Parathyroid Hormone-responsive Smad3-related Factor, Tmem119, Promotes Osteoblast Differentiation and Interacts with the Bone Morphogenetic Protein-Runx2 Pathway*

The mechanisms whereby the parathyroid hormone (PTH) exerts its anabolic action on bone are incompletely understood. We previously showed that inhibition of ERK1/2 enhanced Smad3-induced bone anabolic action in osteoblasts. These findings suggested the hypothesis that changes in gene expression associated with the altered Smad3-induced signaling brought about by an ERK1/2 inhibitor would identify novel bone anabolic factors in osteoblasts. We therefore performed a comparative DNA microarray analysis between empty vector-transfected mouse osteoblastic MC3T3-E1 cells and PD98059-treated stable Smad3-overexpressing MC3T3-E1 cells. Among the novel factors, Tmem119 was selected on the basis of its rapid induction by PTH independent of later increases in endogenous TGF-β. The levels of Tmem119 increased with time in cultures of MC3T3-E1 cells and mouse mesenchymal ST-2 cells committed to the osteoblast lineage by BMP-2. PTH stimulated Tmem119 levels within 1 h as determined by Western blot analysis and immunocytochemistry in MC3T3-E1 cells. MC3T3-E1 cells stably overexpressing Tmem119 exhibited elevated levels of Runx2, osteocalcin, alkaline phosphatase, and β-catenin, whereas Tmem119 augmented BMP-2-induced Runx2 levels in mesenchymal cells. Tmem119 interacted with Runx2, Smad1, and Smad5 in C2C12 cells. In conclusion, we identified a Smad3-related factor, Tmem119, that is induced by PTH and promotes differentiation in mouse osteoblastic cells. Tmem119 is an important molecule in the pathway downstream of PTH and Smad3 signaling in osteoblasts.

The mechanisms whereby the parathyroid hormone (PTH) exerts its anabolic action on bone are incompletely understood. We previously showed that inhibition of ERK1/2 enhanced Smad3-induced bone anabolic action in osteoblasts. These findings suggested the hypothesis that changes in gene expression associated with the altered Smad3-induced signaling brought about by an ERK1/2 inhibitor would identify novel bone anabolic factors in osteoblasts. We therefore performed a comparative DNA microarray analysis between empty vector-transfected mouse osteoblastic MC3T3-E1 cells and PD98059treated stable Smad3-overexpressing MC3T3-E1 cells. Among the novel factors, Tmem119 was selected on the basis of its rapid induction by PTH independent of later increases in endogenous TGF-␤. The levels of Tmem119 increased with time in cultures of MC3T3-E1 cells and mouse mesenchymal ST-2 cells committed to the osteoblast lineage by BMP-2. PTH stimulated Tmem119 levels within 1 h as determined by Western blot analysis and immunocytochemistry in MC3T3-E1 cells. MC3T3-E1 cells stably overexpressing Tmem119 exhibited elevated levels of Runx2, osteocalcin, alkaline phosphatase, and ␤-catenin, whereas Tmem119 augmented BMP-2-induced Runx2 levels in mesenchymal cells. Tmem119 interacted with Runx2, Smad1, and Smad5 in C2C12 cells. In conclusion, we identified a Smad3-related factor, Tmem119, that is induced by PTH and promotes differentiation in mouse osteoblastic cells. Tmem119 is an important molecule in the pathway downstream of PTH and Smad3 signaling in osteoblasts.
Bone remodeling is controlled by bone resorption and subsequent osteoblastic bone formation. Insights into the mechanisms of osteoclastic bone resorption have led to the development of reagents that potently suppress bone resorption for the treatment of osteoporosis. On the other hand, the impairment of osteoblastic bone formation is considered to be the main driving force in age-related, glucocorticoid-induced, and diabetes-related osteoporosis. However, many aspects of osteoblastic bone formation remain to be understood. Intermittent parathyroid hormone (PTH) 3 is one of the most potent anabolic agents; it stimulates de novo osteoblastic bone formation and is superior to bisphosphonates in the treatment of osteoporosis (1,2). The anabolic action of PTH is exerted partly through local growth factors and transcriptional regulators as well by as an anti-apoptotic action in osteoblasts (3,4). However, the precise mechanisms by which PTH exerts its anabolic action on bone are incompletely understood.
We showed previously that Smad3, a crucial TGF-␤-signaling molecule, promotes alkaline phosphatase (ALP) activity in mouse osteoblastic MC3T3-E1 cells (5,6), and Borton et al. (7) have found that mice with targeted disruption of Smad3 exhibit osteopenia caused by decreased bone formation, suggesting that the Smad3 molecule is a promoter of bone formation. Moreover, we demonstrated that TGF-␤-responsive ERK1/2 and c-Jun N-terminal kinase (JNK) cascades negatively regulate Smad3-induced transcriptional activity and ALP activity in osteoblasts (8). In that study, an inhibition of ERK1/2 enhanced Smad3-induced ALP activity in MC3T3-E1 cells.
On the other hand, we demonstrated that PTH rapidly enhances Smad3 expression in osteoblasts and that the PTH-Smad3 axis exerts anti-apoptotic effects in osteoblasts (9). Moreover, we showed that PTH stimulates osteoblast ␤-catenin levels via Smad3 (10,11). In that study, PTH * This work was supported by a grant from Mitsui Life Social Welfare Foundation (to H. K.), Grants-in-aid 17590961 and 21591179 from the Ministry of Science, Education, and Culture of Japan (to H. K.), the Global COE Program F11 from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and Canadian Institutes of Health Research Grant MOP-9315 (to G. N. H.). 1  enhanced ␤-catenin levels at early times independently of endogenous TGF-␤. These findings raised the possibility that further examination of transcriptional changes occurring as a result of the Smad3 signaling rapidly induced by PTH and enhanced by inhibition of ERK1/2 activity might reveal novel bone anabolic factors.
In the present study, we performed a transcriptome evaluation by DNA microarray analysis of untreated vector alone-transfected (control) and ERK1/2 inhibitor-treated stable Smad3-overexpressing mouse osteoblastic MC3T3-E1 cells and identified a novel bone formation-related factor, Tmem119.
Tmem119 Antibody and cDNA-A hexadecapeptide, NH 2 -FSEVPDRAPDSRHEEC-COOH, was synthesized (by solidphase chemistry at the Peptide Synthesis Facility of the Sheldon Biotechnology Centre, McGill University) corresponding to amino acids 145-159 of mouse Tmem119 with an additional cysteine residue at the COOH terminus. The peptide was coupled through the cysteine residue to keyhole limpet hemocyanin, and a rabbit polyclonal antibody was raised by immunization with the conjugate. The antiserum was immunoaffinitypurified before use as described previously (12).
The coding region of mouse Tmem119 was amplified by reverse transcription-PCR of total RNA from MC3T3-E1 cells using 5Ј-ATAGCTCAACATGGTCCCCTGGTTC-3Ј as the forward primer and 5Ј-GGCCTGTTAGACACTGGGGGA-GAC-3Ј as the reverse primer. The cDNA was TA-cloned into the pCR2.1 vector (Invitrogen) according to the manufacturer's specifications. The Tmem119 cDNA insert was cloned into the mammalian expression vector, pcDNA3.1(Ϫ) (Invitrogen). The correctness of all constructs was verified by restriction enzyme analysis and nucleotide sequencing.
Transient or Stable Transfection-Each vector was transfected into MC3T3-E1, ST-2, COS-7, or C2C12 cells with Lipofectamine (Invitrogen) as described previously (13)(14)(15). Six h later, the cells were fed fresh ␣-MEM containing 10% FBS. Forty-eight h later, the transiently transfected cells were used for experiments. To generate stably transfected cells, after incubation in ␣-MEM containing 10% FBS for 48 h, the cells were passaged, and clones were selected in ␣-MEM supplemented with G418 (0.3 mg/ml; Invitrogen) and 10% FBS. Twenty-four clones were picked after 3 weeks of culture in G418. Several clones were selected after Western blotting with anti-Tmem119 antibody and semiquantitative RT-PCR for Tmem119. To rule out the possibility of clonal variation, at least three independent clones for each stable transfection were characterized. Empty vector-transfected cells were used as control.
Protein Extraction and Western Blot Analysis-Cells were lysed with radioimmunoprecipitation buffer containing 0.5 mM phenylmethylsulfonyl fluoride, complete protease inhibitor mixture (Roche Applied Science), 1% Triton X-100, and 1 mM sodium orthovanadate. Cell lysates were centrifuged at 12,000 ϫ g for 20 min at 4°C, and the supernatants were stored at Ϫ80°C. Protein quantitation was performed with bicinchoninic acid protein assay reagent (Pierce). Twenty-microgram protein aliquots were denatured in SDS sample buffer and separated on 10% polyacrylamide-SDS gels. Proteins were transferred in 25 mM Tris, 192 mM glycine, and 20% methanol to polyvinylidene difluoride. Blots were blocked with 20 mM Tris-HCl (pH 7.5) and 137 mM NaCl plus 0.1% Tween 20 containing 3% dried milk powder. The membranes were immunoblotted with each primary antibody. The antigen-antibody complexes were visualized using the appropriate secondary antibodies (Sigma-Aldrich) and the enhanced chemiluminescence detection system as recommended by the manufacturer (Amersham Biosciences). The results depicted in Figs. 1-8 are representative of at least three separate cell preparations. Each experiment was repeated three times.
Small Interfering RNA-Mouse Tmem119 siRNA, mouse Smad3 siRNA, or control siRNA (each from Santa Cruz Biotechnology) were transfected by the procedures recommended by the supplier into cells that were seeded at 5 ϫ 10 5 cells/well using Lipofectamine TM RNAiMAX (Invitrogen).
Microarray Analysis-Total RNA was extracted from stable empty vector (untreated)-or Smad3-transfected, 10 M PD98059-treated (24 h)-confluent MC3T3-E1 cells using TRIzol reagent. Total RNA was purified using an RNeasy MinElute Cleanup Kit (Qiagen, Tokyo, Japan) to yield an A 260/280 Ͼ 1.90. Double-stranded cDNA was synthesized using a T7-oligo(dT) primer with the 3Ј IVT Express Kit (Affymetrix Inc., Santa Clara, CA). Hybridization samples were prepared and processed according to the GeneChip Expression Analysis Technical Manual. The Affymetrix murine genome 430 2.0 set was used to compare gene expression. Data were analyzed using GeneChip operating software, version 1.4 (Affymetrix 690036), according to GeneChip Expression Analysis, Data Analysis Fundamentals, part 701190, rev. 4. Using a DNA microarray viewer (Kurabo, Osaka, Japan), the -fold changes in expression between two groups were calculated, log 2-transformed, and further classified as not changed, increased (signal log ratio change p value Ͻ 0.005), decreased (signal log ratio change p value Ͻ 0.005), or marginally increased or decreased.
Co-immunoprecipitation-For co-immunoprecipitation experiments, cells were lysed with a buffer containing 1% Triton X-100, 1% deoxycholate, 50 mM Tris-HCl (pH 7.4), 100 mM NaCl, 25 mM sodium fluoride, 10 mM sodium pyrophosphate, 2 mM sodium orthovanadate, 1.5 mM MgCl 2 , and 2 mM EGTA plus protease inhibitor mixture for 30 min at 4°C; then insoluble materials were separated by centrifugation at 4°C for 30 min at 15,000 ϫ g. An aliquot of the supernatant containing 1 mg of protein was clarified and incubated with anti-FLAG or anti-Runx2 antibody on a rocking platform at 4°C overnight. The immune complexes were collected with protein G PLUS/ protein A-agarose beads (Calbiochem) for 30 min at 4°C. The beads were washed four times with lysis buffer, resuspended in 2ϫ sample buffer, and boiled for 5 min. Immunoprecipitated proteins were analyzed by SDS-PAGE and subjected to Western blot analysis using anti-Tmem119 antibody as described above.
Luciferase Assay-Cells were seeded at a density of 2 ϫ 10 5 / 6-well plate. Twenty-four h later, cells were transfected with 3 g of reporter plasmid (3GC2 lux or OSC-luci) with the pCH110 plasmid expressing ␤-galactosidase (1 g) in each well using Lipofectamine (Invitrogen) as described (15). Fifteen h later, the medium was changed to one containing 4% FBS, and the cells were incubated for an additional 9 h. Thereafter, cells were cultured for 24 h in the presence or absence of 200 ng/ml BMP-2. Cells were lysed, and luciferase activity was measured and normalized to the relative ␤-galactosidase activity as described (15).
Avidin-biotin-peroxidase (ABC) Complex and Fluorescence Immunocytochemistry-For ABC immunocytochemistry, cells in chamber slides were fixed in 4% paraformaldehyde for 10 min. The slides were incubated in a mixture of 3% H 2 O 2 and methanol for 10 min at room temperature to block endogenous peroxidase activity. After treatment with blocking antibody, the primary antibodies were added, and the slides were incubated for 60 min at room temperature. Non-immunized mouse serum was used for preparing negative controls. The ABC method was used with a MaxiTags TM kit (ThermoShandon, Pittsburgh, PA). Final development of the sections was carried out with 3,3Ј-diaminobenzidine containing 0.03% H 2 O 2 . For fluorescence immunocytochemistry, after treatment with blocking antibody, the primary antibodies were added and the slides incubated for 60 min at room temperature. The samples were incubated with Cy2-conjugated anti-rabbit IgG (HϩL) MARCH 18, 2011 • VOLUME 286 • NUMBER 11 (Jackson ImmunoResearch, West Grove, PA) for 30 min at room temperature and counterstained with 4Ј,6-diamidine-2Јphenylindole dihydrochloride (Roche Applied Science).

Tmem119 Induces Osteoblast Differentiation
Statistics-Data are expressed as mean Ϯ S.E. Statistical analysis was performed using analysis of variance. A p value Ͻ 0.05 was taken to indicate significant difference.

RESULTS
Transcriptional Profiling-We had demonstrated previously that the PTH-Smad3 axis is involved in the bone anabolic action of PTH in osteoblastic cells including MC3T3-E1 cells (9). Moreover, inhibition of ERK1/2 by PD98059 enhances Smad3-induced bone anabolic action in these cells (8). These findings suggested the hypothesis that examination of the altered Smad3-induced signaling brought about by inhibition of ERK1/2 could identify novel bone anabolic factors in osteoblasts. We therefore performed comparative transcriptional profiling between empty vector stably transfected cells (control) and PD98059-treated Smad3 stably overexpressing MC3T3-E1 cells (experimental) using cDNA microarray analysis. From the ϳ45,000 genes on the Affymetrix GeneChip, we selected those genes in which expression was increased at least four times in the experimental versus the control group. Several known bone-related factors such as ALP, the homeodomain protein Dlx3, fritz b (fzb), insulin-like growth factor-binding protein-5 (IGFBP-5), Phospho1, PTH receptor type 1 (PTHR1), OASIS, Osterix, and tenascin-C were identified (Table 1). In addition, two factors of unknown function (Tmem119 and NI) were identified.
Selection of Tmem119-We selected Tmem119 to study further because our previous study had suggested that such a molecule, induced early by PTH but independently of endogenous TGF-␤, might be related to the bone anabolic action of PTH (9). Levels of Tmem119 and OASIS mRNA were increased in stably overexpressing Smad3 relative to vector alone-transfected MC3T3-E1 cells by semiquantitative RT-PCR, although the levels of NI mRNA were equivalent in the two types of cells (Fig.  1A). The levels of Tmem119, OASIS, and NI RNA were not different in vector alone-expressing cells treated without or with the ERK1/2 inhibitor PD98059. However, the levels of Tmem119, OASIS, and NI mRNA were induced further by treatment of the stably overexpressing Smad3 MC3T3-E1 cells with PD98059 relative to the untreated cells (Fig. 1A). Treatment with SB431542, an inhibitor of endogenous TGF-␤ signaling (activin-like kinase 5 inhibitor), did not alter Smad3 induction of Tmem119 mRNA levels, although it inhibited Smad3 enhancement of OASIS mRNA levels. These findings indicated that of these three factors identified by the microarray analysis only Tmem119 is induced by Smad3 independently of endogenous TGF-␤.
PTH elevated the levels of Tmem119 mRNA after both short term (1 h) and long term (24 h) stimulation (Fig. 1B). Although the levels of OASIS and NI mRNA were elevated after 24 h of PTH stimulation, at 1 h the levels were unaltered from time 0 (data not shown). Moreover, SB431542 did not antagonize the early induction (at 1 h) of Tmem119 mRNA by PTH, although it antagonized the later induction (at 24 h) of Tmem119 mRNA by PTH (Fig. 1B). On the basis of these findings, we selected Tmem119 for further study as a putative bone anabolic factor.
Anti-Tmem119 Antibody and the Expression of Tmem119 in Osteoblastic, Mesenchymal, and Kidney Cells-A rabbit polyclonal antibody against Tmem119 was raised to characterize the expression and function of Tmem119 in osteoblasts and  other cells (Fig. 2A). By Western blot analysis of the antibody detected the expression of Tmem119 that migrated as a protein of ϳ30 kDa in MC3T3-E1 cells (Fig. 2B). Transient overexpression of Tmem119 elevated the levels of this molecule in these cells. Tmem119 was detected in mouse mesenchymal C2C12 cells, and transient overexpression of Tmem119 elevated the signal (Fig. 2C). The addition of BMP-2 increased the level of Tmem119 in these cells (Fig. 2C). Moreover, Tmem119 was detected in African green monkey kidney (COS-7) cells, and transient Tmem119 overexpression elevated the level (Fig. 2D).
The levels of Tmem119 protein were elevated in MC3T3-E1 cells stably overexpressing Smad3, and PD98059 enhanced these protein levels in MC3T3-E1 cells (Fig. 2E), showing that the ERK1/2 inhibitor enhances Smad3 induction of Tmem119 at the protein as well as the mRNA level. BMP-2 did not induce the phosphorylation of Smad3 in ST-2 and C2C12 cells in our preliminary study (data not shown).
Tmem119 Expression during Osteoblast Differentiation-MC3T3-E1 cells in culture undergo osteoblastic differentiation with mineralization starting after 2-3 weeks. As we reported previously (14), ALP activity increases with time, and the expression of type 1 collagen and osteopontin peaks at 7 and 14 days of culture, respectively, whereas mRNA expression of OCN, a terminal differentiation marker, is highest at day 21 of culture. The levels of Tmem119 mRNA increased for up to 2 weeks, and the level was maintained up to week 3 in MC3T3-E1 cells (Fig. 3A). The effect of PD98059 in augmenting Smad3induced Tmem119 mRNA expression became less obvious as the culture period progressed and disappeared entirely at day 21. At the protein level, the level of Tmem119 continued to increase during 21 days of culture of the MC3T3-E1 cells (Fig.  3B). Mesenchymal ST2 cells differentiate into osteoblastic cells when stimulated with BMP-2, and in cells cultured in this way, the level of Tmem119 increased with time (Fig. 3C). Tmem119 levels did not increase in the absence of BMP-2 (data not shown). Therefore, the expression level of Tmem119 correlates with the differentiation stage.
Induction of Tmem119 by PTH-We next examined whether PTH and other osteoblast regulators would affect Tmem119 protein levels in MC3T3-E1 cells. PTH induced the level of  Tmem119 protein within 1 h (Fig. 4A). The effect was dose-dependent, and a significant increase was observed at 10 Ϫ9 M PTH (Fig. 4B). The PKA activators forskolin and Bt 2 cAMP induced the levels of Tmem119, but the PKC activator PMA had no effect (Fig. 4C). The PKA inhibitor H89 antagonized the PTH stimulation of Tmem119 in these cells (Fig. 4D). Inactivation of Smad3 by the dominant negative Smad3⌬C antagonized PTHinduced Tmem119 levels (Fig. 4E), indicating that PTH elevates the level of Tmem119 through Smad3 in osteoblasts. Reduction of Tmem119 by siRNA antagonized the level of ␤-catenin induced by PTH at 1 h in MC3T3-E1 cells (Fig. 4F). Moreover, actinomycin D, a transcription inhibitor, antagonized the PTHinduced Tmem119 level at 1 h, although cycloheximide had no effect (data not shown), indicating that PTH rapidly elevates the level of Tmem119 through a transcriptional mechanism but not through new protein synthesis. Neither the active vitamin D metabolite 1,25(OH) 2 D 3 nor the glucocorticoid dexamethasone affected Tmem119 levels under short term (1 h) or longer term (24 h) stimulation (Fig. 4G). In contrast to its effect in MC3T3-E1 cells, PTH did not alter Tmem119 protein levels in C2C12 cells (Fig. 4H). PTHrP elevated the level of Tmem119 in MC3T3-E1 cells (Fig. 4I). These findings suggest that the induction of Tmem119 by PTH is specific for osteoblasts. PTH as well as forskolin and Bt 2 cAMP elevated the levels of Tmem119 (Fig.  5, A and B), and H89 antagonized the PTH stimulation of Tmem119 in mouse primary calvarial cell cultures (Fig. 5C). Reduction of Smad3 by siRNA antagonized the level of Tmem119 induced by PTH at 1 h in these cells (Fig. 5D).
Immunocytochemistry of Tmem119 Expression-In MC3T3-E1 cell lines stably overexpressing Tmem119, the increased levels of Tmem119 were evident by Western blot analysis (Fig.  6A). Next, the cellular expression of Tmem119 was examined in MC3T3-E1 cells by ABC immunocytochemistry using the same antibody. Increased cellular levels of Tmem119 were evident in the stably Tmem119-overexpressing cell lines relative to vector alone-transfected cell lines when examined by immunocytochemistry (Fig. 6B). The low level of expression of Tmem119 in unstimulated MC3T3-E1 cells (control) was increased by PTH stimulation at 1 and 24 h (Fig. 6C). We performed fluorescence immunocytochemistry in Tmem119 stably overexpressing MC3T3-E1 cells. As shown in Fig. 6D, Tmem119 was clearly detected mainly in the cytoplasm. The perinuclear granular

Tmem119 Induces Osteoblast Differentiation
staining is consistent with location in the endoplasmic reticulum.
Tmem119 Promotes Osteoblast Differentiation-We examined the relationship between Tmem119 expression and osteoblast differentiation in MC3T3-E1 cells stably overexpressing Tmem119 versus vector alone-expressing MC3T3-E1 cells. The mRNA and protein levels of the key osteoblast differentiation markers Runx2 and OCN were elevated in the Tmem119-overexpressing cells (Fig. 7, A and B). The protein levels of ALP, an osteoblast differentiation marker (Fig. 7B), and ␤-catenin, a critical regulator of the osteoblast (Fig. 7C), were also elevated in the Tmem119-overexpressing cells. In our preliminary study, the enhancement of Tmem119 overexpression on osteoblast differentiation markers such as ALP and Runx2 levels increased from culture day 7 until day 14 and then seemed to decrease at day 21 in MC3T3-E1 cells (data not shown). A reduction in Tmem119 level by siRNA suppressed the level of ALP, Runx2, Osterix, and ␤-catenin in MC3T3-E1 cells (Fig. 7D).
In addition, we evaluated the role of Tmem119 in mesenchymal cells using transient overexpression of Tmem119 in ST-2   MARCH 18, 2011 • VOLUME 286 • NUMBER 11

JOURNAL OF BIOLOGICAL CHEMISTRY 9793
and C2C12 cells. The BMP-2-induced Runx2 mRNA levels were higher in the Tmem119 versus vector alone-transfected ST-2 cells (Fig. 7E). Moreover, the basal levels and the BMP-2stimulated levels of Runx2 protein were increased in Tmem119 relative to vector alone, transiently transfected C2C12 cells (Fig.  7F). A reduction in Tmem119 by siRNA antagonized the level of OCN mRNA induced by BMP-2 in C2C12 cells (Fig. 7G). In our preliminary data, stable overexpression of Tmem119 did not affect the levels of GPP-1, ANK, and osteopontin mRNA in semiquantitative-RT-PCR in MC3T3-E1 cells. Moreover, Tmem119 overexpression did not affect the levels of DMP-1 (data not shown). However, it suppressed the level of sclerostin in these cells (Fig. 7H).
Interaction between Tmem119 and Smad1/5 or Runx2-Finally, we examined the physical interactions between Tmem119 and BMP-specific Smads, Smad1 and Smad5, and Runx2 in C2C12 and MC3T3-E1 cells. Tmem119 coimmunoprecipitated with Smads 1 and 5 and Runx2 in C2C12 cells (Fig. 8, A-C). PTH rapidly enhanced the coimmunoprecipitation of Tmem119 and Smad5 or Runx2 in MC3T3-E1 cells (Fig. 8D). Reduction of Tmem119 by siRNA significantly antagonized Runx2-induced transcriptional activity as well as BMP-2and Smad5-induced transcriptional activity in C2C12 cells (Fig. 8, E and F). All data are from experiments repeated three times with reproducible results.

DISCUSSION
Multipotential mesenchymal stem cells, derived from mesoderm, become osteoprogenitors and commit to the osteoblast lineage. Thereafter, the cells undergo further differentiation, exhibiting an osteoblast-specific phenotype, and express ALP and OCN. Concomitant with the osteoblast maturation process, extracellular matrix is mineralized or calcified, and bone is formed. These sequential events are regulated by hormones such as PTH and cytokines such as TGF-␤. The intracellular signaling pathways triggered by these factors cross-talk to each other forming a complex network. These signaling pathways drive the activities of transcriptional factors such as Runx2, which are critical for osteoblast differentiation (16 -20).
Several known important bone-related factors were identified by the gene profiling in osteoblastic MC3T3-E1 cells that specifically searched for genes that are up-regulated by increased Smad3 expression in the presence of an ERK1/2 inhibitor. Among them, Dlx3, Fzb, IGFBP-5, Phospho1, PTHR1, OASIS, Osterix, and tenascin-C can be noted. Dlx3 is a homeodomain protein induced by BMP-2 that activates Runx2 gene transcription and is critical for the commitment of mesenchymal stem cells to the osteogenic lineage (21). The frizzledrelated protein, Fzb, modulates Wnt-␤-catenin signaling, important for bone formation (22). IGFBP-5 is related to osteoblast proliferation and differentiation, both dependently on and independently of IGF-1 action (23). OASIS, a member of the CREB/ATF transcription factor family with a transmembrane domain, is highly expressed in osteoblasts (24) and undergoes regulated intramembrane proteolysis in response to endoplasmic reticulum stress inducing Col1a1 gene transcription (25). Osterix is essential for osteoblastic bone formation downstream of Runx2 (26). Phospho1 is involved in mineralization (27), and tenascin-C is a bone matrix protein (28). Therefore, these important factors may function downstream of Smad3 in osteoblastic bone formation.
The Tmem119 transcript identified here as encoding an important bone-related factor came to attention previously as a differentially regulated transcript (of unknown function) in a comparative study of the expression profiles of E14.5 wild-type and Runx2 Ϫ/Ϫ mouse embryonal humerii assessed by microarray analysis (29). Expression of Tmem119 was detected in the perichondrium, the trabecular bone, and the stratum osteogenicum in the periosteum but not in chondrocytes and osteoclasts. The transcript levels are increased in differentiated calvarial cultures (day 17) with increased expression in cells sorted for Col1a1 promoter activity (30) (NCBI DataSet record GDS1632). . D, subconfluent MC3T3-E1 were cultured with or without PTH-(1-34) for 1 h. Cell extracts were made and immunoprecipitated with anti-Smad5 or Runx2 antibodies followed by immunoblotting with anti-Tmem119 antibody. E, the Runx2 promoter-reporter construct, OSC-luci, was transfected with vector or Runx2 into C2C12 cells that had been transfected 24 h earlier with either control (Cont.) or Tmem119 siRNA. Relative luciferase activity was measured as described under "Experimental Procedures." Each value is the mean Ϯ S.E. of four determinations. *, p Ͻ 0.01, compared with vector only-transfected group; **, p Ͻ 0.01, compared with Runx2-and control siRNA-transfected group. F, C2C12 cells were transfected with either control or Tmem119 siRNA; 24 h later they were transfected with the Smad1/5 reporter construct (3GC2-lux), with vector only, or with Smad5 and treated (ϩ) or not (Ϫ) with 200 ng/ml BMP-2. Relative luciferase activity was measured as described under "Experimental Procedures." Each value is the mean Ϯ S.E. of four determinations. *, p Ͻ 0.01, compared with the vector only-transfected group. **, p Ͻ 0.01, compared with the Smad5-and control siRNA-transfected group.
Mouse Tmem119 has 280 amino acids (molecular mass, 29,401 Da) and human Tmem119 has 283 amino acids (29,072 Da). The mouse and human proteins share 75% similarity and 70% identity and are encoded by genes on syntenic loci at mouse chromosome 5 and human 12q24.11, respectively. Tmem119 is predicted to be a single-pass type 1a membrane protein with amino acids 113-280 (in the mouse) comprising a cytoplasmic tail. Consistent with its being a type 1a membrane protein (and hence having a cleavable signal peptide), Tmem119 is predicted to have a signal peptide (which directs the nascent polypeptide to the endoplasmic reticulum) of ϳ18 amino acids (PSORT and SignalP) with residues 90 -112 (TMHMM method) comprising the transmembrane helix. By expressed sequence tag (EST) profiling, Tmem119 mRNA is expressed in several tissues including blood, bone, brain, intestine, kidney, liver, lung, skin, and uterus but apparently not in adipose, bladder, heart, lymph node, muscle, ovary, testis, pancreas, spleen, and spinal cord in mice (UGID: 270244; UniGene Mm.41681).
The present studies show that the expression levels of Tmem119 correlate with the differentiation (commitment) of mesenchymal stem cells into the osteoblast lineage and differentiation of preosteoblasts to osteoblasts with enhanced expression of the differentiation markers Runx2, ALP, OCN, and ␤-catenin. Although Runx2 is critical for the commitment of mesenchymal stem cells into the osteoblast lineage (16,(31)(32)(33), transgenic mice overexpressing Runx2 under the control of the osteoblast-specific Col1a1 promoter exhibit decreased bone mineralization and increased bone resorption (34). Although in the Runx2 osteoblast-overexpressing transgenic mice, the osteoblast number is increased at an early differentiation stage, osteoblast functions (matrix production and mineralization) were impaired in these cells (34). Thus, Runx2 inhibits later osteoblast differentiation, in contrast to its positive action on early osteoblast development, whereas Tmem119 induces both the early and late stages of osteoblast differentiation.
PTH specifically and dose-dependently induced Tmem119 in osteoblastic cells but not in mesenchymal stem cells. The response occurred via the PKA but not the PKC pathway within 1 h. (Other osteoblast regulators like 1,25(OH) 2 D 3 and glucocorticoids were without effect even after 24 h of stimulation.) The roles of the PKA-and PKC-signaling pathways in bone formation are controversial. Some studies have found that PTH exerts anti-apoptotic actions in osteoblasts through PKA signaling (35,36). Others have reported that intermittent treatment with PTH stimulates bone anabolism through both PKA and PKC pathways in osteoblasts (37). However, it has been reported that PKC activity inhibits osteoblast differentiation (38). In the present study, forskolin and Bt 2 cAMP induced the level of Tmem119 in osteoblasts, although PMA, a PKC activator, was without effect. Moreover, a PKA inhibitor antagonized the PTH-induced effects. Hence, these findings indicate that PTH induces Tmem119 through the cAMP pathway.
Our previous studies indicated that PTH rapidly elevates the level of ␤-catenin through Smad3 in osteoblasts (9 -11). In the present study, inactivation of Smad3 antagonized the induction of Tmem119 levels by PTH, and reduction of Tmem119 levels antagonized the induction of ␤-catenin by PTH in osteoblasts These findings indicate that PTH elevates Tmem119 levels through Smad3, resulting in an elevation in ␤-catenin, which might lead to bone anabolic effects.
By immunocytochemistry, Tmem119 was detected in MC3T3-E1 cells, and increased levels of expression were observed after stimulation of the cells for 1 or 24 h with PTH. Clear perinuclear staining was observed consistent with location in the endoplasmic reticulum membrane, with the COOHterminal part (which contains the epitope recognized by the antibody) being cytoplasmic. Further studies will be required to identify the precise cellular location of Tmem119 under different conditions. However, the putative function of Tmem119 might be surmised to be different from that of another transmembrane protein highly expressed in the osteoblast, namely OASIS. OASIS is an endoplasmic reticulum stress transducer. This type 2 membrane protein, without a cleavable signal peptide and oriented in the opposite direction to a type 1 membrane protein (in contrast to Tmem119), is a member of the CREB/ATF transcription factor family (24). In response to endoplasmic reticulum stress, OASIS trafficks to the Golgi where it undergoes intramembrane cleavage, and the released bZip motif-containing NH 2 -terminal part enters the nucleus inducing Col1a1 gene transcription (25). In contrast to OASIS, Tmem119 has no endoplasmic reticulum retention motifs, nuclear localization sequences, or transcriptional regulator motifs.
The BMP-Runx2 pathway is critical for bone formation (39,40). The transcription factors Runx2 and Smad1/5 that are involved in this pathway exist in the cytoplasm as well as the nucleus. We showed that Tmem119 is complexed with Smad1/5 and Runx2 in C2C12 cells. Moreover, a reduction in Tmem119 antagonized Runx2-or Smad5-mediated transcriptional activity. Also, the levels of Tmem119 increased with time in these pluripotent cells treated with BMP-2 and undergoing commitment to the osteoblast lineage. Increased levels of Tmem119 induced the expression of osteoblast differentiation markers. Together these findings suggest that Tmem119 induces the commitment of multipotential mesenchymal cells to become osteoblasts through a mechanism involving the BMP-2-Smad1/5-Runx2 cascade.
In the present study, stable Tmem119 overexpression reduced the level of sclerostin in MC3T3-E1 cells. A constitutive overexpression of Tmem119 at developmentally inappropriate time points might lead to an inhibition of osteocyte differentiation. Alternatively, a recent study has indicated that the suppression of sclerostin expression contributes to the bone anabolic action of intermittent PTH treatment in mice (41). Therefore, reduction of the sclerostin level by Tmem119 might be related to the bone anabolic action of PTH.
While this manuscript was in preparation, Kanamoto et al. (42) reported on Tmem119 as a predominantly expressed gene in osteoblasts, based on microarray screening using a mouse chondroprogenitor cell line, ATDC5. Consistent with our data, Tmem119 stimulated differentiation in osteoblasts (42). Moreover, overexpression and knockdown enhanced and suppressed mineralization in MC3T3-E1 cells, respectively (42).
Taken together, Tmem119 is an important osteoblast differentiation factor.
In conclusion, we have identified Tmem119 as a novel Smad3-related factor in osteoblastic cells. PTH induces Tmem119 through the cAMP-PKA pathway, and Tmem119 induces osteoblast differentiation. The precise mechanism whereby Tmem119 influences osteoblast biology remains to be elucidated.