Acteoside Suppresses RANKL-Mediated Osteoclastogenesis by Inhibiting c-Fos Induction and NF-κB Pathway and Attenuating ROS Production

Animal care and use practices were approved by the Chonbuk National University Committee on Ethics in the Care and Use of Laboratory Animals (Permit No. CBU 2010-0007). All experiments in this study were carried out according to the guidelines of the Animal Care and Use Committee of the University.

Four-week old female ICR mice were purchased from Orient Bio Inc. (Seoul, Korea) and housed at 22±1°C and 55±5% humidity on a 12 h light/dark cycle with free access to food and water. Acteoside (3,4-dihydroxy-β-phenethyl-O-α-rhamnopyranosyl-(1→3)-4-O-caffeoyl-β-D-glucopyranoside; C₂₉H₃₆O₁₅) (Fig. S1) was isolated from the leaves of Rehmannia glutinosa. Acteoside was dissolved in phosphate-buffered saline (PBS) before use. RANKL, TNF-α, IL-1β, IL-6, and M-CSF were purchased from R & D Systems (Minneapolis, MN, USA). Antibodies specific to c-Fos, p65, p-p65, p-ERK, ERK, JNK, p-JNK, NFATc1, IκBα, p-IκBα, and β-actin were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies for p-p38 and p38 were purchased from Cell Signaling Technology (Denvers, MA, USA). Calcium Assay and Osteocalcin (OC) EIA Kits were purchased from BioAssay Systems (Hayward, CA, USA) and Biomedical Technologies (Stoughton, MA, USA), respectively, to determine serum biochemical parameters. Mouse tartrate-resistant acid phosphatase (TRAP) Assay kit (Immunodiagnostic Systems, Scottsdale, AZ, USA) was also used to measure serum TRAP5b level. Unless otherwise specified, additional chemicals were obtained from Sigma Chemical Co. (St. Louis, MO, USA), and laboratory wares were from SPL Life Sciences (Pochun, South Korea).

Bone marrow cells were obtained from the tibiae and femora of 6 week-old female ICR mice according to methods described previously [25]. The bone marrow suspension was incubated in a 100-mm culture dish in the presence of 50 ng/ml M-CSF. After 3 days, adherent cells were used as bone marrow macrophages (BMMs) to induce osteoclastic differentiation. Some bone marrow cells were also incubated for 48 h without M-CSF, and adherent cells were cultured in osteoblast differentiating medium, as described elsewhere [25]. RAW264.7 macrophage cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, and antibiotics. These cells were used as a counterpart cell line for BMMs to explore the effect of acteoside on osteoclastogenesis.

BMMs were pretreated with various concentrations (0–50 µM) of acteoside for 2 h before stimulating with 100 ng/ml RANKL. Culture media was replaced with fresh media on days 2 and 5. After 7 days of incubation, the cultures were fixed in 4% PBS-buffered para-formaldehyde and stained with TRAP using a Sigma Aldrich kit according to the manufacturer's instructions. TRAP-positive cells were counted using optic microscopy, and cells containing 3 or more nuclei were considered to be osteoclasts. RAW 264.7 cells were also exposed to 100 ng/ml RANKL after pretreatment with acteoside for 2 h, and after 7 days of incubation, the cells were processed for TRAP staining.

Cell viability was determined using water-soluble tetrazolium salt (WST)-8 reagent. In brief, BMMs or RAW264.7 cells cultured in a growth medium containing 10% FBS and antibiotics were treated with 10 µM acteoside or a phenolic compound such as quercetin, luteolin, apigenin, or epigallocatechin-3-gallate (EGCG). WST-8 reagent was added into the cultures after 48 h of incubation. After incubating for an additional 4 h, the WST-8-specific absorbance was measured at 450 nm using a microplate reader (Packard Instrument Co., Downers Grove, IL, USA).

BMMs (1×10⁵ cells/ml) were suspended in α-MEM containing 50 ng/ml M-CSF and 100 ng/ml RANKL, then divided across a 24-well plate coated with calcium-phosphate nano crystals (OAAS-24; Osteoclast Activity Assay Substrate, Oscotec Inc., Choongnam, South Korea) at a density of 2×10⁴ cells/cm² with and without acteoside. After incubating for 7 days, the cells were removed from the plates with 5% sodium hypochlorite, and pit formation was observed under an optic microscope. The resorbed area was also measured by image analyzer and expressed as percentage of the control value.

Whole protein lysates were prepared in a lysis buffer as described elsewhere [26]. Cytosolic and nuclear proteins were prepared as described previously [25]. Equal amounts of protein extract were separated by 12–15% SDS-PAGE and blotted onto poly vinyl difluoride membranes. The blots were probed with primary antibodies overnight at 4°C before incubation with secondary antibody in blocking buffer for 1 h. The blots were developed with enhanced chemiluminescence (Amersham Pharmacia Biotech Inc., Buckinghamshire, UK) and exposed on X-ray film (Eastman-Kodak Co., Rochester, NY, USA).

Cells were pretreated with acteoside for 2 h and then stimulated with RANKL for an additional-30 min. MAPK activities were determined using immunometric assay kits, such as the p-p38 kinase assay kit (Assay Designs, Inc., MI, USA), p-ERK enzyme assay kit (Assay Designs), and p-SAPK/JNK sandwich ELISA kit (Cell Signaling Technology, MA, USA). All procedures followed the manufacturer's instructions, and absorbance was measured by a microplate reader.

DNA-protein binding reactions were performed for 30 min at room temperature, with 10–15 µg protein in 20 µl buffer containing 1 µg/ml BSA, 0.5 µg/µl poly (dI-dC), 5% glycerol, 1 mM DTT, 1 mM PMSF, 10 mM Tris-Cl (pH 7.5), 50 mM NaCl, 30,000 cpm of [α-³²P] dCTP-labeled oligonucleotides, and the Klenow fragment of DNA polymerase. The samples were separated on 6% polyacrylamide gels, which were dried and exposed to X-ray film (Eastman Kodak Co.) for 12–24 h at −70°C. The oligonucleotide primer sequences specific for NF-κB were 5′-AAG GCC TGT GCT CCG GGA CTT TCC CTG GCC TGG A-3′ and 3′-GGA CAC GAG GCC CTG AAA GGG ACC GGA CCT GGA A-5′.

RAW264.7 macrophages in 24-well plates were transfected with 0.8 µg κB-luciferase reporter vector using 2 µl of Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. At 24 h after transfection, the cells were stimulated with RANKL in the presence and absence of acteoside for 24 h. Cells were resuspended in 100 µl reporter lysis buffer (Promega, Madison, WI, USA). Equal amounts of protein samples were placed into 96-well microplates and mixed with luciferase substrate. Luminescence was measured by using a microplate luminometer (MicroLumat Plus LB 964, Berthold Technologies, Bad Wildbad, Germany). In this experiment, a permeable NF-κB inhibitor peptide (BIOMOL, Butler Pike, PA, USA) was used as positive κB inhibitor.

BMMs or RAW264.7 cells were stimulated with RANKL in the presence of acteoside in 24-well culture plates. After 48 h of incubation, culture supernatants were collected and assessed by ELISA using TNF-α-, IL-1β- and IL-6-specific OptEIA™ kits according to the manufacturer's instructions.

The mRNA expression of osteoclastic markers, such as c-Fos, NFATc1, and TNF-α, was determined by real-time RT-PCR. In brief, total RNA was extracted from macrophages with Trizol reagent according to the manufacturer's instructions (Invitrogen). cDNA was synthesized with 1 µg of total RNA using SuperScript Reverse Transcripatase II and primers (Invitrogen). Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA) was used to detect the accumulation of PCR product during cycling with the ABI 7500 sequence detection system (Applied Biosystems). After denaturation at 95°C for 10 min, PCR was performed using forty 3-step cycles of denaturation at 95°C for 15 sec, annealing at 60°C for 1 sec, and extension at 72°C for 30 sec. All PCR reactions were performed at least in triplicate, and the expression levels were normalized to the housekeeping gene HPRT in the same reaction. This study used the same primer sequences described elsewhere [7].

A stock solution of 2′,7′-dichlorodihydrofluorescein-diacetate (DCFH-DA) (50 mM; Calbiochem, Darmstadt, Germany) was prepared in DMSO and stored at −20°C in the dark. In brief, BMMs (10⁶ cells/ml in 6-well plates) were cultured with 50 ng/ml M-CSF for 24 h and then treated with various concentrations (0–10 µM) of acteoside 2 h before stimulation with 100 ng/ml RANKL. After 1 h of co-incubation, these cells were subsequently incubated with 25 µM DCFH-DA for 30 min. The green fluorescence of 2′,7′-dichlorofluorescein (DCF) was recorded at 515 nm (FL 1) using a FACS Vantage® system (Becton-Dickinson, San Jose, CA, USA), and 10,000 events were counted per sample.

Female ICR mice (6 week-old) were used for this study. Mice received a sham operation (Sham, n = 10) or surgical ovariectomized (OVX, n = 20) under anesthesia. One week after surgery, the OVX mice were randomly divided into 2 groups of 10 mice each: bilateral OVX and bilateral OVX supplemented with 200 µl PBS containing 1 mM acteoside orally (AC group). Oral acteoside was administrated once every 3 days for 8 weeks after surgery, and the same amount of PBS was administered to the Sham and OVX groups. After 1 day of the last administration, mice were scarified and then biochemical parameters in serum and 3-dimensional bone structure were analyzed.

Blood samples were collected via cardiac puncture and serum was collected by centrifugation. Serum samples were stored at −80°C for analyses of biochemical parameters. The serum levels of IL-1β and IL-6 were estimated by using ELISA kit as described above, while alkaline phosphatase (ALP) activity was determined by using a biochemical colorimetric assay that measures the amount of p-nitrophenol produced from a p-nitrophenol phosphate substrate, as described elsewhere [27]. To estimate the biomarkers of bone formation and resorption, serum OC, calcium, and TRAP5b levels were also determined according to the manufacturers' instructions.

The femora of mice (Sham, OVX, and AC groups) were dissected and filled with physiological saline for mechanical testing. The mechanical strength of the femur was measured as described elsewhere [28]. The fracture load was recorded as the peak force in newtons at the point that the mid shaft of the right femur fractured. In addition, the light femur of each animal was histomorphometrically analyzed using a microcomputer tomography (micro-CT) system (SkyScan 1076 microfocus X-ray system, Kontich, Belgium). In brief, the bones in 4% formaldehyde storage were dried superficially on paper tissue before being wrapped in plastic “cling-film” or in parafilm, to prevent drying during scanning. Each plastic-wrapped bones were placed in plastic/polystyrene foam tubes which were mounted vertically in the horizontally in the 1076 scanner sample chamber for micro-CT imaging. Scanning was carried out using 100 kV source voltage and 140 µA source current with 35 µm resolution. Three-dimensional models of the trabecular bones of femur were reconstructed using SkyScan CT Analyzer version 1.11. The structural parameters such as trabecular bone mineral density (BMD, g/cm³), percent bone volume (bone volume (BV)/tissue volume (TV), %), thickness (Tb.Th, µm), separation (Tb.Sp, mm), and number (Tb.N, 1/mm) were then measured.

Bone marrow cells cultured in 6-well culture plates were treated with DAG (10 nM dexamethasone, 50 µM ascorbic acid, and 20 mM β-glycerophosphate) in the presence of 10 µM acteoside. After 2 weeks of differentiation, the cells were fixed with ice-cold 70% (vol/vol) ethanol for 1 h and stained with 0.2% alizarin red S in distilled water for 30 min at room temperature. After the cells were destained and air-dried, the cell culture plates were evaluated by light microscopy using an inverted microscope (Nikon TS100, Japan). To quantify the amount of red dye, the stain was eluted with 10% acetylpyridinum chloride by shaking for 20 min and the absorbance was measured at 560 nm. The amount of calcium deposited in the cell layers was also measured using a Calcium C kit (Wako Chemical Inc. Osaka, Japan) according to the manufacturer's instructions. In addition, the expression of bone-specific mRNA markers, such as runt-related transcription factor-2 (Runx2), osterix, bone sialoprotein (BSP), and OC was determined by real-time RT-PCR. Oligonucleotide primers of these markers were designed with product sizes less than 200 bp using Primer Express Software 3.0 (Applied Biosystems) as described elsewhere [27].

Unless otherwise indicated, all data are expressed as the mean ± standard deviation (S.D.) of 3 or more independent experiments. A one-way ANOVA was used for multiple comparisons using SPSS version 12.0 software. A p value <0.05 was considered statistically significant.

To verify the effect of acteoside on BMM differentiation into osteoclasts, the cells were cultured with various concentrations (0–20 µM) of acteoside for 7 days in the presence of 50 ng/ml M-CSF and 100 ng/ml RANKL. Acteoside reduced the number of osteoclasts in a dose-dependent manner. When the cells were pretreated with 10 µM acteoside for 2 h, the osteoclast number decreased by 43% compared to cells supplemented with M-CSF and RANKL (Fig. 1A). Fig. 1B shows the RANKL-mediated osteoclast differentiation and its inhibition by combined treatment with acteoside. Consistent with this result, acteoside pretreatment decreased the RANKL-stimulated differentiation of RAW264.7 cells and osteoclast formation (Figs. 1C and D). When the anti-osteoclastic potential of acteoside on BMMs was compared with the anti-osteoclast potentials of several phenolic compounds at the same concentration (10 µM), luteolin showed the highest activity (Fig. 2A). However, luteolin, quercetin, or apigenin itself decreased viability of the cells (Fig. 2B). All the compounds inhibited osteoclastic differentiation by RAW264.7 macrophages with the following relative activities: luteolin > quercetin = apigenin > EGCG = acteoside (Fig. 2C). Luteolin, quercetin, or apigenin itself also showed the decreased viability of the cells (Fig. 2D). In contrast, quercetin treatment only caused a significant reduction of viability in both BMMs and RAW264.7 cells, when these cells were exposed to 10 µM of each compound for 2 days with 50 ng/ml M-CSF, 100 ng/ml RANKL or both (data not shown). When the concentration of these compounds required to inhibit 50% of osteoclast formation in BMMs (IC₅₀) was calculated using concentration-activity curves, the IC₅₀ of acteoside, quercetin, luteolin, apigenin, and EGCG was 5.1, 2.3, 2.6, 4.8 and 6.6 µM, respectively (Fig. 2E). This result was similar to the case that RAW264.7 macrophages were examined. These data suggested that luteolin and quercetin had anti-clastogenic activities higher than acteoside. In contrast, quercetin at the IC₅₀ also had a mild toxic effect on the cells (data not shown).

Acteoside also prevented RANKL-mediated bone resorption in a dose-dependent manner, as measured by an in vitro model system (Fig. 3A). Bone resorption was significantly inhibited when BMMs were incubated with 1 µM acteoside (Fig. 3B). A 10 µM acteoside treatment almost completely attenuated RANKL-induced pit formation by BMMs. Similarly, acteoside decreased bone resorption in RANKL-stimulated RAW264.7 cells (Figs. S2A and B). The ability of acteoside to inhibit bone resorption depended on the timing of the treatment relative to RANKL stimulation. Acteoside (10 µM) added 4 days after RANKL stimulation did not reduce pit formation in BMMs, whereas it suppressed the number of osteoclasts formed (Fig. 3C). This different result was in part due to the pit area already formed after 4 days of RANKL stimulation formation (Fig. 3D).

RANKL induces the activation of 3 well-known MAPKs and NF-κB in osteoclast precursors, and this activation is required for early osteoclast differentiation. To understand the possible mechanisms by which acteoside inhibits osteoclastogenesis, we investigated the effect of acteoside on MAPKs and NF-κB activation in macrophages. BMMs and RAW264.7 cells were pretreated with 10 µM acteoside for 2 h and then stimulated with 100 ng/ml RANKL for 30 min. MAPK phosphorylation was examined by Western blotting and immunometric analysis. RANKL induced phosphorylation of p38, ERK, and JNK in BMMs (Fig. 4A) and RAW264.7 cells (Fig. 4B). Acteoside prevented these RANKL-induced increases in p-p38, p-ERK, and p-JNK. This result was supported by immunometric analysis, which that pretreatment with 10 µM acteoside significantly inhibited the levels of phosphorylated MAPKs in these macrophages (Fig. 4C). RANKL treatment increased the DNA-binding of NF-κB, whereas acteoside inhibited RANKL-induced activation of NF-κB-DNA binding (Fig. 5A). This inhibition was more prominent in BMMs than in RAW264.7 cells. Acteoside also diminished RANKL-stimulated p65 and IκBα phosphorylation in BMMs and RAW264.7 cells (Figs. 5B and C). Adding 10 µM acteoside almost completely inhibited both the degradation and activation of IκBα in BMMs (Fig. 5B). To further confirm that NF-κB activation is involved in the action of acteoside, κB promoter-luciferase constructs were transiently transfected into RAW264.7 cells. The cells incubated with 100 ng/ml RANKL had 3-fold higher κB promoter activity, which was significantly attenuated by 10 µM acteoside (Fig. 5D).

TNF-α, IL-1β, and IL-6 are important in osteoclast formation and function, which is mediated by NF-κB signaling in RANKL-stimulated macrophages. RANKL stimulated the production of these cytokines, and this production was markedly reduced by 10 µM acteoside pretreatment in BMMs (Fig. 6A). Similarly, acteoside attenuated the RANKL-induced production of cytokines, except IL-6, in RAW264.7 macrophages (Fig. S3). To understand the molecular mechanisms of acteoside action in osteoclastogenesis, we further examined the effect of acteoside on TNF-α, c-Fos, and NFATc1 expression. RANKL up-regulated the mRNA expression of these factors in BMMs and RAW264.7 cells (Figs. 6B and C). Pretreatment with 10 µM acteoside significantly inhibited the RANKL-induced expression of these factors in both BMMs and RAW264.7 cells. Acteoside pretreatment also strongly reduced the protein levels of c-Fos and NFATc1 in RANKL-stimulated BMMs (Fig. 6D). These results suggest that acteoside down-regulates RANKL inducing mediators of osteoclast formation at the gene and protein levels.

Since it is known that intracellular ROS production is correlated with RANKL-stimulated osteoclastogenesis, we investigated whether acteoside inhibits ROS production during RANKL-mediated osteoclast differentiation using a cell-permeable, oxidation-sensitive dye, DCFH-DA. Flow cytometry analysis showed that the mean fluorescence signal specific to DCF in BMMs was apparently right-shifted after stimulation with RANKL, compared to the un-treated control cells (Fig. 7A). This shift was similar to the case that RAW264.7 macrophages were observed after RANKL stimulation (data not shown). Treating BMMs with acteoside reduced the signal intensity of the DCF in a dose-dependent manner. Pretreatment with 10 µM acteoside almost completely reduced the levels of intracellular ROS produced during osteoclast differentiation to the un-treated control levels (Fig. 7B).

To explore the effect of acteoside on bone loss, we prepared an osteoporotic animal model by ovariectomy. There was no significant difference of body weight between OVX and Sham mice during the experimental period (data not shown). The OVX group had significant higher serum levels of IL-1β and IL-6 than the Sham group (Fig. 8). Ovariectomy-induced increases in these inflammatory cytokines were attenuated by oral acteoside administration (AC group). The serum levels of bone turnover markers such as ALP, calcium, TRAP5b, and OC were significantly increased in OVX group. Of these osteoporotic markers, the increased levels of calcium, TRAP5b, and OC in OVX mice were apparently inhibited by acteoside treatment, whereas serum level of ALP was not changed by the treatment. The average maximum fracture load to the middle of the right femoral shaft was significantly lower in the OVX group than in the Sham group (Fig. 9A). Acteoside treatment raised the maximum fracture back up to that of the Sham group. When the cortical bone of the femur was dissected and observed by optic microscopy, the osteoporotic features shown in OVX group had almost completely disappeared in AC mice (Fig. 9B). To verify the effect of acteoside on OVX-induced osteoporosis model, BMD and bone morphologic parameters in trabecular of the light proximal femur were analyzed by micro-CT. As shown in Fig. 9C, an alteration of the femoral trabecular architecture was found in OVX mice, whereas this change was diminished by treatment with acteoside. The results from micro-CT analysis revealed that BMD, a measure of bone strength, was dramatically reduced in OVX mice (Fig. 9D). Compared to Sham group, OVX mice also showed significant changes in BV/TV, Tb.Sp, and Tb.N, but not in Tb.Th. Oral acteoside treatment of OVX mice significantly prevented the alteration in BMD as well as BV/TV and Tb.N.

The role of acteoside on osteoblastic differentiation was further investigated using bone marrow cells. As shown in Fig. 10A, DAG treatment increased the number of alizarin red-stained cells and this was not changed by 10 µM acteoside pretreatment. The amount of dye present indicated that the combined acteoside and DAG treatment did not change mineralization (Fig. 10B). Similarly, DAG-induced increases in the intracellular calcium content (Fig. 10C) and mRNA levels (Fig. 10D) of bone specific markers, such as Runx2, osterix, BSP, and OC, were not affected by pretreatment with acteoside.