Evodiamine inhibits RANKL‐induced osteoclastogenesis and prevents ovariectomy‐induced bone loss in mice

Abstract Postmenopausal osteoporosis (PMO) is a progressive bone disease characterized by the over‐production and activation of osteoclasts in elderly women. In our study, we investigated the anti‐osteoclastogenic effect of evodiamine (EVO) in vivo and in vitro, as well as the underlying mechanism. By using an in vitro bone marrow macrophage (BMM)‐derived osteoclast culture system, we found that EVO inhibited osteoclast formation, hydroxyapatite resorption and receptor activator of NF‐κB ligand (RANKL)‐induced osteoclast marker gene and protein expression. Mechanistically, we found that EVO inhibited the degradation and RANKL‐induced transcriptional activity of IκBα. RANKL‐induced Ca2+ oscillations were also abrogated by EVO. In vivo, an ovariectomized (OVX) mouse model was established to mimic PMO, and OVX mice received oral administration of either EVO (10 mg/kg) or saline every other day. We found that EVO can attenuate bone loss in OVX mice by inhibiting osteoclastogenesis. Taken together, our findings suggest that EVO suppresses RANKL‐induced osteoclastogenesis through NF‐κB and calcium signalling pathways and has potential value as a therapeutic agent for PMO.

osteoclast differentiation and function could be a promising therapeutic strategy for attenuating the progression of PMO.
Osteoclasts are multinucleated, specialized bone-resorbing cells that are derived from the monocyte/macrophage lineage. The formation of osteoclasts, also called osteoclastogenesis, is a multi-stage process regulated by a number of genetic, humoural and mechanical factors. Among these factors, macrophage colony-stimulating factor (M-CSF) and receptor activator of nuclear factor-kappa B (NF-κB) ligand (RANKL) are well-known as the key cytokines in osteoclastogenesis. 5,6 The cytokine M-CSF is a prerequisite for osteoclast precursor proliferation and survival, whereas RANKL, a tumour necrosis factor (TNF) family cytokine, controls the function and survival of mature osteoclasts through interaction with its receptor RANK. 1 Following the binding of RANKL to the RANK receptor, multiple intracellular signalling events are activated, including NF-κB and calcium signalling. [7][8][9] Ultimately, these signal transduction pathways lead to the expression and activation of transcription factors such as nuclear factor of activated T cells-c1 (NFATc1) and activator protein-1 (AP-1), both of which are crucial for the differentiation of osteoclast precursors. [9][10][11] Thus, blocking the intracellular signalling stimulated by RANKL is considered a major therapeutic target for the treatment of PMO.
Evodiamine (EVO) is an alkaloidal compound extracted and purified from the unripe fruit of Evodia rutaecarpa, a type of traditional Chinese medicine (TCM) with a long history. 12 EVO has been reported to possess multiple pharmacological activities, including antimicrobial, 13 anti-tumour 14 and anti-inflammatory effects. 15,16 A previous study reported that EVO significantly suppresses zymosaninduced inflammation by inhibiting the NF-κB signalling pathway in the murine RAW264.7 macrophage cell line. 16 EVO also exerts neuroprotective effects via down-regulated NF-κB expression to protect against permanent middle cerebral artery occlusion-induced brain injury in mice. 17 However, the effect of EVO on osteoclastogenesis remains unknown. Therefore, in our study, we investigated the inhibitory effects of EVO on osteoclast differentiation and function, as well as the underlying mechanism of EVO on RANKL-treated osteoclasts. Furthermore, an ovariectomy (OVX)-induced bone loss mouse model was established to mimic PMO, and the protective role of EVO against bone loss in PMO was examined.

| Reagents and antibodies
EVO (purity >98%) was purchased from Nantong Feiyu Biological Technology Co, Ltd. (Nantong, China). EVO was dissolved in DMSO as a 20-mmol/L stock solution and stored at −20°C. Further dilution was performed in cell culture medium. Primary antibodies against NFATc1, c-Fos, integrin-β3, CTSK and β-actin were obtained from Santa Cruz Biotechnology (San Jose, CA). Primary antibodies against IκBα, p65 and phospho-p65 were obtained from Cell Signaling Technologies (Beverly, MA, USA). A V-ATPase d2 antibody was generated as previously described. 18 An MTS assay kit and a luciferase assay system were purchased from Promega (Madison, WI, USA). A leukocyte acid phosphatase staining kit was purchased from Sigma-Aldrich (Sydney, Australia). Recombinant macrophage colony-stimulating factor (M-CSF) was obtained from R&D Systems (Minneapolis, MN).
Recombinant GST-rRANKL protein was synthesized and purified as previously described. 19 The cell culture medium, alpha-modified minimal essential medium (α-MEM) and foetal bovine serum (FBS) were purchased from Thermo Fisher Scientific (Scoresby, Vic., Australia).

| Cell culture
Bone marrow macrophages (BMMs) were isolated from the bone marrow of 6-week-old C57BL/6 mice, which were euthanized according to the procedures approved by the Animal Ethics Committee of the University of Western Australia (RA/3/100/1244). The extracted cells were collected by centrifugation, cultured in a 75 cm 2 culture flask with α-MEM (adding 10% FBS, a 1% antibiotic mixture of penicillin and streptomycin and 50 ng/mL M-CSF) and incubated in an atmosphere of 5% CO 2 at 37°C. The culture medium was changed every 2 days, and cells were passaged when 90%-100% confluence were attained. BMMs from passages one to three were used in this study.

| Cell viability assay
Cell viability was determined using an MTS (3-(4,5-dimethylthiazol-2yl)-2,5-diphenyl tetrazolium bromide) assay following the manufacturer's protocol. Passage 1-3 BMMs were seeded into 96-well plates (5 × 10 3 cell/well) and incubated with 50 ng/mL of M-CSF at 37°C for 24 hours. Then the cells were treated with various concentrations of EVO (1, 2.5, 5, 10 and 20 μmol/L). After 48 hours of treatment, the cells were washed with PBS, 100 μL of non-FBS medium (α-MEM) containing 20 μL of MTS solution was added to each well, and the plate was incubated for an additional 2 hours. The absorbance of the wells was then measured at 490 nm by a micro-plate reader (ThermoFisher, Waltham, MA, USA).

| In vitro osteoclastogenesis assay
BMMs were plated at a density of 5 × Furthermore, the inhibitory effect of EVO (10 μmol/L) at different stages of osteoclast differentiation was also investigated. We examined the effects of EVO on RANKL-induced osteoclast differentiation by treating at different time points from day 1 to day 5 after RANKL stimulation as follows: early stage: EVO was added on day 1 and removed on day 3; middle stage: EVO was added on day 3 and removed on day 5; late stage: EVO was added on day 5 and removed on day 6; and whole stage: EVO was added with RANKL as normal. Finally, cells were fixed and stained for TRAcP as described above.

| Actin ring formation assay
BMMs were seeded into 96-well plates and treated with different concentrations of EVO in the presence of 50 ng/mL M-CSF and 50 ng/mL RANKL, as described above. After 5 days, paraformaldehyde (4%) was used to fix cells for 15 minutes at room temperature.
After being washed with PBS three times, cells were permeabilized with 0.25% Triton X-100 and then blocked with 3% BSA in PBS.
Next, the F-actin rings were stained with rhodamine-conjugated phalloidin (Eugene, OR, USA), and the cell nuclei were stained with DAPI. Images were acquired using confocal laser scanning microscopy (Nikon, Tokyo, Japan). The number of multinucleated cells (>3 nuclei) and the number of nuclei were calculated.

| Resorption pit assay
A resorption pit assay was used to evaluate osteoclast function.
BMMs were seeded at a density of 8 × 10 4 cell/well into 6-well collagen-coated plates and stimulated with M-CSF (50 ng/mL) and RANKL (50 ng/mL) until mature osteoclasts formed. Cells were gently detached from the wells using a cell dissociation solution (Sigma, St. Louis, MO, USA) and then plated into hydroxyapatite-coated 96-well plates (Corning, New York, NY, USA) and bone slices in equal numbers.
The mature osteoclasts were treated with different concentrations of EVO in the presence of M-CSF (50 ng/mL) and RANKL (50 ng/mL).
After 48 hours, half of the hydroxyapatite-coated wells were bleached for 10 minutes to remove the cells and dried for hydroxyapatite resorption visualization using a light microscope, while the remaining hydroxyapatite-coated wells were fixed and stained for TRAcP activity as described above to assess the number of multinucleated cells. Additionally, bone slices were stained with haematoxylin to detect resorption pits. Image J software (NIH, Bethesda, MD, USA) was used to analyse the percentage of hydroxyapatite resorption areas.

| Luciferase reporter gene assays
The transcriptional activities of NF-κB and NFATc1 were measured by luciferase reporter gene assays. RAW264.7 cells were stably transfected with either an NF-κB-responsive luciferase construct or an NFATc1-responsive luciferase reporter construct. 20,21 The cells were seeded in 48-well plates (1.5 × 10 5 cells/well for NF-κB and 5 × 10 4 cells/well for NFATc1). The next day, the cells were pre-treated with or without various concentrations of EVO for 1 hours, and then stimulated by 100 ng/mL RANKL (6 hours for NF-κB and 24 hours for NFATc1). Analysis of luciferase activity was performed following the manufacturer's instructions (Promega, Sydney, Australia).

| Intracellular Ca 2+ measurement
To determine whether EVO can inhibit calcium signalling in the progression of osteoclast differentiation, we performed a measurement of intracellular Ca 2+ oscillation as previously described. Briefly, BMMs  Table 1.

| Western blotting
Western blotting was performed using routine protocols. Treated BMMs were isolated using radioimmunoprecipitation assay (RIPA) lysis buffer (Millipore, Billerica, MA, USA) on ice. Equal amounts of proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes (GE Healthcare, Silverwater, Australia).
Membranes were blocked with 5% skim milk at room temperature for 2 hours and then incubated overnight at 4°C with one of the following primary antibodies: NFATc1, c-Fos, integrin-β3, CTSK, V-ATPase d2, β-actin, IκBα, p-p65 and p65. Subsequently, the membranes were incubated with the corresponding secondary antibodies for 2 hours at room temperature. Protein bands were then visualized using an enhanced chemiluminescence (ECL) system (Amersham Pharmacia Biotech, Sydney, Australia).

| Animal experiments
Seven-week-old female C57BL/6 mice were purchased from the Animal Center of the Chinese Academy of Sciences, Shanghai, China, and randomly divided into three groups (sham group, OVX group and OVX + EVO group, n = 10). All the mice were anaesthetized intraperitoneally with 2% (w/v) pentobarbital (40 mg/kg) and monitored by an assistant during the surgery. The bilateral OVX procedures were performed in the OVX group and OVX + EVO group using a dorsal approach. After the surgery, the animals were allowed to recover from surgery for 1 week prior to the experiments. The OVX + EVO group received EVO (10 mg/kg) dissolved in CMC by intragastric administration every other day for eight consecutive weeks. Mice in the sham group and OVX group were administered an equivalent volume of CMC. All animals were sacrificed 9 weeks after surgery, and femur bone samples were collected for micro-CT scanning and histological examination.

| Statistical analysis
The experiments were performed at least three times. The data obtained are expressed as the mean ± standard error of the mean (SEM). Statistical analysis was performed using GraphPad Prism ver-

| Effects of EVO on BMM viability
The chemical structure of EVO is shown in Figure 1A. The cytotoxic effect of EVO (1, 2.5, 5, 10 and 20 μmol/L) on BMMs was assessed after 48 hours treatment using an MTS assay kit ( Figure 1B). No significant cytotoxicity of EVO was observed at the concentrations used. Indeed, the results also indicated that EVO had no inhibitory effect on the M-CSF-induced proliferation of BMMs.

| EVO inhibits RANKL-induced osteoclastogenesis in vitro
To assess the effect of EVO on RANKL-induced osteoclastogenesis, with EVO at different time phases. We found that EVO strongly inhibited RANKL-induced osteoclast differentiation during the early and middle stages ( Figure 1D, G). These results suggest that EVO acts mainly during the early and middle stages of osteoclast differentiation.   (Figure 4). induced by RANKL. In addition, the transcriptional activity of NF-κB was also assessed by luciferase reporter gene assays. NF-κB transcriptional activity was markedly promoted by RANKL stimulation but decreased by EVO treatment ( Figure 6A). Furthermore, we found that EVO significantly suppressed RANKL-induced Ca 2+ oscillation, indicating that calcium signalling was also involved in the inhibitory effect of EVO ( Figure 6E-H).

| EVO did not affect the differentiation and mineralization of osteoblasts in vitro
Bone remodelling and homeostasis are maintained through a balance between bone resorption by osteoclasts and bone formation by osteoblasts. 1 To determine the effect of EVO on the differentiation of osteoblasts, an ALP assay and alizarin red staining assay were performed. The results showed that EVO had no significant effect on osteoblast differentiation or mineralization compared with the control group ( Figure S1A, B). Additionally, the toxicity of EVO on osteoblasts was evaluated. The results of the MTS assay revealed that the viability of osteoblasts was not affected by EVO at concentrations of 20 μmol/L and lower ( Figure S1C). Taken together, these data indicated that EVO had no effect on the differentiation and mineralization of osteoblasts.

| Effect of EVO on bone loss in OVX mice
To study the potential therapeutic benefits of EVO on bone loss in vivo, an OVX mouse model was established to mimic PMO.
Micro-CT scanning and 3D reconstruction were performed to assess changes in the bone micro-architecture in OVX mice, which revealed a significantly decreased trabecular bone mass in OVX mice compared to that in sham-operated mice. However, compared to vehicle treatment, oral administration of EVO strongly attenuated the bone loss following ovariectomy, with significant increases in BV/TV and Tb.N and a decrease in Tb.Sp. Tb.Th was uniform across all groups in our study ( Figure 7A, B). Meanwhile, these results were further supported by histological assessment of decalcified distal femoral sections stained for H&E and TRAcP. In H&E images, the value of BV/TV in the OVX + EVO group was significantly enhanced compared to that in the OVX group. In addition, the increased values of Oc.S/BS and N.Oc/BS in the femoral metaphysis induced by OVX were also markedly reduced by EVO treatment (Figure 7C, D). Thus, these data indicate that EVO can attenuate bone loss and microarchitecture deterioration in the OVX mouse model.

| DISCUSSION
Oestrogen deficiency in the elderly women is known to increase bone turnover, leading to increased formation and activation of osteoclasts. 4   Data are presented as the mean ± SEM, *P < 0.05, **P < 0.01 relative to the sham group. # P < 0.05, ## P < 0.01 relative to the OVX group. n = 6 critical cytokines for osteoclastogenesis, it is well-known that M-CSF induces the proliferation of early macrophage or osteoclast precursors, whereas RANKL acting via its receptor RANK induces the subsequent differentiation of osteoclast precursors into mature osteoclasts. 5,6 In this study, we found that the M-CSF-induced prolif- in osteoclast differentiation up-regulated by the pathways mentioned above. c-Fos, a member of the activator protein-1 (AP-1) transcription factor family, is involved in the differentiation of precursor cells into bone-resorbing osteoclasts. c-Fos-deficient mice exhibit osteopetrosis due to the blocking of osteoclast formation. 28 NFATc1, an NFAT family member, has been determined to be a master executor of RANKLmediated osteoclast differentiation and regulates the expression of osteoclastic-associated genes, such as TRAcP (Acp5), MMP9, CTSK and V-ATPase d2, through cooperation with MITF and c-Fos. 10 In osteoclast precursors, the AP-1 complex containing c-Fos can trigger the auto-amplification of NFATc1 to accelerate transcriptional processes. 29 NFATc1-deficient mice show the defects of impaired osteoclastogenesis, which result in the symptoms of osteopetrosis. 30 In our study, we found that the RANKL-induced up-regulation of c-Fos and NFATc1 are inhibited by EVO at the protein level. The mRNA levels of TRAcP (Acp5), c-Fos, MMP9, CTSK are reduced by EVO in a dosedependent manner. Taken together, these results suggest that EVO has inhibitory effects on osteoclast differentiation and function, which significantly suppress RANKL-induced osteoclastogenic marker expression and that the NF-κB and calcium signalling pathways are involved in the anti-osteoclastogenic effects of EVO.
Unlike osteoclasts, osteoblasts are responsible for bone formation and also have a main role in the mineralization of bone structures. 31 Therefore, the effect of EVO on osteoblasts was also analysed in our study. However, no significant difference was found between the EVO and control groups. In addition, the result of the MTS assay showed that EVO was not toxic to osteoblasts at concentrations of 20 μmol/L and lower. Collectively, EVO had no effect on the differentiation and mineralization of osteoblasts.
To further investigate the effects of EVO in vivo for PMO, we established an OVX-induced osteoporosis mouse model. The data from the micro-CT analysis and H&E staining showed that the oral administration of EVO significantly suppressed bone loss in OVX mice. The number of activated osteoclasts around the trabecula in the evodiamine-treated group was significantly lower than that in the OVX group. These results and those of the in vitro experiments further suggest that EVO has potential value in preventing the progression of PMO.
In conclusion, based on the in vivo and in vitro results in our study, EVO is suggested to be a safe and effective agent for treating PMO. The protective effect of EVO against PMO was accomplished by inhibiting the RANKL-induced differentiation and function of osteoclasts and inhibiting the activation of the NF-κB and calcium signalling pathways may be the underlying mechanism.