Zinc inhibits osteoclast differentiation by suppression of Ca2+-Calcineurin-NFATc1 signaling pathway

Zinc, an essential trace element, inhibits osteoclast differentiation in vitro and in vivo. The molecular mechanism for the inhibitory effect of zinc, however, is poorly understood. The purpose of this study was to investigate the effect of zinc and determine its molecular mechanism on receptor activator of NF-κB ligand (RANKL)-induced osteoclastogenesis in mouse bone marrow-derived monocyte cells (BMMs) and RAW264.7 cells. In BMMs, zinc treatment during osteoclast differentiation decreased RANKL-induced osteoclast formation in a dose-dependent manner. We show that zinc suppressed the mRNA levels of nuclear factor of activated T-cells, cytoplasmic 1 (Nfatc1). Zinc also accumulated phospho-Nfatc1 (p-Nfatc1) in the cytosol in a dose-dependent manner and inhibited the translocation of Nfatc1 to the nucleus in RAW264.7 cells. Zinc suppressed the activities of Nfatc1 in the nucleus without changing the activities of NF-κB in RAW264.7 cells. In contrast, calcineurin activity decreased in response to zinc but its protein level was unchanged. RANKL-induced Ca2+ oscillations were inhibited by zinc treatment, but phospho-phospholipase Cγ1 (p-PLCγ1), the upstream signaling molecule of Ca2+ oscillations, was unaffected. Moreover, a constitutively active form of Nfatc1 obviously rescued suppression of osteoclastogenesis by zinc. Taken together, these results demonstrate for the first time that the inhibitory effect of zinc during osteoclastogesis is caused by suppressing the Ca2+-Calcineurin-NFATc1 signaling pathway. Thus, zinc may be a useful therapeutic candidate for the prevention of bone loss caused by NFATc1 activation in osteoclasts.

To identify novel genes involved in osteoclastogenesis, we used two sets of microarray data from Gene Expression Omnibus (GEO) DataSets, which were comparative microarrays in mouse bone marrow-derived monocyte cells (BMMs) stimulated with or without receptor activator of NF-κB ligand (RANKL) [12,13]. We performed statistical data analyses using the R program. From these analyses, we found intersections between the two sets of data. Among the intersections, Mt3, which is known to regulate the intracellular level of zinc, and other zincrelated genes were up-regulated (log 2 ratio > 4.0) during osteoclast differentiation (Additional file 1: Table S1). In literatures, one report showed that dietary zinc and Metallothionein (MT) interact in postnatal bone growth [14]. Also, Lee et al. reported that zinc regulates T cell receptor signaling [15]. We thus suggest that zinc may play an important physiologic role in osteoclastogenesis signaling pathways.
Zinc is an important trace element for biological signaling pathways, but also acts as a second messenger in cells [16]. Zinc supplementation has been reported to inhibit bone loss in an adjuvant-induced rheumatoid arthritis rat model, promoting bone formation and suppressing bone resorption [17]. Dietary zinc is also reported to reduce levels of tartrate-resistant acid phosphatase (TRAP), which is a specific marker of osteoclasts in the tibia and calvaria in vivo [18]. In humans, zinc intake negatively correlates with bone loss in postmenopausal women [19] and positively correlates with bone mass in premenopausal women [20,21]. Despite persuasive studies that zinc is involved in bone loss by suppressing osteoclast differentiation, the molecular mechanism for the inhibitory effect of zinc on osteoclast differentiation remains poorly understood.
In this study, to investigate the molecular mechanism of the inhibitory effect of zinc on RANKL-induced osteoclast differentiation, we focused on Nfatc1, a master transcription factor of osteoclastogenesis [9,13,[22][23][24]. Previously, Mackenzie et al. reported that extracellular zinc can regulate NFAT activity in neuronal cells. So, we confirmed previously reported findings that zinc suppresses osteoclast differentiation in vitro. We discerned that its inhibitory mechanism was involved in blocking the Ca 2+ -Calcineurin-NFATc1 signaling pathway.

Zinc inhibits osteoclast formation and fusion from BMMs
To determine whether zinc is cytotoxic to BMMs and RAW264.7 cells, we first examined the cells' viability using an EZcytox cell viability assay kit, which estimates the number of surviving cells using WST-1 (4-[3-(4iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate). BMMs and RAW264.7 cells, which were each treated with up to 200 and 100 μM ZnSO 4 , were viable for 4 days and 24 hours, respectively ( Figure 1). We thus designated 100 μM ZnSO 4 as the maximal concentration that was nontoxic to both BMMs and RAW264.7 cells. We investigated the effects of zinc on osteoclast formation of BMMs in vitro by treating BMMs with M-CSF and RANKL in the presence or absence of zinc. Zinc treatment inhibited osteoclast formation in a dose-dependent manner as shown by a decrease in the number of TRAP-positive multinucleated osteoclasts (Figure 2A). Notably, huge TRAP-positive multinucleated osteoclasts (nuclei ≥ 6) decreased in the zinc-treated group ( Figure 2B). The TRAP activities of BMMs induced by RANKL were also inhibited by zinc in a dose-dependent manner ( Figure 2C). These results suggest that zinc inhibits osteoclast formation and fusion.

Zinc suppresses Nfatc1 expression and transcriptional/ DNA binding activity
To elucidate the molecular mechanism of zinc's inhibition of osteoclast formation and fusion, we analyzed the mRNA levels of genes during osteoclast differentiation in the presence of zinc. Among the many genes related to osteoclast differentiation and fusion, the mRNA levels of Nfatc1, a master regulator of osteoclast formation, and its target genes, such as Acp5, Ctsk, Mmp9, Atp6v0d2, Dcstamp, and Ocstamp [9,25], were decreased ( Figure 3A). During osteoclastogenesis periods in BMMs, the mRNA level of Nfatc1 gradually increased due to auto-amplification. Zinc, however, suppressed Nfatc1 mRNA expression as much as FK506, a known inhibitor of calcineurin-NFATc1 signaling during osteoclastogenesis ( Figure 3B).
To analyze how zinc suppresses Nfatc1 mRNA expression, we evaluated whether zinc inhibits osteoclast differentiation signaling pathways. Calcineurin dephosphorylates cytosolic p-Nfatc1 after which the dephosphorylated Nfatc1 translocates to the nucleus. We thus evaluated the protein levels of cytosolic p-Nfatc1 and nuclear Nfatc1 in RAW264.7 cells. Zinc dose-dependently increased cytosolic p-Nfatc1. In contrast, nuclear Nfatc1 dose-dependently decreased in response to zinc ( Figure 4A). As shown in Figure 4C, the expression and transcriptional activity of Nfatc1 were induced in RAW264.7 cells after exposure for 30 minutes to RANKL. Zinc significantly reduced the protein level of activated Nfatc1 as much as FK506. These results correlated with the transcriptional and DNA binding activities of Nfatc1 ( Figure 4D, 4E, left panel). NF-κB transcriptional and DNA binding activities were also induced by RANKL but were not inhibited by zinc or FK506 ( Figure 4D, 4E, right lower panel).

Zinc inhibits calcineurin activity but not expression
We investigated calcineurin activity and its protein expression in the upstream Nfatc1 signaling pathway during osteoclast differentiation. After exposure to RANKL for 30 minutes in the presence or absence of zinc or FK506 in RAW264.7 cells, PP2B-Aα, the catalytic subunit of calcineurin, was unchanged in terms of protein expression ( Figure 4B). However, zinc and FK506 similarly inhibited RANKL-induced calcineurin activity ( Figure 4F).

Nfatc1 rescues the inhibitory effects of zinc during osteoclastogenesis in RAW264.7 cell
We examined whether Nfatc1 could rescue defects of osteoclastogenesis using zinc. Indeed, when we ectopically expressed a constitutively active form of Nfatc1 (caNfatc1) in RAW264.7 cells, caNfatc1 completely rescued suppression of osteoclastogenesis by zinc ( Figure 6A). TRAP activity was significantly increased compared with the mock ( Figure 6B). These results indicate that impairment of Nfatc1 activation is the cause of suppression during osteoclastogenesis.

Discussion
Here we show that zinc inhibits RANKL-induced Nfatc1 translocation to the nucleus by decreasing calcineurin phosphatase activity during the early period of osteoclastogenesis. This ultimately inhibits osteoclast differentiation. Interestingly, zinc immediately diminished RANKL-induced Ca 2+ oscillations throughout the middle or late period of osteoclastogenesis ( Figure 5A) but did not suppress RANKL-induced PLCγ1 phosphorylation ( Figure 5B), indicating that the inhibition of Ca 2+ oscillations may be independent of PLCγ1. We thought that zinc could be affecting the downstream signaling pathway of the ITAM-containing adaptor-Syk-PLCγ1 axis.
Zinc is known to stimulate osteoclast apoptosis, which is mediated through Ca 2+ signaling [38]. We first defined a concentration of ZnSO 4 (100 μM) that was nontoxic . Cytosolic phospho-Nfatc1 and nuclear Nfatc1 were analyzed using western blot. (D, E) RAW264.7 cells were stimulated with RANKL (R) or RANKL (R) plus ZnSO 4 (30 or 100 μM) for 30 minutes. Nuclear fractions were prepared, and the transcriptional and DNA binding activity of Nfatc1 and NF-κB were measured using luciferase reporter assay and ELISA, respectively. RLU, Relative Light Units (F) RAW264.7 cells were cultured as shown in panel C. Cytosolic PP2B-Aα was examined by western blot and calcineurin activity was compared with the treated groups. Data are presented as the mean ± S.D. of three independent experiments; * p < 0.05 compared to RANKL (R).
to both BMMs and RAW264.7 cells. Although 100 μM ZnSO 4 significantly inhibits TRAP activity in osteoclasts ( Figure 2C), this concentration of zinc does not affect BMM viability ( Figure 1A). As shown in Figure 2B, the number of fused multinucleated osteoclasts significantly decreased upon zinc treatment, indicating that zinc affects the fusion of multinucleated cells in BMMs. These findings suggest that the inhibitory effect of zinc on osteoclast differentiation was not caused by zinc cytotoxicity. In Figure 3B, mRNA levels of Nfatc1 decreased due to zinc and FK506 treatments as osteoclast differentiation progressed. Yet there were still some Nfatc1 inductions at 48 and 72 hours after the zinc and FK506 treatments. Asagiri et al. reported that an autoamplification of Nfatc1 was essential in osteoclast differentiation [39]. As shown in Figure 4D, the inhibitory effects of zinc and FK506 for Nfatc1 transcriptional activity were not 100%. There was some residual activity. We thought that while some autoamplification of Nfatc1 caused part of the Nfatc1 inductions after zinc and FK506 treatments, it might be not enough for osteoclast maturation.
There were many reports that zinc can inhibit calmodulin, which is an important activator of calcineurin. Brewer reported that zinc inhibits calmodulin in the erythrocyte [40]. Zinc inhibits calmodulin by competing with Ca 2+ binding to calmodulin, which has also resulted in a conformational change of the protein [40][41][42][43]. The phosphorylation and activity of calmodulin-dependent protein kinase II is modulated by zinc as well [44]. One in vivo study showed that calmodulin level decreased in epidermal cells after intraperitoneal or intradermal zinc injections [45]. Another study showed that zinc treatment reduced calmodulin in adipocytes of obese mice [46]. We found that zinc decreased the activity of calcineurin in the early period of osteoclastogenesis of RAW264.7 cells ( Figure 4F). Our results were consistent with a previous article that showed Nfatc1 translocation to the nucleus and an activation of calcineurin within 30 to 40 minutes after RANKL stimulation in RAW264.7 cells [47]. In addition, calcium which comes from intracellular calcium storage, such as the sarcoplasmic reticulum, may be working as a second messenger in the early period of osteoclast differentiation. Thus, we thought that zinc might inhibit calcineurin activity by suppression of calmodulin as shown in previous reports [40][41][42][43][44][45][46]. Subsequently, Nfatc1 would not be able to change into its active form and stays in the cytosol. When we overexpressed constitutively active NFATc1 in RAW 264.7 cells, the inhibitory phenotype for osteoclasts rescued ( Figure 6). Since the constitutively active NFATc1 lacks phosphorylation sites in the regulatory domain, it would be expected to effects of zinc on NFATc1 kinases such as calcineurin which is the most important NFATc1 kinase. So, we thought that the rescue of the phenotype was caused by calcineurin-NFATc1 pathway. But, we cannot exclude effects of zinc on other modulators of NFATc1 pathway.
It was previously reported that zinc treatment for 24 hours suppresses RANKL-induced NK-κB luciferase activity in RAW264.7 cells [48]. It was also reported that zinc supplementation for 3 months decreases the DNA binding capacity of NF-κB in mononuclear cells from sickle cell disease patients [49]. These results differed from our own results that zinc did not inhibit NK-κB transcriptional activity in RAW264.7 cells ( Figure 4D). This difference may be controversial. In general, however, an efficiency of DNA transfection in RAW264.7 cells is very poor. Thus, we used electroporation for more efficient gene expression instead of chemical transfections and increased the efficiency up to 65%. We thought that this may be a cause of the difference. Also, Hie et al. demonstrated that Zn treatment could inhibit RANK expression during osteoclast differentiation [50], but its molecular mechanism was unclear. Further investigation is needed in future studies.
Intracellular zinc signaling consists of two signaling pathways. The early zinc signal, which is transcriptionindependent, is rapidly induced by an extracellular stimulus, such as FcεRI [16]. Late zinc signaling involves transcription-dependent changes in expression of zinc transporters, such as ZIP (SLC30A) and ZnT (SLC39A) [16,[51][52][53]. Zinc transporters are ubiquitously expressed and play a role in maintaining the levels of cellular zinc by controlling its influx, efflux, and sequestration. Zinc signaling also modulates numerous cellular processes involved in cell differentiation, proliferation, and growth [54]. Because zinc transporters are expressed in osteoclasts and some are up-regulated during osteoclast differentiation, zinc may play an important role in osteoclast differentiation [55,56].
FK506, an immunosuppressant, is a potent inhibitor of calcineurin phosphatase activity. It inhibits both bone resorption and formation [57]. Overall, FK506 is not beneficial for increasing bone mass and quality. Zinc, on the other hand, inhibits osteoclastogenesis as well as stimulates bone formation in mice and rats [38,48]. In particular, we also found that zinc stimulates osteoblastogenesis in human mesenchymal stem cells (data not shown). Thus, if zinc could be effectively transferred in bone tissue, it may be beneficial for increasing bone mass and quality.

Conclusions
We have shown that zinc is an important inhibitory modulator during osteoclast differentiation that acts on the Ca 2+ -Calcineurin-NFATc1 signaling pathway. We proposed molecular mechanisms through which zinc may inhibit calcineurin in at least the early period of osteoclast differentiation and inhibit calcium oscillations in the middle or late period of osteoclast differentiation (Figure 7). Therefore, zinc might be a good therapeutic candidate for preventing osteoporosis and arthritis caused by NFATc1 activation in osteoclasts.

Preparation of BMMs and in vitro osteoclastogenesis
The femur and tibia were removed from 6-week-old male C57BL/6 mice. Cells derived from the bone marrow were collected and cultured in growth media containing M-CSF (10 ng/ml). After 24 hours, nonadherent cells were collected and seeded in a 100 mm dish and treated with M-CSF (30 ng/ml). After 48 hours, nonadherent cells were washed and the adherent cells were used as BMMs. BMMs were detached from the 100 mm dish using Detachin TM (Genlantis, San Diego, CA). The obtained cell pellet was resuspended and seeded on dishes or plates for osteoclastogenesis. BMMs (1 × 10 5 cells/ml) were cultured for 4 days in growth media containing M-CSF (30 ng/ml) and RANKL (120 ng/ml) with or without ZnSO 4 . Also, RAW264.7 cells were cultured for 4 days in growth media containing RANKL (35 ng/ml) with or without ZnSO 4 for osteoclastogenesis. For rescue experiments, RAW264.7 cells were transfected with constitutively active Nfatc1 plasmid by electroporation using the Amaxa Cell line Nucleofector™ kit V (Lonza).
Cell viability assay RAW264.7 cells were maintained in growth media with or without ZnSO 4 (10, 30, 60, 100, 200, or 500 μM) for 24 hours. Additionally, BMMs were cultured in growth media containing M-CSF (30 ng/ml) in the presence or absence of ZnSO 4 (10, 30, 60, 100, 200, or 500 μM) for 4 days. Cell viability assays were performed using an EZcytox cell viability assay kit (Daeillab Service, South Korea) according to the manufacturer's instructions. Briefly, the cells were plated in 96-well plates at 1 × 10 4 cells per well and cultured in growth media. At the Figure 7 The proposed molecular mechanism for the inhibitory effects of zinc on RANKL-induced osteoclastogenesis. Schematic models of the inhibitory effects of zinc on the Ca 2+ -Calcineurin-NFATc1 signaling pathway; (A) Zinc may inhibit calcineurin in the cytosol in the early period of osteoclastogenesis. (B) In the middle or late period, zinc could be suppressing calcium oscillations by blocking calcium influx from extracellular space. Cn A, Calcineurin A subunit; CaM, Calmodulin; P, phosphorylated.
indicated time points, cells were incubated for 4 h at 37°C with WST-1. The number of viable cells in triplicate wells was measured at an absorbance wavelength of 450 nm.

Measurement of TRAP activity and TRAP staining
TRAP activity was measured from osteoclast culture supernatants using a TRAP Staining kit (Kamiya Biomedical Company). Supernatants (30 μl) were incubated for 3 hours at 37°C with 170 μl of the chromogenic substrates in a tartrate-containing buffer. TRAP activities were measured in terms of the absorbance at a wavelength of 540 nm. TRAP was stained using a similar method as described above. Cultured cells were incubated with a fixative for 5 minutes at room temperature, washed with distilled water, and treated for 20 minutes at 37°C with the chromogenic substrate in tartrate-containing buffer. After staining, TRAP-positive multinucleated (nuclei ≥ 3) cells were counted using manual counting or a nuclei counter plug-in in the Image J program.

Real-time reverse transcription-PCR
RNA was extracted from BMMs on the indicated days using TRIZOL reagent (Invitrogen, Carlsbad, CA). cDNA was reverse transcribed using random hexamers and SuperScript-III reverse transcriptase (Invitrogen). The cDNA was used in real-time PCR with a KAPA SYBR FAST ABI Prism qPCR kit (Kapa Biosystems). The specific primer pairs are shown in Table 1. Nfatc1 and other mRNAs were measured using a StepOne (Applied Biosystems) Real-Time PCR System. The PCR program was initiated for 20 seconds at 95°C, followed by 40 thermal cycles of 3 seconds at 95°C and 30 seconds at 60°C, and terminated for 15 seconds at 95°C, 1 minute at 60°C, and 15 seconds at 95°C. Data were analyzed according to the comparative cycle threshold (Ct) method [59] and were normalized to GAPDH in each sample. We examined individual gene expression in triplicate and repeated each experiment more than three times.

Cell fractionation
RAW264.7 cells at 70-80% confluence were incubated in α-MEM containing RANKL (35 ng/ml), with or without ZnSO 4 , for the indicated times, washed, and scraped in cold PBS. Cell pellets were fractionated into cytoplasmic and nuclear fractions using a NE-PER Nuclear and Cytoplasmic Extraction Reagents kit (Pierce, Rockford, IL).

Calcineurin activity
Cellular calcineurin phosphatase activity was measured in cell extracts using a Calcineurin Cellular Activity assay kit (Enzo Life Sciences, Farmingdale, NY). In brief, cells were lysed in a lysis buffer containing protease inhibitors and centrifuged. The same amount of protein (5 μg) was used in the calcineurin activity assays. Colorimetric measurements were performed at 620 nm. The amount of phosphate released by calcineurin was calculated using a standard curve.
Intracellular Ca 2+ measurement RAW264.7 cells were seeded on a cover glass in a 35 mm dish (1 × 10 5 cells per dish) and activated for 48 hours with RANKL (35 ng/ml). Cells were incubated for 30 minutes at room temperature with 5 μM Fura-2/AM and 0.05% Pluronic F-127 (Invitrogen) and washed with a bath solution (140 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 10 mM HEPES, 1 mM CaCl 2 , 10 mM glucose, 310 mOsm, pH 7.4). The cover glass was transferred to a perfusion chamber, and the cells were continuously perfused with prewarmed (37°C) bath solution. The excitation wavelengths for Fura-2 fluorescence were 340 and 380 nm and the emission wavelength was 510 nm. The fluorescence intensity was measured by the ratio of emitted fluorescence (F 340 /F 380 ), which was monitored using a CCD camera (Universal Imaging Co., Downingtown, PA) every 2 seconds. CCD camera images were analyzed using MetaFluor software (Universal Imaging, Downingtown, PA). For the inhibition assays, ZnSO 4 or a known store-operated Ca 2+ (SOC) channel inhibitor, Gd 3+ was added 10 minutes after RANKL-induced Ca 2+ oscillations. At the end of the assay, 5 μM ionomycin (Sigma) was added.

Statistical analysis
The results are shown as the mean ± standard deviation (S.D.) from at least three independent experiments. The differences between groups were analyzed using Student's t-tests and p < 0.05 was considered statistically significant.

Additional file
Additional file 1: Table S1. List of Zinc-related genes that were upregulated (log 2 ratio > 4.0) during osteoclastogenesis.

Competing interests
The authors declare that they have no competing interests.
Authors' contributions KHP wrote the manuscript, performed most of the experiments, analyzed data, and obtained IACUC approval. BP and DMS performed calcium oscillation assays and analyzed the data. DSY, SK and JWL reviewed and helped in the data analyses. HGL, JS and JHP critiqued with the drafting and preparing of the manuscript for publication. JML conceived and designed the study and wrote the manuscript. All authors read and approved the final manuscript.