Differentiation of osteoclasts, the cells primarily responsible for bone resorption, is controlled by a variety of osteotropic hormones and cytokines. Of these factors, receptor activator of NF-κB (RANK) ligand (RANKL) has been recently cloned as an essential inducer of osteoclastogenesis in the presence of M-CSF. Here, we isolated a stroma-free population of monocyte/macrophage (M/Mφ)-like hemopoietic cells from mouse unfractionated bone cells that were capable of differentiating into mature osteoclasts by treatment with soluble RANKL (sRANKL) and M-CSF. However, the efficiency of osteoclast formation was low, suggesting the requirement for additional factors. The isolated M/Mφ-like hemopoietic cells expressed TGF-β and type I and II receptors of TGF-β. Therefore, we examined the effect of TGF-β on osteoclastogenesis. TGF-β with a combination of sRANKL and M-CSF promoted the differentiation of nearly all M/Mφ-like hemopoietic cells into cells of the osteoclast lineage. Neutralizing anti-TGF-β Ab abrogated the osteoclast generation. These TGF-β effects were also observed in cultures of unfractionated bone cells, and anti-TGF-β blocked the stimulatory effect of 1,25-dihydroxyvitamin D3. Translocation of NF-κB into nuclei induced by sRANKL in TGF-β-pretreated M/Mφ-like hemopoietic cells was greater than that in untreated cells, whereas TGF-β did not up-regulate the expression of RANK, the receptor of RANKL. Our findings suggest that TGF-β is an essential autocrine factor for osteoclastogenesis.

Osteoclasts are the cells primarily responsible for bone resorption, and are of hemopoietic stem cell origin. Precursors of osteoclasts have been demonstrated to share common properties with those of a monocyte/macrophage (M/Mφ)3 cell lineage (1, 2). Recent extensive studies have increased our understanding of osteoclast biology, particularly osteoclastogenesis. Many systemic hormones and local cytokines pathophysiologically participate in regulating osteoclast differentiation, including M-CSF, IL-1, IL-6, IL-11, TNF-α, 1,25-dihydroxyvitamin D3 (1,25(OH)2D3), parathyroid hormone, and PGs (3, 4). Osteoclast differentiation factor/osteoprotegerin ligand/TNF-related activation-induced cytokine/receptor activator of NF-κB (RANK) ligand (RANKL) has recently been identified as the most important and critical molecule for osteoclast development (5, 6). Bone marrow stroma/osteoblasts produce this molecule on the plasma membrane in response to several osteotropic factors, and osteoclast precursors express a receptor of RANKL (RANK). Most recently, mice with a disrupted RANKL gene were found to have severe osteopetrosis (7). Therefore, the RANKL/RANK system is considered to be an essential signal for osteoclast differentiation in the interaction between stromal cells and cells of the osteoclast lineage. Like the interaction of osteoclast precursors/stromal cells, RANKL is expressed on activated T cells and activates mature dendritic cells that express RANK on their plasma membrane, implying a role for T cell-dendritic cell interaction during an immune response (7, 8, 9).

Soluble RANKL (sRANKL) lacking transmembrane and intracellular regions is now available and has allowed us to elucidate the role of RANKL in osteoclast development and function in more detail. Osteoclasts have been shown to be formed from spleen cells, nonadherent bone marrow cells, or peripheral blood-derived monocytes in the presence of M-CSF and sRANKL in the absence of stromal cells (5, 6, 10, 11). In addition, a macrophage-like cell line has been demonstrated to potentially differentiate into osteoclasts when treated with M-CSF and sRANKL (12). However, the efficiency for osteoclast formation was low in the above cultures, suggesting the requirement for other factors for osteoclastogenesis. Furthermore, due to the lack of a population of osteoclast progenitors that synchronously differentiate into osteoclasts, the molecular mechanisms regulating the process of osteoclastogenesis have remained uncertain.

In this study we developed a new isolation method for obtaining osteoclast progenitors. By this procedure, we isolated M/Mφ-like hemopoietic cells from mouse unfractionated bone cells; these isolated cells are potentially capable of differentiating into osteoclasts in response to M-CSF and sRANKL. Surprisingly, neutralizing Ab against TGF-β completely blocked osteoclast formation from the precursors induced by sRANKL/M-CSF signaling. In addition, exogenous TGF-β induced the further commitment and maturation of osteoclast progenitors into mature osteoclasts in the absence of stromal cells. In contrast, many studies using in vitro culture systems containing stromal cells have demonstrated that TGF-β inhibited the differentiation and function of osteoclasts (13, 14). Thus, TGF-β possesses multifunctional biological activities. Target cells of TGF-β are heterogeneous in bone, including bone-forming and -resorbing cells, hemopoietic cells, and bone marrow stromal cells (15, 16, 17). Therefore, it has been difficult to elucidate the precise and direct action of TGF-β on osteoclast development. Here we report that endogenous production of TGF-β by M/Mφ-like hemopoietic cells and the derived osteoclast precursors is essential for osteoclastogenesis induced by a combination of RANKL and M-CSF. Our findings expand the established roles of TGF-β in osteoclastogenesis and provide a novel insight into bone metabolism.

Neutralizing mAb (clone, 1D11) against TGF-β1, -2, and -3 and isotype control mouse IgG1 were obtained from R&D Systems (Minneapolis, MN). Polyclonal rabbit anti-RANK Ab was provided by Snow Brand Milk Products Co. Ltd. (Tochigi, Japan). Unlabeled anti-CD16/32 (clone 2.4G2) Ab, biotinylated anti-CD11b (Mac-1α-chain; clone M1/70), anti-CD11a (LFA-1 α-chain; clone 2D7), anti-CD44 (clone IM-7) and anti-CD61 (integrin β3; clone C9.G2) Abs and PE-labeled anti-CD14 (clone rmC5-3) Ab were purchased from PharMingen International (San Diego, CA). Biotinylated anti-F4/80 (clone A3-1), FITC-labeled anti-integrin αv, and unlabeled anti-DEC-205 (clone NLDC-145) were obtained from Serotec (Kidlington, U.K.), Sumitomo Electronic (Osaka, Japan), and BMA Biomedicals (Augst, Switzerland), respectively. Anti-p50 (sc-114X) and anti-p65 (sc-109X) Abs were purchased from Santa Cruz Biotechnology (San Diego, CA).

Mouse unfractionated bone cells were prepared from femora and tibiae of 4- to 5-wk-old ICR mice (Shizuoka Laboratories Animal Center, Shizuoka, Japan). After removal of connective soft tissues, the bones were minced into small pieces in α-MEM (ICN Biomedicals, Aurora, OH) supplemented with 10% FBS (Intergen, Purchase, NY) and 100 U/ml of penicillin. The cells were dissociated from the bone fragments by vortexing and were filtered through a nylon mesh with a 70-μm pore size. The cells obtained in suspension were used as mouse unfractionated bone cells. The unfractionated bone cells (108 cells) were seeded and cultured for 6 days in α-MEM containing 10% FBS and PGE2 (10−8 M; Sigma, St. Louis, MO) in 100-mm tissue culture dishes in a humidified atmosphere of 5% CO2. The medium was exchanged on day 4 of culture. During the 6 days in culture, the stromal cells derived from the unfractionated bone cells proliferated to become overconfluent, forming a stromal cell layer sheet. Poking at the end of the stromal cell layer caused the layer to spontaneously roll up and detach from the dish. When the unfractionated cells were precultured in the presence of high concentration of PGE2 (10−6 M), numerous tartrate-resistant acid phosphatase (TRAP)-positive mononuclear cells and TRAP-positive multinucleate cells (MNCs) were generated in the culture, consistent with the previous studies (18). However, in the preculture pretreated with the lower dose of PGE2 (10−8 M), the cells remaining on the bottom of the dishes consisted of a large population of M/Mφ-like cells, a small population of nonadherent cells and stromal cells, and few TRAP-positive cells. After removal of nonadherent cells and stromal cells by washing with PBS and incubating in 0.25% trypsin/0.05% EDTA, M/Mφ-like hemopoietic cells were harvested in PBS by vigorously pipetting. In the population of isolated hemopoietic cells, contaminating stromal cells and TRAP-positive cells represented <0.01% of the total cells.

Isolated M/Mφ-like hemopoietic cells were seeded at an initial density of 1 or 2.5 × 104 cells/cm2 and cultured in α-MEM/10% FBS with or without several cytokines and/or other agents. The culture medium was exchanged every 3 days. After a culture period of the desired length, the cells were fixed in 10% formalin and stained for TRAP activity with a leukocyte acid phosphatase kit (Sigma). The numbers of total cells, TRAP-positive mononuclear cells, and TRAP-positive MNCs were counted under a microscope. TRAP-positive mononuclear cells and MNCs were considered to be preosteoclastic and osteoclastic cells, respectively. Thereafter, nuclei of these cells were again stained with propidium iodide (50 μg/ml) in 0.1% sodium citrate, and the numbers of nuclei of total cells, TRAP-positive mononuclear cells, and TRAP-positive MNCs in the culture were counted under a fluorescence microscope. The total nuclei number represents the rate of cell division. The number of nuclei in TRAP-positive cells (mononuclear cells plus MNCs) was used as an indicator of the commitment to the osteoclast lineage. A fusion index was calculated as the percentage of nuclei in TRAP-positive MNCs per those in total cells, and the value was considered the percentage of cells that participated in the cell fusion, indicating osteoclast maturation.

Osteoclastic cells were generated in α-MEM/10% FBS containing M-CSF (10 ng/ml; Chemicon International, Temecula, CA) with sRANKL (40 ng/ml; PeproTech EC, London, U.K.) and/or various concentrations of TGF-β1 (Austral Biologicals, San Ramon, CA) for 5 days. Then, after treatment with trypsin/EDTA, the cells in the culture were pipetted off and harvested. The cells obtained (800 cells) were seeded on each dentine slice and incubated for 1 day. At the end of the incubation, the cells on the dentine slices were stained for TRAP activity to confirm their survival. Then, the cells were scraped off the dentine slices, and the slices were stained with acid hematoxylin (Sigma). The number of the stained pits was counted under a microscope.

After isolation of M/Mφ-like hemopoietic cells, the cells were suspended in ice-cold PBS containing 0.5% BSA, 0.1% sodium azide, and 1 mM glucose. Before being stained for cell surface Ags, the progenitors were preincubated with anti-CD16/32 Ab or an excess of mouse IgG (Sigma) to reduce nonspecific binding of Abs. The pretreated cells were stained for 30 min with biotinylated anti-CD11b, anti-CD11a, anti-CD44, anti-F4/80, or anti-integrin β3 Abs; with FITC-labeled anti-integrin αv; with PE-labeled anti-CD14; or with unlabeled anti-DEC-205 Abs. For staining with the biotinylated and the unlabeled Abs, the stained cells were secondarily incubated with avidin-FITC (PharMingen) and FITC-conjugated anti-rat IgG (PharMingen), respectively, for 30 min in PBS containing anti-CD16/32 Ab or an excess of mouse IgG. Then, the cells were analyzed without gating on a FACStar (Becton Dickinson, San Jose, CA).

Total RNA (1 μg) extracted from cells in the culture was used as a template for cDNA synthesis. cDNA was prepared by use of a Superscript II preamplification system (Life Technologies, Gaithersburg, MD). Primers were synthesized on the basis of the reported mouse cDNA sequences for TRAP, integrin αv, integrin β3, calcitonin receptor, cathepsin K, RANK, CD14, TGF-β1, TGF-β2, TGF-β3, TGF-β receptor I (TGFR-I), and TGFR-II. Sequences of the primers used for PCR were as follows: TRAP forward, 5′-CACGATGCCAGCGACAAGAG-3′; TRAP reverse, 5′-TGACCCCGTATGTGGCTAAC-3′; integrin αv forward, 5′-GCCAGCCCATTGAGTTTGATT-3′; integrin αv reverse, 5′-GCTACCAGGACCACCGAGAAG-3′; integrin β3 forward, 5′-TTACCCCGTGGACATCTACTA-3′; integrin β3 reverse, 5′-AGTCTTCCATCCAGGGCAATA-3′; cathepsin K forward, 5′-GGAAGAAGACTCACCAGAAGC-3′; cathepsin K reverse, 5′-GTCATATAGCCGCCTCCACAG-3′; calcitonin receptor forward, 5′-ACCGACGAGCAACGCCTACGC-3′; calcitonin receptor reverse, 5′-GCCTTCACAGCCTTCAGGTAC-3′. CD14 forward, 5′-AAGTTCCCGACCCTCCAAGTT-3′; CD14 reverse, 5′-CTGCCTTTCTTTCCTTACATC-3′; RANK forward, 5′-CTCTGCGTGCTGCTCGTTCC-3′; RANK reverse, 5′-TTGTCCCCTGGTGTGCTTCT-3′; TGF-β1 forward, 5′-GGACCGCAACAACGCCATCTA-3′; TGF-β1 reverse, 5′-CGCACACAGCAGTTCTTCTCT-3′; TGF-β2 forward, 5′-CATCCCGAATAAAAGCGAAGA-3′; TGF-β2 reverse, 5′-AAAACTCCCTCCCTCCTGTCA-3′. TGF-β3 forward, 5′-TTTTCCTCCCCCTTTCTACTG-3′; TGF-β3 reverse, 5′-GGTTCCATTTTTCTCCACTGA-3′; TGFR-I forward, 5′-GAAGGGCTCATCACCACCAAT-3′; TGFR-I reverse, 5′-AGGCAGCTAACCGTATCCAGA-3′; TGFR-II forward, 5′-GGCATCGCTCATCTCCACAGT-3′; TGFR-II reverse, 5′GCCCTCGGTCTCTCAGCACAC-3′; β-actin forward, 5′-TCACCCACACTGTGCCCATCTAC-3′; and β-actin reverse, 5′-GAGTACTTGCGCTCAGGAGGAGC-3′. Amplification was conducted for 22–32 cycles, each of 94°C for 30 s, 58°C (TGF-β2 and TGF-β3, 56°C) for 30 s, and 72°C for 1 min in a 25-μl reaction mixture containing 0.5 μl of each cDNA, 25 pmol of each primer, 0.2 mM dNTP, and 1 U of Tap DNA polymerase (Qiagen, Valencia, CA). After amplification, 15 μl of each reaction mixture was analyzed by 1.5% agarose gel electrophoresis, and the bands were then visualized by ethidium bromide staining.

After the isolated M/Mφ-like hemopoietic cells had been treated with M-CSF and/or TGF-β for 2 days, the cells were washed with PBS; scraped into a solution consisting of 10 mM sodium phosphate (pH 7.5), 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, 1 mM aminoethylbenzenesulfonyl fluoride, 10 μg/ml leupeptin, and 10 μg/ml aprotinin; and sonicated for 15 s. The protein concentration in the cell lysate was measured with a bicinchoninic acid protein assay kit (Pierce, Rockford, IL). Each sample containing equal amounts of protein was subjected to 10% SDS-PAGE, and the proteins separated in the gel were subsequently electrotransferred onto a polyvinylidene difluoride membrane. After having been blocked with 5% skim milk, the membrane was incubated with anti-RANK Abs or nonimmune rabbit IgG and subsequently with peroxidase-conjugated anti-rabbit IgG Ab. Immunoreactive proteins were visualized with Western blot chemiluminescence reagents (DuPont-New England Nuclear Products, Boston, MA) following the manufacturer’s instructions.

Nuclear extracts were prepared from M/Mφ-like hemopoietic cells pretreated for 2 days with M-CSF alone or with M-CSF and TGF-β as previously described (19). Double-stranded oligonucleotides containing an NF-κB binding site (5′-AGTTGAGGGGACTTTCCCAGGC-3′) were radiolabeled with [γ-32P]ATP and combined with 1 μg of nuclear extracts for 20 min at room temperature using a gel shift assay system (Promega, Madison, WI). The specificity of the reaction was confirmed by competition with a 50-fold molar excess of nonlabeled oligonucleotides. The protein-DNA complexes were resolved by 7.2% PAGE in 0.5 × TBE buffer and visualized by autoradiography. In the supershift experiment, the nuclear extracts were incubated with anti-p50 or anti-p65 Ab for 30 min on ice after binding to the oligonucleotides, and then were subjected to PAGE.

Significant differences between means of group were analyzed by one-way ANOVA and Dunnett’s test.

When unfractionated bone cells prepared from 4- to 5-wk-old mice were cultured in the presence of a low dose of PGE2 (10−8 M) for 6 days, the stromal cells proliferated to overconfluence in the culture and formed a sheet of cells. Morphologically appearing macrophage-like cells adhered to the substratum under the stromal cell layer. After the stromal cell sheet was detached, the remaining cells were isolated. As shown in Fig. 1, A and B, these cells revealed a mononuclear macrophage-like shape with a relatively large cytoplasm, and all the cells were capable of phagocytosing latex beads. These cells required M-CSF for their survival (data not shown). In addition, they were TRAP negative. Cell surface molecules expressed on the isolated cells were analyzed by flow cytometry using various Abs (Fig. 1 C). The cells were positive for CD11b (Mac-1 α-chain), CD44, F4/80, and CD11a (LFA-1) and weakly positive for CD14 and integrin αv, but negative for integrin β3, which is expressed on mature osteoclasts, and DEC205, which is expressed on dendritic cells. Taken together, these findings indicate that these cells belonged to the M/Mφ lineage.

FIGURE 1.

Phase-contrast micrographs of isolated M/Mφ-like hemopoietic cells. After removal of stromal cell layer, the remaining cells in the culture dishes were isolated, replated, and cultured in α-MEM containing 10% FBS for 3 h (A). The isolated cells morphologically appeared to be M/Mφ. The isolated cells were incubated with 0.2% latex beads for 3 h and were washed with PBS to remove unphagocytosed beads. Most cells showed phagocytotic activity (B). The bar in each photograph indicates 100 μm. The isolated cells were analyzed for their cell surface Ags by flow cytometry (C). After Fc block or excessive mouse IgG pretreatment, the cells were stained with various Abs as shown in the unshaded area under the curve. The shaded area under the curve shows control staining with FITC-avidin or FITC-anti-rat IgG.

FIGURE 1.

Phase-contrast micrographs of isolated M/Mφ-like hemopoietic cells. After removal of stromal cell layer, the remaining cells in the culture dishes were isolated, replated, and cultured in α-MEM containing 10% FBS for 3 h (A). The isolated cells morphologically appeared to be M/Mφ. The isolated cells were incubated with 0.2% latex beads for 3 h and were washed with PBS to remove unphagocytosed beads. Most cells showed phagocytotic activity (B). The bar in each photograph indicates 100 μm. The isolated cells were analyzed for their cell surface Ags by flow cytometry (C). After Fc block or excessive mouse IgG pretreatment, the cells were stained with various Abs as shown in the unshaded area under the curve. The shaded area under the curve shows control staining with FITC-avidin or FITC-anti-rat IgG.

Close modal

In vivo and in vitro studies have demonstrated that RANKL in cooperation with M-CSF is essential for generation of osteoclasts from hemopoietic cells (5, 6). When the cells isolated by the above procedure were incubated for 6 days in the presence of M-CSF and sRANKL, TRAP-positive MNCs were formed in the culture in a dose-dependent manner, as shown in Fig. 2. The total number of nuclei in the cultures was increased by the addition of M-CSF, indicating that the proliferation of the cells was dependent on M-CSF. M-CSF also increased the percentage of nuclei in TRAP-positive cells as an indicator of commitment to the osteoclast lineage and the fusion index of TRAP-positive MNCs, defined as the percentage of cells participating in the fusion. These values at 20 ng/ml of M-CSF were 28 and 17%, respectively, producing a maximal effect; and no further stimulation was observed at higher concentrations. On the other hand, although the total nuclear number was not changed by the addition of sRANKL, the percentage of nuclei in TRAP-positive cells and the fusion index of TRAP-positive MNCs were increased in a concentration-dependent fashion, with a maximal stimulation at 40 ng/ml. When TRAP-positive MNCs generated in the presence of M-CSF and sRANKL were replated on dentine slices, these cells resorbed dentine and formed pits on the surface (Fig. 2,D). In addition, semiquantitative PCR analysis revealed that treatment with M-CSF and sRANKL for 6 days induced an increase in the levels of mRNA for TRAP, cathepsin K, calcitonin receptor, and integrin β3, all of which are abundantly expressed in osteoclasts (Fig. 3). On the other hand, the addition of sRANKL decreased the level of CD14 mRNA. These effects of M-CSF and sRANKL indicate that the isolated hemopoietic cells were potentially capable of differentiating into mature osteoclasts. In addition, these results demonstrate that M-CSF and RANKL play crucial roles in survival, proliferation, and differentiation of osteoclast progenitors, consistent with the conclusion of previous studies (5, 6). However, the percentages of nuclei in TRAP-positive cells and the fusion index were not very high, indicating that all the cells did not differentiate into the osteoclast lineage.

FIGURE 2.

Isolated M/Mφ-like hemopoietic cells potentially differentiate into mature osteoclasts. The isolated cells (2 × 104 cells) were cultured with various concentrations of M-CSF (A) in the presence of sRANKL (40 ng/ml), or the cells (5 × 104 cells) were cultured with various doses of sRANKL (B) in the presence of M-CSF (10 ng/ml) for 6 days. At the end of the culture period, the cells were stained for nuclei and TRAP activity. The numbers of total nuclei and of TRAP-positive mononuclear cells and MNCs were counted. The total number of nuclei (□) represents the proliferation of the cells in the culture. The percentages (•) of nuclei in TRAP-positive mononuclear cells and of MNCs per total nuclear number show the rate of commitment of the M/Mφ-like hemopoietic cells into cells of the osteoclast lineage. Fusion index (○) indicates the rate of TRAP-positive osteoclast precursors that participate in cell fusion resulting in mature osteoclasts. Values are the mean ± SE for three cultures in a representative experiment. C, Photomicrograph of the TRAP-stained cells cultured for 6 days with sRANKL (40 ng/ml) and M-CSF (10 ng/ml). D, The cells treated for 5 days with sRANKL and M-CSF were replated on dentine slices and incubated for 1 day. Resorption pits formed by the cells were stained with acid hematoxylin. The bar in each photograph indicates 100 μm.

FIGURE 2.

Isolated M/Mφ-like hemopoietic cells potentially differentiate into mature osteoclasts. The isolated cells (2 × 104 cells) were cultured with various concentrations of M-CSF (A) in the presence of sRANKL (40 ng/ml), or the cells (5 × 104 cells) were cultured with various doses of sRANKL (B) in the presence of M-CSF (10 ng/ml) for 6 days. At the end of the culture period, the cells were stained for nuclei and TRAP activity. The numbers of total nuclei and of TRAP-positive mononuclear cells and MNCs were counted. The total number of nuclei (□) represents the proliferation of the cells in the culture. The percentages (•) of nuclei in TRAP-positive mononuclear cells and of MNCs per total nuclear number show the rate of commitment of the M/Mφ-like hemopoietic cells into cells of the osteoclast lineage. Fusion index (○) indicates the rate of TRAP-positive osteoclast precursors that participate in cell fusion resulting in mature osteoclasts. Values are the mean ± SE for three cultures in a representative experiment. C, Photomicrograph of the TRAP-stained cells cultured for 6 days with sRANKL (40 ng/ml) and M-CSF (10 ng/ml). D, The cells treated for 5 days with sRANKL and M-CSF were replated on dentine slices and incubated for 1 day. Resorption pits formed by the cells were stained with acid hematoxylin. The bar in each photograph indicates 100 μm.

Close modal
FIGURE 3.

Semiquantitative RT-PCR analysis for the expression of various mRNAs. Total RNA of the M/Mφ-like hemopoietic cells was immediately extracted after isolation of the cells (lanes 1, 5, and 9), or the isolated M/Mφ-like hemopoietic cells were treated with M-CSF (10 ng/ml) alone (lanes 2, 6, and 10), M-CSF plus sRANKL (40 ng/ml; lanes 3, 7, and 11), or M-CSF, sRANKL, and TGF-β1 (10 ng/ml; lanes 4, 8, and 12) for 6 days, and then total RNA was extracted from each culture. The primers used were designed for mouse genes of TRAP, cathepsin K, CD14, integrins αv and β3, calcitonin receptor, and β-actin. Each cDNA was amplified for the indicated number of PCR cycles.

FIGURE 3.

Semiquantitative RT-PCR analysis for the expression of various mRNAs. Total RNA of the M/Mφ-like hemopoietic cells was immediately extracted after isolation of the cells (lanes 1, 5, and 9), or the isolated M/Mφ-like hemopoietic cells were treated with M-CSF (10 ng/ml) alone (lanes 2, 6, and 10), M-CSF plus sRANKL (40 ng/ml; lanes 3, 7, and 11), or M-CSF, sRANKL, and TGF-β1 (10 ng/ml; lanes 4, 8, and 12) for 6 days, and then total RNA was extracted from each culture. The primers used were designed for mouse genes of TRAP, cathepsin K, CD14, integrins αv and β3, calcitonin receptor, and β-actin. Each cDNA was amplified for the indicated number of PCR cycles.

Close modal

Cells of the M/Mφ lineage are known to produce several cytokines and growth factors, and their proliferation and differentiation are regulated in an autocrine and paracrine manner (20, 21). Of these factors, TGF-β is expressed not only by monocyte/macrophages but also by MNCs (22). The expression of two types (TGFR-I and TGFR-II) of TGF-β receptor on isolated hemopoietic cells was confirmed by RT-PCR analysis (Fig. 4). These results indicate that the isolated hemopoietic cells are potentially responsive to TGF-β. Simultaneous addition of TGF-β with sRANKL (40 ng/ml) and M-CSF (10 ng/ml) dose dependently increased the number of TRAP-positive MNCs among the cells cultured for 6 days, with a maximal effect of 12-fold at 1.25–20 ng/ml. At 20 ng/ml of TGF-β, the fusion index was 60%, i.e., 12-fold greater than that in the absence of TGF-β. Besides the TRAP-positive MNCs, almost all the mononuclear cells were TRAP positive (Fig. 5, A and B). However, since the cells died in the presence of TGF-β and sRANKL without M-CSF, TGF-β could not replace M-CSF for the survival of the osteoclast progenitors (data not shown). In addition, the combination of TGF-β and M-CSF without sRANKL supported the survival of the cells, but did not induce the formation of osteoclastic TRAP-positive MNCs. Associated with the enhancement of differentiation into osteoclasts, TGF-β further increased the mRNA levels of TRAP, cathepsin K, calcitonin receptor, and integrins αv and β3 and further decreased the CD14 mRNA level (Fig. 3). The number of pits excavated by the cells cultured with TGF-β, M-CSF, and sRANKL was much greater than that with M-CSF and sRANKL (Fig. 5,C). As shown in Fig. 6, the stimulatory effect of TGF-β was dose dependent, consistent with the increase in osteoclastic cell formation. Taken together, the data show exogenous TGF-β to be a potent, but additive, inducer of osteoclastogenesis.

FIGURE 4.

Expression of TGF-βs and TGF-β receptors. Total RNA of the M/Mφ-like hemopoietic cells was immediately extracted after isolation of the cells (lane 1), or the isolated M/Mφ-like hemopoietic cells were treated with M-CSF (10 ng/ml) alone (lane 2), M-CSF plus sRANKL (40 ng/ml; lane 3), or M-CSF, sRANKL, and TGF-β1 (10 ng/ml; lane 4) for 6 days, and then total RNA was extracted from each culture. The primers used were designed for mouse genes of TGF-β1, TGF-β2, TGF-β3, and types I and II of TGFR. The numbers in parentheses indicate the numbers of PCR cycles.

FIGURE 4.

Expression of TGF-βs and TGF-β receptors. Total RNA of the M/Mφ-like hemopoietic cells was immediately extracted after isolation of the cells (lane 1), or the isolated M/Mφ-like hemopoietic cells were treated with M-CSF (10 ng/ml) alone (lane 2), M-CSF plus sRANKL (40 ng/ml; lane 3), or M-CSF, sRANKL, and TGF-β1 (10 ng/ml; lane 4) for 6 days, and then total RNA was extracted from each culture. The primers used were designed for mouse genes of TGF-β1, TGF-β2, TGF-β3, and types I and II of TGFR. The numbers in parentheses indicate the numbers of PCR cycles.

Close modal
FIGURE 5.

Effect of exogenous addition of TGF-β on osteoclastogenesis from isolated M/Mφ-like hemopoietic cells. The isolated M/Mφ-like hemopoietic cells were cultured for 6 days with various concentrations of TGF-β1 in the presence of M-CSF (10 ng/ml) and sRANKL (40 ng/ml). A, At the end of the culture, the cells were stained for nuclei and TRAP activity. Total nuclear number, percentage of that number in TRAP-positive cells, and fusion index were calculated. Values are the mean ± SE for three cultures in a representative experiment. B, Photomicrograph of TRAP-positive cells cultured for 6 days with TGF-β1 (10 ng/ml), sRANKL (40 ng/ml), and M-CSF (10 ng/ml). C, The cells treated for 5 days with TGF-β, sRANKL, and M-CSF were replated on dentine slices and incubated for 1 day. Resorption pits formed by the cells were stained with acid hematoxylin. The bar in each photograph indicates 100 μm.

FIGURE 5.

Effect of exogenous addition of TGF-β on osteoclastogenesis from isolated M/Mφ-like hemopoietic cells. The isolated M/Mφ-like hemopoietic cells were cultured for 6 days with various concentrations of TGF-β1 in the presence of M-CSF (10 ng/ml) and sRANKL (40 ng/ml). A, At the end of the culture, the cells were stained for nuclei and TRAP activity. Total nuclear number, percentage of that number in TRAP-positive cells, and fusion index were calculated. Values are the mean ± SE for three cultures in a representative experiment. B, Photomicrograph of TRAP-positive cells cultured for 6 days with TGF-β1 (10 ng/ml), sRANKL (40 ng/ml), and M-CSF (10 ng/ml). C, The cells treated for 5 days with TGF-β, sRANKL, and M-CSF were replated on dentine slices and incubated for 1 day. Resorption pits formed by the cells were stained with acid hematoxylin. The bar in each photograph indicates 100 μm.

Close modal
FIGURE 6.

Pits excavated by osteoclastic cells generated by treatment with M-CSF, sRANKL, and TGF-β1. Isolated M/Mφ-like hemopoietic cells were cultured for 5 days without or with various concentrations of TGF-β1 in the presence of M-CSF (10 ng/ml) and sRANKL (40 ng/ml). Then, after treatment with trypsin/EDTA, the cells obtained (800 cells) were seeded on each dentine slice and incubated for 1 day. At the end of the incubation, the cells were scraped off the dentine slices, the slices were stained with acid hematoxylin, and the number of resorption pits on the dentine slice was counted under a microscope. Values are the mean ± SE for four cultures in a representative experiment.

FIGURE 6.

Pits excavated by osteoclastic cells generated by treatment with M-CSF, sRANKL, and TGF-β1. Isolated M/Mφ-like hemopoietic cells were cultured for 5 days without or with various concentrations of TGF-β1 in the presence of M-CSF (10 ng/ml) and sRANKL (40 ng/ml). Then, after treatment with trypsin/EDTA, the cells obtained (800 cells) were seeded on each dentine slice and incubated for 1 day. At the end of the incubation, the cells were scraped off the dentine slices, the slices were stained with acid hematoxylin, and the number of resorption pits on the dentine slice was counted under a microscope. Values are the mean ± SE for four cultures in a representative experiment.

Close modal

As shown in Fig. 4, the isolated hemopoietic cells expressed TGF-β1 and -β2 as well as their receptors. Therefore, we next examined whether endogenous TGF-β is involved in osteoclast generation in an autocrine fashion. Addition of neutralizing Ab against TGF-β abrogated the stimulation of osteoclast-like cell formation induced by M-CSF and sRANKL, whereas the nonimmune IgG had no effect (Fig. 7). This result shows that osteoclastogenesis induced by M-CSF and RANKL requires the endogenous production of TGF-β by the osteoclast progenitors. Next, to ascertain the action point of endogenous TGF-β in osteoclast development, we examined the effects of different treatment periods with anti-TGF-β on the osteoclast-like cell formation from the isolated hemopoietic cells pretreated with M-CSF and/or TGF-β. When the isolated hemopoietic cells were cultured for 6 days in the presence of M-CSF and sRANKL without TGF-β after pretreatment with TGF-β and M-CSF for the first 2 days, the fusion index of TRAP-positive osteoclastic MNCs formed in the cultures was equivalent to that in the cultures treated for the last 6 days with a combination of TGF-β, sRANKL, and M-CSF (Fig. 8,B; lanes 2 and 3 from left). The cultures pretreated only with M-CSF required the continuous presence of TGF-β for the high efficiency of osteoclastic cell formation (Fig. 8,A; lanes 2 and 3 from left). These data suggest that 2-day pretreatment with TGF-β allows osteoclast progenitors to prime to commit to an osteoclast lineage. However, the expression of the RANK receptor (RANK) in the isolated hemopoietic cells was not up-regulated by 2-day pretreatment with TGF-β at mRNA and protein levels, whereas the expression was enhanced by M-CSF (Fig. 9, A and B). Instead, the TGF-β pretreatment synergistically stimulated activation of NF-κB evoked by sRANKL as determined by direct EMSA (Fig. 9,C). Following the TGF-β pretreatment, treatment for 6 days with anti-TGF-β greatly reduced the osteoclast generation induced by the combination of sRANKL and M-CSF. Likewise, the inhibition was seen in the cultures treated with anti-TGF-β for the last 5 and 3 days, implying that endogenous production of TGF-β is involved in the processes of osteoclast differentiation, including priming and maturation (Fig. 8, A and B; lanes 4–6 from left). Finally, as in the cultures of osteoclast progenitors, anti-TGF-β Ab abolished the formation of TRAP-positive osteoclastic MNCs induced by M-CSF plus sRANKL or 1,25(OH)2D3 in cultures of unfractionated bone cells (Fig. 10), indicating that the requirement of endogenous TGF-β for osteoclastogenesis is not restricted to cultures of isolated hemopoietic cells.

FIGURE 7.

Neutralizing anti-TGF-β Ab abolishes generation of osteoclasts induced by M-CSF and sRANKL. The isolated M/Mφ-like hemopoietic cells were treated for 6 days with neutralizing anti-TGF-β Ab (20 μg/ml) or isotype control IgG1 (20 μg/ml) in the presence of various combinations of M-CSF (10 ng/ml), TGF-β1 (10 ng/ml), and/or sRANKL (40 ng/ml). Then, the fusion index in each culture was calculated. Values are the mean ± SE for three cultures in a representative experiment. ∗, p < 0.01 vs cultures treated with M-CSF and sRANKL, or cultures treated with M-CSF, TGF-β1, and sRANKL, by Dunnett’s analysis.

FIGURE 7.

Neutralizing anti-TGF-β Ab abolishes generation of osteoclasts induced by M-CSF and sRANKL. The isolated M/Mφ-like hemopoietic cells were treated for 6 days with neutralizing anti-TGF-β Ab (20 μg/ml) or isotype control IgG1 (20 μg/ml) in the presence of various combinations of M-CSF (10 ng/ml), TGF-β1 (10 ng/ml), and/or sRANKL (40 ng/ml). Then, the fusion index in each culture was calculated. Values are the mean ± SE for three cultures in a representative experiment. ∗, p < 0.01 vs cultures treated with M-CSF and sRANKL, or cultures treated with M-CSF, TGF-β1, and sRANKL, by Dunnett’s analysis.

Close modal
FIGURE 8.

Effect of neutralizing anti-TGF-β Ab on priming and maturation of osteoclast differentiation. The isolated M/Mφ-like hemopoietic cells were pretreated for 2 days with M-CSF (10 ng/ml) alone (A) or with TGF-β1 (10 ng/ml) plus M-CSF (B). Then, after the factors had been removed by washings, the pretreated cells were further treated for 6 days with various combinations of M-CSF, TGF-β, and/or sRANKL (40 ng/ml). Neutralizing anti-TGF-β Ab (20 μg/ml) or isotype control IgG1 (20 μg/ml) was added to the cultures at day 0 (D-0), day 1 (D-1), or day 3 (D-3) after the start of treatment. Values are the mean ± SE for three cultures in a representative experiment. ∗, p < 0.01 vs culture treated with M-CSF and sRANKL, by Dunnett’s analysis.

FIGURE 8.

Effect of neutralizing anti-TGF-β Ab on priming and maturation of osteoclast differentiation. The isolated M/Mφ-like hemopoietic cells were pretreated for 2 days with M-CSF (10 ng/ml) alone (A) or with TGF-β1 (10 ng/ml) plus M-CSF (B). Then, after the factors had been removed by washings, the pretreated cells were further treated for 6 days with various combinations of M-CSF, TGF-β, and/or sRANKL (40 ng/ml). Neutralizing anti-TGF-β Ab (20 μg/ml) or isotype control IgG1 (20 μg/ml) was added to the cultures at day 0 (D-0), day 1 (D-1), or day 3 (D-3) after the start of treatment. Values are the mean ± SE for three cultures in a representative experiment. ∗, p < 0.01 vs culture treated with M-CSF and sRANKL, by Dunnett’s analysis.

Close modal
FIGURE 9.

Expression of RANKL receptor (RANK) and activation of NF-κB in M/Mφ-like hemopoietic cells. A and B, Total RNA and membrane proteins were immediately extracted from the isolated untreated M/Mφ-like cells (A, lanes 1 and 4; B, lanes 1, 4, and 7) and from the cells treated for 2 days with M-CSF at 10 ng/ml (A, lanes 2 and 5; B, lanes 2, 5, and 8), or with M-CSF and TGF-β1 at 10 ng/ml (A, lanes 3 and 6; B, lanes 3, 6 and 9). Western blotting (A) and RT-PCR (B) analyses for RANK expression were performed. Each cDNA was amplified for the indicated number of PCR cycles. C, EMSA. The isolated M/Mφ-like cells were precultured in the absence (lanes 1–5) or presence (lanes 6–12) of TGF-β1 (10 ng/ml) with M-CSF (10 ng/ml) for 2 days. After preculture, the cells were treated with sRANKL (40 ng/ml) for 0 h (lanes 1 and 6), 0.5 h (lanes 2 and 7), 1 h (lanes 3 and 8), 2 h (lanes 4 and 9), and 4 h (lanes 5 and 10). Then, nuclear proteins in the cells were extracted and subjected to EMSA. Lanes 11 and 12, EMSAs using a sample from the cells treated with sRANKL for 1 h after TGF-β pretreatment (lane 8) were performed in the presence of anti-p65 and anti-p50 Abs, respectively.

FIGURE 9.

Expression of RANKL receptor (RANK) and activation of NF-κB in M/Mφ-like hemopoietic cells. A and B, Total RNA and membrane proteins were immediately extracted from the isolated untreated M/Mφ-like cells (A, lanes 1 and 4; B, lanes 1, 4, and 7) and from the cells treated for 2 days with M-CSF at 10 ng/ml (A, lanes 2 and 5; B, lanes 2, 5, and 8), or with M-CSF and TGF-β1 at 10 ng/ml (A, lanes 3 and 6; B, lanes 3, 6 and 9). Western blotting (A) and RT-PCR (B) analyses for RANK expression were performed. Each cDNA was amplified for the indicated number of PCR cycles. C, EMSA. The isolated M/Mφ-like cells were precultured in the absence (lanes 1–5) or presence (lanes 6–12) of TGF-β1 (10 ng/ml) with M-CSF (10 ng/ml) for 2 days. After preculture, the cells were treated with sRANKL (40 ng/ml) for 0 h (lanes 1 and 6), 0.5 h (lanes 2 and 7), 1 h (lanes 3 and 8), 2 h (lanes 4 and 9), and 4 h (lanes 5 and 10). Then, nuclear proteins in the cells were extracted and subjected to EMSA. Lanes 11 and 12, EMSAs using a sample from the cells treated with sRANKL for 1 h after TGF-β pretreatment (lane 8) were performed in the presence of anti-p65 and anti-p50 Abs, respectively.

Close modal
FIGURE 10.

Effects of exogenous TGF-β and neutralizing anti-TGF-β Ab on osteoclast generation in cultures of mouse unfractionated bone cells. A, Unfractionated bone cells (1 × 105) were cultured in each well of 96-well plates for 5 days in the presence of M-CSF (10 ng/ml) with various combinations of TGF-β1 (10 ng/ml), sRANKL (40 ng/ml), anti-TGF-β Ab (20 μg/ml), and/or nonimmune IgG1 (20 μg/ml). B, Unfractionated bone cells (5 × 105) were cultured for 7 days with or without 1,25(OH)2D3 (10 nM) in the absence or the presence of neutralizing anti-TGF-β. Then, the cells in the culture were stained for TRAP activity, and the TRAP-positive MNCs were counted. Values are the mean ± SE for three cultures in a representative experiment. A: ∗, p < 0.05; ∗∗, p < 0.01 (vs cultures treated with M-CSF and sRANKL, by Dunnett’s analysis). B: ∗∗, p < 0.01 (vs cultures treated with 1,25(OH)2D3, by Dunnett’s analysis).

FIGURE 10.

Effects of exogenous TGF-β and neutralizing anti-TGF-β Ab on osteoclast generation in cultures of mouse unfractionated bone cells. A, Unfractionated bone cells (1 × 105) were cultured in each well of 96-well plates for 5 days in the presence of M-CSF (10 ng/ml) with various combinations of TGF-β1 (10 ng/ml), sRANKL (40 ng/ml), anti-TGF-β Ab (20 μg/ml), and/or nonimmune IgG1 (20 μg/ml). B, Unfractionated bone cells (5 × 105) were cultured for 7 days with or without 1,25(OH)2D3 (10 nM) in the absence or the presence of neutralizing anti-TGF-β. Then, the cells in the culture were stained for TRAP activity, and the TRAP-positive MNCs were counted. Values are the mean ± SE for three cultures in a representative experiment. A: ∗, p < 0.05; ∗∗, p < 0.01 (vs cultures treated with M-CSF and sRANKL, by Dunnett’s analysis). B: ∗∗, p < 0.01 (vs cultures treated with 1,25(OH)2D3, by Dunnett’s analysis).

Close modal

In this study we succeeded in developing a new isolation method for obtaining a homogenous population of osteoclast progenitors. The progenitors possessed common phenotypes of monocyte/macrophages, representing M/Mφ-like hemopoietic cells. Using the stroma-free culture system of isolated M/Mφ-like hemopoietic cells, we demonstrated that TGF-β directly acts on the hemopoietic cells to enhance the osteoclast formation elicited by a combination of sRANKL and M-CSF. The isolated hemopoietic cells expressed mRNAs of TGF-β1 and -β2 and TGFR-I and -II throughout the culture period, suggesting that osteoclast progenitors are both the TGF-β-producing cells and cells responsive to TGF-β. Because various hemopoietic cells express TGF-β and TGF-β receptors and their proliferation and differentiation are widely regulated by TGF-β (23, 24), and because osteoclasts are of hemopoietic origin (25), these expressions in the isolated cells are not surprising. In fact, the production of TGF-β and the expression of TGF-β receptors have been previously reported in chick osteoclasts (22) and in osteoclastic MNCs derived from human giant cell tumors of bone (26). In addition, since anti-TGF-β Ab greatly suppressed the osteoclast formation from isolated cells, endogenous production of autocrine-acting TGF-β by hemopoietic cells appears to be required for osteoclastic differentiation. Furthermore, TGF-β induced both the priming of hemopoietic cells to differentiate into the cells of osteoclast lineage and the maturation of these cells.

We isolated M/Mφ-like hemopoietic cells from cultures of unfractionated bone cells treated with PGE2. These cells required M-CSF for their survival and growth. In addition, the isolated cells expressed various monocyte/macrophage-phenotypic surface Ags, and showed phagocytotic activity. GM-CSF also supported the survival, but neither stimulated the proliferation nor induced osteoclast generation even in the presence of sRANKL (data not shown). It has been recently demonstrated that human cells sharing monocyte/macrophage phenotypes were capable of differentiating into dendritic cells and osteoclasts dependent on GM-CSF and M-CSF, respectively (27). In addition, when the isolated M/Mφ-like hemopoietic cells were cultured with LPS (28), these cells did not differentiate into an osteoclastic cell lineage even with sRANKL, M-CSF, and TGF-β (data not shown). Taken together, the available data indicate that the isolated cells represent bipotential immature monocytes/macrophages.

TGF-βs are multifunctional cytokines that widely regulate the proliferation and differentiation of a variety of cell types, including epithelial and mesenchymal cells (29, 30). Numerous studies on bone cells have indicated that TGF-β stimulates the growth and differentiation of osteoprogenitors to become bone matrix-producing cells (31, 32). Thus, TGF-βs are positive regulators of bone formation. However, the effect of TGF-β on bone resorption is controversial. A stimulatory effect of TGF-β on bone resorption was observed in organ cultures of mouse calvariae (33). In contrast, TGF-β inhibited the osteoclastic bone resorption in fetal rat long bones (34). Furthermore, Hughes et al. (35) showed that TGF-β promoted the apoptosis of osteoclasts in culture of bone marrow cells consisting of a heterogeneous population. Therefore, the inhibition of bone resorption by TGF-β may in part be attributed to the induced osteoclast apoptosis, although we did not observe such a stimulatory effect of TGF-β on osteoclast apoptosis in our culture system. Regarding osteoclastogenesis, the inhibition by TGF-β was demonstrated in cultures of bone marrow cells, which contained stromal cells, and in cocultures of bone marrow cells or spleen cells and stromal cells (13). On the other hand, TGF-β was reported to stimulate the formation of osteoclast-like cells in cultures of a human leukemia cell line, FLG 29.1, in an autocrine manner (36). Transgenic mice overexpressing TGF-β2 exhibited an osteoporosis-like phenotype due to the increased osteoclastic function (37, 38), and transgenic mice expressing dominant negative type II TGF-β receptor decreased osteoclastic bone resorption (39), suggesting a locally positive participation of TGF-β in osteoclast development. Sells et al. (40) recently demonstrated that TGF-β in combination with RANKL and M-CSF enhanced osteoclast-like formation in cultures of bone marrow cells and spleen cells containing few osteoblastic/stromal cells. In addition, TGF-β was demonstrated to stimulate osteoclast formation in cocultures of spleen cells and T lymphocytes expressing RANKL (41). Our findings are consistent with those results, although target cells of TGF-β were not defined due to the heterogeneity of hemopoietic cells in those culture systems (40, 41). The isolated cells examined in this study consisted of a homogenous population with monocyte/macrophage phenotypes, and all of them differentiated into cells of osteoclast lineage by TGF-β treatment in the presence of sRANKL and M-CSF. Therefore, TGF-β directly acts on osteoclast progenitors to stimulate their differentiation into osteoclasts. Taken together, the overall effects of TGF-β on osteoclastogenesis are dependent on the cell population.

We also indicated that TGF-β in combination with sRANKL and M-CSF stimulated osteoclast formation in the cultures of unfractionated bone cells. These cultures contained stromal cells, but the stromal cells somehow did not expansively proliferate in the cultures. The inhibitory effect of TGF-β seems to be observed when a large number of osteoblastic/stromal cells are present. It was recently demonstrated that TGF-β increased the expression of osteoprotegerin (identical with osteoclastogenesis-inhibitory factor), which strongly inhibits osteoclastogenesis as a decoy receptor of RANKL (42). Therefore, osteoprotegerin may be at least in part a mediator of the TGF-β inhibitory effect via stromal cells. However, endogenous TGF-β is intrinsically essential for osteoclast development, and the stimulatory effect of exogenous TGF-β is seen under the condition of a minimal number of stromal cells.

PGE2 has recently been reported to cooperate with RANKL and M-CSF in the promotion of osteoclast formation from hemopoietic cells (43). In a variety of cell types, TGF-β induces the production of PGs mediated by up-regulation of prostaglandin G/H synthase-2 (44, 45). Those studies suggest that the enhancement of osteoclast formation by TGF-β presented in this study is mediated by endogenous synthesis of PGs. However, since NS-398, a selective inhibitor of PGG/H synthase-2, did not affect the stimulation of osteoclast generation by TGF-β (data not shown), the stimulatory effect of TGF-β is not endogenous PG dependent.

RANKL has been demonstrated to activate NF-κB and c-Jun N-terminal protein kinase (JNK) through RANK in osteoclastic cells as well as in dendritic cells (12, 46, 47). Recent studies indicate that binding of RANKL to RANK caused association of the receptor with several TNF receptor-associated factors (TRAFs), resulting in the activation of NF-κB (46, 47, 48). Knockout mice of both NF-κB1 and NF-κB2, and of TRAF6, exhibited severe osteopetrosis due to impaired osteoclast differentiation (49, 50). Therefore, TRAF6 and NF-κB seem to be involved in osteoclastogenesis. In this study we demonstrated that TGF-β synergistically increased the translocation of NF-κB into nuclei induced by RANKL in M/Mφ-like osteoclast progenitors, although TGF-β did not affect the expression of RANK, suggesting an intracellular cross-talk in signalings of TGF-β and RANKL. At present, the detailed molecular signalings of TGF-β that strongly promote osteoclast formation are not known. It was recently reported that TGF-β-activated kinase 1 functionally interacted with IκB kinase to stimulate NF-κB (51). Such an interaction of TGF-β receptor downstream signaling molecules with RANK-associated molecules may at least in part account for the synergistic induction of osteoclastogenesis by TGF-β and RANKL.

Involvement of TGF-β in the pathogenesis of osteopenic disorders has been suggested. It was demonstrated in rheumatoid arthritis that the synovium contained a large number of macrophage-like cells that have a strong ability to produce TGF-β as well as other inflammatory cytokines, such as IL-1 and TNF-α (52, 53), and a high level of endogenous TGF-β was also detectable in other types of arthritides, including osteoarthritis (54). In addition, the TGF-β concentration in serum was shown to be elevated in osteoporotic women, with good correlation with the bone loss (55). Taken together with our findings, TGF-β may contribute to destruction of bone as well as bone formation in vivo.

In conclusion, TGF-β is intrinsically required for osteoclastogenesis in combination with RANKL and M-CSF. The results presented here expand our knowledge about the multifunctional roles of TGF-β in bone metabolism.

We thank Drs. L. G. Raisz (University of Connecticut Health Center), K. Higashio (Snow Brand Milk Products Co. Ltd.), and T. Kuriki (Aventis Pharma Ltd.) for suggestions and critical review, for the generous gift of anti-RANK Ab, and for operating the FACStar, respectively.

1

This work was supported in part by a grant-in aid for scientific research from the Ministry of Education, Science, and Culture of Japan.

3

Abbreviations used in this paper: M/Mφ, monocyte/macrophage; RANK, receptor activator of NF-κB; RANKL, receptor activator of NF-κB ligand; sRANKL, soluble RANKL; 1,25(OH)2D3, 1,25-dihydroxyvitamin D3; TGFR, TGF-β receptor; TRAP, tartrate-resistant acid phosphatase; MNCs, multinucleate cells; JNK, c-Jun N-terminal protein kinase; TRAF, TNF receptor-associated factor.

1
Felix, R., M. G. Cecchini, W. Hofstetter, P. R. Elford, A. Stutzer, H. Fleisch.
1990
. Impairment of macrophage colony-stimulating factor production and lack of resident bone marrow macrophages in the osteopetrotic op/op mouse.
J. Bone Miner. Res.
5
:
781
2
Kodama, H., A. Yamasaki, M. Nose, S. Niida, Y. Ohgame, M. Abe, M. Kumegawa, T. Suda.
1991
. Congenital osteoclast deficiency in osteopetrotic (op/op) mice is cured by injections of macrophage colony-stimulating factor.
J. Exp. Med.
173
:
269
3
Chambers, T. J..
1985
. The pathobiology of the osteoclast.
J. Clin. Pathol.
38
:
241
4
Suda, T., N. Udagawa, I. Nakamura, C. Miyaura, N. Takahashi.
1995
. Modulation of osteoclast differentiation by local factors.
Bone
17
:
87S
5
Lacey, D. L., E. Timms, H. L. Tan, M. J. Kelley, C. R. Dunstan, T. Burgess, R. Elliott, A. Colombero, G. Elliott, S. Scully, et al
1998
. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation.
Cell
93
:
165
6
Yasuda, H., N. Shima, N. Nakagawa, K. Yamaguchi, M. Kinosaki, S. Mochizuki, A. Tomoyasu, K. Yano, M. Goto, A. Murakami, et al
1998
. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL.
Proc. Natl. Acad. Sci. USA
95
:
3597
7
Kong, Y. Y., H. Yoshida, I. Sarosi, H. L. Tan, E. Timms, C. Capparelli, S. Morony, A. J. Oliveira-dos-Santos, G. Van, A. Itie, et al
1999
. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis.
Nature
397
:
315
8
Wong, B. R., J. Rho, J. Arron, E. Robinson, J. Orlinick, M. Chao, S. Kalachikov, E. Cayani, F. S. Bartlett, W. N. Frankel, et al
1997
. TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-Jun N-terminal kinase in T cells.
J. Biol. Chem.
272
:
25190
9
Anderson, D. M., E. Maraskovsky, W. L. Billingsley, W. C. Dougall, M. E. Tometsko, E. R. Roux, M. C. Teepe, R. F. DuBose, D. Cosman, L. Galibert.
1997
. A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function.
Nature
390
:
175
10
Shalhoub, V., J. Faust, W. J. Boyle, C. R. Dunstan, M. Kelley, S. Kaufman, S. Scully, G. Van, D. L. Lacey.
1999
. Osteoprotegerin and osteoprotegerin ligand effects on osteoclast formation from human peripheral blood mononuclear cell precursors.
J. Cell. Biochem.
72
:
251
11
Faust, J., D. L. Lacey, P. Hunt, T. L. Burgess, S. Scully, G. Van, A. Eli, Y. Qian, V. Shalhoub.
1999
. Osteoclast markers accumulate on cells developing from human peripheral blood mononuclear precursors.
J. Cell. Biochem.
72
:
67
12
Hsu, H., D. L. Lacey, C. R. Dunstan, I. Solovyev, A. Colombero, E. Timms, H. L. Tan, G. Elliott, M. J. Kelley, I. Sarosi, et al
1999
. Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand.
Proc. Natl. Acad. Sci. USA
96
:
3540
13
Chenu, C., J. Pfeilschifter, G. R. Mundy, G. D. Roodman.
1988
. Transforming growth factor β inhibits formation of osteoclast-like cells in long-term human marrow cultures.
Proc. Natl. Acad. Sci. USA
85
:
5683
14
Pfeilschifter, J., S. M. Seyedin, G. R. Mundy.
1988
. Transforming growth factor β inhibits bone resorption in fetal rat long bone cultures.
J. Clin. Invest.
82
:
680
15
Shinar, D. M., G. A. Rodan.
1990
. Biphasic effects of transforming growth factor-β on the production of osteoclast-like cells in mouse bone marrow cultures: the role of prostaglandins in the generation of these cells.
Endocrinology
126
:
3153
16
Mundy, G. R., L. F. Bonewald.
1990
. Role of TGF-β in bone remodeling.
Ann. NY Acad. Sci.
593
:
91
17
Centrella, M., T. L. McCarthy, E. Canalis.
1991
. Transforming growth factor-β and remodeling of bone.
J. Bone Joint Surg. Am.
73
:
1418
18
Collins, D. A., T. J. Chambers.
1991
. Effect of prostaglandins E1, E2, and F on osteoclast formation in mouse bone marrow cultures.
J. Bone Miner. Res.
6
:
157
19
Schreiber, E., P. Matthias, M. M. Muller, W. Schaffner.
1989
. Rapid detection of octamer binding proteins with ‘mini-extracts,’ prepared from a small number of cells.
Nucleic Acids Res.
17
:
6419
20
Tushinski, R. J., I. T. Oliver, L. J. Guilbert, P. W. Tynan, J. R. Warner, E. R. Stanley.
1982
. Survival of mononuclear phagocytes depends on a lineage-specific growth factor that the differentiated cells selectively destroy.
Cell
28
:
71
21
Assoian, R. K., B. E. Fleurdelys, H. C. Stevenson, P. J. Miller, D. K. Madtes, E. W. Raines, R. Ross, M. B. Sporn.
1987
. Expression and secretion of type β transforming growth factor by activated human macrophages.
Proc. Natl. Acad. Sci. USA
84
:
6020
22
Oursler, M. J..
1994
. Osteoclast synthesis and secretion and activation of latent transforming growth factor β.
J. Bone Miner. Res.
9
:
443
23
Bursuker, I., K. M. Neddermann, B. A. Petty, B. Schacter, G. L. Spitalny, M. A. Tepper, R. D. Pasternak.
1992
. In vivo regulation of hemopoiesis by transforming growth factor β1: stimulation of GM-CSF- and M-CSF-dependent murine bone marrow precursors.
Exp. Hematol.
20
:
431
24
Zhang, Y., Y. Y. Zhang, M. Ogata, P. Chen, A. Harada, S. Hashimoto, K. Matsushima.
1999
. Transforming growth factor-β1 polarizes murine hematopoietic progenitor cells to generate Langerhans cell-like dendritic cells through a monocyte/macrophage differentiation pathway.
Blood
93
:
1208
25
Ash, P., J. F. Loutit, K. M. Townsend.
1980
. Osteoclasts derived from haematopoietic stem cells.
Nature
283
:
669
26
Zheng, M. H., Y. Fan, S. J. Wysocki, A. T. Lau, T. Robertson, M. Beilharz, D. J. Wood, J. M. Papadimitriou.
1994
. Gene expression of transforming growth factor-β1 and its type II receptor in giant cell tumors of bone: possible involvement in osteoclast-like cell migration.
Am. J. Pathol.
145
:
1095
27
Akagawa, K. S., N. Takasuka, Y. Nozaki, I. Komuro, M. Azuma, M. Ueda, M. Naito, K. Takahashi.
1996
. Generation of CD1+RelB+ dendritic cells and tartrate-resistant acid phosphatase-positive osteoclast-like multinucleated giant cells from human monocytes.
Blood
88
:
4029
28
Hailman, E., T. Vasselon, M. Kelley, L. A. Busse, M.C. Hu, H. S. Lichenstein, P. A. Detmers, S. D. Wright.
1996
. Stimulation of macrophages and neutrophils by complexes of lipopolysaccharide and soluble CD14.
J. Immunol.
156
:
4384
29
Cashman, J. D., A. C. Eaves, E. W. Raines, R. Ross, C. J. Eaves.
1990
. Mechanisms that regulate the cell cycle status of very primitive hematopoietic cells in long-term human marrow cultures. I. Stimulatory role of a variety of mesenchymal cell activators and inhibitory role of TGF-β.
Blood
75
:
96
30
Masui, T., L. M. Wakefield, J. F. Lechner, M. A. LaVeck, M. B. Sporn, C. C. Harris.
1986
. Type β transforming growth factor is the primary differentiation-inducing serum factor for normal human bronchial epithelial cells.
Proc. Natl. Acad. Sci. USA
83
:
2438
31
Antosz, M. E., C. G. Bellows, J. E. Aubin.
1989
. Effects of transforming growth factor β and epidermal growth factor on cell proliferation and the formation of bone nodules in isolated fetal rat calvaria cells.
J. Cell Physiol.
140
:
386
32
Hock, J. M., E. Canalis, M. Centrella.
1990
. Transforming growth factor-β stimulates bone matrix apposition and bone cell replication in cultured fetal rat calvariae.
Endocrinology
126
:
421
33
Lerner, U. H..
1996
. Transforming growth factor-β stimulates bone resorption in neonatal mouse calvariae by a prostaglandin-unrelated but cell proliferation-dependent pathway.
J. Bone Miner. Res.
11
:
1628
34
Hattersley, G., T. Chambers.
1991
. Effects of transforming growth factor β 1 on the regulation of osteoclastic development and function.
J. Bone Miner. Res.
6
:
165
35
Hughes, D. E., A. Dai, J. C. Tiffee, H. H. Li, G. R. Mundy, B.F. Boyce.
1996
. Estrogen promotes apoptosis of murine osteoclasts mediated by TGF-β.
Nat. Med.
2
:
1132
36
Fiorelli, G., R. T. Ballock, L. M. Wakefield, M. B. Sporn, F. Gori, L. Masi, U. Fredian, A. Tanini, P. A. Bernabei, M. L. Brandi.
1994
. Role for autocrine TGF-β1 in regulating differentiation of a human leukemic cell line toward osteoclast-like cells.
J. Cell. Physiol.
160
:
482
37
Erlebacher, A., R. Derynck.
1996
. Increased expression of TGF-β2 in osteoblasts results in an osteoporosis-like phenotype.
J. Cell Biol.
132
:
195
38
Erlebacher, A., E. H. Filvaroff, J. Q. Ye, R. Derynck.
1998
. Osteoblastic responses to TGF-β during bone remodeling.
Mol. Biol. Cell
9
:
1903
39
Filvaroff, E., A. Erlebacher, J. Ye, S. E. Gitelman, J. Lotz, M. Heillman, R. Derynck.
1999
. Inhibition of TGF-β receptor signaling in osteoblasts leads to decreased bone remodeling and increased trabecular bone mass.
Development.
126
:
4267
40
Sells, R. J., C. L. Galvin, J. W. Horn Gatlin, T. R. Fuson.
1999
. TGF-β enhances osteoclast differentiation in hematopoietic cell cultures stimulated with RANKL and M-CSF.
Biochem. Biophys. Res. Commun.
265
:
233
41
Horwood, N. J., V. Kartsogiannis, J. M. Quinn, E. Romas, T. J. Martin, M. T. Gillespie.
1999
. Activated T lymphocytes support osteoclast formation in vitro.
Biochem. Biophys. Res. Commun.
265
:
144
42
Takai, H., M. Kanematsu, K. Yano, E. Tsuda, K. Higashio, K. Ikeda, K. Watanabe, Y. Yamada.
1998
. Transforming growth factor-β stimulates the production of osteoprotegerin/osteoclastogenesis inhibitory factor by bone marrow stromal cells.
J. Biol. Chem.
273
:
27091
43
Wani, M. R., K. Fuller, N. S. Kim, Y. Choi, T. Chambers.
1999
. Prostaglandin E2 cooperates with TRANCE in osteoclast induction from hemopoietic precursors: synergistic activation of differentiation, cell spreading, and fusion.
Endocrinology
140
:
1927
44
Pilbeam, C., Y. Rao, O. Voznesensky, H. Kawaguchi, C. Alander, L. G. Raisz, H. Herschman.
1997
. Transforming growth factor-β1 regulation of prostaglandin G/H synthase-2 expression in osteoblastic MC3T3–E1 cells.
Endocrinology
138
:
4672
45
Diaz, A., K. P. Chepenik, J. H. Korn, A. M. Reginato, S. A. Jimenez.
1998
. Differential regulation of cyclooxygenases 1 and 2 by interleukin-1β, tumor necrosis factor-α, and transforming growth factor-β1 in human lung fibroblasts.
Exp. Cell Res.
241
:
222
46
Wong, B. R., R. Josien, S. Y. Lee, M. Vologodskaia, R. M. Steinman, Y. Choi.
1998
. The TRAF family of signal transducers mediates NF-κB activation by the TRANCE receptor.
J. Biol. Chem.
273
:
28355
47
Darnay, B. G., V. Haridas, J. Ni, P. A. Moore, B. B. Aggarwal.
1998
. Characterization of the intracellular domain of receptor activator of NF-κB (RANK): interaction with tumor necrosis factor receptor-associated factors and activation of NF-κB and c-Jun N-terminal kinase.
J. Biol. Chem.
273
:
20551
48
Darnay, B. G., J. Ni, P. A. Moore, B. B. Aggarwal.
1999
. Activation of NF-κB by RANK requires tumor necrosis factor receptor-associated factor (TRAF) 6 and NF-κB-inducing kinase: identification of a novel TRAF6 interaction motif.
J. Biol. Chem.
274
:
7724
49
Franzoso, G., L. Carlson, L. Xing, L. Poljak, E. W. Shores, K. D. Brown, A. Leonardi, T. Tran, B. F. Boyce, U. Siebenlist.
1997
. Requirement for NF-κB in osteoclast and B-cell development.
Genes Dev.
11
:
3482
50
Lomaga, M. A., W. C. Yeh, I. Sarosi, G. S. Duncan, C. Furlonger, A. Ho, S. Morony, C. Capparelli, G. Van, S. Kaufman, et al
1999
. TRAF6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling.
Genes Dev.
13
:
1015
51
Sakurai, H., H. Miyoshi, W. Toriumi, T. Sugita.
1999
. Functional interactions of transforming growth factor β-activated kinase 1 with IκB kinases to stimulate NF-κB activation.
J. Biol. Chem.
274
:
10641
52
Cutolo, M., A. Sulli, A. Barone, B. Seriolo, S. Accardo.
1993
. Macrophages, synovial tissue and rheumatoid arthritis.
Clin. Exp. Rheumatol.
11
:
331
53
van den Berg, W. B..
1999
. The role of cytokines and growth factors in cartilage destruction in osteoarthritis and rheumatoid arthritis.
Z. Rheumatol.
58
:
136
54
Gravallese, E. M., Y. Harada, J. T. Wang, A. H. Gorn, T. S. Thornhill, S. R. Goldring.
1998
. Identification of cell types responsible for bone resorption in rheumatoid arthritis and juvenile rheumatoid arthritis.
Am. J. Pathol.
152
:
943
55
Grainger, D. J., J. Percival, M. Chiano, T. D. Spector.
1999
. The role of serum TGF-β isoforms as potential markers of osteoporosis.
Osteoporos. Int.
9
:
398