SEARCH

SEARCH BY CITATION

Keywords:

  • DICAM;
  • OSTEOCLAST;
  • INTEGRIN αVβ3;
  • p38 MAP KINASE

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Dual immunoglobulin (Ig) domain-containing adhesion molecule (DICAM) is involved in cell–cell adhesion through a heterophilic interaction with αVβ3 integrin, which suggests that DICAM may participate in osteoclast differentiation. DICAM was localized in the plasma membrane of RAW264.7 and THP-1 cells, and its expression gradually increased during osteoclastogenesis in mouse bone marrow-derived macrophages (BMMs) treated with receptor activator of nuclear factor κ-B ligand (RANKL) and macrophage colony-stimulating factor (M-CSF). Forced expression of DICAM in BMMs and RAW264.7 cells blocked the generation of tartrate-resistant acid phosphatase (TRAP)-positive osteoclasts. Conversely, knockdown of DICAM by small hairpin RNA (shRNA) increased osteoclast formation in RAW264.7 cells. DICAM-mediated suppression of osteoclast differentiation was in part due to the inhibition of the p38 mitogen-activated protein (MAP) kinase pathway, which was corroborated by a decrease in the expression of c-Fos and nuclear factor of activated T cells (NFAT)c1. Mechanistically, DICAM directly interacted with integrin β3, which inhibited heterodimerization between integrin αV and β3. Exogenous expression of integrin β3 or high-dose M-CSF rescued DICAM-mediated inhibition of osteoclastogenesis, suggesting crosstalk between the integrin β3 and c-Fms pathways. Finally, recombinant DICAM ectodomain suppressed the RANKL- and M-CSF–induced osteoclastogenesis of BMMs. Collectively, these results indicate that DICAM acts as a negative regulator of osteoclast differentiation by suppressing the integrin αVβ3 pathway. © 2012 American Society for Bone and Mineral Research.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Osteoclastogenesis is a complicated process regulated by various systemic hormones and cytokines produced locally in the bone microenvironment. Much progress has been made in understanding the major molecular mechanisms, and it is now known that receptor activator of nuclear factor κ-B ligand (RANKL) and macrophage colony-stimulating factor (M-CSF) are necessary and sufficient for osteoclastogenesis.1, 2 RANKL and M-CSF are required to induce the expression of genes that typify the osteoclast lineage including those encoding tartrate-resistant acid phosphatase (TRAP), cathepsin K, calcitonin receptor, and integrin β3.3 Among these genes, the integrins, a superfamily of heterodimeric transmembrane receptors, mediate cell–matrix and cell–cell interactions and play a central role in the capacity for osteoclasts to degrade bone.4

Osteoclasts express several integrin heterodimers including αVβ3, α2β1, αVβ1, and α9β1.4, 5 Of these, integrin αVβ3 is the principal and most abundant integrin mediating bone resorption.4, 6, 7 Several lines of evidence indicate that blocking integrin αVβ3 function by genetic ablation or using pharmacologic antagonists, such as monoclonal antibodies or arginine-glycine-aspartate (RGD) peptides, severely attenuates differentiation as well as the resorptive functions of osteoclasts.8–10 Integrin β3-deficient bone marrow macrophages (BMMs) show impaired differentiation into osteoclasts.11 In addition, integrin β3-knockout mice reveal an osteopetrotic phenotype as a consequence of osteoclast dysfunction, due to cytoskeletal abnormalities resulting in failure to form actin rings and ruffled membranes.12 Interestingly, despite their increased bone mass, integrin β3-knockout mice exhibit an apparent increase in the number of osteoclasts per trabecular bone surface.12 This increase of osteoclasts may result from compensatory effects via upregulation of the other key regulators of osteoclastogenesis. In fact, integrin β3 null mice have increased levels of circulating and bone-marrow microenvironmental M-CSF.12 These observations suggest that integrin β3 has a common signaling pathway or crosstalk with M-CSF/c-Fms in osteoclastogenic processes.13

Several receptor tyrosine kinases are known to associate with integrins. These interactions enhance integrin-dependent cellular events, via “inside-out” signaling, inducing either conformational changes in the extracellular regions and/or integrin diffusion and clustering.14 M-CSF, the major growth factor of monocyte lineage cells including osteoclasts, is also involved in the conformational change of integrin αVβ3 into its high-affinity state via targeting the integrin β3 cytoplasmic domain.15, 16 In addition, c-Fms activation induces a direct interaction with integrin αVβ3 at the podosomal actin ring of osteoclasts.17 This complex containing c-Fms and integrin αVβ3 also contains signaling molecules such as Pyk2, p130Cas, and c-Cbl, which mediate integrin-stimulated signaling; ie, “outside-in” signaling.17, 18 These results demonstrate that integrin αVβ3 and c-Fms collaborate during osteoclast differentiation by involving both “inside-out” and “outside-in” integrin signaling.

We have recently identified a cortical thymocyte marker of Xenopus (CTX) protein family, dual immunoglobulin (Ig) domain-containing adhesion molecule (DICAM), which is a type I transmembrane protein with two V-type Ig domains in the extracellular region and a short cytoplasmic tail.19 Phylogenetic analysis of DICAM indicates a high homology with the junctional adhesion molecule (JAM) protein family. JAM family proteins are major components of junctional complexes in endothelial and epithelial cells as well as mediators of leukocyte-endothelial cell interaction due to their ability to undergo heterophilic binding with integrins. Functional blocking assays with integrin antibodies demonstrate that the extracellular domain of DICAM exerted cell adhesion activity, in part through binding to integrin αVβ3.19 Therefore, it is plausible to hypothesize that DICAM may be involved in osteoclastogenic processes through integrin αVβ3.

In this study, we focused on the function of DICAM during osteoclastogenesis. DICAM inhibited integrin αVβ3 signaling through the suppression of integrin β3 expression and heterodimerization between αV and β3. DICAM-mediated inhibition of osteoclastogenesis was rescued by forced expression of integrin β3 or high doses of M-CSF, suggesting the presence of crosstalk between integrin β3 and M-CSF/c-Fms pathways. These findings indicate that DICAM is an endogenous regulator of osteoclastogenesis.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Reagents and antibodies

Recombinant human M-CSF and mouse RANKL were obtained from PeproTech (Rocky Hill, NJ, USA) and R&D Systems (Minneapolis, MN, USA), respectively. The DICAM polyclonal antibody (pAb) was developed as described.19 Antibodies against αV integrin, p-Src (Y416), Src, p-p38 (T180/Y182) mitogen-activated protein kinase (MAPK), p38 MAPK, and c-Fos were obtained from Cell Signaling Technology (Beverly, MA, USA). Antibodies against nuclear factor of activated T cells (NFAT)c1 (from rabbit; Abcam, Cambridge, MA, USA), c-Myc (from mouse; Invitrogen, Camarillo, CA, USA), and β-actin (Sigma-Aldrich, St. Louis, MO, USA) were used for Western blot analyses.

Cell culture, transfection, and osteoclast differentiation

Primary cultured BMMs were obtained from 8-week-old mice as described.20 RAW264.7 and HEK-293 cells were maintained in Dulbecco's modified Eagle medium (DMEM) containing 10% fetal bovine serum (FBS) and antibiotics. RAW264.7 cells were plated at a density of 1.0 × 106 cells/well and transfected using FuGENE HD reagent (Roche, Indianapolis, IN, USA) with a total of 2 µg of pcDNA3.1-myc/his-C-DICAM and DICAM small hairpin RNA (shRNA). Clones were selected using G418 (Invitrogen). HEK-293 cells were also transfected using FuGENE HD reagent. RAW264.7 cells were plated at 8.0 × 104 cells/well in a six-well plate and were allowed to differentiate for 4 to 5 days into mature osteoclasts in phenol-free α modified essential medium (α-MEM) containing 10% FBS in the presence of recombinant RANKL (30 ng/mL). The formation of mature osteoclasts was evaluated using TRAP staining. For retrovirus construction, the full-length cDNA of mouse DICAM was amplified by PCR and inserted into pMX vectors. Plat-E packaging cells (2 × 106) were transfected with 20 µg of the DICAM retroviral vectors using FuGENE. After 48 hours, retrovirus was collected and introduced into BMMs for 24 hours in the presence of 30 ng/mL M-CSF and 4 µg/mL polybrene (Sigma-Aldrich). BMMs were selected in the presence of 30 ng/mL M-CSF and 1 µg/mL blasticidine S (Calbiochem, San Diego, CA, USA) for 3 days, and osteoclastogenesis was induced by addition of M-CSF (30 ng/mL) and RANKL (30 ng/mL) for 6 days. Mature osteoclasts were washed and fixed for TRAP staining as described.21 Resorption pit formation assays were performed using the Osteo Assay kit (Corning Inc., Corning, NY, USA).

Fluorescence-activated cell sorting analyses and immunofluorescent staining

To determine surface expression of DICAM, we conducted fluorescence-activated cell sorting (FACS) analyses. For labeling, cells were incubated with the DICAM antibody for 1 hour on ice, followed by fluorescent anti-rabbit Ig in phosphate-buffered saline (PBS) supplemented with 1% bovine serum albumin (BSA, pH7.5). Isotype-matched irrelevant polyclonal antibodies were used as controls for background nonspecific staining. Cells were analyzed on a FACS Aria II (BD Biosciences, Franklin Lakes, NJ, USA).

For immunofluorescent staining, cells were washed with PBS, fixed for 10 minutes with 4% paraformaldehyde in PBS, and permeabilized for 5 minutes on ice with 0.25% Triton X-100 in PBS. After blocking with 2% BSA/PBS for 1 hour, the slide was incubated with the DICAM polyclonal antibody (1:200 dilution) or normal rabbit IgG (1:200 dilution) in 1% BSA at room temperature (RT) for 1 hour. Cells were washed three times with PBS, incubated for 40 minutes at RT with Alexa Fluor 488 donkey anti-rabbit IgG antibody (1:200) (Invitrogen) in 1% BSA/PBS, and followed by three washes in PBS. The slides were mounted with anti-fade mounting solution (Invitrogen) and photographed using a fluorescence microscope (Olympus, Tokyo, Japan).

RNA preparation and RT-PCR

Total RNA was extracted from cultured cells using Trizol reagent (Invitrogen). Complementary DNA was synthesized, and PCR was performed using the following primers: DICAM: 5′-CCTGCTTGACCTGTATGCA (forward) and 5′-GCTACAGCAAACTCCACCA (reverse); GAPDH: 5′-TGAGAACGGGAAGCTTGTCA (forward) and 5′-GGAAGGCCATGCCAGTGA (reverse); αV integrin: 5′-ATTGTACCACTGGAGGACTGA (forward) and 5′-AAGTCCTTGCTGCTCTTGGA (reverse); β3 integrin: 5′-TGACTCGGACTGGACTGGCTA (forward) and 5′-CACTCAGGCTCTTCCACCACA (reverse); TNF receptor associated factor 6 (TRAF6): 5′-GTTTGACCCACCTCTGGAGA (forward) and 5′-TATGCCATGGACACAGCACA (reverse); NFATc1: 5′-TTGGTGTAGGCAGGAGAGGA (forward) and 5′-GTTTGACCCACCTCTGGAGA (reverse); Oscar: 5′-CTGGCTGAGTTCTTCCTGGA (forward) and 5′-AGCCAGAACCTTCGAAACTGA (reverse); c-Src: 5′-TCTGCAAAGGAGATGTGCTCA (forward) and 5′-TTTGATGAGGCGAGTGCAGA (reverse); and cathepsin K: 5′-CTTGTGGACTGTGTGACT (forward) and 5′-AACACTGCATGGTTCACA (reverse). The number of cycles was selected so that the amplification was remained in the linear range.

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphyenyltetrazolium bromide assay and Bromodeoxyuridine ELISA cell proliferation assay

Cell viability was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphyenyltetrazolium bromide (MTT) assay. After 48 hours of treatment with or without RANKL and in the presence of M-CSF, 50 µL MTT was added to each well, and the plate was incubated for 4 hours at 37°C. The media was removed and the cells were solubilized with 150 µL dimethylsulfoxide. The spectrophotometric absorbance of each sample was then measured at 570 nm.

The proliferative capacity of BMMs was determined by a bromodeoxyuridine (BrdU) ELISA cell proliferation assay according to the manufacturer's protocol (Cell Signaling Technology). In brief, transduced BMMs were incubated for 24 hours with 10 µM BrdU solution and fixed with fixing/denaturing solution at RT for 30 minutes. After removing the solution, cells were incubated for 1 hour with BrdU monoclonal antibody (100 µL). After incubation with horseradish peroxidase (HRP)-conjugated secondary antibody, 100 µL 3,3′,5,5′-tetramethylbezidine substrate was added for 30 minutes at RT. The reaction was stopped by adding 100 µL of stop solution. The absorbance was measured at 450 nm using a spectrophotometric microplate reader.

Western blotting and immunoprecipitation

Western blot analysis was performed to assess the levels of RANKL-mediated signaling proteins in total cell lysates using specific antibodies. Cell lysates from each day of osteoclastogenesis were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA). The transferred membrane was blocked using 5% skim milk at RT for at least 1 hour, and RANKL-mediated signaling proteins and DICAM were detected using specific as antibodies described above. After incubation with HRP-conjugated secondary antibodies, the blots were visualized using enhanced chemiluminescence (ECL) solution (Amersham Biosciences, Buckinghamshire, UK).

For immunoprecipitation assays, semi-confluent HEK-293 cells were transfected using FuGENE HD reagent with pcDNA-DICAM and/or integrin β3. After 1 day, transfected HEK-293 cell lysates were precleared using protein A-sepharose suspension and incubated with c-myc monoclonal antibody, integrin αV polyclonal antibody, or integrin β3 polyclonal antibody for 1 hour. Fifty microliters of a 50% protein A-sepharose suspension was added and cells were incubated for 1 hour. Beads were collected by centrifugation at 3000 rpm for 30 seconds, washed three times with lysis buffer, and then analyzed by immunoblotting.

Purification of recombinant DICAM protein

Recombinant DICAM (rDICAM) protein was purified, as described.19 Briefly, pET28(a)-DICAM-transformed E. coli strain BL21 were cultured overnight and recombinant protein expression was induced using 1 mM of isopropyl-β-D-thiogalactopyranoside (IPTG) for 2 hours. Cells were pelleted by centrifugation, suspended in lysis buffer (50 mM Tris-HCl, 100 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride [PMSF], 0.5 mM dithiothreitol [DTT]), and sonicated 5 times for 10 seconds each, with 50 seconds rest between sonications. After centrifugation, nickel-nitrilotriacetic acid (Ni-NTA) agarose resin was added to the supernatant and incubated at 4°C for 1 hour. The mixture was added to the chromatograph column and the resin was washed with five volumes of binding buffer (20 mM Tris-HCl, 0.5 M NaCl, 5 mM imidazole). After another wash with washing buffer (20 mM Tris-HCl, 0.5 M NaCl, 20 mM imidazole), the recombinant protein was eluted five times in 200 µL of elution buffer (20 mM Tris-HCl, 0.5 M NaCl, 0.3 M imidazole). The E. coli endotoxin (lipopolysaccharide) was removed from the recombinant protein using the EndoClean kit (Viovintage, San Diego, CA, USA) according to the manufacturer's protocol.

Recombinant DICAM-ecto-Fc (rDICAM-Fc) protein (amino acids 1–301) was purified, as described.19 Briefly, the DICAM cDNA containing two Ig domain regions (DICAM-ecto) was inserted into pcDNA3.1-Fc (human IgG1 Fc) vector and transfected into HEK293 cells. After 1 day, protein A-sepharose beads were used to obtain purified rDICAM-Fc protein. Signal peptide-conjugated Fc recombinant protein was used as a control.

Statistical analyses

Data were analyzed statistically using the Mann-Whitney U test to determine the difference between the means of two groups. Values of p < 0.05 were considered statistically significant. All analyses were conducted using SPSS version 14.0 software (SPSS, Chicago, IL, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

DICAM increases during osteoclastogenesis

The expression of DICAM was evaluated in various myeloid cells, including RAW264.7, THP-1, and primary cultured BMMs. Using flow cytometry, DICAM was expressed in the membrane of the myeloid cell lines and BMMs (Fig. 1A). Expression of the 70-kDa DICAM protein was confirmed by Western blot analysis using protein extracts from RAW264.7, THP-1, and BMMs (Fig. 1B). At the single-cell level, DICAM was localized in the plasma membrane in both RAW264.7 and THP-1 cell lines (Supplemental Fig. 1).

thumbnail image

Figure 1. DICAM expression in osteoclast precursor cells and during osteoclastogenesis. (A) FACS analysis of DICAM surface expression in myeloid cell lines RAW264.7 and TPH-1, and primary mouse bone marrow macrophages (BMMs) (filled histograms). Open histograms represent binding of fluorescent nonspecific IgG. (B) Western blot analysis of DICAM protein expression in myeloid cell lines RAW264.7 and TPH-1, and mouse BMMs. Total protein was isolated from unstimulated RAW264.7 and TPH-1, and mouse BMMs, and then subjected to SDS-PAGE for analysis of DICAM expression. Molecular weight of the DICAM protein is 70 kDa. β-actin is used as an internal control. (C, D) DICAM expression in BMMs during osteoclast differentiation. Mouse BMMs obtained from C57BL/6 mice were cultured for 3 days with M-CSF (30 ng/mL), followed by culture in the presence of M-CSF (30 ng/mL) and RANKL (30 ng/mL) for 6 days. In C, The mRNA expression levels of DICAM and other markers such as integrin αvβ3, Oscar, cathepsin K, c-Src, and DC-stamp were analyzed by RT-PCR. GAPDH mRNA is used as an internal control. In D, whole-cell lysates were subjected to immunoblotting with DICAM or integrin β3 antibodies.

Download figure to PowerPoint

We then analyzed the expression level of DICAM during osteoclast differentiation. BMMs were treated for 3 days with M-CSF alone to allow differentiation into macrophages, and subsequently treated with both RANKL and M-CSF to induce formation of multinucleated TRAP+ osteoclasts. During osteoclast differentiation, DICAM mRNA and protein gradually increased until late osteoclastogenesis, when the expression of integrin β3 reached its peak (Fig. 1C, D). Other osteoclastogenic markers, integrin αV, Oscar, Cathepsin K, c-Src, and DC-Stamp, increased during osteoclastogenesis (Fig. 1C). These results indicate that DICAM is gradually upregulated during osteoclastogenesis but decreases at later stages of differentiation.

DICMA inhibits osteoclast differentiation

To identify whether DICAM regulates osteoclast differentiation, we examined the effects of gain- or loss-of-function of DICAM. For the gain-of-function study, RAW264.7 cells were transfected with a myc/His-tagged DICAM plasmid or empty vector. Exogenous expression of DICAM strongly decreased the generation of TRAP+ osteoclasts, which was not restored by high concentrations of RANKL (Fig. 2A). To gain insight into the mechanism of DICAM-mediated inhibition of osteoclastogenesis, the expression of osteoclastogenic genes was measured. DICAM-stable RAW264.7 cells revealed reduced mRNA expression of M-CSF at day 1 and integrin β3, Oscar, and c-Src at day 5 (Fig. 2B). Levels of integrin αV, TRAP6, cathepsin K, and NFATc1 mRNAs were not changed in DICAM-stable cells.

thumbnail image

Figure 2. DICAM inhibits RANKL-induced osteoclastogenesis in RAW264.7 cells. (A) Inhibition of osteoclast differentiation by DICAM. To establish stable cell lines expressing DICAM, RAW264.7 cells were transfected with pcDNA3.1-myc/His-c-DICAM or pcDNA3.1-vector and selected with G418 for 5 days. Stable cell lines were cultured in the presence of different concentrations of RANKL (10, 30, or 50 ng/mL) for 5 days to generate mature multinucleated osteoclasts. The cells were fixed and stained with TRAP activity. TRAP+ multinucleated (>3 nuclei) cells (MNCs) were counted for quantification. (B) RT-PCR analysis of DICAM and differentiation markers of osteoclasts. At day 1 and day 5, total RNA was prepared and expression of the various transcripts was evaluated by RT-PCR. (C) Enhanced RANKL-induced osteoclast differentiation by knockdown of DICAM. RAW264.7 cells transfected with DICAM shRNA were treated with 30 ng/mL RANKL for 5 days. Cells were fixed and stained for TRAP activity, and TRAP+ MNCs were quantified. (D) DICAM knockdown changes the expression pattern of osteoclast differentiation markers. After stimulation with 30 ng/mL RANKL for the indicated time, total RNA was prepared from DICAM knockdown RAW264.7 cells and the expression of osteoclast markers was analyzed by RT-PCR. *p < 0.01.

Download figure to PowerPoint

For the loss-of-function study, DICAM shRNA was introduced during osteoclastogenesis. As expected, RAW264.7 cells transfected with DICAM shRNA plasmid showed a prominent increase in TRAP+ mature osteoclasts (Fig. 2C). Knockdown of DICAM led to increased expression of Oscar and integrin β3. Among osteoclastogenic genes, integrin β3 was the most strongly regulated by DICAM. The expression of c-Src was slightly increased in DICAM knockdown cells at day 2 and returned to basal level at day 4 (Fig. 2D). Although the knockdown was less effective at day 4 than day 2, the DICAM-mediated inhibitory effect on osteoclastogenic genes was still present, as reflected by the upregulation of integrin β3 and Oscar expression at day 4. These results indicate that DICAM negatively regulates osteoclast differentiation in RAW264.7 cells by suppressing integrin β3 expression.

DICAM-mediated inhibition of osteoclastogenesis is rescued by high-dose M-CSF

Next, we tested whether DICAM has an inhibitory role in osteoclast differentiation of primary cultured mouse BMMs, which requires M-CSF and RANKL for the differentiation process. To evaluate this hypothesis, BMMs were transduced with the pMX retroviral DICAM expression construct or empty vector, and selected using blasticidin S (Fig. 3A). BMMs were then cultured with 30 ng/mL RANKL and various concentrations of M-CSF for 6 days, to compensate for the inhibition of M-CSF expression by DICAM (Fig. 2B). When the formation of multinucleated cells was visualized by TRAP staining, DICAM transduction suppressed osteoclast differentiation at low concentrations of M-CSF (5 and 20 ng/mL). Interestingly, however, the suppressive effect of DICAM on osteoclastogenesis was completely rescued by the addition of high-concentration M-CSF (50 ng/mL; Fig. 3B). Additionally, pit formation was decreased at low concentrations but comparable at high concentrations of M-CSF, which correlated with TRAP+ osteoclast formation (Fig. 3C). With regard to osteoclastogenic transcription factors, DICAM reduced c-Fos and NFATc1 expression levels, which were partially rescued by the high concentration of M-CSF (Fig. 3D). DICAM-mediated inhibition of osteoclastogenesis is rescued by high concentrations of M-CSF, suggesting the presence of common interacting pathways between DICAM- and M-CSF–mediated signaling during osteoclastogenesis.

thumbnail image

Figure 3. High-dose M-CSF rescues DICAM-mediated inhibition of osteoclastogenesis in BMMs. (A) BMMs were transduced with retrovirus containing pMX-empty vector or pMX-DICAM in the presence of M-CSF (20 ng/mL) for 1 day. The retrovirally transduced cells were selected with blasticidin S-deaminase (1 µg/mL) for 3 days in the presence of M-CSF (20 ng/mL). Then mRNA and total cell lysates were subjected to RT-PCR and Western blotting, respectively. (B) DICAM inhibited osteoclast formation at low doses of M-CSF, but this was rescued at high-dose M-CSF. Transduced BMMs were cultured in the presence of various concentrations of M-CSF (5, 20, or 50 ng/mL) and 30 ng/mL RANKL for 6 days. The cells were fixed and stained for TRAP activity, and TRAP+ multinucleated (>3 nuclei) cells (MNCs) were counted. (C) Retrovirally-transduced BMMs were cultured on calcium-phosphate-coated plates as described above. Cells were detached with a bleaching solution and images were taken of the resorption pits formed by functional osteoclasts. (D) Transduced BMMs were cultured as described for 2 days. Total cell lysates were subjected to Western blotting for c-Fos and NFATc1. The experiment was performed in triplicate, and representative data from three experiments are shown. Data are presented as the mean ± SD (n = 3). *p < 0.01.

Download figure to PowerPoint

Effect of DICAM on the proliferation of BMMs

The inhibitory effects of DICAM on osteoclastogenesis may be related to the proliferation of osteoclast precursors. Cell density and proliferation are important for the differentiation of monocytes into osteoclasts.22 To test this possibility, MTT and BrdU incorporation assays were performed (Fig. 4). In the presence of M-CSF alone, BMMs overexpressing DICAM showed lower viability and proliferation, indicating that the M-CSF–mediated proliferative effects on BMMs are inhibited by DICAM. When BMMs were treated with both RANKL and M-CSF, however, the exogenous expression of DICAM had no effect on the viability or proliferative activity of BMMs (Fig. 4). These data suggest that RANKL-mediated signaling masks DICAM's suppression of BMMs proliferation mediated by M-CSF and further indicate that the inhibition of osteoclastogenesis is not due to the suppression of cell proliferation.

thumbnail image

Figure 4. DICAM does not attenuate viability and proliferation of osteoclastogenic BMMs. (A) BMMs transduced with DICAM or empty vector were starved overnight, and treated with M-CSF (30 ng/mL) alone or M-CSF (30 ng/mL) plus RANKL (30 ng/mL) for 48 hours. Cell viability was measured by the MTT assay. (B) Transduced BMMs were starved overnight and cultured with BrdU in the presence of M-CSF (30 ng/mL) and RANKL (30 ng/mL) for 24 hours. The level of proliferation of BMMs was assessed by a BrdU-incorporation ELISA assay. Data are presented as the mean ± SD (n = 6). *p < 0.05.

Download figure to PowerPoint

DICAM attenuates p38 MAP kinase signaling in response to M-CSF and RANKL

To elucidate the molecular mechanisms responsible for the inhibition of osteoclastogenesis mediated by DICAM, we evaluated M-CSF and RANKL signaling molecules by immunoblotting. Activation of M-CSF signaling was achieved with varying concentrations of M-CSF and changes in well-known osteoclastogenic genes were determined at the macrophage stage.23 Despites DICAM overexpression, MAP kinases and NF-κB pathway molecule, I-κB, showed no differences in expression upon stimulation with 20 or 50 ng/mL M-CSF (Fig. 5A). Interestingly, when a low dose of M-CSF (5 ng/mL) was applied, p38 MAP kinase activity was attenuated, but others were not changed by DICAM (Fig. 5A). Next, we evaluated changes in signaling cascades by RANKL stimulation (100 ng/mL) at the preosteoclast stage (Fig. 5B). In response to RANKL stimulation, DICAM inhibited p38 MAP kinase phosphorylation. However, other MAP kinases, including extracellular signal-related kinase (ERK) and c-Jun-N-terminal kinase (JNK), as well as I-κB, were not affected (Fig. 5B). Taken together, these results indicate that the inhibitory role of DICAM during osteoclast differentiation is mediated by attenuation of the p38 MAP kinase pathway in response to M-CSF and/or RANKL.

thumbnail image

Figure 5. DICAM attenuates p38 MAP kinase signaling. (A) MAP kinase and NF-κB signaling after M-CSF treatment in BMMs. BMMs infected with retrovirus expressing DICAM or empty vector were selected with blasticidin S-deaminase (1 µg/mL) for 3 days in the presence of M-CSF (20 ng/mL). Transduced BMMs were starved overnight and treated with various concentration of M-CSF (5, 20, or 50 ng/mL) for 5 minutes, and cell lysates were prepared and subjected to immunoblot analyses with the indicated antibodies. (B) MAP kinase and NF-κB signaling after RANKL treatment in preosteoclasts. Transduced BMMs were cultured with RANKL (30 ng/mL) and M-CSF (30 ng/mL) for 3 days, starved overnight, and then stimulated with 100 ng/mL RANKL for the indicated time period (0, 5, or 30 minutes). Activation of p38 MAP kinase, ERK and JNK, and I-κB were assessed by Western blot analyses.

Download figure to PowerPoint

Role of integrin β3 interaction with DICAM during osteoclast differentiation

The functional rescue of DICAM-mediated inhibition of osteoclast differentiation by high-dose M-CSF suggests that DICAM may be associated with M-CSF signaling or with other pathways which affect M-CSF signaling, such as integrin αVβ3. It has been reported that β3 integrin collaborates with M-CSF to enhance osteoclast differentiation.13 In a previous study, we determined that DICAM was involved in cell-to-cell adhesion by interacting with integrin αVβ3.19 Based on these data, we hypothesized that interaction of DICAM with integrin αVβ3 affects M-CSF signaling during osteoclast differentiation. Therefore, we investigated whether DICAM and integrin β3 directly interact with each other. As shown in Fig. 6A, integrin β3 clearly associated with DICAM, as determined by coimmunoprecipitation and Western blot analyses. In addition, the interaction between integrin αV and β3 was suppressed in the presence of DICAM (Fig. 6A).

thumbnail image

Figure 6. Role of integrin β3 in DICAM-mediated inhibition of osteoclastogenesis. (A) DICAM interacts with integrin β3 and attenuates heterodimeric bonds between integrin αV and β3. pcDNA 3.1-myc/His-DICAM and/or pcDNA3-ITGB3 were transfected into HEK293 cells and cultured for 48 hours. The cell lysates were immunoprecipitated with c-myc, integrin αV, or integrin β3 antibodies, and followed by immunoblot with the indicated antibodies. (B) DICAM reduces the duration of ERK signaling. Transduced BMMs were treated with various concentration of M-CSF (5, 20, or 50 ng/mL) and 30 ng/mL RANKL for 2 days. Cell lysates were prepared and subjected to immunoblot analysis. (C) Integrin β3 rescues DICAM-mediated inhibition of c-Fos and NFATc1. After selection for DICAM transduced BMMs, cells were then transfected with pcDNA3-ITGB3. These BMMs were cultured for another 3 days with 30 ng/mL RANKL and M-CSF. Total cell lysates were subjected to Western blotting with the indicated antibodies.

Download figure to PowerPoint

It is known that integrin β3 signaling influences the duration of ERK activity during RANKL-directed osteoclast differentiation.24 Based on the interaction of DICAM with integrin β3, we further investigated whether DICAM affects the duration of ERK activity. At lower concentrations of M-CSF (5 ng/mL and 20 ng/mL), DICAM reduced the expression of integrin β3 as well as ERK activity (Fig. 6B). At high concentrations of M-CSF (50 ng/mL), however, the ERK activity and integrin β3 expression recovered (Fig. 6B). Moreover, exogenous expression of integrin β3 rescued DICAM-mediated suppression of c-Fos and NFATc1 expression (Fig. 6C), confirming the involvement of the integrin β3 signaling cascade in DICAM-mediated inhibition of osteoclastogenesis. These results suggest that DICAM inhibits integrin signaling by suppressing integrin β3 expression and attenuating the interaction between integrin αV and β3.

Effect of DICAM ecto-domain on osteoclast differentiation

Based on the observation that the role of DICAM in inhibiting osteoclast differentiation depends on integrin β3 signaling, we further investigated whether the extracellular ecto-domain of DICAM is sufficient for DICAM's suppressive effects on osteoclastogenesis. When BMMs, in the presence of M-CSF and RANKL, were treated with various doses of recombinant DICAM-Fc (rDICAM-Fc) purified from HEK293 cells, the formation of TRAP+ multinucleated cells was attenuated in a dose-dependent manner (Fig. 7A, B). The rDICAM-Fc protein also suppressed the expression of integrin β3 and NFATc1, as well as suppressed p38 MAP kinase activation. However, ERK phosphorylation was not changed (Fig. 7C). Because the suppression of osteoclastogenesis by exogenous DICAM expression was rescued by treatment with high-dose M-CSF, we tested osteoclast differentiation with rDICAM-Fc in the presence of various doses of M-CSF. The formation of TRAP+ multinucleated cells was reduced by rDICAM-Fc in the presence of low-dose M-CSF (5 ng/mL and 20 ng/mL), and this was partially rescued by high-dose M-CSF (50 ng/mL; Fig. 7D, E). The rDICAM-mediated suppressive effect was not rescued by high doses of RANKL in the RAW264.7 cell culture system (Supplemental Fig. 2). These results suggest that the inhibitory effect of DICAM on osteoclastogenesis is due to an interaction between the DICAM ecto-domain and integrin αVβ3.

thumbnail image

Figure 7. Recombinant DICAM (rDICAM) inhibits osteoclast differentiation. (A) BMMs were cultured in the presence of M-CSF (20 ng/mL) and RANKL (30 ng/mL) with indicated concentrations of rDICAM-Fc for 6 days. Cells were fixed and stained for TRAP activity. (B) The number of TRAP+ multinucleated (>3 nuclei) cells (MNCs) was counted and graphed. (C) BMMs were cultured with M-CSF (20 ng/mL) and RANKL (30 ng/mL) in the presence or absence of rDICAM-Fc (1 µg/mL) for 3 days, and Western blotting was performed to detect integrin β3 and NFATc1 expression. (D) BMMs were cultured with RANKL (30 ng/mL) and indicated concentrations of M-CSF in the presence of 1 µg/mL of rDICAM-Fc or control-Fc. The cells were fixed and TRAP-stained. (E) The number of TRAP+ NMCs was counted and graphed. All figures are representative of three independent experiments. Data are presented as the mean ± SD (n = 3). *p < 0.05.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

In this study, we identified that DICAM, a protein with homology to the JAM family of proteins, has an inhibitory effect on M-CSF– and RANKL-mediated osteoclastogenesis through the suppression of integrin αVβ3 signaling. Exogenous expression of DICAM inhibited osteoclast differentiation and conversely, knockdown of DICAM accelerated osteoclastogenesis. Mechanistically, DICAM attenuated integrin β3 expression as well as attenuated heterodimerization between integrin β3 and αV. This resulted in decreased the p38 MAP kinase activity and c-Fos and NFATc1 expression. This inhibition of osteoclastogenesis by DICAM was rescued by overexpression of integrin β3 or high doses of M-CSF. The inhibitory function of DICAM in osteoclastogenesis was also confirmed by the addition of recombinant DICAM. These results indicate that DICAM acts as an endogenous negative regulator of osteoclastogenesis through the attenuation of integrin αVβ3 signaling.

Integrin αVβ3 is a key player for osteoclastogenesis.7 Integrin αVβ3 mediates the differentiation and function of osteoclasts to polarize, spread, and degrade bone, and can be regulated by integrin ligands in the extracellular matrix containing RGD domains.25 Ligand binding to integrin αVβ3 evokes c-Src phosphorylation and recruitment of Syk bound to the immunoreceptor tyrosine-based activation motif (ITAM) proteins, which eventually form the integrin/c-Src/Syk/ITAM complex, the major machinery of integrin-mediated “outside-in” signaling.26 DICAM is an adhesion molecule that binds integrin αVβ3 through the second Ig domain and inhibits integrin β3 expression.19 Interestingly, expression of integrin β3 was also downregulated by rDICAM. Because NFATc1 as an upstream transcription factor to regulate integrin β3 expression, DICAM-mediated inhibition of integrin β3 expression may be due to delayed osteoclast differentiation.27, 28 However, we cannot rule out the possibility of accelerated endocytosis and degradation of integrin β3 by DICAM. More studies are needed to understand the precise molecular mechanism of DICAM's gene expression during osteoclastogenesis.

In this study, DICAM inhibited p38 MAP kinase activity in response to RANKL and M-CSF without altering of the ERK, JNK, or NF-κB pathways. The molecular mechanism involved in the inhibition of p38 MAP kinase signaling by DICAM is unclear. It may depend on a direct effect of DICAM per se or an indirect effect via the modulation of integrin β3 signaling. DICAM has one putative binding site for protein kinase C (PKC)-µ in the extracellular juxta-transmembrane region19 and recent studies have demonstrated the involvement of PKC-µ in p38 MAP kinase-mediated signaling.29, 30 However, rDICAM containing only the extracellular domain inhibited osteoclast differentiation, which was rescued by high-dose M-CSF, implying that DICAM-mediated inhibition of osteoclastogenesis is mostly through integrin β3 signaling. In the “outside-in” signaling model that connects integrin β3 and p38 MAP kinase, integrin first activates c-Src and consequently phosphorylates PLC-γ.4 When we assessed c-Src phosphorylation induced by RANKL, the phosphorylation of c-Src was not attenuated, but rather was slightly upregulated by DICAM (data not shown). In addition to the c-Src-PLC-γ pathway, integrin-linked kinase (ILK) has been shown to be an additional link between integrin and p38 MAP kinase during osteoclastogenesis.31, 32 Recently, it has been demonstrated that osteoclast-specific ILK ablation results in an increased bone volume and an augmented number of osteoclasts, which have a similar phenotype as integrin β3-deficient mice.33 Future studies on the complex interaction between DICAM and p38 MAP kinase during osteoclast differentiation will provide new insights into the regulatory mechanisms of osteoclast differentiation and function.

The expression of c-Fos and NFATc1 was significantly reduced in osteoclasts overexpressing DICAM. The RANKL-RANK signaling pathway is known to selectively induce NFATc1 through TRAF6 and the c-Fos pathway during osteoclast differentiation.34 The expression of c-Fos in response to RANKL is regulated by the calcium/calmodulin-dependent protein kinase-cyclic adenosine monophosphate (AMP)-responsive element-binding protein system and the NF-κB pathway.35, 36 In addition, p38 MAP kinase signaling is also important for the induction of c-Fos and NFATc1 during RANKL-stimulated osteoclast differentiation.37, 38 When osteoclast precursor cells are treated with p38 MAP kinase inhibitors or dominant-negative forms of MAP kinase kinase 3 (MKK3) and MKK6 are overexpressed, RANKL induced c-Fos and NFATc1 expressions is dramatically suppressed. Overexpression of c-Fos rescues the decreased NFATc1 expression in cells treated with a p38-specific inhibitor, implying that p38 MAP kinase dependent NFATc1 expression is downstream of c-Fos.37, 38 These data are further evidence of the role of p38 MAP kinase in the activation of c-Fos and NFATc1.

DICAM inhibited M-CSF/RANKL-mediated osteoclast differentiation that was rescued by high-dose M-CSF, but not by RANKL. The crosstalk between integrin αVβ3 and c-Fms is crucial during osteoclastogenesis.11, 15 Faccio and colleagues11 reported an interaction between M-CSF and integrin β3 signaling that is similar to the results presented here in several respects. First, attenuated osteoclast differentiation from integrin β3−/− BMMs was rescued by high-dose M-CSF. Second, high-dose M-CSF also restored the duration of ERK activity and expression of c-Fos. Taken together, these data provide consistent evidence that both M-CSF and αVβ3 signaling cooperatively involve c-Fos and NFATc1 induction.

DICAM is a unique cell-surface protein that inhibits integrin signaling through its extracellular domain. This study supports the concept of a functional interaction between integrin αVβ3 and M-CSF–dependent signaling pathways in osteoclast differentiation. Our study, in addition to other studies, enables us to provide a schematic model of the role of DICAM during osteoclast differentiation (Fig. 8). DICAM directly interacts with integrin αVβ3 and attenuates the heterodimerization of αV and β3. DICAM suppresses p38 MAP kinase phosphorylation by RANKL and low concentrations of M-CSF, leading to downregulation of c-Fos and NFATc1. DICAM also decreases the duration of ERK activation. High-dose M-CSF rescues the inhibition of osteoclast differentiation by DICAM, which can be mediated by “inside-out” activation of integrin. Therefore, DICAM acts as a negative modulator of osteoclastogenesis and may be a suitable therapeutic target for resorptive bone diseases.

thumbnail image

Figure 8. Model of DICAM function during osteoclast differentiation. DICAM interacts with integrin αVβ3 and attenuates the heterodimerization between αV and β3. DICAM suppresses phosphorylation of p38 MAP kinase by RANKL and low concentrations of M-CSF, thus leading to downregulation of c-Fos and NFATc1. DICAM also decreases the duration of ERK activation. High-dose M-CSF or overexpression of integrin β3 rescues the inhibition of osteoclast differentiation mediated by DICAM.

Download figure to PowerPoint

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

This work was supported by the Korea Research Foundation Grant (KRF-2008-313-E00096) and WCU program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (R32-10064).

Author's roles: Study design: YJ and JC. Study conduct: YJ, SH, EL, GK, and JJ. Data analysis: YJ, HK, and JC. Manuscript writing: YJ, SH, and JC. Approving final revision of manuscript: YJ, SH, EL, GK, JJ, HK, and JC. YJ and JC take responsibility for the integrity of the data analyses.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
jbmr_1632_sm_SupplFig1.tif846KSupplementary Figure 1
jbmr_1632_sm_SupplFig2.tif5182KSupplementary Figure 2
jbmr_1632_sm_SupplFigsLegend.doc108KSupplementary Figures Legend

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.