Diphenylhydantoin Inhibits Osteoclast Differentiation and Function Through Suppression of NFATc1 Signaling¹*

Abstract

Diphenylhydantoin (DPH) is widely used as an anticonvulsant drug. We examined the effects of DPH on osteoclast differentiation and function using in vivo and in vitro assay systems. Transgenic mice overexpressing a soluble form of RANKL (RANKL Tg) exhibited increased osteoclastic bone resorption. Injection of DPH into the subcutaneous tissue overlying calvaria of RANKL Tg mice suppressed the enhanced resorption in the calvaria. In co‐cultures of mouse osteoblasts and bone marrow cells, DPH inhibited lipopolysaccharide (LPS)‐induced osteoclast formation. DPH affected neither the mRNA expression of RANKL and osteoprotegerin nor the growth of mouse osteoblasts in culture. On the other hand, DPH inhibited the RANKL‐induced formation of osteoclasts in cultures of mouse bone marrow–derived macrophages (BMMϕs) and of human peripheral blood‐derived CD14⁺ cells. DPH concealed LPS‐induced bone resorption in mouse calvarial organ cultures and inhibited the pit‐forming activity of mouse osteoclasts cultured on dentine slices. DPH suppressed the RANKL‐induced calcium oscillation and expression of nuclear factor of activated T cells c1 (NFATc1) and c‐fos in BMMϕs. Moreover, DPH inhibited the RANKL‐induced nuclear localization and auto‐amplification of NFATc1 in mature osteoclasts. Both BMMϕs and osteoclasts expressed mRNA of a T‐type calcium channel, Cav3.2, a target of DPH. Blocking the expression of Cav3.2 by short hairpin RNAs significantly suppressed RANKL‐induced osteoclast differentiation. These results suggest that DPH inhibits osteoclast differentiation and function through suppression of NFATc1 signaling. The topical application of DPH may be a therapeutic treatment to prevent bone loss induced by local inflammation such as periodontitis.

diphenylhydantoin, osteoclasts, nuclear factor of activated T cells c1, bone resorption, calcium channel

INTRODUCTION

Osteoclasts, bone‐resorbing multinucleated cells, are differentiated from the monocyte macrophage lineage under the tight regulation of osteoblasts.^(1,2) Osteoblasts express two cytokines essential for osteoclast differentiation: macrophage‐colony stimulating factor (M‐CSF)⁽³⁾ and RANKL.^(4,5) M‐CSF is constitutively expressed by osteoblasts, whereas RANKL is inducibly expressed by osteoblasts in response to osteotropic hormones and factors including 1α,25‐dihydroxyvitamin D₃ [1α,25(OH)₂D₃], prostaglandin E₂ (PGE₂), and lipopolysaccharide (LPS).⁽⁶⁾ Osteoblasts also produce osteoprotegerin (OPG), a soluble decoy receptor for RANKL, which inhibits osteoclastogenesis by blocking RANKL–RANK interaction.^(7,8) Osteoclast precursors such as bone marrow–derived macrophages (BMMϕs) express c‐Fms (M‐CSF receptors) and RANK (RANKL receptors), and differentiate into osteoclasts in the presence of M‐CSF and RANKL.^(4,6)

The RANKL–RANK interaction in osteoclast precursors specifically and strongly induced the expression of nuclear factor of activated T cells c1 (NFATc1), a master transcription factor for osteoclast differentiation.⁽⁹⁾ NFATc1 is activated by the calcium‐regulated protein phosphatase, calcineurin. RANK‐mediated signaling evokes calcium oscillations and activates the calcinurin‐NFATc1 signaling pathway.^(9,10) The dephosphorylated NFATc1 translocates into nuclei and induces transcription of targets including cathepsin K, TRACP, and calcitonin receptors (CTR). The transcription of NFATc1 itself is also induced by NFATc1 in osteoclast precursors.⁽¹⁰⁾ Recent studies have shown that c‐fos, a transcription factor, plays an important role in NFATc1‐induced osteoclast differentiation.^(1,9)

Activated osteoclasts form ruffled borders and sealing zones toward bone surfaces during bone resorption.^(11,12) The sealing zone, which serves for the attachment of osteoclasts to the bone surface, is observed as a ringed structure of F‐actin dots (actin ring).^(13,14) The disruption of sealing zones in osteoclasts by cytoskeletal disrupting agents results in the suppression of bone‐resorbing activity of osteoclasts.^(15–17) Nuclear translocation of NFATc1 has been observed in osteoclasts activated by acidosis in the absence of RANKL.⁽¹⁸⁾ This finding suggests that activation of the calcium–calcineurin–NFATc1 pathway is also prerequisite for osteoclastic bone resorption. However, we do not fully understand how this pathway regulates osteoclast functions.

Diphenylhydantoin (DPH, also termed phenytoin) is widely used as an anticonvulsant drug for epileptic patients.⁽¹⁹⁾ Several candidates have been identified as targets of DPH, such as sodium‐potassium ATPase, the γ aminobutyric acid _A (GABA_A) receptor complex, inotropic glutamate receptors, calcium channels, and sigma binding sites. The best evidence hinges on the inhibition of voltage‐sensitive Na⁺ channels in the plasma membrane of neurons undergoing seizure activity.^(19–21) DPH often causes osteomalacia, characterized by an increase in osteoid in epileptic patients.⁽²²⁾ The bone loss induced by the administration of DPH is thought to be mainly caused by a drug‐induced vitamin D deficiency rather than direct effects of the treatment on bone.^(22–24) On the other hand, it was shown that DPH added to cultures of human and rat bone cells stimulated osteocalcin secretion, alkaline phosphatase (ALP) activity, and type I collagen synthesis,^(25,26) suggesting that DPH has a direct osteogenic action on osteoblasts. However, the effect of DPH on bone resorption is still controversial: some reports indicated DPH to be an inhibitory factor,^(27,28) and the other reports, a stimulator of bone resorption in vitro and in vivo.^(29,30) Our study focused on the functional role of DPH in osteoclastic bone resorption in vivo and in vitro.

MATERIALS AND METHODS

Mice and reagents

Seven‐week‐old male mice and newborn mice of the ddY strain were obtained from Japan SLC (Shizuoka, Japan) for the in vitro experiments. C57BL/6 mice obtained from Japan SLC were used for the in vivo experiments. RANKL Tg mice (genetic background of C57BL/6), which express mouse a soluble form of RANKL using the human serum amyloid P component promoter, were generated in one of the author's laboratories.⁽³¹⁾ All procedures for animal care were approved by the Matsumoto Dental University Experimental Animal Management Committee and performed accordingly. LPS (Escherichia coli O55:B5) and DPH were purchased from Sigma‐Aldrich (St. Louis, MO, USA). Recombinant human RANKL, a fusion protein comprising GST and the extracellular domain of human RANKL (amino acid residues140–317), was from Oriental Yeast (Tokyo, Japan). Recombinant human M‐CSF (Leukoprol) was obtained from Kyowa Hakko (Tokyo). 1α,25(OH)₂D₃ and PGE₂ were purchased from Wako Pure Chemical Industries (Osaka, Japan). Other chemicals and reagents were of analytical grade.

In vivo experiments

Vehicle (dimethylsulfoxide) or DPH (2 mg/kg body weight) was daily injected into subcutaneous tissue overlying calvaria of 7‐to 8‐wk‐old wildtype (WT) and RANKL Tg mice for 7 days. The mice were killed on day 8. Calvariae and tibias were collected, fixed in 4% paraformaldehyde (PFA), decalcified with 10% EDTA, and embedded in paraffin. Histological sections were prepared and stained for TRACP (a marker enzyme of osteoclasts). The osteoclast surface was histomorphometrically measured and expressed as the percentage of bone marrow interface covered by osteoclasts. Blood samples were also collected from mice for the measurements of serum parameters. Four to five mice were used in each group. Osteoclast numbers were counted in a double‐blind manner. Serum activities of TRACP5b and ALP were measured using a mouse TRACP5b assay kit (SBA Sciences, Turku, Finland) and an ALP kit (Wako), respectively.

Formation of mouse osteoclasts in culture

Primary osteoblasts prepared from newborn mouse calvaria (1 × 10⁴ cells/well) were co‐cultured with bone marrow cells (2 × 10⁵ cells/well) in αMEM (Sigma) containing 10% FBS (JRH Biosciences, Lenexa, KS, USA) in 48‐well plates (0.5 ml/well) as described. Co‐cultures were treated with or without LPS (100 ng/ml) together with various concentrations of DPH. After culture for 7 days, cells were fixed and stained for TRACP. TRACP⁺ cells containing three or more nuclei were counted as osteoclasts. To obtain functional osteoclasts, co‐cultures of mouse osteoblasts and bone marrow cells were performed in the presence of 1α,25(OH)₂D₃ (10⁻⁸ M) and PGE₂ (10⁻⁶ M) in 10‐cm‐diameter dishes precoated with type I collagen gel (Nitta Gelatin, Osaka, Japan) as described. After culturing for 6 days, all the cells including osteoclasts were recovered by treatment with 0.2% collagenase (Wako), suspended in 10 ml of αMEM containing 10% FBS, and used for osteoclast function assays. To obtain purified osteoclast preparations, the recovered cells were cultured for 6 h in 6‐well plates. Osteoblasts were removed by treatment with trypsin‐EDTA. The purity of osteoclasts in this preparation was ∼95%.⁽³²⁾

Mouse bone marrow macrophages (BMMϕs) were prepared as osteoclast precursors as described previously.⁽³²⁾ BMMϕs were cultured in 96‐well plates (1 × 10⁴ cells/well) with or without RANKL (100 ng/ml) and M‐CSF (50 ng/ml) in the presence of various concentrations of DPH. After culture for 3 days, cells were fixed and stained for TRACP. TRACP⁺ cells containing five or more nuclei were counted as osteoclasts.

Cell viability was determined with the Alamar blue assay. Osteoblasts (1 × 10⁴ cells/well), BMMϕs (1 × 10⁴ cells/well), or human CD14⁺ cells (3 × 10⁴ cells/well) were cultured for 1 day in 96‐well plates and treated with DPH (100 and 200 μM) for specific periods. The viability of cells was measured using an Alamar Blue assay kit (Biosource, Camarillo, CA, USA).

Formation of human osteoclasts in culture

Human CD14⁺ cells isolated from human peripheral blood were used as human osteoclast precursors. Human CD14⁺ cells were isolated from peripheral blood of three male volunteers, 27–36 yr of age, who were not receiving drug therapy. Informed consent for all procedures was obtained from all volunteers. Human CD14⁺ cells were isolated from peripheral blood mononuclear cells using MACS CD14 MicroBeads (Miltenyi Biotec, Auburn, CA, USA) as described.⁽³³⁾ CD14⁺ cells were cultured in αMEM supplemented with 10% FBS in 96‐well plates (1 × 10⁵ cells/well) in the presence of RANKL (100 ng/ml) and M‐CSF (50 ng/ml). After culture for 6 days, cells were fixed and stained for TRACP. TRACP⁺ cells containing five or more nuclei were counted as osteoclasts.

Calvarial organ cultures

Bone‐resorbing activity was determined using an organ culture system of mouse calvaria as described. Pairs of half calvaria obtained from 2‐day‐old ddY mice were preincubated for 24 h in BGJb (Gibco, New York, NY, USA) containing penicillin (100 U/ml) and streptomycin (100 μg/ml). One half of each pair was transferred to 1 ml of BGJb and the other to 1 ml of BGJb containing LPS (100 ng/ml) with increasing concentrations of DPH. After culture for 3 days, calcium concentrations in the medium were measured using a calcium determination kit (Ca E; Wako). Bone‐resorbing activity was expressed as the net amounts of calcium released from bone (amounts of calcium released in the presence of agents − amounts of calcium in the absence of agents).⁽³⁴⁾

Pit formation and actin ring formation assays

Osteoclasts were obtained from mouse co‐cultures performed on collagen gel‐coated dishes as described above. Aliquots of the osteoclast preparation (0.1 ml) were plated on dentine slices (4 mm diameter) and cultured for 24 h with increasing concentrations of DPH. The cells were removed from the dentine slices, and the slices were stained with Mayer's hematoxylin (Sigma) to identify resorption pits. The number of resorption pits was counted. In some experiments, cells on dentine slices were fixed and stained for TRACP. For the actin ring formation assay, aliquots of the osteoclast preparation (0.1 ml) were cultured on dentine slices. After culture for 24 h, cells were fixed with 4% PFA in PBS for 10 min. The cells were permeabilized with 0.1% Triton‐X 100 in PBS for 5 min and incubated with rhodamine‐conjugated phalloidin (Molecular Probes, Eugene, OR, USA) to visualize F‐actin.

PCR amplification of reverse‐transcribed mRNA

For the semiquantitative RT‐PCR analysis, total cellular RNA was extracted from cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). First‐strand cDNA was synthesized from the total RNA with oligo (dT)_12–18 primers and subjected to PCR amplification with EX Taq polymerase (Takara Biochemicals, Shiga, Japan) using the following specific PCR primers: mouse COX‐2, 5′‐GGGTTGCTGGGGGAAGAAATGTG‐3′ (forward) and 5′‐GGTGGCTGTTTTGGTAGGCTGTG‐3′ (reverse); mouse RANKL, 5′‐CGCTCTGTTCCTGTACTTTCGAGCG‐3′ (forward) and 5′‐TCGTGCTCCCTCCTTTCATCAGGTT‐3′ (reverse); mouse OPG, 5′‐CAGAGACTAATAGATCAAAGGCAGG‐3′ (forward) and 5′‐ATGAAGTCTCACCTGAGAAGAACC‐3′ (reverse); mouse NFATc1, 5′‐TGGAGAAGCAGAGCACAGAC‐3′ (forward) and 5′‐GCGGAAAGGTGGTATCTCAA −3′ (reverse); mouse cathepsin K, 5′‐TCAGAAGATGACGGGACTCA‐3′ (forward) and 5′‐TCTTGAGTTGGCCCTCCA‐3′ (reverse); mouse CTR, 5′‐TTTCAAGAACCTTAGCTGCCAGAG −3′ (forward) and 5′‐CAAGGCACGGACAATGTTGAGAG −3′ (reverse); mouse T‐type calcium channel Cav3.1, 5′‐TTCATCGCCCTCATGACTTT‐3′ (forward) and 5′‐TTCTTCCTGTCCCCATCACC‐3′ (reverse); mouse T‐type calcium channel Cav3.2, 5′‐ATGTACTCACTGGCTGTGACC‐3′ (forward) and 5′‐GAGTCCAAAAGAGTGTGGGC‐3′ (reverse); and mouse glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH), 5′‐ACCACAGTCCATGCCATCAC‐3′ (forward) and 5′‐TCCACCACCCTGTTGCTGTA‐3′ (reverse). The PCR products were separated on 2% agarose gels and visualized by ethidium bromide staining. The sizes of the PCR products for mouse COX‐2, RANKL, OPG, NFATc1, cathepsin K, CTR, T‐type calcium channels (Cav3.1 and Cav3.2), and GAPDH were 490, 587, 630, 150, 200, 450, 224, 146, and 452 bp, respectively.

Western blot analysis for c‐fos

Cells were lysed in 0.1% NP‐40 lysis buffer (20 mM Tris [pH 7.5], 50 mM β‐glycerophosphate, 150 mM NaCl, 1 mM EDTA, 25 mM NaF, 1 mM Na₃VO₄, 1× protease inhibitors cocktail [Sigma]). Whole cell extracts were electrophoresed on a 10% SDS‐polyacrylamide gel and transferred onto a PVDF membrane (Clear blot P membrane; Atto, Tokyo, Japan). After blocking with 5% skin milk in Tris‐buffered saline containing 0.1% Tween 20 (TBS‐T), the membrane was incubated with the rabbit polyclonal antibodies against mouse c‐fos (Santa Cruz Biotechnology, Santa Cruz, CA, USA) (1:500) in TBS‐T containing 5% skim milk, and the bound antibodies were visualized using ECL (Amersham), followed by exposure to X‐ray film.

Localization of NFATc1

The localization of NFATc1 in cultured mouse BMMϕ and osteoclasts was examined using an immunocytochemical technique as described.⁽¹⁸⁾ BMMϕ and osteoclasts treated with or without DPH at 200 μM were fixed with 4% PFA for 15 min. Cells were washed with 0.2% Triton X‐100 in PBS for 10 min, blocked with 1% BSA in PBS, and incubated with monoclonal anti‐NFATc1 antibody (7A6; Santa Cruz), followed by biotinylated goat anti‐mouse IgG antibody (Vector Laboratories, Burlingame, CA, USA) and fluorescein‐conjugated streptavidin (Vector). Cells were counterstained with propidium iodide (Vector).

Measurement of cytosolic calcium oscillations

The effect of DPH on cytosolic calcium oscillations in BMMϕs was measured using a confocal laser scanning microscope (LSM510; Carl Zeiss, Jena, Germany) according to the methods described previously.⁽⁹⁾ BMMϕs were cultured in a 35‐mm‐diameter glass bottom dish in the presence of RANKL (100 ng/ml) and M‐CSF (50 ng/ml) for 72 h. BMMϕs were incubated with or without DPH (200 μM) for 1 h and treated with 5 μM fluo‐4 AM, 5 μM Fura Red AM, and 0.05% Pluronic F127 for 30 min in serum‐free DMEM (Sigma). Cells were washed with DMEM, incubated in DMEM containing 10% FBS and M‐CSF (10 ng/ml) for 20 min, and further washed with Hank's balanced salt solution. Cells were excited at 488 nm, and emission at 505–530 nm for fluo‐4 and 600–680 nm for Fura red were acquired simultaneously at 10‐s intervals. The ratio of the fluorescence intensity of the fluo‐4 to Fura red was calculated to estimate intracellular calcium concentrations in single cells. The ratio of fluorescence intensity under baseline conditions was divided by the maximum ratio increase obtained by adding ionomycin (10 μM) and expressed as the percent maximum ratio increase.

RNA interference analysis

To repress Cav3.2 expression, short hairpin RNA (shRNA) for Cav3.2 was expressed in BMMϕ using a retroviral vector system (Clontech Laboratories, Mountain View, CA, USA). The short hairpin DNA oligos (Cav3.2 no. 1: GCCTCTAATACTAGAAACTGAACTCGAGTTCAGTTTCTAGTATTAGAGGC; Cav3.2 no. 2: GCTAGAATGCAGCGAGGATAACTCGAGTTATCCTCGCTGCATTCTAGC) were synthesized and ligated to the pSIREN‐RetroQ vector. The cloned constructs were transfected with virus vectors into Plate‐E cells to produce viruses. Two days later, the supernatant of cultures was harvested. BMMϕs were infected with appropriate amounts of the shRNA‐expressing virus for 1 day. The cells were further cultured in the presence of RANKL (100 ng/ml) and M‐CSF (50 ng/ml) for 3 days, fixed, and stained for TRACP. TRACP⁺ cells containing three or more nuclei were counted as osteoclasts. To determine expression of Cav3.1 and Cav3.2, BMMϕs infected with virus were cultured in the presence of M‐CSF (50 ng/ml) for 3 days and subjected to RT‐PCR analysis.

Statistical analysis

In vivo experiments were repeated twice, and similar results were obtained. In vitro experiments were performed at lease three times, and similar results were obtained. The results were expressed as the means ± SD for three or more cultures. The significance of differences was determined using Student's t‐test.

RESULTS

DPH inhibits osteoclastic bone resorption in vivo

To elucidate effects of DPH on bone resorption in vivo, we used RANKL Tg mice as a bone resorption model.⁽³¹⁾ RANKL Tg mice exhibit severe osteoporosis because of increased osteoclastic bone resorption. In preliminary experiments, we confirmed that the number of osteoclasts was significantly increased in the calvarial sections prepared from RANKL Tg mice in comparison with the number in wildtype (WT) mice. DPH or vehicle was daily injected into the subcutaneous tissue overlying calvaria of RANKL Tg mice and WT mice for 7 days (2 mg/kg body weight/d). Calvaria and blood were collected on day 8. Calvarial sections prepared from RANKL Tg mice exhibited enhanced osteoclastic bone resorption (Fig. 1A). The osteoclast surface of RANKL Tg mice was significantly greater than that of WT mice. Daily injections of DPH decreased the osteoclast surface in RANKL Tg mice to the control level in WT mice. The osteoclast surface in WT mice was slightly decreased by the DPH injection, although this effect was not significant. Daily injections of DPH significantly decreased the serum activity of TRACP5b in RANKL Tg mice (Fig. 1B). In contrast, there was no significant difference in serum ALP activities among the four groups of mice (Fig. 1B).

Figure 1

Daily injection of DPH suppresses osteoclastic bone resorption in RANKL Tg mice. DPH (2 mg/kg body weight) or vehicle was daily injected into the subcutaneous tissue overlying calvaria of 7‐ to 8‐wk‐old WT and RANKL Tg mice for 7 days (1 injection/d). The mice were killed on day 8. (A) Histological analysis of calvaria. Calvaria were processed for TRACP staining. TRACP⁺ cells appeared as red cells. Osteoclast surface was measured. Scale bar, 1 mm. (B) Serum parameters. Serum TRACP5b and ALP activities were determined. Results were expressed as the mean ± SD for five animals. Significantly different from WT mice treated with vehicle and from RANKL Tg mice treated with DPH, ^ap < 0.01.

DPH inhibits osteoclast formation in vitro

DPH strongly inhibited the osteoclastic bone resorption in vivo. We next examined whether DPH acts on osteoblasts, osteoclast precursors, or both in vitro. Mouse osteoblasts and bone marrow cells were co‐cultured in the presence of LPS with or without increasing concentrations of DPH (Fig. 2A). TRACP⁺ osteoclasts were formed within 7 days in response to LPS. DPH added to the co‐culture for the entire 7 days dose‐dependently inhibited the osteoclast formation induced by LPS (Fig. 2A). We previously reported that LPS stimulated osteoclasts to form in co‐culture through the induction of RANKL expression and suppression of OPG expression in osteoblasts. PGE₂ produced by osteoblasts in response to LPS was involved in the formation of osteoclasts in the co‐culture. We examined the effects of DPH on the expression of COX‐2, RANKL, and OPG in osteoblasts treated with LPS (Fig. 2B). Treatment of osteoblasts with LPS enhanced the expression of COX‐2 and RANKL and suppressed that of OPG.⁽³⁵⁾ DPH showed no effect on the expression of COX‐2, RANKL, or OPG. Primary osteoblasts were cultured with increasing concentrations of DPH, and cell growth was monitored by Alamar blue assay (Fig. 2C). The growth of osteoblasts was not affected by DPH at 100 and 200 μM at the time points measured. Consistent with previous studies, treatment of mouse osteoblasts for 6 days with DPH at 200 μM enhanced the ALP activities (data not shown). These results suggest that functions of osteoblasts including the osteoclast regulatory function is not impaired by the treatment with DPH.

Figure 2

DPH inhibits osteoclast formation in mouse co‐cultures. (A) Effects of DPH on osteoclast formation in mouse co‐cultures. Primary mouse osteoblasts and mouse bone marrow cells were co‐cultured for 7 days in the presence of LPS (100 ng/ml) with increasing concentrations of DPH. Cells were stained for TRACP. TRACP⁺ multinucleated cells were counted as osteoclasts. Results were expressed as the mean ± SD for five cultures. Significantly different from the culture treated with LPS alone, ^ap < 0.01. (B) Effects of DPH on the expression of COX‐2, RANKL, and OPG mRNAs in mouse osteoblasts. Primary osteoblasts were preincubated with or without DPH (200 μM) for 15 min. Cells were further cultured for 24 h with or without DPH (200 μM) in the presence or absence of LPS (100 ng/ml). Total RNA was extracted from cells, and the expression of COX‐2, RANKL, OPG, and GAPDH was detected by RT‐PCR. (C) Effects of DPH on growth of mouse osteoblasts. Primary osteoblasts were cultured for 2 days in the presence or absence of DPH at 100 and 200 μM. Cell viability was determined by Alamar Blue assay. Results were expressed as the mean ± SD for four cultures.

We examined the effect of DPH on osteoclast formation in BMMϕ cultures treated with RANKL and M‐CSF (Fig. 3A). RANKL stimulated osteoclasts to form in BMMϕ cultures in the presence of M‐CSF. DPH dose‐dependently inhibited the RANKL‐induced formation of osteoclasts. Human CD14⁺ cells prepared from peripheral blood have been reported to differentiate into osteoclasts in the presence of RANKL and M‐CSF. Treatment of human CD14⁺ cells with RANKL and M‐CSF stimulated the formation of TRACP⁺ osteoclasts, which was dose‐dependently inhibited by adding DPH (Fig. 3B). M‐CSF supported the growth of both mouse BMMϕ and human CD14⁺ cells (Fig. 3C). DPH at 200 μM showed no significant inhibitory effect on the M‐CSF–induced proliferation of mouse BMMϕs. The M‐CSF–supported growth of human CD14⁺ cells was slightly inhibited by DPH (Fig. 3C). However, the inhibitory effect of DPH on osteoclastic differentiation of human CD14⁺ cells was much stronger than that on the growth of the human cells supported by M‐CSF. These results suggest that DPH directly acts on osteoclast precursors and inhibits their differentiation into osteoclasts.

Figure 3

DPH inhibits osteoclast formation in mouse BMMϕ cultures and in human CD14⁺ cells cultures. (A) Effects of DPH on osteoclast formation in mouse BMMϕ cultures. Mouse BMMϕs were cultured with in the presence of RANKL (100 ng/ml) and M‐CSF (50 ng/ml) together with increasing concentrations of DPH. After 3 days, the cells were stained for TRACP. TRACP⁺ multinucleated cells were counted as osteoclasts (right panels). Scale bar, 100 μm. (B) Effects of DPH on osteoclast formation in human CD14⁺ cell cultures. Human CD14⁺ cells were cultured in the presence of RANKL (100 ng/ml) and M‐CSF (50 ng/ml) together with increasing concentrations of DPH. After 6 days, the cells were stained for TRACP. TRACP⁺ multinucleated cells were counted as human osteoclasts. Results were expressed as the mean ± SD of four cultures. Significantly different from the culture treated with RANKL, ^ap < 0.01. Scale bar, 100 μm. (C) Effects of DPH on growth of mouse BMMϕ and human CD14⁺ cells. Mouse BMMϕs and human CD14⁺ cells were cultured in the presence of M‐CSF (50 ng/ml) with or without DPH (200 μM). After 3 and 6 days, cell viability was determined by Alamar blue assay. Results were expressed as the mean ± SD for four cultures. Significantly different from the culture treated with vehicle, ^ap < 0.01.

DPH inhibits osteoclast function in vitro

We next examined effects of DPH on bone resorption induced by LPS in paired cultures of neonatal mouse calvaria. Treatment of mouse calvaria with LPS significantly increased the release of calcium from bone into the culture medium, suggesting that LPS enhanced osteoclastic bone resorption in the calvaria.⁽³⁴⁾ DPH suppressed the LPS‐induced bone resorption in a dose‐dependent manner, and a significant inhibition was observed at 200 μM (Fig. 4A). However, DPH showed no effect on the static bone resorption in the calvaria in the absence of LPS.

Figure 4

DPH inhibits osteoclast function in cultures. (A) Effect of DPH on bone resorption in mouse calvarial cultures. Paired half calvaria were cultured for 3 days in the presence of LPS (100 ng/ml) together with DPH (50 and 200 μM). Bone‐resorbing activity was expressed as the net concentration of calcium released from bone. Results were expressed as the mean ± SD of five cultures. Significantly different from the culture treated with LPS alone, ^ap < 0.01. (B and C) Effect of DPH on function of mouse osteoclasts. Osteoclast preparations were cultured on dentine slices with increasing concentrations of DPH. (B) After 24 h, the cells were stained for TRACP (right panels, TRACP). After removal of cells, dentine slices were stained with Mayer's hematoxylin (right panels, Pits). The number of resorption pits was counted (left panel). Results were expressed as the mean ± SD for five cultures. Significantly different from the culture treated with vehicle, ^ap < 0.01. Scale bar, 100 μm. (C) After 24 h, some dentine slices were processed for F‐actin staining (right panels). Osteoclasts having actin rings were counted (left panel). Results were expressed as the mean ± SD for five cultures. Significantly different from the culture treated with vehicle, ^ap < 0.01. Scale bar, 100 μm.

We next examined the effects of DPH on the pit‐forming activity and actin ring‐forming activity of osteoclasts placed on dentine slices. When the osteoclast preparation was cultured on dentine slices for 24 h, many resorption pits formed on the dentine slices (Fig. 4B). DPH added at 100 μM significantly inhibited the pit‐forming activity of osteoclasts (Fig. 4B). The inhibitory effect of DPH seemed not to be caused by the toxic effect, because TRACP⁺ osteoclasts were similarly observed on dentine slices treated with DPH even at 200 μM (Fig. 4B). The sealing zones were observed as actin rings in pit‐forming osteoclasts on dentine slices (Fig. 4C). DPH at 100 μM almost completely disrupted actin rings in osteoclasts. These results suggest that DPH inhibits not only the differentiation but also the function of osteoclasts.

DPH inhibits NFATc1 signaling in osteoclast precursors

RANK‐mediated signaling stimulates mRNA expression of NFATc1 in osteoclast precursors.⁽⁹⁾ We examined the effect of DPH on the expression of NFATc1 in BMMϕs (Fig. 5A). Treatment of BMMϕs with RANKL for 1 day sharply induced NFATc1 expression in BMMϕs. The expression of c‐fos protein in BMMϕs was also upregulated by RANKL (Fig. 5A). Treatment of BMMϕs with RANKL for 4 days induced expression of CTR and cathepsin K, markers of osteoclasts. The expression of NFATc1 was still strong in BMMϕs treated with RANKL for 4 days (Fig. 5A). RANKL‐induced expression of NFATc1 mRNA and c‐fos protein was strongly inhibited by the simultaneous treatment with DPH for 1 day. Expression of CTR and cathepsin K in BMMϕs was also inhibited by DPH on day 4 (Fig. 5A). The expression of NFATc1 is enhanced by calcium oscillations.⁽⁹⁾ RANKL elicited cytosolic calcium oscillations in BMMϕs (Fig. 5B). The RANKL‐induced calcium oscillation was blocked by treatment with DPH (Fig. 5B). NFATc1 protein induced by RANKL was detected in nuclei in BMMϕs (Fig. 5C). In contrast, the RANKL‐induced nuclear localization of NFATc1 was not observed in BMMϕs, when they were simultaneously treated with DPH (Fig. 5C). RT‐PCR analysis showed that BMMϕs, mature osteoclasts, and osteoblasts expressed T‐type calcium channels, Cav3.1 and Cav3.2, targets of DPH (Fig. 5D). The expression level of Cav3.2 was higher than that of Cav3.1 in all types of cells examined. To determine the role of Cav3.2 in RANKL‐induced osteoclastogenesis, BMMϕs were transfected with retroviral vectors carrying two different shRNAs for Cav3.2 (Cav3.2 no.1 or Cav3.2 no. 2) or control (Fig. 5E). Expression of Cav3.2 was markedly suppressed by either Cav3.2 no.1 or Cav3.2 no.2 shRNA but not control siRNA (Fig. 5E). Blocking the expression of Cav3.2 by shRNAs significantly but not completely suppressed the RANKL‐induced osteoclast differentiation (Fig. 5F). These results suggest that DPH inhibits osteoclast differentiation through the suppression of calcium signaling and T‐type calcium channels are the targets of DPH.

Figure 5

DPH suppresses the expression of osteoclastogenesis‐related genes in mouse BMMϕs. (A) Effects of DPH on expression of osteoclastogenesis‐related markers in BMMϕs. Mouse BMMϕs were preincubated with or without DPH (200 μM) for 15 min and cultured with or without RANKL (100 ng/ml) in the presence of M‐CSF (50 ng/ml). After culture for 1 or 4 days, the expression of NFATc1, CTR, cathepsin K, and GAPDH was determined by RT‐PCR. For analysis of c‐fos protein, total cell lysates were prepared form BMMϕs cultured for 1 day and processed for Western blotting. (B) Effects of DPH on RANKL‐induced calcium oscillations in BMMϕ. BMMϕ were cultured in the presence of RANKL (100 ng/ml) and M‐CSF (50 ng/ml) for 3 days. Cells were subjected to single‐cell calcium measurements. Each color indicates an individual cell in the same field. (C) Effects of DPH on cellular localization of NFATc1 in BMMϕs. Mouse BMMϕs were cultured with or without RANKL (100 ng/ml) in the presence of M‐CSF (50 ng/ml). DPH (200 μM) was added to some of the cultures treated with RANKL. After 24 h, the cells were stained with anti‐NFATc1 antibody (green) and propidium iodide (nuclear staining, red) (left panels). The nuclear localization of NFATc1 was confirmed in merged images of BMMϕs treated with RANKL (yellow, arrows). NFATc1‐positive nuclei were counted in BMMϕs (right panel). Results were expressed as the mean ± SD for five cultures. Significantly different from the culture treated with RANKL, ^ap < 0.01. Scale bar, 50 μm. (D) Expression of Cav3.1, Cav3.2, and CTR in osteoblasts, BMMϕs, and purified osteoclasts. Total RNA was extracted, and the expression of Cav3.1, Cav3.2, CTR, and GAPDH was determined by RT‐PCR using specific primers. (E) Depletion of Cav3.2 by shRNAs in BMMϕs. BMMϕs were infected with retroviral vectors carrying two different shRNAs for Cav3.2 (cav3.2 no. 1 and Cav3.2 no.2) or control. After infection for 72 h, total RNA was extracted, and the expression of Cav3.1, Cav3.2, CTR, and GAPDH was determined by RT‐PCR using specific primers. (F) Effects of shRNA‐mediated depletion of Cav3.2 on RANKL‐induced osteoclast differentiation. BMMϕs were infected with retroviral vectors as described in E. After infection for 24 h, BMMϕs were cultured for 3 days in the presence of RANKL (100 ng/ml) and M‐CSF (50 ng/ml). Cells were stained for TRACP. TRACP⁺ multinuclear cells were counted as osteoclasts. Results were expressed as the mean ± SD for five cultures. Significantly different from the culture treated with control shRNA, ^bp < 0.05.

DPH inhibits NFATc1 signaling in mature osteoclasts

RANKL stimulates the nuclear localization of NFATc1 and induces auto‐amplification of NFATc1 in osteoclast precursors.^(9,10) We examined effects of DPH on the distribution of NFATc1 in mature osteoclasts treated with and without RANKL. BMMϕs were cultured for 4 days with RANKL and M‐CSF to produce osteoclasts. Osteoclasts were incubated for 60 min in the absence of RANKL. Osteoclasts were preincubated with or without DPH at 200 μM for 15 min and further cultured for 90 min with and without RANKL in the presence of M‐CSF (Fig. 6A). Treatment with RANKL induced the nuclear localization of NFATc1 in osteoclasts, which was strongly suppressed by pretreatment and simultaneous treatment with DPH. DPH added to osteoclast cultures for 3 h strongly inhibited the expression of NFATc1 (Fig. 6B). Expression of CTR and cathepsin K in osteoclasts was not affected at 3 h but decreased thereafter by the treatment with DPH (Fig. 6B). These results suggest that DPH also blocks NFATc1 signaling in mature osteoclasts through the inhibition of calcium signaling.

Figure 6

DPH suppresses nuclear localization of NFATc1 in mature osteoclasts. (A) Effects of DPH on nuclear localization of NFTAc1 in mature osteoclasts. Osteoclasts were preincubated with or without DPH (200 μM) for 15 min and cultured for 90 min with or without RANKL (100 ng/ml) in the presence of M‐CSF (50 ng/ml). Cells were stained with anti‐NFATc1 antibody (green) and propidium iodide (red) (left panels). Bottom two panels are high‐power views of the portions in the middle panels. The nuclear localization of NFATc1 (yellow, arrows) was increased in osteoclasts in response to RANKL treatment. NFATc1‐positive nuclei were counted in osteoclasts (right panel). Results were expressed as the mean ± SD for five cultures. Significantly different from the culture treated with RANKL, ^ap < 0.01, ^bp < 0.05. Scale bar, 50 μm. (B) Effects of DPH on osteoclastogenesis‐related markers in osteoclasts. Osteoclasts were cultured with or without DPH (200 μM) for 3, 5, and 12 h in the presence of RANKL (100 ng/ml) and M‐CSF (50 ng/ml). Total RNA was extracted, and the expression of NFATc1, CTR, cathepsin K, and GAPDH was determined by RT‐PCR using specific primers.

DISCUSSION

We showed here that DPH inhibits osteoclastic bone resorption but has no adverse effects on osteoblastic bone formation in vivo. DPH at serum concentrations of 40–80 μM was shown to act as an anticonvulsant.⁽²⁰⁾ In our experiments, DPH even at 200 μM affected neither the proliferation of osteoblasts nor the expression of RANKL OPG, and COX‐2 in osteoblasts. ALP activities of osteoblasts in culture were rather enhanced by the treatment with DPH at 200 μM (data not shown). Local administration of DPH to RANKL Tg mice significantly suppressed the serum activity of TRACP5b but not that of ALP. These results suggest that DPH inhibits osteoclastic bone resorption but not osteoblastic bone formation in vivo. Osteomalacia found in epileptic patients systemically treated with DPH is proposed to be caused by the expression of the drug‐metabolizing enzyme cytochrome p450 3A4 (CYP3A4).⁽²²⁾ CYP3A4 is involved in the degradation of 1α,25(OH)₂D₃, in hepatocytes and the intestine, thereby causing a vitamin D deficiency in epileptic patients, suggesting that DPH‐induced osteomalacia is not mediated by the direct action of DPH on osteoblasts. Taken together, these data imply that DPH inhibits osteoclastic bone resorption but not osteoblastic bone formation in vivo.

In disagreement with our findings, Onodera et al.⁽³⁰⁾ reported that DPH administered at high concentrations (20 mg/kg) in rats stimulated bone resorption and induced bone loss. It was also shown that DPH stimulated bone resorption in cultured neonatal mouse calvaria.⁽²⁹⁾ In our experiments, DPH suppressed LPS‐induced bone resorption in neonatal mouse calvarial cultures. The pit‐forming activity of osteoclasts was also inhibited by adding DPH. This discrepancy may be explained by the different experimental conditions used. DPH at high concentrations may induce the reverse effect on bone resorption through unknown mechanisms.

Recent studies have shown that NFATc1 signaling plays a role in osteoclast function as well as osteoclast differentiation.^(18,36) Komarova et al.⁽¹⁸⁾ reported that acidosis and RANKL caused the nuclear accumulation of NFATc1 in authentic osteoclasts, which in turn enhanced their pit‐forming activity. Calcineurin mediated NFATc1 activation in both cases. We have shown that stimulation with RANKL induces the nuclear accumulation of NFATc1 in mature osteoclasts. DPH inhibited the RANKL‐induced nuclear accumulation of NFATc1 and NFATc1 expression, followed by the suppression of CTR and cathepsin K expression. Cyclosporine A, a calcineurin inhibitor, inhibited the pit‐forming activity of osteoclasts (data not shown). These results suggest that DPH inhibits osteoclast function through suppression of the activation and auto‐amplification of NFATc1.

c‐fos plays important roles in NFATc1‐induced osteoclastogenesis. Takayanagi et al.⁽⁹⁾ showed that c‐fos is required for NFATc1‐induced TRACP promoter activity. Matsuo et al.⁽³⁷⁾ reported that NFATc1 is a transcriptional target of c‐fos, and NFATc1 transfection rescued osteoclastogenesis in c‐fos–deficient macrophages in the presence of RANKL. In our experiments, DPH inhibited both activation of T‐type calcium channels and RANKL‐induced expression of c‐fos protein in BMMϕs. Calcium influx through L‐ and N‐type channels was shown to increase c‐fos expression in sympathetic neurons by electrical stimulation.⁽³⁸⁾ It was also reported that L‐type calcium channel blockers suppressed haloperidol‐induced c‐fos expression in the lateral part of the neostriatum in rats.⁽³⁹⁾ These results suggest that calcium signals play a role in c‐fos expression in several types of cells including BMMϕs. Therefore, it is possible that DPH may inhibit NFATc1 signals through suppression of c‐fos expression. The initial interaction between NFATc1 and c‐fos in osteoclast precursors seems to be important to determine the differentiation pathway to osteoclasts.

Calcium‐dependent dephosphorylation of NFATc1 by calcineurin is prerequisite for the nuclear accumulation of this transcription factor.^(9,10) Several potential targets for DPH have been identified in the central nervous system.^(19–21) Among them, the T‐type calcium channel⁽⁴⁰⁾ is now proposed to be a target of DPH to inhibit osteoclast differentiation and function. Blocking the expression of Cav3.2 by shRNAs in BMMϕs suppressed their differentiation into osteoclasts. DPH also inhibited RANKL‐induced calcium oscillations and nuclear accumulation of NFATc1. We also examined whether DPH action is rescued by the overexpression of NFATc1 in BMMϕs. Transfection of BMMϕs with retroviral vectors carrying NFATc1 stimulated their differentiation into osteoclasts. DPH slightly inhibited the osteoclast formation induced by NFATc1 (control vector, 24 ± 8; NFATc1 vector, 152 ± 9; NFATc1 vector + DPH, 102 ± 26 [TRACP⁺ cells/well, the means ± SD, n = 4]). Thus, the inhibitory effect of DPH on the NFATc1‐induced osteoclast formation was much weaker than that on RANKL‐induced osteoclastogenesis (Fig. 3A). These results suggest that DPH blocks calcium signaling by inhibiting T‐type calcium channels in BMMϕs, which in turn suppresses activation of NFATc1.

The calcium oscillations are generated mainly by influx of extracellular calcium through multiple channels including the voltage‐gated calcium channels.^(41,42) Yang and Li⁽⁴³⁾ recently isolated a new RANKL‐induced signaling protein, regulator of G signaling protein 12 (RGS12), in osteoclasts using genome‐wide screening. RGS12 was shown to directly interact with the N‐type calcium channels. Silence of RGS12 expression by shRNA in osteoclast precursors blocked RANKL‐induced calcium oscillations.⁽⁴³⁾ These results suggest that N‐type calcium channels are also involved in RANKL‐induced osteoclastogenesis. In addition, transient receptor potential vanilloid (TRPV) members play as the gatekeepers, facilitating calcium influx in bone as well as kidney and small intestine.^(44),TRPV5 knockout mice showed increased osteoclast numbers, but TRPV5‐deficient osteoclasts failed to resorb bone.⁽⁴⁵⁾ Using TRPV4 knockout mice, Masuyama et al.⁽⁴⁶⁾ showed that TRPV4‐mediated calcium influx is required to sustain NFATc1‐dependent gene expression in osteoclast differentiation. The inositol 1,4,5‐triphosphate type 2 receptor (IP₃R2) releases calcium from intracellular storage sites in the endoplasmic reticulum. Recently, Kuroda et al.⁽⁴⁷⁾ reported that BMMϕs obtained from IP₃R2 knockout mice did not exhibit calcium oscillation or differentiation into osteoclasts in response to RANKL plus M‐CSF. However, IP₃R2‐deficient BMMϕs could differentiate into osteoclasts in co‐culture with osteoblasts in the absence of calcium oscillation. These results including our own suggest that both calcium channels and IP₃R2 are involved in the intracellular calcium oscillations and that DPH may act on osteoblasts to inhibit the process of calcium oscillation‐independent osteoclastogenesis.

Periodontal diseases are induced by chronic bacterial infections.⁽⁴⁸⁾ Severe alveolar bone loss caused by excessive bone resorption is observed in periodontal diseases. LPS has been proposed to be a potent stimulator of alveolar bone resorption.⁽³⁴⁾ DPH inhibited the LPS‐induced bone resorption in organ cultures and LPS‐induced osteoclast formation in co‐cultures. These results suggest that topical application of DPH is a therapeutic treatment to prevent alveolar bone loss induced by periodontitis. However, there may be one problem in using DPH as a therapeutic drug in case of periodontitis. Gingival overgrowth is often observed as an adverse side effect in patients receiving anticonvulsants including DPH, cyclosporine A, and calcium channel blockers.⁽⁴⁹⁾ The pathogenesis of drug‐induced gingival overgrowth is not entirely clear. A disturbance of calcium ion influx into specific cell populations may be involved in the tissue reaction with alterations in collagen metabolism.⁽⁴⁹⁾ The regeneration of periodontal ligaments is important for preventing periodontal diseases, and DPH treatment can be expected to stimulate the growth of periodontal ligaments. Although it is not known whether the topical administration of DPH induces gingival overgrowth, this treatment is worth considering as a future therapeutic approach for prevention of alveolar bone loss. Thus, topical application of DPH may be a possible therapeutic treatment to prevent local bone loss cased by inflammation diseases.

Acknowledgements

This work was supported in part by Grants‐in‐Aid 18791604, 18390495, 18390557, and 18659618 from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

Author notes

The authors state that they have no conflicts of interest

Published online on March 16, 2009