SEARCH

SEARCH BY CITATION

Keywords:

  • OSTEOCLAST;
  • ADENYLATE CYCLASE;
  • CA2+/CALMODULIN-DEPENDENT PROTEIN KINASE;
  • PROTEIN KINASE A

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

Nuclear factor of activated T cells c1 (NFATc1) is a transcription factor crucial for the differentiation of osteoclasts. In this study we discovered new signaling pathways involving cAMP regulators that modulate NFATc1 during osteoclastogenesis. The osteoclast differentiation factor receptor activator of NF-κB ligand (RANKL) increased the expression of adenylate cyclase 3 (AC3), accompanied by a rise in the intracellular cAMP level in osteoclasts. The knockdown of AC3 enhanced in vitro osteoclastogenesis and in vivo bone resorption, whereas cAMP-elevating agents showed opposite effects. The antiosteoclastogenic effect of the AC3-cAMP pathway was mediated by the inhibition of NFATc1 nuclear translocation and its autoamplification via a protein kinase A (PKA)–dependent mechanism. RANKL has been shown to activate Ca2+/calmodulin-dependent protein kinases (CaMKs). Knockdown or catalytic inhibition of CaMKs elevated intracellular cAMP levels in RANKL-treated osteoclast precursors and suppressed the activation of NFATc1. Taken together, our results demonstrate a pivotal role for the cAMP-PKA-NFATc1 signaling pathway during osteoclast differentiation, suggesting a mechanism by which osteoclastogenesis is fine-tuned by a balance between AC3 and CaMKs activities. © 2011 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

Disorders of skeletal homeostasis, such as osteopetrosis and osteoporosis, are caused by an imbalance in the activities between osteoblasts and osteoclasts.1, 2 The generation of active osteoclasts involves a series of steps: the commitment of hematopoietic precursors to the osteoclastic lineage, the proliferation of committed precursor cells, and the fusion of mononuclear osteoclasts.3, 4 This osteoclastogenesis process is characterized experimentally by the formation of tartrate-resistant acid phosphatase (TRACP)–positive multinuclear cells. Under physiologic conditions, osteoclastogenesis is controlled by the differentiation factor receptor activator of NF-κB ligand (RANKL) in addition to macrophage colony-stimulating factor (M-CSF), which supports cell proliferation and survival.5, 6

RANKL has been known to activate NF-κB and mitogen-activated protein kinase (MAPK) signaling pathways through tumor necrosis factor (TNF) receptor–associated factor (TRAF) family proteins.7–9 RANKL also stimulates the expression of nuclear factor of activated T cells c1 (NFATc1), which induces osteoclastic genes such as TRACP and cathepsin K.10, 11 Initial induction of NFATc1 leads to stimulation of further increased expression of NFATc1 because this transcription factor binds to its own promoter, constituting an autoamplification loop.12 In RANKL-treated osteoclast precursors, a Ca2+ oscillation response is elicited, which leads to the activation of two distinct downstream signaling molecules: Ca2+/calmodulin (CaM)–dependent phsophatase calcineurin and Ca2+/CaM-dependent protein kinases (CaMKs). Calcineurin dephosphorylates cytoplasmic NFATc1 proteins, allowing their subsequent nuclear translocation and execution of transcriptional activity.13, 14 It was shown previously that Ca2+-activated CaMKII and CaMKIV enhance the expression of NFATc1 through c-Fos induction in osteoclasts.15–17

Adenylate cyclases (ACs) catalyze the formation of cyclic adenosine monophosphate (cAMP) from ATP, which activates target proteins such as protein kinases and transcription factors. The cellular concentration of cAMP is regulated by ACs and phosphodiesterases (PDEs). One soluble isoform and nine membrane-bound AC isoforms that have distinct patterns of responses to Ca2+/CaM and protein kinases have been identified.18–20 Among these isotypes, AC3 is inhibited by calmodulin and Ca2+ increases. There are several reports on the regulation of ACs by Ca2+-activated CaMKs.21, 22 CaMKII was shown to inhibit AC3 activity by phosphorylating Ser-1076.23 CaMKIV also was shown to reduce AC1 activity.24

It was reported previously that cAMP enhances RANKL expression in osteoblasts, thereby augmenting osteoclast differentiation. However, the direct effect of cAMP on osteoclast differentiation has not been addressed. In this study we discovered that AC3 was upregulated by RANKL in osteoclast precursors. We provide evidence that AC3 inhibits RANKL-induced osteoclast differentiation through cAMP-PKA-NFATc1, whereas CaMK counteracts this process by limiting intracellular cAMP levels.

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

Lipofectamine 2000 was obtained from Invitrogen Life Technologies (Carlsbad, CA, USA). Human soluble RANKL and human M-CSF were purchased from Pepro-Tech (Rocky Hill, NJ, USA). Anti-β-actin antibody was purchased from Sigma (St Louis, MO, USA). Antibodies against AC3, CaMKIIα, and CaMKIV were purchased from Abcam (Cambridge, UK). Anti-NFATc1 and anti-lamin B antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-phosphorylated PKA substrate antibody was purchased from Cell Signaling Technology (Beverly, MA, USA). All other chemicals were from Sigma. (See also Supplemental Materials and Methods on the Web site.)

Bone marrow macrophage culture and osteoclast differentiation

Bone marrow macrophages (BMMs) were generated as described previously.25, 26 Isolated bone marrow cells were cultured in α modified essential medium (α-MEM) containing 10% fetal bovine serum (FBS) overnight on culture dishes. Nonadherent cells were further cultured for 3 days in the presence of 5 ng/mL of M-CSF. These BMMs were seeded in 48-well plates (2 × 104 cells per well) and cultured with 30 ng/mL of M-CSF and 100 ng/mL of RANKL for 5 days to induce osteoclast differentiation.

Gene knockdown by small interfering RNA (siRNA) oligonucleotides

The 22-nucleotide siRNA duplexes targeting CaMKIIα, CaMKIV, and nonspecific control were designed and purchased from Invitrogen Life Technologies. The target sequences were 5'-AATAGAATCGATGAAAGTCCAGGCC-3' for CaMKIIα and 5'-TCCACAATCCTGTCAAACAGTTCTC-3' for CaMKIV. The siRNA oligonucleotides were transfected into BMMs using Lipofectamine 2000 following the manufacturer's instructions.

Retroviral transduction

For retrovirus-mediated gene knockdown, oligonucleotides for siRNAs were generated by targeting the 21-base sequence of AC3 and cloned into pSUPER-retro vector (Oligoengine, Seattle, WA, USA). The oligonucleotide sequences were 5'-GATCCCGCTGTGCCTGCCTCGCTTTATTCAAGAGATAAAGCGAGGCAGGCACAGTTTTTTCCAAA-3' (sense) and 5'-AGCTTTTGGAAAAAACTGTGCCTGCCTCGCTTTATCTCTTGAATAAAGCGAGGCAGGCACAGCGG-3' (antisense). Plat-E retroviral packaging cells grown in DMEM supplemented with 10% FBS were transfected with siRNA constructs using Lipofectamine 2000 according to the manufacturer's instructions. The medium was changed with α-MEM containing 10% FBS after overnight culture. After further culture for 48 hours, viral supernatants were collected and filtered through a 0.45-µm syringe filter (Sartorius Stedim Biotech, Goettingen, Germany). BMMs were infected with viral supernatants in the presence of polybrene (10 µg/mL) and M-CSF (30 ng/mL) for 12 hours.

Reverse-transcriptase polymerase chain reaction (RT-PCR) analysis

Total RNA was prepared using Trizol (Invitrogen) and reverse transcribed using SuperScript II reverse transcriptase (Invitrogen). One microliter of cDNA synthesized from 1 µg of total RNA was amplified with the specific primers described in Supplemental Table S1.

Nuclear fractionation

Cells were washed with PBS, lysed in lysis buffer (10 mM HEPES, pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM dithiothreitol (DTT), 0.1% NP-40, 2 mM Na3VO4, 5 mM NaF, 1 µg/mL of aprotinin, 1 µg/mL of leupeptin, 1 µg/mL of pepstatin, and 0.2 mM phenylmethylsulfonyl fluoride) by pipetting up and down. After microcentrifugation at 16,000 g for 5 minutes at 4 °C, the supernatant was collected and considered as the cytosolic fraction. The pellet was resuspended in the lysis buffer containing 20 mM HEPES (pH 7.9), 1.5 mM MgCl2, 420 mM NaCl, 0.5 mM DTT, 25% glycerol, 2 mM Na3VO4, 5 mM NaF, 1 µg/mL of aprotinin, 1 µg/mL of leupeptin, 1 µg/mL of pepstatin, and 0.2 mM phenylmethylsulfonyl fluoride (PMSF). After incubation for 20 minutes on ice and microcentrifugation at 16,000 g for 5 minutes at 4 °C, the supernatant was collected and referred to as the nuclear fraction.

PKA kinase activity assay

BMMs were cultured with RANKL (200 ng/mL) and M-CSF (30 ng/mL) for 2 days and then treated with forskolin (1 µM) or prostaglandin E2 (PGE2, 1 µM) for 2 hours. PKA kinase activity was measured with the PKA Kinase Activity Assay Kit (Assay Designs, Ann Arbor, MI, USA) following the manufacturer's instructions.

Western blotting

Cells were washed with ice-cold PBS before collection by scraping with a rubber policeman and lysis in RIPA buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.5 mM PMSF, 1 µg/mL of aprotinin, 1 µg/mL of leupeptin, and 1 µg/mL of pepstatin). After protein quantification with a protein assay kit (Biorad, Hercules, CA, USA), whole-cell extracts were separated on polyacrylamide gels and transferred onto nitrocellulose membranes. After blocking for 12 hours with 5% skim milk in Tris-buffered saline containing 0.1% Tween 20 (TBST), membranes were incubated overnight at 4 °C with primary antibodies in TBST containing 2% skim milk. Membranes were washed, incubated for 1 hour with appropriate secondary antibodies conjugated with horseradish peroxidase, and developed using an enhanced chemiluminescence system.

TRACP staining

Cells were fixed in 3.7% formaldehyde solution for 10 minutes and permeablized with 0.1% Triton X-100 for 1 minute. After washing with PBS, cells were stained using the Leukocyte Acid Phosphatase Assay Kit (Sigma) following the manufacturer's instructions. TRACP-stained multinuclear osteoclasts containing three or more nuclei were counted under a light microscope and photographed.

Measurement of cAMP

Intracellular cAMP concentrations were measured with a cAMP EIA system (GE Healthcare, Little Chalfont, UK) following the manufacturer's instructions.

In vitro bone erosion experiment

BMMs were cultured on dentin slices in the presence of RANKL (100 ng/mL) and M-CSF (30 ng/mL) for 6 days. Dentin slices were sonicated and stained with hematoxylin.

In vivo bone-resorption assay

Forskolin dissolved in DMSO (30 µL of 1 µM forskolin per mouse) or vehicle were injected onto calvaria (n = 5 per group) of 5-week-old male ICR mice three times with 2-day intervals. One day after the first injection, collagen sheets soaked with 10 µg of RANKL or PBS were implanted onto the center of the calvaria. At 5 days after the first injection, mice were euthanized, and calvaria were collected. Calvaria then were stained for TRACP activity. The TRACP-stained area was measured using the Image J program (Version 1.40; available on the National Institutes of Health Web site). Micro–computated tomography (µCT) also was performed with a Skyscan1072 scanner (Skyscan, Aartselaar, Belgium; 40 kV, 250 µA, 11.55-µm pixel size). Images were reconstituted using CT-volume software (Version 1.11; Skyscan).

Immunofluorescence microscopy

BMMs were cultured with 30 ng/mL of M-CSF and 100 ng/mL of RANKL for 2 days. After fixation in 3.7% formaldehyde for 5 minutes, permeabilization in 0.25% Triton X-100 for 5 minutes, and blocking with 2% bovine serum albumin (BSA) in PBS, cells were stained with anti-NFATc1 antibody followed by Cy3-conjugated secondary antibody and counterstained with 4,6-diamidino-2-phenylindole (DAPI). Cells were observed under a Zeiss LSM 5 PASCAL laser-scanning microscope (Carl Zeiss Microimaging GmbH, Goettingen, Germany) with an × 400 objective (C-Apochromat/1.2WCorr).

Statistics

To determine the significance of results, Student's t test was used. Differences with p < .05 were regarded as significant.

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

Expression of AC3 following RANKL treatment

To identify genes expressed differentially during osteoclast differentiation, we performed a cDNA microarray analysis with RANKL-stimulated and -unstimulated BMMs. We found that the mRNA level of AC3 was more than threefold upregulated after RANKL treatment for 3 days. We next examined the expression of all nine isoforms of ACs in BMMs using RT-PCR. As shown in Fig. 1A, BMMs expressed AC isoforms 2, 3, 5, 6, 7, and 9, whereas isoforms 1, 4, and 8 were undetectable. Consistent with the microarray results, the mRNA level of AC3 was markedly elevated by RANKL treatment. In contrast, the mRNA levels of AC5 and AC7 were reduced by RANKL. AC3 mRNA level started to increase at 1 day after exposure to RANKL (Fig. 1B). A significant increase in the AC3 protein level also was observed in RANKL-treated BMMs (Fig. 1C). Since the expression of AC isoforms varied during the course of osteoclast differentiation, we next measured the intracellular cAMP levels. We found that the intracellular cAMP level was elevated by more than twofold in mature osteoclasts (day 5) compared with BMMs (Fig. 1D). These results suggested that the increased expression of the AC3 isoform by RANKL may contribute to the increased cAMP level in osteoclasts.

thumbnail image

Figure 1. Expression of AC isotypes during osteoclastogenesis. (A) BMMs were cultured with or without 100 ng/mL of RANKL in the presence of 30 ng/mL of M-CSF for the indicated periods. Total RNA was extracted, reverse-transcribed, and amplified with isotype-specific primers. The expression of GAPDH served as a control. (B) The mRNA expression of AC3 was examined in BMMs cultured with M-CSF and RANKL for the indicated times. (C) Whole-cell lysates of BMMs cultured with M-CSF and RANKL for the indicated times were subjected to Western blotting using anti-AC3 antibody. (D) BMMs were cultured with M-CSF and RANKL for the indicated times. Intracellular cAMP was measured with a cAMP EIA kit. All results are representative of at least three experiments. ap < .05 versus control.

Download figure to PowerPoint

Effect of cAMP on RANKL-induced osteoclast differentiation

Since the levels of AC3 and cAMP increased during osteoclastogenesis, agents known to elevate intracellular cAMP concentrations were assessed for their effects on osteoclast differentiation. When BMMs were stimulated with RANKL in the presence of AC activator forskolin, the generation of RANKL-induced TRACP+ osteoclasts was significantly inhibited in a dose-dependent manner (Fig. 2A, B). Another cAMP-elevating agent, prostaglandin E2 (PGE2), also reduced RANKL-induced osteoclastogenesis (Fig. 2D, E). The inhibitory effects of forskolin and PGE2 correlated with the cAMP production in BMMs, as shown in Fig. 2C, F, respectively. Furthermore, sp-cAMP, a PDE-resistant cell-permeable cAMP analogue, efficiently suppressed RANKL-induced osteoclastogenesis in a dose-dependent manner (Fig. 2G, H). These results indicated that the elevation of intracellular cAMP has an inhibitory effect on RANKL-induced osteoclast differentiation.

thumbnail image

Figure 2. Effect of cAMP modulation on osteoclast differentiation. (A, B) BMMs were cultured with increasing concentrations of forskolin in the presence of 30 ng/mL of M-CSF and 100 ng/mL of RANKL for 5 days. TRACP-stained cells with more than three nuclei were counted as osteoclasts. (C) BMMs were treated with forskolin for 1 hour in the presence of 30 ng/mL of M-CSF and 100 ng/mL of RANKL. Intracellular cAMP concentration was measured using a cAMP EIA kit. (D, E) BMMs were differentiated with RANKL (100 ng/mL) and M-CSF (30 ng/mL) in the presence of increasing concentrations of PGE2. After 5 days, cells were stained for TRACP activity, and the number of osteoclasts was counted. (F) BMMs were treated with PGE2 for 1 hour in the presence of M-CSF (30 ng/mL) and RANKL (100 ng/mL). Intracellular cAMP concentration was measured as in panel C. (G, H) BMMs were cultured with increasing concentrations of sp-cAMP in the presence of M-CSF and RANKL for 5 days. (I) BMMs were infected with retroviruses harboring luciferase (Luci)–specific siRNA or AC3-specific siRNA. The protein levels of AC3 and NFATc1 in infected cells were evaluated 2 days after RANKL treatment by Western blotting. (J) BMMs infected with retroviruses as in panel I were cultured with M-CSF (30 ng/mL) and RANKL (100 ng/mL) for 3 days before measuring intracellular cAMP levels with a cAMP EIA kit. (K) Control or AC3-silenced BMMs were cultured with M-CSF and RANKL for 5 days in the presence of forskolin or vehicle (DMSO). Osteoclasts were stained for TRACP activity and counted. All experiments were repeated three times with similar results. ap < .01 versus control.

Download figure to PowerPoint

We next evaluated the effect of AC3 silencing on cAMP level and osteoclastogenesis because only the expression of AC3 isoform was significantly elevated during osteoclast differentiation. AC3 mRNA and protein levels were reduced significantly in BMMs infected with retroviruses harboring siRNA specific for AC3 compared with cells infected with the control luciferase (Luci) siRNA retroviruses (Fig. 2I and data not shown). The expression level of NFATc1, a key osteoclastogenic transcription factor, was markedly elevated in AC3-silenced RANKL-treated cells (Fig. 2I), in which the intracellular cAMP level also was significantly lower compared with control cells (Fig. 2J). Consistent with the effect of AC3 silencing on NFATc1 expression, the formation of TRACP+ multinuclear osteoclasts by RANKL was enhanced dramatically by AC3 silencing in BMMs (Fig. 2K). In addition, the inhibitory effect of forskolin on osteoclast differentiation was completely abrogated by AC3 silencing. These results suggested that the AC3-dependent cAMP production negatively regulates osteoclastogenesis.

Effect of cAMP-modulation on RANKL-induced bone loss

Since the increase in intracellular cAMP inhibited osteoclastogenesis, we assessed whether treatment with the cAMP stimulators forskolin and PGE2 also could result in a reduction in bone-resorption activity of osteoclasts. As shown in Fig. 3A, forskolin and PGE2 strongly inhibited bone resorption. Next, we evaluated the in vivo effect of cAMP modulation on bone resorption and osteoclastogenesis. Forskolin was injected for 5 days into the supracalvarial region of mice implanted with RANKL-soaked sponges. In Fig. 3B, TRACP staining of whole calvaria showed massive osteoclastogenesis by RANKL administration in the absence of forskolin. However, forskolin treatment completely blocked the increase in the number of TRACP+ osteoclasts by RANKL (Fig. 3C). In agreement, µCT analysis revealed intense bone resorption (shown by the dark area in reconstituted images) induced by RANKL treatment that was effectively suppressed by forskolin administration (Fig. 3D). Analyses of bone volume showed a protective effect of forskolin against RANKL-induced bone loss in mice calvaria (Fig. 3E).

thumbnail image

Figure 3. Effect of cAMP-modulating agents on bone resorption. (A) BMMs were cultured on dentin slices with or without forskolin and PGE2 in the presence of 200 ng/mL of RANKL and 30 ng/mL of M-CSF. After 6 days, cells were removed by sonication, and dentin slices were stained with hematoxylin to observe resorption pits. The resorbed area was measured by densitometry. (B, C) Mice calvaria were implanted with collagen sponges soaked in PBS or RANKL. Forskolin was injected onto mice calvaria for 5 days with 2-day intervals. Calvaria were collected and stained for TRACP activity (B). The TRACP-stained area in panel B was measured by densitometry (C). (D) Calvaria in panel B were analyzed by µCT. 3D images were reconstituted using CT volume software as described under “Materials and Methods.” (E) Calvarial bone volumes were calculated using a CT analyzer program. Representative images are shown. ap < .05 versus control.

Download figure to PowerPoint

Inhibition of NFATc1 autoamplification by cAMP elevation

To delineate the mechanism by which cAMP inhibits osteoclastogenesis, we investigated the effect of forskolin on the induction of NFATc1 by RANKL. Forskolin treatment greatly attenuated the RANKL-mediated NFATc1 induction (Fig. 4A). Forskolin did not affect RANKL-dependent p38, Erk, Akt, and NF-κB activation (Supplemental Fig. S1), suggesting that these signaling pathways are not involved in the forskolin-dependent inhibition of NFATc1 induction. Next, we examined the nuclear localization of NFATc1 in forskolin- or PGE2-treated BMMs after RANKL stimulation. BMMs were differentiated with RANKL for 2 days, treated with forskolin or PGE2 for 2 hours, and immunostained with a NFATc1-specific antibody. Figure 4B shows that RANKL strongly elevated the nuclear localization of NFATc1 in addition to its expression. Forskolin and PGE2 treatment substantially decreased not only the number of total NFATcl+ cells but also the proportion of cells with nuclear NFATc1 in the presence of RANKL (Fig. 4B). In addition, the nuclear fraction of forskolin- or PGE2-treated BMMs contained reduced level of NFATc1 compared with vehicle-treated cells (Fig. 4C, D). These observations suggested that increased cAMP might be involved in disengagement from the NFATc1 autoamplification loop by interfering with the nuclear translocation of NFATc1 in BMMs, reducing the RANKL-induced NFATc1 expression. In order to further clarify the role of cAMP-dependent NFATc1 inhibition in the forskolin-mediated inhibition of osteoclastogenesis, a constitutively active form of NFATc1 (caNFATc127) was used in BMMs. After infection of BMMs with caNFATc1-containing retroviruses, cells were treated with RANKL for 5 days with or without forskolin. The overexpression of caNFATc1 protein was confirmed in cells infected with caNFATc1 virus (Fig. 4E). As shown in Fig. 4F, forskolin strongly inhibited the formation of TRACP+ osteoclasts from control cells. However, in caNFATc1-overexpressed cells, the forskolin-mediated inhibition of osteoclastogenesis was substantially reversed.

thumbnail image

Figure 4. Effect of cAMP-modulating agents on RANKL-induced NFATc1 autoamplification. (A) BMMs were cultured with forskolin (1 µM) or vehicle (DMSO) in the presence of 30 ng/mL of M-CSF and 100 ng/mL of RANKL for the indicated days. Whole-cell lysates were subjected to immunoblotting with anti-NFATc1 antibody. (B) BMMs were cultured with 200 ng/mL of RANKL and 30 ng/mL of M-CSF for 2 days to induce NFATc1. After treating cells with foskolin (1 µM) or PGE2 (1 µM) for 2 hours, the intracellular localization of NFATc1 (red) was examined by immunostaining using anti-NFATc1 antibody followed by Cy3-conjugated secondary antibody. After staining the nuclei with DAPI (blue), the percentage of cells with nuclear NFATc1 was determined. (C) BMMs were differentiated with RANKL (200 ng/mL) and M-CSF (30 ng/mL) for 2 days and then treated with forskolin or PGE2 for 2 hours. Nuclear fractions were separated and subjected to Western blotting an antibody for NFATc1. (D) The NFATc1 levels in panel C were measured by densitometry. (E) BMMs were infected with control retroviruses (pMSCV) or retroviruses containing a constitutive active form of NFATc1 (pMSCV-caNFATc1). At 2 days after infection, whole-cell lysates were examined for NFATc1 expression by immunoblotting. (F) Control BMMs (pMSCV) or cells overexpressing caNFATc1 (pMSCV-caNFATc1) were cultured with forskolin or vehicle (DMSO) for 5 days in the presence of M-CSF (30 ng/mL) and RANKL (100 ng/mL). Osteoclasts were stained for TRACP activity and counted. ap < .05 versus DMSO-treated control. bp < .05 versus forskolin-treated control.

Download figure to PowerPoint

Inhibition of osteoclastogenesis by PKA-mediated NFATc1 phosphorylation

The subcellular localization of NFATc1 is regulated by multisite phosphorylation. Previously, it was reported that the NFATc1 is negatively regulated by cAMP-dependent protein kinase A (PKA) in 293T cells.28 Since PKA is a major cAMP-responsive kinase, the role of PKA during the cAMP-mediated inhibition of osteoclastogenesis was investigated. First, we determined the kinase activity of PKA during osteoclast differentiation. The kinase activity was significantly elevated in mature osteoclasts, whereas the protein level of PKA was maintained constant during osteoclastogenesis (Fig 5A). When BMMs were treated with a cell-permeable PKA inhibitor H89, RANKL-induced osteoclast formation was markedly enhanced in a dose-dependent manner (Fig. 5B). The number as well as the size of osteoclasts was significantly increased by H89 treatment. The cAMP elevating agents forskolin and PGE2 increased PKA activity in differentiating osteoclasts (Fig. 5C). Furthermore, the inhibitory effect of forskolin on osteoclast differentiation was abolished by PKA inhibition (Fig. 5D). Next, we investigeted the role of PKA in cAMP-dependent NFATc1 inhibition by evaluating the phosphorylation status of NFATc1 after forskolin or PGE2 treatment. NFATc1 was immunoprecipitated from lysates of RANKL-primed BMMs and then immunoblotted with an antibody specific for phosphorylated PKA substrates. As shown in Fig. 5E, forskolin and PGE2 prominently enhanced NFATc1 phosphorylation at its PKA phsophorylation sites. In agreement with enhanced NFATc1 phosphorylation by PKA, treatment of BMMs with H89 augmented the nuclear translocation of NFATc1 on RANKL stimulation (Fig. 5F). In the presence of 10 nM H89, the proportion of osteoclast precursors with nuclear NFATc1 was about twofold higher than control. Finally, the RANKL-induced NFATc1 protein level also was markedly elevated by H89 treatment (Fig. 5G). These data suggest that PKA negatively regulates osteoclastogenesis by directly phosphorylating NFATc1, thus inhibiting nuclear translocation and subsequent autoamplification.

thumbnail image

Figure 5. Regulation of osteoclastogenesis and NFATc1 phoshporylation by PKA. (A) BMMs were differentiated with RANKL (100 ng/mL) and M-CSF (30 ng/mL) for the indicated days. Whole-cell extracts were subjected to PKA kinase assay or Western blotting with PKA antibody. (B) BMMs were cultured with increasing concentrations of PKA inhibitor H89 .The number of TRACP+ osteoclasts was counted. (C) BMMs were cultured with RANKL (100 ng/mL) and M-CSF (30 ng/mL) for 2 days. After treating these cells with forskolin or PGE2 for 2 hours, PKA kinase activity was measured from cell lysates. (D) BMMs were differentiated into osteoclasts in the presence of forskolin (0.1 µM) with increasing concentration of H89. TRACP+ cells after 5-day incubation with M-CSF and RANKL were counted. (E) BMMs were cultured with M-CSF and RANKL for 2 days and then treated with forskolin (1 µM) or PGE2 (1 µM) for 1 hour. Whole-cell lysates were immunoprecipitated with NFATc1 antibody and subjected to Western blotting using a phosphorylated PKA substrate-specific antibody. (F) BMMs were treated with M-CSF and RANKL for 2 days in the presence of H89. The intracellular localization of NFATc1 (red) was examined by immunostaining using anti-NFATc1 antibody followed by Cy3-conjugated secondary antibody. After staining the nuclei (blue), the percentage of cells with nuclear NFATc1 was determined. (G) BMMs were cultured with M-CSF and RANKL for 2 days in the presence of H89. Whole-cell lysates were subjected to Western blotting using anti-NFATc1 antibody. ap < .05 versus control.

Download figure to PowerPoint

Regulation of AC activity by CaMKs during osteoclast differentiation

CaMKII has been reported to inhibit AC3 activity in macrophage cells.18, 29 To test whether CaMKs add another layer of regulatory control on the cellular cAMP level, intracellular cAMP concentrations were measured after the treatment of osteoclast precursors with KN93 and STO609, CaMK and calmodulin-dependent protein kinase kinase (CaMKK) inhibitors, respectively. Both KN93 and STO609 significantly increased intracellular cAMP levels (Fig. 6A) while inhibiting RANKL-induced osteoclast formation in a dose-dependent manner (Supplemental Fig. S2). This inhibitory effect of KN93 on osteoclastogenesis was rescued by H89 treatment (Fig. 6B, C). Among tested CaMK isoforms, the mRNA expression of CaMKIγ, CaMKIIα, and CaMKIV each was elevated significantly on stimulation of BMMs with RANKL (Supplemental Fig. S3). Since CaMKII and CaMKIV have been reported to enhance osteoclastogenesis,15, 17 we tested whether CaMKIIα and CaMKIV have any intrinsic regulatory role for cAMP modulation and osteoclastogenesis. BMMs were transfected with siRNA oligonucleotides specific for CaMKIIα and CaMKIV. The mRNA and protein levels of CaMKIIα and CaMKIV were reduced significantly in the specific siRNA-transfected BMMs compared with control cells transfected with scrambled oligonucleotides (Fig. 6D and data not shown). RANKL-mediated NFATc1 induction also markedly reduced by CaMKIIα and CaMKIV silencing compared with control (Fig. 6D). Next we determined whether intracellular cAMP levels were changed by CaMKIIα or CaMKIV silencing. Notably, intracellular cAMP level was markedly elevated in CaMKIIα- or CaMKIV-silenced BMMs (Fig. 6E). The CaMK silencing clearly reduced the formation of TRACP+ osteoclasts (Fig. 6F, G). Finally, the effect of CaMK modulation on NFATc1 translocation was examined. CaMKIIα or CaMKIV silencing conspicuously suppressed NFATc1 nuclear accumulation compared with controls (Fig. 6H, I). Furthermore, the nuclear level of NFATc1 was reduced significantly by CaMK inhibitors KN93 and STO609 (Fig. 6J, K, and Supplemental Fig. S4). Taken together, these results suggest that the cAMP-PKA-NFATc1 signaling pathway is counterbalanced by CaMKs during osteoclastogenesis.

thumbnail image

Figure 6. Regulation of cAMP by CaMKs. (A) BMMs were cultured with 30 ng/mL of M-CSF and 100 ng/mL of RANKL for 3 days in the presence of CaMK inhibitors KN93 and STO609. Intracellular cAMP levels were measured using a cAMP EIA kit. (B, C) BMMs were incubated in the presence of M-CSF and RANKL for 5 days with 1 µM KN93 and the indicated concentrations of H89. After staining for TRACP activity, the number of osteoclasts was counted. (D) BMMs transfected with CaMK isoform-specific siRNA were differentiated with RANKL and M-CSF for 2 days. The protein levels of CaMKIIα, CaMKIV, and NFATc1 was evaluated by Western blotting. (E) Intracellular cAMP levels were measured from the cells in panel D. (F, G) Control (Scr siRNA) or CaMK-silenced (CaMKIIα and CaMKIV siRNA) BMMs were cultured with M-CSF and RANKL for 5 days. Cells were stained for TRACP activity, and the number of osteoclasts formed was counted. (H, I) Control or CaMK-silenced BMMs were cultured with M-CSF and RANKL for 2 days. The intracellular localization of NFATc1 (red) was examined by immunostaining using anti-NFATc1 antibody. After staining the nuclei (blue), the percentage of cells with nuclear NFATc1 was determined. (J, K) BMMs were cultured with RANKL and M-CSF for 2 days. After treating the cells with KN93 (1 µM) or STO609 (1 µM) for 2 hours, nuclear fractions were separated and subjected to Western blotting. The NFATc1 level in panel J was measured by densitometry (K). All images are representative of at least three experiments with similar results. ap < .05 versus control.

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

Several in vitro experiments using osteoblast-osteoclast cocultures showed that cAMP-elevating agents stimulate osteoclastogenesis by inducing RANKL expression in osteoblasts.30–32 However, the direct effect of cAMP modulation in osteoclast precursors in the absence of osteoclastogenesis-supporting cells has never been thoroughly investigated. In a microarray approach to discover novel regulators of osteoclastogenesis, we observed a significant increase in AC3 mRNA on RANKL stimulation of BMMs. Interestingly, among nine known isoforms of membrane-bound ACs, only the expression of AC3 was significantly upregulated in osteoclast precursors by RANKL (Fig. 1). Other isoforms were either absent in BMMs (AC1, -4, and -8) or slightly downregulated by RANKL treatment (AC5, -7, and -9). Thus the increased intracellular cAMP level induced by RANKL (Fig. 1D) correlated with the enhanced expression of AC3 during the course of osteoclast differentiation. We observed that RANKL-induced osteoclastogenesis was significantly elevated by reducing intracellular cAMP levels by AC3 silencing. To our knowledge, this is the first report showing that AC3 plays a key regulatory role in osteoclast differentiation. In addition to the synthesis by ACs, cellular cAMP level can be modulated by cyclic nucleotide phosphodiesterase (PDE)–dependent decomposition. We observed that the A and B isoforms of PDE4, the major cAMP-specific PDE, are expressed predominantly in osteoclasts (Supplemental Fig. S5A). Notably, knockdown of these two isoforms resulted in a significant increase in cellular cAMP concentrations and a clear inhibition of osteoclastogenesis (Supplemental Fig. S5B–D). Thus the modulation of cAMP plays a pivotal role in RANKL-dependent osteoclast differentiation.

It was reported previously that the cAMP-activated protein kinase A (PKA) could directly phosphorylate NFATc1, suppressing nuclear localization.28 The dephosphorylation-dependent nuclear translocation of NFATc1 and resulting induction of NFATc1 (ie, NFATc1 autoamplification) on RANKL stimulation is an essential event for osteoclast differentiation.33 Therefore, the effect of PKA inhibitor on the cAMP-mediated suppression of osteoclastogenesis was of particular interest. Notably, PKA inhibition with H89 not only stimulated RANKL-dependent osteoclastogenesis by itself but also rescued the forskolin-induced reduction of osteoclast differentiation by increasing nuclear translocation and induction of NFATc1 (Fig. 5). These results suggest that the PKA-dependent NFATc1 phosphorylation is an intrinsic negative regulatory signal that inhibits osteoclastogenesis. In the course of osteoclast differentiation, increased cAMP would further stimulate PKA activity to limit excessive osteoclastogenesis.

Interestingly, we also discovered a RANKL-dependent signaling mechanism that limits cAMP accumulation in differentiating osteoclasts. In RANKL-stimulated BMMs, the mRNA expression of CaMKIIα and CaMKIV isoforms increased significantly (Supplemental Fig. S3). The CaMK inhibitor KN93 and knockdown of CaMKIIα or CaMKIV by specific siRNA oligonucleotides significantly upregulated the intracellular cAMP level while greatly inhibiting NFATc1 nuclear translocation and osteoclastogenesis (Fig. 6). CaMKIV is CaMKK-dependent, whereas CaMKIIs are CaMKK-independent.17 Inhibition of CaMKK activity with STO609 showed similar effects on cAMP and osteoclastogenesis (Fig. 6 and Supplemental Fig. S2). Thus both CaMKK-dependent and -independent mechanisms participate in the CaMK-mediated regulation of intracellular cAMP in differentiating osteoclasts to counteract the cAMP increase by AC3 on RANKL stimulation.

It is likely that the intracellular levels of cAMP are intricately regulated by ACs, PDEs, and CaMKs at various stages during osteoclastogenesis. Both the expression level and the extent of catalytic activity of these enzymes would affect cAMP levels. Since cAMP apparently has an inhibitory role for osteoclast differentiation, the level of cAMP should be kept low until the final stage of differentiation. In fact, cAMP level was relatively low until the middle stage (day 3) of osteoclastogenesis despite substantial induction of AC3 (Fig. 1D). This apparent paradox might be explained by the downregulation of other AC isoforms (mainly AC5), the suppression of AC3 activity by CaMKs, and/or the induction of PDE4B (Supplemental Fig. 5A). At the late stage (day 5) of osteoclastogenesis, the cAMP level became high, whereas the levels of AC3, CaMKs, and PDE4 remained similar to those in the middle stage. We reason that this is possibly due to cessation of the negative regulation of CaMKs on the cAMP pathway because the Ca2+ oscillation required for CaMK activity disappears at the late stage.34 One remaining question then would be for what purpose high cAMP levels are achieved during osteoclastogenesis. With regard to this query, we observed that cAMP in already differentiated mature osteoclasts has a positive role for cell survival (Supplemental Fig. S6). Therefore, cAMP is likely to play a dual role for osteoclast differentiation and survival.

CaMK isoforms have been shown to accelerate osteoclastogenesis through the activation of AP-1 transcription factors.15, 21, 24 The cAMP regulation by CaMKs demonstrated in our study adds another mechanism by which osteoclastogenesis is regulated by these kinases. Thus it is likely that both cAMP-PKA-dependent and -independent mechanisms are operating for CaMKs to regulate osteoclastogenesis. In addition, although we could show clearly that CaMK activities reduced cAMP levels in osteoclast precursors, the target of CaMKs that modulates cAMP is yet to be defined. Previous reports suggested that CaMKs could phosphorylate and inhibit ACs directly in neuronal cells, as well as in human embryonic kidney cells.18, 29 Further studies are needed to clarify whether AC3 is the direct target of CaMKs in the downregulation of cAMP levels in osteoclasts.

Our data showing the direct regulation of osteoclastogenesis by cAMP suggest the possibility of using cAMP-modulating agents as drugs to treat skeletal diseases that result from unregulated osteoclast activity. In fact, forskolin administration significantly inhibited bone resorption induced by RANKL treatment in our mouse calvarial model (Fig. 3). Interestingly, many studies reported that cAMP-increasing agents could stimulate osteoblast-mediated osteoclast differentiation in vitro by upregulating M-CSF and RANKL expression on osteoblasts. However, the reduced bone resorption in vivo caused by forskolin showed that the direct inhibitory effect on osteoclasts seemingly overrode the indirect stimulatory effect on osteoblasts. Thus our results suggest that depending on the context of treatment, cAMP-increasing agents could be used to reduce osteoclastogenesis and concomitant pathologic bone resorption in vivo.

To conclude, we showed a novel signaling network involved in the regulation of osteoclast differentiation. We found that AC3-dependent cAMP modulation results in the PKA-dependent inhibition of NFATc1 autoamplification and reduces osteoclastogenesis. In addition, a CaMK-dependent mechanism that inhibits cAMP accumulation counteracted the cAMP-dependent negative regulation of osteoclast differentiation (Fig. 7). Thus a balance between these two pathways might be crucial for fine-tuning of osteoclastogenesis.

thumbnail image

Figure 7. The role of AC3- and CaMK-dependent cAMP modulation in osteoclast differentiation. AC3 induction by RANKL results in the accumulation of intracellular cAMP, which culminates in the PKA-mediated phosphorylation and subsequent inactivation of NFATc1. On the other hand, CaMK activities evoked by RANKL curtail intracellular cAMP elevation, limiting the inhibition of osteoclastogenesis by the AC3-cAMP-PKA pathway. Ultimately, the AC3 and CaMK pathways constitute a delicate signaling network for the regulation of osteoclast differentiation.

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 21C Frontier Functional Proteomics Project (FPR08B1-170), Science Research Center (20100001744), and the Basic Research Promotion [KRF-2007-E00001 (I00145)] grants.

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_310_sm_SuppFig1.jpg554KSupplementary Figure 1
jbmr_310_sm_SuppFig2.jpg2825KSupplementary Figure 2
jbmr_310_sm_SuppFig3.jpg699KSupplementary Figure 3
jbmr_310_sm_SuppFig4.jpg823KSupplementary Figure 4
jbmr_310_sm_SuppFig5.jpg1098KSupplementary Figure 5
jbmr_310_sm_SuppFig6.jpg3236KSupplementary Figure 6

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.