Cbfa1 was purchased from Oncogene Research Product, Cambridge, MA. Luciferase Assay—Oligonucleotides containing 6 copies of Cbfa1 response element, OSE2, (6×OSE2) and 12 copies of Smad1 binding element (12×SBE) were synthesized in the DNA laboratory at The University of Texas Health Science Center at San Antonio. The oligonucleotides were cloned in front of the osteocalcin basal promoter (–155/+1), which was amplified by RT-PCR, and cloned in pGL3 vector. C2C12 and 2T3 cells were co-transfected with Cbfa1 expression plasmid and 6×OSE2-OC-pGL3 reporter plasmid in the presence or absence of Smurf1 or mutant Smurf1 expression plasmid. caBMPR-IB and Smad1 expression plasmids were co-transfected with 12×SBE-OC-luc or 12×SBE-SV40-luc in the presence or absence of mutant Cbfa1. Cell lysates were extracted 48 h later, and luciferase activity was measured and normalized by β-galactosidase activity. Alkaline Phosphatase (ALP) Activity and Osteocalcin Production— C2C12 and 2T3 cells were plated into 6-well culture plates grown to 60% confluency, transfected transiently with Smurf1 or mutant Smurf1 expression plasmids using LipofectAMINE Plus reagents (Invitrogen), or treated with proteasome inhibitors. 48 h after transfection, the medium was collected; the cells were washed twice with ice-cold phosphate-buffered saline, and cell lysates were extracted with 0.05% Triton X-100. ALP activity in cell lysates was measured using a Sigma ALP assay kit (Sigma), and osteocalcin in the medium was measured with a mouse osteocalcin immunoradiometric assay (IRMA) kit (Immutopics, Inc., San Clemente, CA). Calvarial Periosteal Bone Formation Assay—Proteasome inhibitor 1 (PS1) was injected subcutaneously over the parietal bone of the calvariae of 2-month-old ICR Swiss mice (14Chen D. Harris M.A. Rossini G. Dunstan C.R. Dallas S.L. Feng J.Q. Mundy G.R. Harris S.E. Calcif. Tissue Int. 1997; 60: 283-290Crossref PubMed Scopus (204) Google Scholar). Mice received either vehicle (20 μl of phosphate-buffered saline) alone or PS1 (0.2, 1, and 5 mg/kg/day, daily for 5 days). Mice were sacrificed 14 days after commencing injections, and calvarial bones were removed. The bones were decalcified in 14% EDTA, bisected coronally midway between the coronal and lambdoid sutures, dehydrated through graded alcohols, and embedded in paraffin. Transverse 3-μm-thick sections were cut and stained with Hematoxylin and Eosin. Newly formed bone was identified histologically using an E400 microscope (Nikon Inc., Melville, NY) linked to a color video monitor (PVM-14M2MDU Trinitron, Sony Corp., Japan) and image capture techniques. Smurf1 Mediates Cbfa1 Degradation in Osteoblasts—Because Cbfa1 interacts with Smad1 protein (7Hanai J.I. Chen L.F. Kanno T. Ohtani-Fujita N. Kim W.Y. Guo W.-H. Imamura T. Ishidou Y. Fukuchi M. Shi M.J. Stavnezer J. Kawabata M. Miyazono K. Ito Y. J. Biol. Chem. 1999; 274: 31577-31582Abstract Full Text Full Text PDF PubMed Scopus (410) Google Scholar) whose degradation is mediated by Smurf1 (8Zhu H. Kavsak P. Abdollah S. Wrana J. Thomsen G.H.A. Nature. 1999; 400: 687-693Crossref PubMed Scopus (688) Google Scholar), we examined the effect of Smurf1 on Cbfa1 degradation in myoblast/osteoblast precursor C2C12 cells and osteoblast precursor 2T3 cells. FLAG-tagged Cbfa1 expression plasmid was co-transfected with different amounts of Smurf1 expression plasmid into C2C12 and 2T3 cells. Expression of Smurf1 greatly reduced steady-state levels of Cbfa1 in a dose-dependent manner (Fig. 1a). Expression of an unrelated Hect domain E3 ligase, Itch, with Cbfa1 had no effect on the degradation of Cbfa1 (data not shown), suggesting that Smurf1-mediated Cbfa1 degradation is a specific effect of the ligase. To determine whether Smurf1 mediates endogenous Cbfa1 degradation, 2T3 cells were transfected with Smurf1 expression plasmid. Expression of endogenous Cbfa1 protein was detected by Western blot analysis using an antibody against human Cbfa1. Transfection of Smurf1 reduced Cbfa1 protein levels in 2T3 cells (Fig. 1b), suggesting that Smurf1 mediates degradation of endogenous Cbfa1 protein. To determine whether Smurf1-mediated degradation of Cbfa1 is ubiquitin-dependent, we co-transfected Smurf1, Cbfa1, and ubiquitin expression plasmids in C2C12 cells. After immunoprecipitating the Cbfa1 protein, a typical ubiquitinated Cbfa1 protein ladder was observed (Fig. 1c), indicating that Smurf1-mediated Cbfa1 degradation is ubiquitin-dependent. To examine whether Smurf1-mediated degradation of Cbfa1 is proteasome-dependent, we treated transfected C2C12 cells with 1 μm PS1 for 2 h. Expression of Smurf1 resulted in total degradation of Cbfa1, an effect completely abolished by treatment with PS1 (Fig. 1d). Similar results were obtained when another structurally different proteasome inhibitor, epoxomicin, was used (data not shown). These results demonstrate that Smurf1-mediated Cbfa1 degradation is proteasome-dependent. To determine whether Smurf1 physically interacts with Cbfa1, Cbfa1, and mutant Smurf1 (ΔSmurf1, C710A), expression plasmids were co-transfected into C2C12 cells. Because it is known that Smurf1 mediates the degradation of proteins with which it interacts (8Zhu H. Kavsak P. Abdollah S. Wrana J. Thomsen G.H.A. Nature. 1999; 400: 687-693Crossref PubMed Scopus (688) Google Scholar), we attempted to stabilize the protein complex with Smurf1 using an expression plasmid of the catalytic point mutant of Smurf1 (ΔSmurf1) as described by Zhu et al. (8Zhu H. Kavsak P. Abdollah S. Wrana J. Thomsen G.H.A. Nature. 1999; 400: 687-693Crossref PubMed Scopus (688) Google Scholar). Using epitope-tagged proteins in co-immunoprecipitation and Western blot analyses, we found that Cbfa1 co-precipitates with ΔSmurf1 (Fig. 1e). A putative PY motif, presumably bound with the WW domain of Smurf1 protein, was found in the carboxyl-terminal of Cbfa1, suggesting that Smurf1 directly binds Cbfa1 and mediates its degradation. To determine whether Smurf1-mediated Cbfa1 degradation causes functional changes in Cbfa1, we examined the ability of Cbfa1 to activate a specific reporter gene, 6×OSE2-OC-Luc (15Ducy P. Karsenty G. Mol. Cell. Biol. 1995; 15: 1858-1869Crossref PubMed Scopus (527) Google Scholar), in C2C12 cells. A Cbfa1 expression plasmid and 6×OSE2-OC-Luc reporter construct were co-transfected into C2C12 cells in the presence and absence of Smurf1 or ΔSmurf1. Expression of Cbfa1 significantly increased activity of 6×OSE2-OC-Luc reporter, whereas co-transfection of Smurf1 with Cbfa1 significantly reduced Cbfa1-induced luciferase activity of this reporter gene. In contrast, co-transfection of ΔSmurf1, lacking in catalytic activity, had no significant effect on Cbfa1-induced luciferase activity of the reporter gene (Fig. 1f). These results demonstrate that Smurf1 mediates Cbfa1 degradation and leads to a decrease in Cbfa1 activity. That there was no significant effect of ΔSmurf1 on Cbfa1-induced luciferase activity of the reporter gene may be because activation of the reporter gene by expressed Cbfa1 had already plateaued. Because it has been reported that Cbfa1 degradation is cAMP-dependent in osteoblasts (10Tintut Y. Parhami F. Le V. Karsenty G. Demer L.L. J. Biol. Chem. 1999; 274: 28875-28879Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar), in the present studies we also treated C2C12 cells with forskolin, an agent that stimulates adenylyl cyclase and examined expression of endogenous Smurf1, Cbfa1 degradation, and the effect of ΔSmurf1 on forskolin-induced Cbfa1 degradation. Although forskolin treatment increased endogenous Smurf1 level only slightly (Fig. 2a), Cbfa1 protein level was significantly reduced by forskolin; this reduction was blocked in cells transfected with ΔSmurf1 (Fig. 2b). These results suggest that Smurf1 may be involved in the cAMP-induced Cbfa1 degradation. Smurf1 Mediates Smad1 Degradation in Osteoblasts—Although Smurf1 has been shown to mediate Smad1 degradation in COS cells transfected with a Smad1 expression plasmid, the effects of Smurf1 on Smad1 degradation in osteoblasts, and in particular degradation of endogenous Smad1 protein by proteasomal-dependent mechanism in osteoblasts, have not been examined. In the present studies, we confirmed the effect of Smurf1 on Smad1 degradation in osteoblast precursor cells and examined the regulatory roles of proteasome inhibitors on Smad1 protein. We first examined the effect of Smurf1 on steady-state protein levels of Smad1 in C2C12 and 2T3 cells and found that Smurf1 mediates Smad1 degradation in a dose-dependent manner (Fig. 3a). The effect of Smurf1 on Smad1 degradation is ubiquitin-proteasome-dependent because the catalytic mutant of Smurf1 (C710A) had only minor effects on Smad1 degradation (Fig. 3b) and the proteasome inhibitor epoxomicin reversed the effect of Smurf1 on Smad1 degradation in a dose-dependent manner (Fig. 3c). Similar effects were observed when other structurally different proteasome inhibitors such as PS1, lactacystin, and MG-132 were utilized (data not shown). To further determine the effects of proteasome inhibitors on the steady-state protein levels of endogenous Smad1, the C2C12 cells were treated with different proteasome inhibitors for 6 h. The treatment with proteasome inhibitors increased Smad1 levels 2– 6-fold (Fig. 3d), demonstrating that Smad1 is under Smurf1-mediated ubiquitin-proteasome regulation in osteoblasts. Cbfa1 Is Required for Activation of BMP-2 Signaling—Cbfa1 has been reported to interact with Smad1, but details of the transactivation mechanism of these two signaling molecules have not been clearly defined. Therefore, we examined whether Cbfa1 is specifically required for BMP signaling by first confirming the interaction of Cbfa1 with Smad1 in osteoblast precursor cells. Epitope-tagged Cbfa1 and Smad1 expression plasmids were co-transfected into C2C12 and 2T3 cells. Co-precipitation of Cbfa1 and Smad1 was detected in both cells (Fig. 4a). To determine the role of Cbfa1 in BMP signaling, we generated multiple copies of Smad1 response element (12×SBE) and cloned those upstream of an osteocalcin and a SV40 basal promoter. Treatment of BMP-2 or expression of caBMPR-IB stimulated luciferase activity of the 12×SBE-OC-Luc reporter but not the 12×SBE-SV40-Luc reporter (Fig. 4b). Because the osteocalcin basal promoter contains one copy of the Cbfa1 response element (OSE2), we next examined whether binding of Cbfa1 to this response element in the osteocalcin basal promoter is required for BMP-2 signaling. A mutant Cbfa1 (ΔCbfa1) expression plasmid was generated as described by Ducy et al. (5Ducy P. Starbuck M. Priemel M. Shen J. Pinero G. Geoffroy V. Amling M. Karsenty M. Gene Dev. 1999; 13: 1025-1036Crossref PubMed Scopus (709) Google Scholar), and C2C12 cells were transfected with caBMPR-IB expression plasmid and the 12×SBE-OC-Luc reporter construct with or without ΔCbfa1. Expression of ΔCbfa1 completely blocked the effect of caBMPR-IB on the 12×SBE-OC-Luc reporter (Fig. 4c). caBMPR-IB or BMP-2 had only minor effects on the osteocalcin basal promoter (data not shown). To further define the role of Cbfa1 in BMP signaling, the OSE2 site in the osteocalcin basal promoter was mutated. The 12×SBE-OC-Luc reporter with OSE2 mutation (12×SBE-ΔOC) completely lost its response to BMP-2 or caBMPR-IB (Fig. 4d). These results suggest that Cbfa1 is required for BMP-2 and Smad1 to activate downstream target genes in osteoblasts. Function of Smurf1 in Osteoblast Differentiation—To examine the function of Smurf1 in osteoblast differentiation, a Smurf1 or ΔSmurf1 expression plasmid was transfected into C2C12 cells, and ALP activity and osteocalcin production were measured. Expression of Smurf1 significantly reduced ALP activity as well as osteocalcin production in C2C12 cells (data not shown). In contrast, transfection of ΔSmurf1 increased basal as well as BMP-2-induced ALP activity and osteocalcin production in these cells (Fig. 5, a and b). Similar results were obtained when the same expression plasmids were transfected into 2T3 cells (data not shown). These results suggest that Smurf1 mediates Cbfa1 and Smad1 degradation and inhibits function of Cbfa1 and Smad1. ΔSmurf1 acts in a dominant-negative fashion to inhibit the degradation of Cbfa1 and Smad1. To establish a correlation between Smurf1-mediated Smad1 degradation and osteoblast differentiation, we analyzed changes in endogenous Smad1 protein after C2C12 and 2T3 cells were transfected with mutant Smurf1 or treated with proteasome inhibitors. We found that proteasome inhibitors increase endogenous Smad1 levels 2– 6-fold (Fig. 3d). Transfection of ΔSmurf1 causes a 5-fold increase in protein levels of endogenous Smad1. Overexpression of Smad1 (6-fold higher than endogenous Smad1) significantly enhanced ALP activity and osteocalcin production in C2C12 cells (Fig. 5, a and b). These results suggest that enhanced Smad1 levels by ΔSmurf1 or proteasome inhibitors promote osteoblast differentiation. Proteasome Inhibitors Induce Osteoblast Differentiation and Bone Formation—To further determine the role of proteasome inhibitors in osteoblast function and bone formation, we examined the effects of PS1, a synthetic proteasome inhibitor, on ALP activity and the BMP signaling reporter in C2C12 cells and on periosteal bone formation in rodents in vivo. We found that in C2C12 cells, treatment with PS1 increased ALP activity (Fig. 6a) and luciferase activity of the 12×SBE-OC-Luc reporter (Fig. 6b) in a dose-dependent manner. Increasing concentrations of PS1 (0.2, 1, and 5 mg/kg/day) were injected daily for 5 days into subcutaneous tissues over the calvariae of mice. The mice were sacrificed 2 weeks after the injections, and calvariae were processed for histology. PS1 induced significant new bone formation (Fig. 6, c and d). Similar effects were also obtained when epoxomicin, a naturally occurring proteasome inhibitor, was administered in vivo (data not shown). Moreover, these proteasome inhibitors administered systemically stimulated bone formation in intact and ovariectomized mice (16Garrett I.R. Chen D. Gutierrez G. Zhao M. Harris S.E. Rossini G. Escobedo A. Kim K.B. Hu S. Crews C.M. Mundy G.R. J. Clin. Invest. 2003; 111: 1771-1782Crossref PubMed Scopus (281) Google Scholar). These results demonstrate that inhibition of Smurf1 and proteasome degradation can lead to increased osteoblast function and suggest that part of the mechanism of proteasome inhibitor on new bone formation may be due to regulation of intracellular protein levels of Cbfa1 and Smad1. We have demonstrated for the first time that 1) E3 ubiquitin ligase Smurf1 mediates Cbfa1 degradation through the ubiquitin-proteasome pathway, 2) Cbfa1 is required for BMP-2 signaling, and 3) expression of mutant Smurf1 stimulates osteoblast differentiation. These findings suggest that Smurf1 plays an important role in osteoblast differentiation by regulating Cbfa1 and Smad1 function. Cbfa1 is a critical transcription factor in osteoblast differentiation and function (2Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Abstract Full Text Full Text PDF PubMed Scopus (3647) Google Scholar). BMPs have been shown to stimulate osteoblast differentiation in vitro and in vivo (1Wozney J.M. Mol. Reprod. Dev. 1992; 32: 160-167Crossref PubMed Scopus (613) Google Scholar), in part by up-regulating Cbfa1 expression (2Ducy P. Zhang R. Geoffroy V. Ridall A.L. Karsenty G. Cell. 1997; 89: 747-754Abstract Full Text Full Text PDF PubMed Scopus (3647) Google Scholar, 6Chen D. Ji X. Harris M.A. Feng J.-Q. Karsenty G. Celeste A.J. Rosen V. Mundy G.R. Harris S.E. J. Cell Biol. 1998; 142: 295-305Crossref PubMed Scopus (341) Google Scholar) and/or through direct interaction with Smad1 and Cbfa1 (7Hanai J.I. Chen L.F. Kanno T. Ohtani-Fujita N. Kim W.Y. Guo W.-H. Imamura T. Ishidou Y. Fukuchi M. Shi M.J. Stavnezer J. Kawabata M. Miyazono K. Ito Y. J. Biol. Chem. 1999; 274: 31577-31582Abstract Full Text Full Text PDF PubMed Scopus (410) Google Scholar). Smad1 is a downstream mediator of BMP receptors and plays a central role in BMP receptor signaling (17Hoodless P.A. Haerry T. Abdollah S. Stapleton M. O'Connor M.B. Attisano L. Wrana J.L. Cell. 1996; 85: 489-500Abstract Full Text Full Text PDF PubMed Scopus (626) Google Scholar). In the present studies, we found that Cbfa1 is required to activate BMP-2 signaling. It has been reported (18Leboy P. Grasso-Knight G. D'Angelo M. Volk S.W. Lian J.V. Drissi H. Stein G.S. Adams S.L. J. Bone Jt. Surg. Am. Vol. 2001; 83: S15-S22PubMed Google Scholar, 19Chikazu D. Li X. Kawaguchi H. Sakuma Y. Voznesensky O.S. Adams D.J. Xu M. Hoshi K. Katavic V. Herschman H.R. Raisz L.G. Pilbeam C.C. J. Bone Miner. Res. 2002; 17: 1430-1440Crossref PubMed Scopus (100) Google Scholar) that Smad1 and Cbfa1 binding sites co-localize in the promoters of several bone-related genes. Taken together with our results, this suggests that Smad1 and Cbfa1 co-activate downstream target genes in osteoblasts. Expression of mutant Smurf1 inhibits forskolin-induced Cbfa1 degradation in C2C12 cells, suggesting that Smurf1 may be involved in Cbfa1 degradation induced by a cAMP-dependent protein kinase pathway. Parathyroid hormone (PTH) induces proteasomal degradation of protein substrates in osteoblasts (20Murray E.J.B. Bentley G.V. Grisanti M.S. Murray S.S. Exp. Cell Res. 1998; 242: 460-469Crossref PubMed Scopus (38) Google Scholar), and it has recently been reported that the anabolic effect of PTH requires Cbfa1-dependent signaling (21Krishnan V. Moore T.L. Ma Y.L. Helvering L.M. Frolik C.A. Valasek K.M. Ducy P. Geiser A.G. Mol. Endocrinol. 2003; 17: 423-435Crossref PubMed Scopus (148) Google Scholar). We speculate that PTH may induce Cbfa1 degradation via cAMP-dependent signaling and that Smurf1 is involved in this process. However, this remains to be determined. To maintain efficient signal transduction and gene activation, cells must precisely regulate levels of signaling proteins and transcription factors during cellular processes. Under physiological conditions, Smurf1 may mediate degradation of the Cbfa1-Smad1 protein complex, but further investigation is required for unequivocal proof. Because Smurf1 mediates degradation of two critical proteins in the BMP signaling pathway and overexpression of Smurf1 modulates osteoblast differentiation and function, we propose that Smurf1 plays an important role in bone formation in vivo. It has been shown that Smurf mutations in Drosophila lead to enhanced decapentaplegic (a homologue of BMP-2/4) signaling and downstream target gene expression (22Podos S.D. Hanson K.K. Wang Y. Ferguson E.L. Dev. Cell. 2001; 1: 567-578Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar). Our findings suggest that Smurf1 may be a critical protein regulating Cbfa1 and Smad1 functions and, ultimately, osteoblast differentiation and bone formation. We sincerely thank our colleagues, Dr. Jeffrey Wrana (Program in Molecular Biology and Cancer, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Canada) for the constitutively active HA-BMPR-IB, Myc-Smurf1, and Smad1 expression vectors, Dr. Gerald Thomsen (Department of Biochemistry and Cell Biology and Institute for Cell and Developmental Biology, State University of New York, Stony Brook, NY) for the wild-type and mutant FLAG-Smurf1 expression vectors, Dr. Xu Cao (Department of Pathology, University of Alabama at Birmingham, AL) for the HA-ubiquitin expression vector, and Dr. Tongwen Wang (Department of Immunology, Virginia Mason Research Center, Seattle, WA) for the Itch expression vector.