The authors state that they have no conflicts of interest.
Induction of DC-STAMP by Alternative Activation and Downstream Signaling Mechanisms†
Article first published online: 2 APR 2007
Copyright © 2007 ASBMR
Journal of Bone and Mineral Research
Volume 22, Issue 7, pages 992–1001, July 2007
How to Cite
Yagi, M., Ninomiya, K., Fujita, N., Suzuki, T., Iwasaki, R., Morita, K., Hosogane, N., Matsuo, K., Toyama, Y., Suda, T. and Miyamoto, T. (2007), Induction of DC-STAMP by Alternative Activation and Downstream Signaling Mechanisms. J Bone Miner Res, 22: 992–1001. doi: 10.1359/jbmr.070401
- Issue published online: 4 DEC 2009
- Article first published online: 2 APR 2007
- Manuscript Accepted: 23 MAR 2007
- Manuscript Revised: 15 MAR 2007
- Manuscript Received: 12 SEP 2006
- cell–cell fusion;
- macrophage giant cells;
- transcriptional regulation
DC-STAMP is essential for fusion of osteoclasts and foreign body giant cells; however, it is not known whether dc-stamp expression in these two cell types is differentially regulated. Here, we show that dc-stamp expression and cell–cell fusion are regulated in a cell type–specific manner.
Introduction: The transcription factors c-Fos and NFATc1 cooperate to regulate osteoclast differentiation, whereas PU.1 and NF-κB are activated in macrophages and osteoclasts or in both cell types. Thus, we asked what role c-Fos, NFATc1, PU.1, and NF-κB played in regulating dendritic cell–specific transmembrane protein (dc-stamp) expression and fusion of osteoclasts and macrophage giant cells.
Materials and Methods: Transcriptional activation by c-Fos and NFATc1 was examined by dc-stamp promoter analysis. Multinuclear cell formation was analyzed in cells from c-Fos–deficient mice or in wildtype cells treated with the NFAT inhibitor FK506. The role of DC-STAMP in cell fusion was examined in vitro in a macrophage giant cell formation assay using DC-STAMP–deficient cells. Recruitment of c-Fos, NFATc1, PU.1, and NF-κB to the dc-stamp promoter in osteoclasts and macrophage giant cells was analyzed by chromatin-immunoprecipitation analysis.
Results: Both activator protein-1 (AP-1) and NFAT binding sites in the dc-stamp promoter were needed for dc-stamp expression after RANKL stimulation of osteoclasts. dc-stamp expression was induced in osteoclasts and macrophage giant cells, and cells from DC-STAMP–deficient mice failed to form either multinuclear osteoclasts or macrophage giant cells. In contrast, c-Fos is indispensable for dc-stamp expression and cell–cell fusion under conditions favoring in vitro and in vivo induction of osteoclasts but not macrophage giant cells. Consistently, an NFAT inhibitor suppressed multinuclear osteoclast formation but not macrophage giant cell formation. In addition, PU.1 and NF-κB binding sites were detected in the dc-stamp promoter, and both PU.1 and NF-κB were recruited to the dc-stamp promoter after granulocyte-macrophage colony stimulating factor (GM-CSF) + interleukin (IL)-4 stimulation.
Conclusions: dc-stamp expression is regulated differently in osteoclasts and macrophage giant cells. c-Fos and NFATc1, both of which are essential for osteoclast differentiation, are needed for dc-stamp expression and cell–cell fusion in osteoclasts, but both factors are dispensable for giant cell formation by macrophages. Because PU.1 and NF-κB are recruited to the dc-stamp promoter after stimulation with GM-CSF + IL-4, dc-stamp transcription is regulated in a cell type–specific manner.
Despite their morphological similarity, osteoclasts and macrophage giant cells have distinct functions, namely, to resorb bone and reject foreign bodies, respectively.(1,2) Both cell types are, however, derived from common monocyte/macrophage lineage precursor cells. Both also express the transmembrane receptor dendritic cell–specific transmembrane protein (DC-STAMP), which we reported as essential for fusion of both cell types in vivo and in vitro.(3) Interestingly, DC-STAMP–deficient mononuclear osteoclastic cells have bone-resorbing activity in vivo and in vitro, although this activity is less efficient in knockout than in wildtype multinuclear osteoclasts.(3) Thus, multinucleation of osteoclasts is not essential for bone resorption but enhances bone-resorbing efficiency. Osteoclast differentiation is induced by macrophage-colony stimulating factor (M-CSF) and RANKL,(4) suggesting that DC-STAMP is downstream of RANKL-RANK signals in the osteoclast lineage. RANKL binds its cognate receptor RANK and induces expression of Fos and Nfatc1.(4,5) c-Fos is a component of transcription factor activator protein-1 (AP-1). It has been shown that c-Fos induces Nfatc1 expression and that c-Fos and NFATc1 cooperatively regulate osteoclastogenesis in response to RANKL stimulation.(6,7) M-CSF alone induces mononuclear macrophages,(8,9) whereas macrophage giant cells are formed from bone marrow mononuclear cells in the presence of interleukin (IL)-3 + IL-4, granulocyte-macrophage colony stimulating factor (GM-CSF) + IL-4, or M-CSF + IL-4, all of which inhibit osteoclastogenesis in vitro.(10,11) GM-CSF inhibits osteoclastogenesis by downregulating Fos, which is an essential transcription factor for osteoclastogenesis, in the presence of M-CSF and RANKL,(4) and IL-4 inhibits both RANK expression and activation of Nfatc1.(12–14) Thus, osteoclasts and macrophage giant cells are considered different cell types, although both cells form multinuclear cells. PU.1 and NF-κB are transcription factors activated after specific stimulation.(15,16) Currently, it is unclear whether and how molecules common to both osteoclasts and macrophage giant cells—such as DC-STAMP—are induced by different stimuli.
Osteoclasts derived from DC-STAMP–deficient mice fail to fuse, even though these cells express osteoclast-specific molecules such as TRACP and cathepsin K, indicating that DC-STAMP is specifically needed for fusion but not differentiation.(3) Because DC-STAMP–deficient mice suffer from mild osteopetrosis, fusion of osteoclasts is likely needed for effective bone resorption and maintenance of bone quantity. On the other hand, macrophage giant cells are known to form in response to chronic inflammatory conditions such as biomaterial implantation, sarcoidosis, and tuberculosis. These cells likely contribute to rejection of biomaterials, and their activity determines the duration of the inflammation response.(1,2) The precise regulation of cell–cell fusion through DC-STAMP is, however, not known. Here, we used promoter assays and analyzed regulation of dc-stamp expression and cell–cell fusion using c-Fos–deficient mice and the NFAT inhibitor FK506.
MATERIALS AND METHODS
Mice were born and maintained under pathogen-free conditions and cared for in accordance with the guidelines of Keio University School of Medicine.
In vitro osteoclastogenesis assay
Bone marrow mononuclear cells were isolated from wildtype or DC-STAMP–deficient mice and cultured in the presence of 50 ng/ml M-CSF (R&D Systems, Minneapolis, MN, USA) for 3 days. Spleen mononuclear cells were isolated from wildtype or c-Fos–deficient mice and cultured in the presence of 50 ng/ml M-CSF for 3 days. M-CSF–dependent adherent cells were harvested, and 5 × 104 cells were plated in each well of 96-well culture plates. Cells were cultured in α-MEM (Invitrogen, Carlsbad, CA, USA) containing 10% FBS (JRH Bioscience, Lenexa, KS, USA) with M-CSF (50 ng/ml) or M-CSF (50 ng/ml) + RANKL (50 ng/ml, R&D Systems) for 6 days. Calcineurin inhibitor FK506 (1 μg/ml; Calbiochem, Torrance, CA, USA) or 0.1% DMSO as vehicle was added to the culture. Osteoclastogenesis was evaluated by TRACP staining as previously described.(9) For retroviral transduction, retrovirus expressing c-Fos or ΔNFAT, a constitutively active form of NFAT, was prepared as described and incubated with osteoclast progenitor cells for 3 days with M-CSF (50 ng/ml).(4,7) Cells were cultured in the presence of M-CSF alone or M-CSF + RANKL for 6 days.
In vitro giant cell formation assay
M-CSF–dependent adherent cells were harvested as described above, and 5 × 104 cells were plated in each well of 96-well culture plates. Cells were cultured in α-MEM containing 10% FBS in the presence of GM-CSF (50 ng/ml; R&D Systems) + IL-4 (100 U/ml; R&D Systems), IL-3 (100 U/ml; R&D Systems) + IL-4 (100 U/ml), or M-CSF (50 ng/ml) + IL-4 (100 U/ml) for 6 days. FK506 (1 μg/ml) or vehicle (0.1% DMSO) was added to the culture. Cells were stained with May-Gruenwald and Giemsa as described.(3)
Total RNA was isolated from osteoclasts or macrophage giant cells using an RNeasy mini kit (Qiagen, Hilden, Germany). RT-PCR analysis was performed as described.(3,9) Primers for RT-PCR were as follows:
gapdh sense: 5′-TGAAGGTCGGTGTGAACGGATTTGGC-3′
gapdh antisense: 5′-CATGTAGGCCATGAGGTCCACCAC-3′
Fos sense: 5′-GAGCTGACAGATACACTCCAAGCG-3′
Fos antisense: 5′-CAGTCTGCTGCATAGAAGGAACCG-3′
Nfatc1 sense: 5′-CCCAGTATACCAGCTCTGCCATTG-3′
Nfatc1 antisense: 5′-GGAGCCTTCTCCACGAAAATGACT-3′
dc-stamp sense: 5′-GAATTCATGAGGCTCTGGACCTTGGGCACCAGT-3′
dc-stamp antisense: 5′-CTCGAGTCATAGATCATCTTCATTTGCAGGGAT-3′
PU.1 sense: 5′-CTGAGAACCACTTCACAGAGCTGCAGAG-3′
PU.1 antisense: 5′-GCGCCATCTTCTGGTAGGTCATCTTCTT-3′
Relb sense: 5′-GGGAAGGTCTAAATGCCATCCACATAGC-3′
Relb antisense: 5′-CTCGTGTCTTCTGTCAGCTGCTTCATGT-3′
β-actin sense: 5′-TGAGAGGGAAATCGTGCGTGAC-3′
β-actin antisense: 5′-AAGAAGGAAGGCTGGAAAAGAG-3′.
Luciferase reporter gene assay
Four- (−3672 to +237) and 0.2-kb (−213 to −1) regions from the dc-stamp promoter region were cloned or PCR-amplified and ligated into the PicaGene luciferase reporter plasmid, pGL3 (Promega, Madison, WI, USA). Mutations in NFAT and AP-1 binding sites were: NFAT (GAAAC/T → Gcccg) and AP-1 (TGATTCA → TgcacCc).
RAW 264.7 cells were plated at 1 × 105 cells/well the day before in 24-well plates and transfected with reporter plasmids and pCMV-RHL using FuGene 6 (Roche). RANKL was added 3 h after transfection. After 48 h, cells were washed twice with PBS and lysed in reporter lysis buffer (Promega). Luciferase activity was measured with a Dual-Luciferase Reporter Assay System (Promega) as described.(17)
Osteoclast and macrophage giant cells were generated as above. Cells were fixed in 4% paraformaldehyde/PBS for 10 minutes, washed twice at 4°C, and incubated in 1% normal goat serum/0.2%BSA/2% skim milk/0.1% Triton-X100 in PBS for 30 min at room temperature. For NFATc1 staining, cells were stained with 1 μg/ml mouse anti-NFATc1 monoclonal antibody (7A6; Santa Cruz Biotechnology, Santa Cruz, CA, USA) for 60 min at room temperature followed by 1 μg/ml Alexa 488 anti-mouse IgG antibody (Invitrogen) for 60 min at room temperature. Cells were observed using a fluorescence microscope (model OX70; Olympus, Tokyo, Japan).
Chromatin immunoprecipitation (ChIP) was performed with the ChIP Assay Kit (Upstate Biotechnology, Billerica, MA, USA) according to the manufacturer's instructions, using antibodies against c-Fos (K-25; Santa Cruz), NFATc1 (7A6; Santa Cruz), PU.1 (T-21; Santa Cruz), and NF-κB (C-20; Santa Cruz). Purified DNA was analyzed by PCR using primers that detect sequences containing the dc-stamp promoter, specifically for c-Fos and NFATc1, 5′-GGGGGTCC TCATTTCTACAACTCAT-3′ (sense) and 5′-GCCACATCACCCTGAATCAATCTT-3′ (antisense); for PU.1, 5′-GAGCTATGGGCTCATCCAGAAATC-3′ (sense) and 5′-AACTGGAAAAAGGAGCCACAGGTT-3′ (antisense); and for NF-κB, 5′-AATGCT ATCCCCAAA GTCCCCTAT-3′ (sense) and 5′-CTCGCGAAGATCATACGCTCTAGT-3′ (antisense).
Foreign body giant cells (FBGCs) were induced into c-Fos–deficient mice by implantation of polyvinyl sponges in vivo as described.(3) Implanted sponges were extracted and fixed in acetone for 24 h, embedded in paraffin, and cut into 4-μm sections. Sections were stained with H&E.
AP-1 and NFAT cooperatively regulate dc-stamp expression during osteoclastogenesis
To clarify how cell–cell fusion is regulated in osteoclasts, we analyzed transcriptional regulation of dc-stamp in osteoclasts. Because dc-stamp is a target of RANKL stimulation, dc-stamp expression is likely regulated by transcription factors downstream of RANKL-RANK signals, such as c-Fos and NFATc1. We identified a 4- (−3672 to +237 bp) and 0.2-kb (−213 to −1 bp) region flanking the 5′ end of dc-stamp (Fig. 1A). One potential AP-1 site, three putative NFAT binding sites, and a TATA box were observed within 0.2-kb upstream of the transcriptional start site (TSS), all of which are conserved between human and mouse (Fig. 1A). We performed promoter analyses in RAW264.7 osteoclast precursor cells, which upregulate dc-stamp expression in the presence of RANKL (Fig. 1B). dc-stamp has at least two splice variants (accession numbers AY517483 and AY517484), and both were upregulated in RAW264.7 cells by RANKL (Fig. 1B). RANKL activated both the 4- and 0.2-kb promoter constructs 20-fold higher than a promoter-less control in RAW264.7 cells (data not shown), suggesting that the 0.2-kb promoter is sufficient for RANKL induction of dc-stamp. Thus, the 0.2-kb promoter was used for the remaining experiments. Mutation of the most proximal NFAT (−88 to −82) or AP-1 (−102 to −98) site abolished efficient activation by RANKL, indicating that these tandem NFAT and AP-1 sites are needed for dc-stamp expression in osteoclasts (Fig. 1C). Therefore, NFATc1and AP-1 transcription factors such as c-Fos are likely positive regulators of dc-stamp expression in osteoclasts.
To analyze the requirement for c-Fos in dc-stamp expression, cells derived from c-Fos–deficient mice were cultured in the presence of M-CSF and RANKL. dc-stamp was not induced in c-Fos–deficient cells treated with M-CSF and RANKL; however, its expression was efficiently rescued by retroviral transduction of c-Fos but not of ΔNFAT (Fig. 1D). NFAT family proteins reportedly form a transcriptional complex with AP-1 heterodimers to activate target gene expression.(18) Moreover, recent studies report that Nfatc1 expression is abrogated in osteoclasts from c-Fos–deficient mice and that Nfatc1 is transcriptionally regulated by c-Fos in vitro.(6,7) Taken together, these results strongly suggest that c-Fos and NFAT cooperatively regulate dc-stamp expression during osteoclastogenesis.
DC-STAMP is needed for macrophage fusion
It has been shown that cytokine combinations that include IL-4, such as GM-CSF + IL-4, M-CSF + IL-4, and IL-3 + IL-4, induce macrophage giant cells in vitro. dc-stamp was also cloned from an IL-4–induced macrophage library and is an IL-4–induced (FIND) gene.(19) To determine whether IL-4 regulates macrophage fusion through dc-stamp expression, bone marrow cells derived from wildtype and DC-STAMP–deficient mice were cultivated with M-CSF for 3 days, and M-CSF–dependent adherent cells were cultivated in the presence of M-CSF, M-CSF + RANKL, M-CSF + IL-4, IL-3 + IL-4, or GM-CSF + IL-4 for 6 days. Both RANKL- and IL-4–induced cell fusion were completely abrogated in DC-STAMP–deficient cells (Figs. 2A and 2B). DC-STAMP–deficient spleen cells also failed to fuse (data not shown). Because enhanced green fluorescent protein (EGFP) had been knocked into the dc-stamp locus in knockout mice, dc-stamp transcriptional activation was detected as EGFP expression in DC-STAMP–deficient cells by immunohistochemistry or in wildtype cells by RT-PCR under fusion-inducing culture conditions described above (Figs. 2A and 2C). The two splice variants of dc-stamp isolated in osteoclasts were also detected under all macrophage giant cell–inducing conditions tested (namely, M-CSF + IL-4, IL-3 + IL-4, or GM-CSF + IL-4). Therefore, DC-STAMP mediates macrophage giant cell formation and multinuclear osteoclast formation.
c-Fos is essential for osteoclast fusion but not for formation of macrophage giant cells
c-Fos regulates dc-stamp expression in osteoclastogenesis (Fig. 1), and Fos expression was detected in macrophage giant cells induced by M-CSF + IL-4, IL-3 + IL-4, and GM-CSF + IL-4 (Fig. 2C). To examine the role of c-Fos in formation of multinuclear cells, osteoclasts (M-CSF + RANKL) and macrophage giant cells (M-CSF + IL-4, IL-3 + IL-4, and GM-CSF + IL-4) were induced in spleen cells derived from wildtype and c-Fos–deficient mice (Fig. 3). Consistent with previous reports,(6,7,20) no multinuclear osteoclasts were generated in spleen cells derived from c-Fos–deficient mice in the presence of M-CSF and RANKL (Fig. 3A). Surprisingly, however, as many macrophage giant cells were generated in spleen cells derived from c-Fos–deficient mice as in wildtype mice, indicating that c-Fos is essential for osteoclast cell fusion but not macrophage giant cell fusion (Figs. 3A and 3B). The separate activities of c-Fos in osteoclast versus macrophage giant cell fusion and DC-STAMP expression were confirmed by in vivo analysis, which showed that macrophage giant cells formed in c-Fos–deficient mice as they did in wildtype mice (Fig. 3C) and that dc-stamp expression was also induced in c-Fos–deficient mice (Fig. 3D). c-Fos–deficient macrophage giant cells also have a similar potential for phagocytosis as wildtype cells (Fig. 3E), suggesting that lack of c-Fos does not affect their phagocytotic activities. Fos expression was induced interchangeably with M-CSF + RANKL or IL-4 + GM-CSF in osteoclasts and macrophage giant cells (Fig. 3F), suggesting that downstream signaling from a variety of receptors converge to induce Fos expression in osteoclasts and macrophage giant cells.
NFATc1 is dispensable for macrophage giant cell fusion
We showed that NFATc1 is needed for dc-stamp expression in osteoclasts (Fig. 1). In addition, expression of Nfatc1 was observed during macrophage giant cell formation (Fig. 2C). Thus, we examined the role of NFATc1 by inhibiting NFATc1 activation during macrophage giant cell formation. Because NFATc1 activated by RANKL stimulation undergoes efficient nuclear translocation through Ca2+-dependent activation of calcineurin,(21) FK506, a calcineurin inhibitor, was added to giant cell–inducing cell cultures (Fig. 4). Interestingly, although multinuclear osteoclast formation in the presence of M-CSF plus RANKL was abrogated by FK506, giant cell formation was insensitive to FK506 (Figs. 4A and 4B). Under these conditions, we measured the surface area occupied by macrophage giant cells or osteoclasts and counted the number of multinuclear cells (Fig. 4B). As expected, the larger the multinucleated cells, the fewer their numbers in vitro, particularly in M-CSF + RANKL– or GM-CSF + IL-4–treated cells in the absence of FK506. Fluorescent immunohistochemical staining revealed that nuclear translocation of NFATc1 was seen in osteoclasts in the presence of M-CSF + RANKL. In contrast, NFATc1 nuclear translocation was not observed in macrophage giant cells cultured in the presence of GM-CSF + IL-4 (Fig. 4C), although Nfatc1 mRNA was detected during macrophage giant cell formation (Fig. 2C). These results indicate that NFATc1 is dispensable for cell fusion during formation of macrophage giant cells.
c-Fos and NFATc1 are not needed for dc-stamp expression in macrophage giant cells
Finally, we examined dc-stamp expression in osteoclasts and macrophage giant cells derived from spleen cells of c-Fos–deficient mice by RT-PCR. Interestingly, dc-stamp was induced in macrophage giant cells even in the absence of c-Fos, indicating that c-Fos is dispensable for both cell fusion and dc-stamp expression in macrophage giant cells (Fig. 5A). Because various AP-1 family transcription factors such as JunB and JunD were expressed in spleen cells derived from c-Fos–deficient and wildtype mice (data not shown), other AP-1 family members may substitute for c-fos in activating dc-stamp expression in macrophage giant cells. Moreover, FK506 did not inhibit dc-stamp expression in giant cell–inducing culture conditions (Fig. 5B). To understand differences in the requirement for c-Fos and NFATc1 for cell–cell fusion and dc-stamp expression in osteoclasts and macrophage giant cells, ChIP analysis was performed targeting the dc-stamp promoter in a region containing the most proximal NFAT and AP-1 sites (Fig. 5c). ChIP analysis with c-Fos and NFATc1 antibodies showed that both endogenous c-Fos and NFATc1 were recruited to the dc-stamp promoter in cells grown in osteoclast differentiation conditions (M-CSF + RANKL), whereas neither factor was recruited to the promoter in macrophage giant cell–inducing conditions (GM-CSF + IL-4). Thus, other transcription factors may induce dc-stamp expression in macrophage giant cells. PU.1 and NF-κB are transcription factors activated under myeloid differentiation and inflammatory conditions, and we found a putative PU.1 binding site between –284 and −262, as well as a putative NF-κB binding site at –674 to −665 in the dc-stamp promoter. Because PU.1 and NF-κB are both expressed in macrophage giant cell–inducing conditions (Fig. 5D), ChIP analysis targeting the dc-stamp promoter region containing these sites was performed. As shown in Figs. 5E and 5F, both endogenous PU.1 and NF-κB were recruited to the dc-stamp promoter in cells cultured in macrophage giant cell–inducing conditions much more efficiently than in cells grown in osteoclast differentiation conditions. These results showed that DC-STAMP is needed for cell–cell fusion in both osteoclasts and macrophage giant cells, but that transcriptional regulation of dc-stamp differs between osteoclasts and macrophage giant cells.
We previously reported that DC-STAMP is a membrane protein essential for fusion of osteoclasts and FBGCs.(3) Genetic ablation of dc-stamp leads to complete lack of cell fusion of osteoclasts and FBGCs, whereas osteoclasts derived from DC-STAMP–deficient mice express typical osteoclast markers such as TRACP and cathepsin K and exhibit ruffled border formation. Thus, DC-STAMP is specifically needed for cell–cell fusion rather than differentiation; however, how fusion is mediated by DC-STAMP in osteoclasts and FBGCs is unclear. We show here that, although cell fusion of osteoclasts and macrophages is regulated by the common receptor DC-STAMP, expression of that receptor is differentially regulated in these two cell types.
Osteoclasts and macrophage giant cells are derived from macrophage precursors. Several reports suggest that IL-4 suppresses RANKL-induced osteoclastogenesis,(12–14) suggesting that IL-4 suppresses c-Fos–NFATc1 signaling, a key regulator of differentiation, dc-stamp expression, and cell–cell fusion in osteoclasts. On the other hand, IL-4 induces macrophage giant cell formation,(10,11) suggesting that IL-4 may not inhibit signaling pathways functioning in cell–cell fusion in macrophages. Because NFATc1 is not needed for dc-stamp expression and fusion in macrophage giant cells induced by IL-4, dc-stamp expression must be regulated in a cell type–specific manner. Indeed, RANKL induced recruitment of c-Fos and NFATc1 to the dc-stamp promoter, whereas IL-4 induced PU.1 and NF-κB to the dc-stamp promoter in macrophage giant cells. Fos and NFATc1, both of which are essential transcription factors for osteoclast differentiation, are not needed for macrophage giant cell formation, showing that osteoclasts and macrophage giant cells are different cell types. On the other hand, both osteoclasts and macrophage giant cells activate Fos transcription interchangeably with GM-CSF + IL-4 and M-CSF + RANKL. This suggests that signaling downstream of several receptors may converge on dc-stamp expression. However, DC-STAMP is absolutely needed for fusion in both osteoclasts and macrophages after stimulation with either M-CSF + RANKL or GM-CSF + IL-4, and thus, cell–cell fusion is not rescued in DC-STAMP–deficient cells by the interchangeable signaling.
Although osteoclasts and macrophage giant cells are induced by distinct pathways, DC-STAMP is needed for fusion of both cell types. Fusion of osteoclasts increases the efficiency of bone resorption.(3) Thus, it is reasonable that dc-stamp expression is regulated by master regulators of osteoclastogenesis, namely, c-Fos and NFATc1. In contrast, macrophage giant cells are detected at sites of chronic inflammation; therefore, dc-stamp expression in these cells might be controlled by factors upregulated in inflammation. Our results suggest that differences in NFATc1 dependency may underlie differences in TRACP activity of osteoclasts and macrophage giant cells, because TRACP expression is regulated by nuclear translocation of activated NFATc1 in vitro and in vivo.(6,7,21,22)
It is likely that a putative DC-STAMP ligand is either expressed on the membrane or secreted from fusing osteoclasts and macrophages. Macrophage fusion receptor (MFR), CD47, CD44, CD9, and CD81 are reported to participate in cell–cell fusion in macrophage lineage cells.(23–26) The structural similarity between DC-STAMP and chemokine receptors, all of which are seven-membrane-spanning proteins, suggests that a chemokine is a potential DC-STAMP ligand. Monocyte chemoattractant protein-1 (MCP-1), a CC-type chemokine expressed in macrophage/monocyte lineage cells, functions in chronic inflammatory reactions.(27) Genetic ablation of McP1 causes compromised fusion of macrophages in foreign body reactions,(28) but multinuclear osteoclast formation has not been described. Moreover, it has been shown that MCP-1 is induced by RANKL and promotes human osteoclast fusion with or without RANKL.(29) Because cell fusion and dc-stamp expression are distinctly regulated in osteoclasts and macrophage giant cells, DC-STAMP ligand expression might also be differentially regulated in both cell types. We detected Mfr, CD47, CD44, McP1, and Meltrinα expression by RT-PCR in both osteoclasts and macrophage giant cells (Fig. 6), suggesting these molecules are candidate DC-STAMP ligands. Further studies are necessary to identify that ligand.
Myoblasts derived from mesenchymal stem cells fuse to form multinuclear myotubes. It has been shown that the gene encoding IL-4 is an NFATc2 target in muscle and IL-4 induces myoblast fusion. Disruption of genes encoding IL-4 and the IL-4α receptor subunit reduces myotube size and myonuclear number in vivo.(30,31) Therefore, DC-STAMP, a downstream regulator of IL-4–mediated fusion, potentially regulates myoblast fusion. Thus, similar mechanisms may function in myoblast and macrophage fusion.
In conclusion, our results reveal that, although DC-STAMP is a common membrane protein essential for fusion of macrophage lineage cells, dc-stamp expression and cell–cell fusion are induced through distinct set of transcription factors.
The authors thank Y Sato and A Kumakubo for technical support. TS was supported by a grant-in–aid from Specially Promoted Research of the Ministry of Education, Science, Sports and Culture, Japan; TM was supported by a grant-in-aid for Young Scientists (B) and the Takeda Science Foundation, Japan; and MY was supported by a grant-in-aid from the 21st century COE Program of the Ministry of Education, Culture, Sports, Science and Technology, Japan, to Keio University.