The polymethoxy flavonoid sudachitin suppresses inflammatory bone destruction by directly inhibiting osteoclastogenesis due to reduced ROS production and MAPK activation in osteoclast precursors

Inflammatory bone diseases, including rheumatoid arthritis, periodontitis and peri-implantitis, are associated not only with the production of inflammatory cytokines but also with local oxidative status, which is defined by intracellular reactive oxygen species (ROS). Osteoclast differentiation has been reported to be related to increased intracellular ROS levels in osteoclast lineage cells. Sudachitin, which is a polymethoxyflavone derived from Citrus sudachi, possesses antioxidant properties and regulates various functions in mammalian cells. However, the effects of sudachitin on inflammatory bone destruction and osteoclastogenesis remain unknown. In calvaria inflamed by a local lipopolysaccharide (LPS) injection, inflammation-induced bone destruction and the accompanying elevated expression of osteoclastogenesis-related genes were reduced by the co-administration of sudachitin and LPS. Moreover, sudachitin inhibited osteoclast formation in cultures of isolated osteoblasts and osteoclast precursors. However, sudachitin rather increased the expression of receptor activator of NF-κB ligand (RANKL), which is an important molecule triggering osteoclast differentiation, and the mRNA ratio of RANKL/osteoprotegerin that is a decoy receptor for RANKL, in the isolated osteoblasts, suggesting the presence of additional target cells. When osteoclast formation was induced from osteoclast precursors derived from bone marrow cells in the presence of soluble RANKL and macrophage colony-stimulating factor, sudachitin inhibited osteoclastogenesis without influencing cell viability. Consistently, the expression of osteoclast differentiation-related molecules including c-fos, NFATc1, cathepsin K and osteoclast fusion proteins such as DC-STAMP and Atp6v0d2 was reduced by sudachitin. In addition, sudachitin decreased activation of MAPKs such as Erk and JNK and the ROS production evoked by RANKL in osteoclast lineage cells. Our findings suggest that sudachitin is a useful agent for the treatment of anti-inflammatory bone destruction.

Introduction animal care committee and conformed to relevant guidelines and laws. Animal sacrifice was humanely performed by cervical dislocation for adult mice and decapitation for neonatal mice.

In vivo LPS-induced calvarial bone destruction model
The in vivo LPS-induced inflammatory calvarial bone destruction model was established as previously described [37]. Eight-week-old male C57BL/6J mice were injected with 100 μg LPS subperiosteally into the calvarial bone daily for 5 days. After 6 days, computed tomography (CT) scanning of the calvariae was performed using μCT (Skyscan 1172, Bruker, Billerica, MA) and reconstructed into a three-dimensional image. In addition, the calvarial bones were crushed in Buffer RLT Plus (Qiagen, Valencia, CA) using Polytron PT3100 (Kinematica AG), and the total RNA was subsequently extracted and used for cDNA synthesis using an RNeasy Mini plus kit (Qiagen).

Osteoclast formation in a co-culture of osteoblasts from calvariae and osteoclast precursors from bone marrow cells
Osteoblasts were obtained from the calvariae of 3-to 7-day-old C57BL/6J mice by sequential digestion with 0.1% collagenase/0.2% dispase II in α-MEM, as previously described [37]. The cells released from the 3 rd -5 th digestion were cultured for expansion in α-MEM/10% FBS and stored in liquid nitrogen as calvaria-derived osteoblasts. Osteoblasts (4000/well in a 96-multiwell plate) and osteoclast precursors (8000/well) derived from bone marrow cells pretreated with M-CSF (100 ng/ml) for 3 days were co-cultured in the presence of IL-1β (10 ng/ml) and PGE 2 (10 μM) for 5-6 days. After culturing, the cells were fixed in 10% formalin and stained to detect the TRAP activity using a leukocyte acid phosphatase kit (Sigma-Aldrich). The osteoclast formation in the co-culture was evaluated by counting the TRAP-positive multi-nucleated cells (MNCs) per well. In addition, TRAP activity in the conditioned medium was measured using p-nitrophenyl phosphate as a substrate, as previously described [38].

In vitro assay of osteoclastogenesis in bone marrow cells
The in vitro osteoclast formation was measured as previously described [39]. Briefly, bone marrow cells isolated from femora and tibiae are cultured for 3 days in α-MEM (ICN Biomedicals, Aurora, OH) containing 10% FBS and M-CSF (100 ng/ml) in a humidified atmosphere at 5% CO 2 . Following the removal of the non-adherent cells and the small population of stromal cells by washing the dishes with PBS and a subsequent incubation for 3 min in 0.25% trypsin/ 0.05% EDTA, the adherent monocytes were harvested by vigorous pipetting and used as osteoclast precursors. We used the high concentration (100 ng/ml) of M-CSF to obtain more osteoclast precursors efficiently, as previously reported [40]. The harvested osteoclast precursors were seeded in various tissue culture dishes and plates at an initial density of 2.5×10 4 /cm 2 and cultured in α-MEM/10% FBS/M-CSF (20 ng/ml)/sRANKL (10 ng/ml) with or without various concentrations of sudachitin. The culture medium was exchanged with fresh medium every 2 days. TRAP-positive MNCs with more than 3 nuclei were considered osteoclastic cells and counted under a microscope. In addition, the enzymatic activity of TRAP in the conditioned medium was measured.

Quantitative real-time PCR
The total RNA from the calvarial bones and cultured cells was reverse-transcribed using a High-capacity RNA-to-cDNA kit (Life Technologies) to produce cDNA. Quantitative realtime PCR was performed using the TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA) or Quantitative Tech SYBR Green PCR Master Mix (QIAGEN) on a Gen-eAmp 5700 Sequence Detection System (Applied Biosystems). The TaqMan primers for the indicated genes were obtained from Applied Biosystems and are listed in S1 Table. The relative quantification of the target mRNA expression was calculated and normalized to the amount of 18S rRNA. The primer sequences used for the quantitative RT-PCR performed with the SYBR Green PCR Master Mix are also listed in S1 Table.

Western blotting analysis
After washing the cells with PBS, the cells were lysed in whole-cell lysis buffer [10 mM sodium phosphate (pH 7.5), 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, 1 mM p-(aminoethyl)benzenesulfonyl fluoride (p-ABSF), 10 μg/ml leupeptin, 10 μg/ml pepstatin, and 10 μg/ml aprotinin] at 4˚C, as previously described [37]. The whole-cell lysates were centrifuged at 12,000 × g for 10 min, and the supernatants were used for western blotting analysis. Samples of whole-cell lysates containing equal amounts of protein were subjected to SDS-PAGE, and the proteins that separated in the gel were subsequently electrotransferred onto PVDF membranes. After blocking with 5% skim milk, the membranes were incubated with the indicated antibodies, followed by a peroxidase-conjugated anti-mouse or anti-rabbit IgG antibody. The immunoreactive proteins were visualized, as previously described [37].

Measurement of viable cells
The number of viable cells in the cultures of the osteoclast linage cells was estimated using a Cell Counting Kit-8 (CCK-8, Dojindo Molecular Technologies, Inc., Kumamoto), according to the manufacturer's instructions. The osteoclast precursors were cultured with M-CSF and sRANKL for 24 h and 48 h. Then, 10 μl of CCK-8 solution was added to the cultures. After a 1 h incubation, the absorbance of the culture was measured at 450 nm.

Measurement of intracellular ROS
After culturing the osteoclast precursors in the presence of M-CSF and sRANKL with or without sudachitin for 24 h, we measured the concentration of intracellular ROS using an OxiSelect TM Intracellular ROS Assay Kit (Green Fluorescence, Cell Biolabs, Inc., San Diego, CA), according to the manufacturer's instructions. The assay employs the cell-permeable fluorogenic probe 2', 7'-Dichlorodihydrofluorescin diacetate (DCFH-DA). In brief, DCFH-DA is diffused into cells and is deacetylated by cellular esterases to non-fluorescent 2', 7'-Dichlorodihydrofluorescin (DCFH), which is rapidly oxidized to highly fluorescent 2', 7'-Dichlorodihydrofluorescein (DCF) by ROS [41]. After washing the cells twice with α-MEM, we added a DCFH-DA probe (100 μl)/α-MEM/10% FBS to the culture and incubated the cells for 30 min. Then, the probes were removed and washed with PBS. Subsequently, the cells were lysed in the cell lysis buffer. The concentration of DCF in the cell lysates was measured at an excitation of 480 nm and an emission of 530 nm using a standard solution of DCF.

Statistical analysis
The data are presented as the mean ± standard error of the mean (SEM). The mean group values were compared by unpaired Student's t-test, one-way ANOVA or two-way ANOVA; the significance of the observed differences was subsequently determined by post hoc testing using Tukey's method or Bonferroni's method. A P-value < 0.05 was considered significant.

Sudachitin suppressed LPS-induced inflammatory bone destruction
The anti-inflammatory actions of sudachitin have been demonstrated by many in vivo and in vitro studies. Therefore, we first examined the effects of sudachitin on inflammatory bone loss using an LPS-induced in vivo model of inflammatory bone destruction. Following a subperiosteal LPS injection into the calvarial bones in mice, severe bone destruction was observed on both sides of the sagittal suture using three-dimensional μCT ( Fig 1A). An injection of up to 50 μM sudachitin alone did not result in any of the irregular bone lesions observed following the LPS injection. However, an injection of sudachitin along with LPS markedly inhibited the bone loss induced by the LPS injection at a dose ranging from 10 and 50 μM (Fig 1A). Consistently, the LPS injection markedly increased the mRNA levels of osteoclast-differentiation related molecules, such as TRAP (acp5) and cathepsin K (ctsk), in the locally inflamed calvariae, and the simultaneous administration of sudachitin with LPS reduced these enhanced mRNA levels ( Fig 1B). Thus, sudachitin strongly suppresses inflammatory bone destruction.

Sudachitin inhibits osteoclastogenesis in a co-culture of osteoblasts and osteoclast precursors in the presence of pro-inflammatory factors
Osteoclast formation proceeds following an interaction between osteoclast lineage cells and osteoblasts [37,42]. Thus, we examined the effect of sudachitin on osteoclastogenesis in a coculture of isolated osteoblasts and osteoclast precursors derived from bone marrow cells. In the co-culture, pro-inflammatory factors, such as IL-1β and PGE 2, induced osteoclast formation (Fig 2A and 2B). However, the simultaneous addition of 10 to 50 μM sudachitin with IL-1β and PGE 2 inhibited osteoclast formation. The inhibitory effect of sudachitin was consistent with the effect observed following the LPS-induced inflammatory bone destruction in the calvariae.
To examine the mechanism underlying the inhibitory effect of sudachitin on osteoclast formation in the co-culture, we examined the mRNA expression levels of RANKL (rankl), which is a molecule that triggers osteoclast differentiation, and OPG (opg), which is an anti-osteoclast differentiation cytokine, in isolated osteoblasts (Fig 2C-2E). The RANKL mRNA expression level in the osteoblasts increased in response to the IL-1β and PGE 2 treatment, and the enhanced level was maintained in the presence of sudachitin; however, sudachitin increased the mRNA level compared with that in the osteoblasts treated with IL-1β and PGE 2 (Fig 2C). The opg mRNA expression level did not differ between the sudachitin-treated and untreated osteoblasts, while 50 μM sudachitin decreased the opg mRNA expression level (Fig 2D). In addition, sudachitin dose-dependently increased the mRNA ratio of rankl to opg (Fig 2E). Thus, the inhibitory effect of sudachitin on osteoclast formation in inflammatory bone destruction and the co-culture of osteoblasts and osteoclast precursors cannot be attributed to its action on osteoblasts.

Sudachitin directly acts on osteoclast precursors and inhibits osteoclastogenesis
The above-mentioned results suggest that sudachitin plays other roles in addition to its role in osteoblasts. Thus, we next examined the effects of sudachitin on in vitro osteoclast formation from osteoclast precursors derived from bone marrow cells. In our culture system for in vitro osteoclast formation, the generation of TRAP-positive mononuclear preosteoclasts begins on ) was subperiosteally injected into the calvariae of mice once a day for 5 days in vivo. On day 6 after the initiation of the injections, the calvariae were removed, and the three-dimensional structures were determined by a μCT analysis (A). The total RNA was extracted from the injected region in calvariae, and the mRNA levels of osteoclast differentiation-related genes, such as acp5 and ctsk, were measured by quantitative real-time RT-PCR (B). The presented values represent the mean ± SD (n = 3). Ã P < 0.05 vs. calvaria injected without LPS and sudachitin and ÃÃ P < 0.05 vs. calvaria injected only with LPS.
https://doi.org/10.1371/journal.pone.0191192.g001 day 1 after the addition of RANKL to the osteoclast precursor culture, and the preosteoclasts begin to fuse with each cell to generate TRAP-positive multinucleated osteoclasts between day 1.5 and day 3 (Fig 3A). Sudachitin at a concentration ranging from 2 and 10 μM time-and dose-dependently inhibited osteoclast formation and maturation (Fig 3A). In particular, at the 10 μM concentration, the TRAP-positive multinucleate cells barely formed in the culture. The presented values represent the mean ± SD (n = 4). Ã P < 0.05 vs. co-culture in the absence of IL-1β and PGE 2 . ÃÃ P < 0.05 vs. co-culture with IL-1β and PGE 2 in the absence of sudachitin. In addition, the isolated osteoblasts were treated with various concentrations of sudachitin in the absence or presence of IL-1β (10 ng/ml) and PGE 2 (10 μM) for 6 h. Then, the total RNA was extracted; the mRNA levels of RANKL (rankl, C) and OPG (opg, D) were measured by quantitative real-time RT-PCR, and the ratio of rankl mRNA/opg mRNA was calculated (E). The presented values represent the mean ± SD (n = 3). In C and E, Ã P < 0.05 vs. culture with IL-1β and PGE 2 in the absence of sudachitin; ÃÃ P < 0.05 vs. culture with IL-1β and PGE 2 alone. In D, ÃÃ P < 0.05 vs. culture without IL-1β, PGE 2 and sudachitin. NS indicates that the difference is not significant. Consistently, sudachitin decreased the enzymatic activity of TRAP in the conditioned medium. Thus, sudachitin directly inhibits osteoclast differentiation. However, the number of viable cells in the presence of up to 10 μM sudachitin did not change, while a higher dose of sudachitin (30 μM) decreased cell viability (Fig 3B). Therefore, inhibition of osteoclastogenesis by sudachitin at low doses is not caused by a decrease in the cell viability. The effective sudachitin doses that inhibited in vitro osteoclast formation were lower than those that prevented inflammatory bone destruction in vivo. This discrepancy might have resulted from the local diffusion of sudachitin in vivo.
We next examined the expression of osteoclast differentiation-related molecules and the signaling pathways involved in osteoclastogenesis. The mRNA expression levels of c-fos and NFATc1, which are crucial transcription factors for osteoclast differentiation, TRAP (acp5) and cathepsin K (ctsk), which are functional enzymes for bone resorption, and DC-STAMP, OC-STAMP and Atp6v0d2, which are integral proteins for the cell-cell fusion of preosteoclasts, markedly increased in response to the sRANKL/M-CSF treatment during osteoclast formation (Fig 4). The enhanced mRNA levels decreased in the presence of sudachitin (10 μM) at all timepoints of the culture (Fig 4).
Consistent with the inhibitory effects of sudachitin on the mRNA expression levels in the osteoclast lineage cells, sudachitin decreased the increased protein levels of these osteoclast differentiation-related molecules including c-Fos, NFATc1, cathepsin K, DC-STAMP and Atp6v0d2 during osteoclast formation stimulated by sRANKL and M-CSF (Fig 5).

Sudachitin suppresses the production of intracellular ROS and activation of MAPKs in osteoclast lineage cells
Osteoclast differentiation has been associated with an elevated production of intracellular ROS [23,43,44]. Furthermore, sudachitin has been shown to exert antioxidant effects in the body [34]. Therefore, to explore whether the inhibitory effect of sudachitin on osteoclastogenesis is related to the intracellular production of ROS, we determined the concentration of intracellular ROS in osteoclast lineage cells treated with sRANKL and/or sudachitin. As shown in Fig  6A, when osteoclast precursors were treated with sRANKL for 24 h, the intracellular ROS content was 2.5-fold higher than that in the cells treated with M-CSF alone. However, the simultaneous addition of sudachitin with sRANKL dose-dependently attenuated the elevated intracellular ROS content. The ROS content in the cells treated with sRANKL plus 10 μM sudachitin was equivalent to that in the cells treated with M-CSF alone. Thus, sudachitin inhibited the intracellular production of ROS during osteoclast formation.
Finally, we examined the effects of sudachitin on the activation of osteoclast differentiationsignaling pathways, such as NF-κB and MAPKs, including Erk and JNK. The activation of these pathways was evoked by the addition of sRANKL in the short term. However, pretreatment with sudachitin reduced the activation of Erk and JNK (Fig 6B). The activation of NF-κB stimulated by sRANKL is also slightly decreased by the sudachitin-pretreatment, whereas the reduction was less than those of Erk and JNK (Fig 6B).

Discussion
In this study, we demonstrated that sudachitin, which is a polymethoxy flavonoid, suppresses inflammatory bone destruction induced by an LPS injection in the calvariae. The inhibition by sudachitin could be attributed to the direct inhibition of osteoclastogenesis from osteoclast precursors because sudachitin reduced osteoclast formation from osteoclast precursors in response to RANKL in vitro and sudachitin increased the ratio of RANKL to Opg in the osteoblasts. Furthermore, sudachitin suppressed the intracellular ROS production and the activation of MAPKs, including Erk and JNK, both of which are involved in osteoclast differentiation. These inhibitory activities could be associated with a decrease in osteoclast formation. Thus, sudachitin could be a useful agent for the treatment of inflammatory bone destruction.
ROS are mainly classified into the following four types: hydroxyl radicals, nitric oxide, superoxide anions and hydrogen peroxide. Although the role of ROS in osteoclast formation has been reported since 1990, the ROS type that is important for osteoclastogenesis has not been determined. The ROS production in osteoclast lineage cells has been demonstrated to be mediated by NADHP oxidases, which consist of five isoforms, i.e., Nox1 to Nox5. Furthermore, many NOX organizers, NOX activators and small GTPases participate in regulating the enzymatic activities of the NADPH oxidases [45]. Although NOX1 -/-, NOX2 -/and NOX3 mutant mice do not demonstrate bone abnormalities [44,46,47], global Nox4-knockout mice display a higher trabecular bone density and reduced numbers and markers of osteoclasts in vivo [44]. In particular, ex vivo experiments using NOX4 -/osteoclast precursors showed a reduction in osteoclastogenesis that is consistent with the down-regulation of intracellular ROS production. In addition, the NOX inhibitors GKT137928 and GKT137831 both rescued the bone loss induced by ovariectomy [44]. These observations highlight the pivotal role of intracellular ROS catalyzed by NOX4 in osteoclast differentiation and function. The present study also indicates the importance of intracellular ROS in osteoclastogenesis because sudachitin, which has antioxidant properties, strongly inhibited inflammatory bone destruction and osteoclast formation, changes accompanied by a reduction in the intracellular ROS content.
Although many studies have demonstrated the effects of antioxidant compounds on osteoclast differentiation associated with the intracellular redox status, the effects have not always been consistent and remain controversial. Glutathione (GSH) is a representative antioxidant, and cellular redox status is defined by the balance between oxidants and antioxidants, in particular the GSH/oxidized GST (GSSG) ratio. Romagnoli et al demonstrated that a low GSH/ GSSG ratio downregulated OPG expression in human osteoclasts, resulting in an increase in the ratio of RANKL to OPG and indicating that GSH/GSSG redox coupling could affect osteoclastogenesis [48]. However, these authors did not directly determine the influence of the GSG/GSSG change on osteoclast formation. Le Nihouannen et al found that the addition of ascorbic acid to a culture of osteoclast precursors in the presence of RANKL decreased osteoclast formation, which was consistent with the decrease in the ratio of GSH/GSSG [49]. However, several studies have reported an inhibitory effect of ascorbic acid on osteoclastogenesis [50], indicating the difficulty of controlling the intracellular redox status.
The efficacy of natural food-derived antioxidants in osteoclastic bone resorption has also been extensively studied [51]. Recent epidemiological studies have shown that flavonoid consumption has a stronger association with bone integrity than general fruit and vegetable consumption. Epigallocatechin gallate and its polymerized theaflavin digallate, which are contained in green and black tea, exert suppressive effects on all processes of osteoclast differentiation, including osteoclast precursor generation from bone marrow cells, osteoclast The presented values represent the mean ± SD (n = 5). Ã P < 0.05 and NS (not significant) vs. culture with M-CSF alone; ÃÃ P < 0.05 vs. culture with sRANKL and M-CSF. In addition, osteoclast precursors were cultured in the absence or presence of sudachitin (10 μM) in α-MEM/ 10% FBS for 3 h. Then, the cells were treated with sRANKL (20 ng/ml) with or without sudachitin in α-MEM/10% FBS for the indicated times. Subsequently, whole-cell lysates were prepared and subjected to a western blotting analysis to measure phosphorylated and unphosphorylated Erk1/2, SAPK/JNK, and IκB-α levels. The values presented below the images indicate the relative ratios of the total band intensity of phosphorylated Erk1/2 (p44 and p42) and SAPK/JNK (p46 and p54) to the total band intensity of non-phosphorylated Erk and SAPK/JNK, respectively. In case of IκB-α, the values are the relative ratio of phosphorylated IκB-α non-phosphorylated IκB-α. In addition, total amounts of Erk1/2 (p44 and p42), SAPK/JNK (p46 and p54) and IκB-α were also compared with the level of β-actin.
https://doi.org/10.1371/journal.pone.0191192.g006 differentiation from osteoclast precursors and the maturation of multinucleated functional osteoclasts [52][53][54]. More recently, polymethoxy flavonoids contained in citrus fruits, such as nobiletin and tangeretin, have attracted attention due to their diverse physiological activities, such as the repression of carcinogenesis and cancer growth, anti-atherosclerosis activity, and the improvement of dyslipidemia and anti-inflammatory diseases. Although the sudachitin used in this study is also a polymethoxy flavonoid and has previously shown stronger physiological activities than nobiletin and tangeretin, its effects on bone metabolism were revealed for the first time in this study. Our findings show that the action of sudachitin in the suppression of osteoclast formation is more prominent than that of other polymethoxy flavonoids. In addition, this suppression was attributed to the direct inhibition of osteoclast differentiation from osteoclast precursors without acting on osteoblasts. Therefore, although the inhibition of osteoclast formation by nobiletin and tangeretin has been previously reported to be associated with the suppression of intracellular PGE 2 production in osteoblasts [31], our study indicates that the primary cells targeted by sudachitin are osteoclast lineage cells.
The elevation of intracellular ROS content causes the activation of MAPKs, such as Erk, p38 MAPK and JNK, in various cells [55][56][57]. Consistently, in this study, RANKL, which is a molecule that triggers osteoclast differentiation, induced the activation of Erk and JNK and increased ROS; simultaneously, sudachitin also inhibited the elevation of ROS and the activation of MAPKs. In contrast, sudachitin hardly affects the activation of NF-κB evoked by RANKL in osteoclast precursors, suggesting that the primary action target of sudachitin is MAPK rather than NF-κB.
Various types of antibodies targeting pro-inflammatory cytokines have currently been used for the treatments of inflammatory bone destruction such as RA, resulting in significant improvements in clinical scores [58][59][60]. However, in some cases of treatment with an anti-RANKL antibody, hypocalcemia developed in patients with severe renal dysfunction [61]. Moreover, treatments using these antibodies are extremely expensive. As previously mentioned, sudachitin has a variety of useful biological activities and is relatively inexpensive. In addition, as bone loss induced by not only inflammation but also ovariectomy [44] is associated with the elevation of intracellular ROS production, our findings suggest that use of sudachitin may lead to a possible therapeutic approach for various bone diseases including postmenopausal osteoporosis and inflammatory bone destruction.

Conclusions
Sudachitin, which is a polymethoxy flavonoid, blocked LPS-induced inflammatory bone destruction by directly inhibiting osteoclast differentiation from osteoclast precursors. Furthermore, sudachitin repressed the activation of Erk and JNK, which are pivotal signaling pathways for osteoclast differentiation associated with a decrease in intracellular ROS production. Therefore, sudachitin is a useful therapeutic agent for inflammatory bone resorption.
Supporting information S1 Table. Primers and probes used in quantitative RT-PCR. (DOCX)