Dulbecco's Modified Eagle's Medium (DMEM) and Dulbecco's PBS (DPBS) were acquired from Welgene, Inc. α-Minimum essential medium (α-MEM) and penicillin/streptomycin (P/S) were acquired from Gibco (Thermo Fisher Scientific, Inc.). Fetal bovine serum (FBS) was purchased from Atlas Biologicals, Inc. RANKL was purchased from PeproTech EC Ltd. Cell Counting Kit-8 (CCK-8) was acquired from Dojindo Laboratories, Inc. TRAP staining kits, bicinchoninic acid (BCA) solution, 17b-estradiol (E₂), alendronate (ALN) and 4′,6-diamidino-2-phenylindole (DAPI) were obtained from Sigma-Aldrich (Merck KGaA). Acti-stain™ 488 Fluorescent Phalloidin was obtained from Cytoskeleton, Inc. Osteo assay strip well plates were acquired from Corning, Inc. PCR primers were purchased from GenoTech Corp. Primary and secondary antibodies were as follows: β-actin (sc-8432; Santa Cruz Biotechnology, Inc.), NFATc1 (cat. no. 556602; BD Biosciences), c-Fos (cat. no. sc-447; Santa Cruz Biotechnology, Inc.), matrix metalloprotease 9 (MMP-9; cat. no. ab38898; Abcam), cathepsin K (CTK; cat. no. ab19027; Abcam), peroxidase AffiniPure Goat Anti-Mouse IgG (cat. no. 115-035-062; Jackson ImmunoResearch Laboratories, Inc.) and peroxidase AffiniPure Goat Anti-Rabbit IgG (cat. no. 115-035-144; Jackson ImmunoResearch Laboratories, Inc.).

GF (100 g) purchased from Omni Herb, Co., Ltd. was immersed in a liter of 30% ethanol at 4°C for 3 weeks and sonicated (40 kHz) once every 2 days at room temperature for 1 h. The resulting mixture was filtered using no. 3 filter paper (Whatman plc; Cytiva), and the filtrate was concentrated using a rotary evaporator. Finally, the concentrated extract was freeze-dried into a powder with a yield of 15.9% (15.9 g powder). The resulting powder was stored at −20°C until further use.

RAW 264.7 cells were obtained from The Korean Cell Line Bank (Korean Cell Line Research Foundation) and MC3T3-E1 cells from the American Type Culture Collection (cat. no. ATCC-CRL-2593). RAW 264.7 cells were cultured in DMEM supplemented with 10% FBS and 1% P/S and subcultured every 2 days in a humidified incubator (Thermo Fisher Scientific, Inc.) at 37°C and 5% CO₂. MC3T3-E1 cells were cultured in α-MEM (without ascorbic acid), supplemented with 10% FBS and 1% P/S and subcultured every 3 days in a humidified incubator at 37°C and 5% CO₂. To evaluate the cytotoxicity of GF, 5,000 RAW 264.7 and 5,000 MC3T3-E1 cells were seeded in 96-well plates and allowed to stabilize for 24 h. Various concentrations of GF (0, 5, 10, 20 and 40 µg/ml) were then added to the cells and the cells were incubated for a further 24 h. After treatment, 20 µl CCK-8 solution was added to each well and the plates were incubated for an additional 2 h. Cell viability was determined by measuring the absorbance at 450 nm using an enzyme-linked immunosorbent assay (ELISA) plate reader. The results were expressed as a percentage of the untreated cell control and a survival rate ≤90% compared with the control was considered indicative of toxicity.

To evaluate the inhibitory effect of GF on osteoclast formation, RAW 264.7 cells were seeded at a density of 5,000 cells/well in 96-well plates and incubated for 24 h. The cells were then treated with 100 ng/ml RANKL, which is an osteoclastogenesis-inducing cytokine, and various concentrations of GF (0, 5, 10, 20 and 40 µg/ml) for 5 days with medium changes on days 2 and 4. After fixation with 4% formalin for 1 h at the room temperature, TRAP staining was performed for 1 h at the room temperature and TRAP (+) cells with ≥3 nuclei were counted using ImageJ software v1.46 (National Institutes of Health) under an inverted microscope (magnification, ×100). In addition, the activity of TRAP in the culture medium was measured by mixing an equal volume of TRAP solution (4.93 mg p-nitrophenyl phosphate in 750 ml 0.5 M acetate solution and 150 ml tartrate acid solution) with the medium on day 5 and then incubated for 1 h. Finally, the reaction was stopped with NaOH and the absorbance at 405 nm was measured using an ELISA plate reader.

To confirm the effect of GF on the inhibition of bone resorption ability, 5,000 RAW 264.7 cells were seeded into osteo assay strip well plates and stabilized for 24 h at 37°C. The cells were treated with RANKL (100 ng/ml) and various concentrations of GF (0, 5, 10, 20 and 40 µg/ml) for 5 days at 37°C. The culture medium was exchanged on days 2 and 4 with the same medium. Then, the cells were removed using 4% sodium hypochlorite. The area of the pit in the plate was measured using ImageJ software v1.46 and the area in the plate absorbed by osteoclasts was expressed as a percentage of the total area.

To determine the effect of GF on actin ring formation, 5,000 RAW 264.7 cells were seeded into 96-well plates and stabilized for 24 h. The cells were treated with RANKL (100 ng/ml) and various concentrations of GF (0, 5, 10, 20 and 40 µg/ml) for 5 days at 37°C. The culture medium was exchanged on days 2 and 4 with the same medium. After the formation of osteoclasts, the cells were fixed using 4% paraformaldehyde for 10 min at room temperature and then permeabilized with 0.1% Triton™ X-100 in PBS for 5 min at room temperature. The cells were then incubated with Acti-stain™ 488 fluorescent phalloidin and DAPI in the dark for 30 min at room temperature. The actin rings were visualized using a fluorescence microscope (magnification, ×200) and the number of actin rings was quantified using ImageJ software v1.46.

RAW 264.7 cells were seeded at a density of 5×10⁵ cells per 60 mm dish and allowed to stabilize for 24 h. The cells were then treated with RANKL (100 ng/ml) and different concentrations of GF (0, 5, 10, 20 and 40 µg/ml) for 24 h to confirm the effect of GF on transcription factors related to osteoclastogenesis. After treatment, the cells underwent a DPBS wash, and the total protein was extracted from the cells using RIPA buffer (50 mM Tris-Cl, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate and 0.1% SDS) containing protease and phosphatase inhibitors (1:100; Sigma-Aldrich; Merck KGaA), to obtain a comprehensive protein profile. The protein concentration was quantified using the BCA assay and 30 µg protein/lane was separated by SDS-PAGE on 10% gels, before being transferred onto nitrocellulose membranes. The membranes were blocked with TBS-0.5% Tween 20 containing 5% skim milk to minimize non-specific protein binding for 1 h at room temperature, and then incubated with primary antibodies overnight at 4°C (NFATc1, 1:1,000; c-Fos, 1:200; β-actin, 1:1,000) and then with secondary antibodies (1:10,000) for 1 h at room temperature. Finally, the bands were visualized using ECL solution (Cytiva) and their density was measured using ImageJ v1.46. The protein expression levels were then normalized to those of the loading control (β-actin).

To confirm the effect of GF on osteoclastogenesis-related transcription factors, 2×10⁵ RAW 264.7 cells were seeded in 6-well plates and stabilized for 24 h. Cells were treated with RANKL (100 ng/ml) and various concentrations of GF (0, 5, 10, 20 and 40 µg/ml) for 4 days. Next, the cells were washed with DPBS three times and total RNA was extracted using TRIzol^® reagent (Takara Bio, Inc.). The NanoDrop 2.0 spectrophotometer (NanoDrop Technologies; Thermo Fisher Scientific, Inc.) was utilized to quantify 2 µg of extracted RNA. Subsequently, cDNA was synthesized in accordance with the manufacturer's protocol provided by the RT kit (Invitrogen; Thermo Fisher Scientific, Inc.), followed by its amplification with Taq polymerase (Kapa Biosystems; Roche Diagnostics). The PCR conditions were as follows: Denaturation for 30 sec at 94°C, annealing for 30 sec at 53–58°C and extension for 30 sec at 72°C. Primer sequences and number of cycles for each gene are listed in Table I. mRNA expression levels were normalized to that of the loading control (β-actin). The RT-PCR results were confirmed using a 2% agarose gel stained with SYBR Green (1:10,000; Invitrogen; Thermo Fisher Scientific, Inc.). The density of the bands was measured using ImageJ v1.46.

A total of 48 Sprague Dawley 11-week-old female rats were obtained from Nara Biotech (mean weight 240±10 g) and the animals were stabilized in a new environment for 1 week. All animals were raised in a specific pathogen-free environment, including a temperature of 20–24°C and humidity of 50–60%, under a 12 h light/dark cycle. All animals had ad libitum access to food and water and the animal experiments were approved by The Kyung Hee University Institutional Animal Care and Use Committee (Seoul, Republic of Korea; approval no. KHSASP-21-185). The postmenopausal osteoporosis model was induced by OVX. For this, all animals were subjected to deep anesthesia in 100% oxygen with 5% isoflurane before undergoing bilateral OVX. The concentration of isoflurane was maintained at 2–3% during surgery. Animals in the sham group underwent the same surgical procedure with identical stress levels but without bilateral OVX. After surgery, 4 mg/kg gentamycin was injected for 3 days to prevent infection at the wound site. The animal groups (n=8/group) comprised the following: i) Sham group, animals subjected to sham operation and orally administered distilled water; ii) OVX group, animals subjected to OVX and orally administered distilled water; iii) E₂ group, animals subjected to OVX and orally administered 100 µg/kg E₂; iv) ALN group, animals subjected to OVX and orally administered 3 mg/kg ALN; v) low GF (GF-L) group, animals subjected to OVX and orally administered 21.147 mg/kg GF; and vi) high GF (GF-H) group, animals subjected to OVX and orally administered 135.34 mg/kg GF. The concentration of administered GF was determined as per the following procedure. In Korean medicine, the average amount of GF consumed by a 60 kg adult per day is 8 g (27). The yield of GF is 15.9%. The calculated dose to be administered in the GF-L group was 21.147 mg/kg. It is known that the metabolism of rats is 6.4× faster than that of humans (28). Therefore, the GF-H group was administered 135.34 mg/kg. In the present study, two drugs were used as positive controls to compare the anti-osteoporotic effects of GF. The first is E₂, which is commonly used to treat osteoporosis in postmenopausal women. The second is ALN, which is a bisphosphonate agent used to treat osteoporosis (29), preventing bone destruction and increasing bone thickness. The change in body weight of animals was measured once a week and drug administration was performed for a total of 8 weeks. The health status of the rats was checked twice daily. The humane endpoints of this study were set as follows: i) Inability to eat, ii) rapid weight loss of 20% or more in 1 week, iii) immobility. During the experiment, no abnormal changes in health were observed in all of the rats. After 8 weeks, all animals were sacrificed by lethal cardiac puncture up to 10 ml (30), under deep anesthesia in 100% oxygen and 5% isoflurane. Subsequently, after confirming that the hearts of all animals had completely stopped, cervical dislocation was performed.

Following whole blood extraction and reaction in an serum separator tube for 30 min at room temperature, the serum was separated through centrifugation (14,310 × g, 4°C, 10 min) and TRAP activity of the serum was assessed as aforementioned. The expression of alkaline phosphatase (ALP), aspartate aminotransferase (AST) and alanine transferase (ALT) was analyzed by DkKorea, an animal serum analysis company.

The femurs of all animals were harvested 8 weeks after surgery and immersed in 10% neutral buffered formalin (NBF) at room temperature for 24 h to fix the bone tissue. Bone mass and microstructure were then evaluated by micro-CT using SkyScan1176 (Skyscan; Bruker Corporation) under the following conditions: An aluminum filter of 0.5 mm, 8.9 µm pixel size and a rotation angle of 180° with rotation steps of 0.4°. The analyzed bone parameters included bone mineral density (BMD), bone volume/total volume (BV/TV), trabecular thickness (Tb.Th) and trabecular separation (Tb.Sp).

The femurs of all animals were harvested 8 weeks after surgery and fixed in 10% NBF at room temperature for 24 h. Subsequently, the femurs were demineralized in ethylenediaminetetraacetic acetic acid (pH 7.4) for 3 weeks and embedded in paraffin. Tissue sections of 5 µm thickness were obtained using a Zeiss rotary microtome and the sections were air-dried for 1 day. Hematoxylin-eosin (H&E) staining was used to analyze histomorphological changes. In brief, paraffin-embedded tissue sections were subjected to deparaffinization using xylene and ethanol, followed by counterstaining with eosin for 10 sec and hematoxylin for 5 min. For immunohistochemistry (IHC) staining, initially, the paraffin-embedded tissue underwent deparaffinization using a combination of xylene and ethanol. After that, antigen retrieval was performed using proteinase K (0.4 mg/ml) at 37°C for 30 min. Subsequently, 0.3% H₂O₂ (w/v) in Mt-OH was added at room temperature for 30 min to inhibit the activity of endogenous peroxidases, and 5% goat serum (Gibco; Thermo Fisher Scientific, Inc.) in PBS was used to prevent the binding of non-specific proteins for 1 h at room temperature. The tissues were incubated overnight at 4°C with primary antibodies NFATc1 (diluted 1:100) and CTK (diluted 1:100). The sections were then incubated with the secondary antibody (cat. no. BA-1000; Vector Laboratories, Inc.; Maravai LifeSciences) for 1 h at room temperature. The signal was detected using the ABC kit (Vector Laboratories, Inc.; Maravai LifeSciences) at room temperature for 30 min and visualized with 3,3-diaminobenzidine (DAB solution; cat. no. SK-4100; Vector Laboratories, Inc.; Maravai LifeSciences) at room temperature for 10 min. The stained tissues were analyzed using a light microscope (Olympus Corporation) and the trabecular area and positive cells were measured using ImageJ software v1.46.

Alizarin red S staining was conducted to investigate the impact of GF on osteoblast formation and the formation of calcified nodules. A 6-well plate was seeded with 3×10⁵ MC3T3-E1 cells and incubated for 24 h. In order to differentiate MC3T3-E1 cells into osteoblasts, the cells were treated with ascorbic acid (A.A; 25 µg/ml), β-glycerol phosphate (B.G.P; 10 mM) and various concentrations of GF (0, 5, 10, 20 and 40 µg/ml) at 37°C for 2 weeks. The culture medium was replaced every 3 days. Once calcified nodules were visible on the plate, the medium was removed and the cells were washed three times with cold PBS. The cells were then fixed with 80% ethanol at 4°C for 1 h and stained with alizarin red S dye at room temperature for 30 min. Finally, images of the plate were captured using a digital camera (Canon, Inc.). The degree of dye staining was measured by extracting the dye using 10% (v/w) cetylpyridinium chloride in sodium phosphate and measuring the absorbance at 570 nm.

Echinocystic acid and oleanolic acid are active ingredients of GF (20) and were purchased from Sigma-Aldrich (Merck KGaA). To analyze the GF extract, the absorbance was measured using a UV detector (2695 HPLC system with 2996 PDA detector; Waters Corp.), the flow rate was 1 ml and the operating time was 45 min using an Xbridge C18 (250×4.6 mm, 5 µm; Waters Corp.) column at the room temperature. The sample insertion volume was 10 µl. Solvent A of the mobile phase was acetonitrile and solvent B was water with 0.1% formic acid. The gradient program was as follows. 0 min, 5% B; 0–30 min, 5–50% B; 30–60 min, 50–95% B.

All data are presented as the mean ± standard error of mean. All experiments were repeated at least 3 times. Comparisons were conducted with one-way ANOVA and Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference. Statistical analysis was performed using GraphPad PRISM v5.01 (Dotmatics).

To determine the effect of GF on RAW 264.7 cell cytotoxicity, a CCK-8 assay was performed that showed that GF did not induce RAW 264.7 cell cytotoxicity (Fig. 1A). The impact of GF on osteoclastogenesis in RAW 264.7 cells was also investigated by analyzing TRAP activity and performing TRAP staining. The results showed that RANKL-treated cells had a greater number of TRAP (+) cells and higher TRAP activity compared with untreated cells, while GF reduced both the number of TRAP (+) cells and TRAP activity in a dose-dependent manner (Fig. 1B-D).

To investigate the impact of GF on osteoclastogenesis and bone resorption, experiments on F-actin ring and pit formation were conducted. The results, as illustrated in Fig. 2A and C, showed that RANKL-treated cells had a higher pit area compared with the untreated cells and GF dose-dependently decreased the pit area. Furthermore, RANKL-treated cells displayed more F-actin rings than the untreated cells, but GF inhibited both the number and structure of F-actin rings, as depicted in Fig. 2B and D.

To determine the effect of GF on osteoclastogenesis-related transcription factors, western blotting and RT-PCR were used. As shown in Fig. 3A and B, the RANKL-treated cells exhibited increased expression of NFATc1 compared with the untreated cells. However, GF significantly reduced the protein and mRNA levels of NFATc1 at concentrations of 20 and 40 µg/ml. Furthermore, the results also indicated that treatment with RANKL led to increased protein and mRNA expression levels of c-Fos compared with the untreated cells. Conversely, treatment with 40 µg/ml GF resulted in a decrease in c-Fos protein and mRNA expression levels compared with the untreated cells (Fig. 3C and D).

To explore the effects of GF treatment on osteoclast-specific genes, RT-PCR was performed. The results revealed that RANKL-induced cells showed upregulation in the mRNA expression of TRAP (Acp5), RANK (Tnfrsf11a), MMP-9 (Mmp9), CA2 (Ca2), osteoclast-associated receptor (OSCAR; Oscar), ATP6vod2 (Atp6v0d2) and dendritic cell-specific transmembrane protein (DC-STAMP; Dcstamp), compared with the non-treated cells (Fig. 4A and B). However, GF treatment (20 and 40 µg/ml) resulted in a decrease in the mRNA expression of TRAP, RANK, MMP-9, CA2, OSCAR, ATP6vod2 and DC-STAMP, compared with the RANKL-treated cells.

To further demonstrate the suppressive effect of GF on osteoclast differentiation, formation and bone resorption, an OVX-induced rat model was used. As shown in Fig. 5A, compared with the sham group, the OVX group showed a significant increase in body weight from 3 weeks. The ALN and GF groups did not show significant changes in body weight compared with the OVX group. Uterine weight was significantly reduced in the OVX group compared with the sham group (Fig. 5B). In the E₂ and ALN groups, there was a significant increase in uterine weight compared with that of the OVX group. However, the GF groups showed no changes in uterine weight compared with the sham group.

Next, serum analysis was performed to confirm the effect of GF on bone formation and bone resorption markers. Compared with the sham group, the OVX group exhibited a slight increase in TRAP activity (not significant), as indicated in Fig. 5C. The E₂ and GF-L groups demonstrated a reduction in TRAP activity compared with the OVX group, but this difference was not significant. The ALN group had no significant change in TRAP activity. In the GF-H group, TRAP activity was significantly decreased compared with that in the OVX group. In addition, the ALP level increased in the OVX group compared with the sham group, although the difference was not statistically significant (Fig. 5D). The E₂ group did not show a significant change in ALP levels, while the ALN and GF-L groups demonstrated lower ALP levels compared with the OVX group, although this difference was not statistically significant. The ALP level significantly decreased in the GF-H group compared with the OVX group.

To evaluate the potential impact of GF administration on hepatotoxicity, an analysis of serum AST and ALT levels was conducted. Both factors showed no difference between the SHAM and OVX groups. The E₂ group showed no change in AST levels compared with OVX group and the ALN and GF-L groups appeared to have decreased AST levels compared with the OVX group, but these differences were not significant (Fig. 5E). However, AST levels were significantly decreased in the GF-H group compared with the OVX group. In the case of ALT, there were no significant differences between the OVX and drug administered groups (Fig. 5F).

To further confirm the in vitro results, the ability of GF to inhibit postmenopausal osteoporosis caused by OVX induction was investigated. The OVX models were divided into six groups of 8 rats: sham, OVX, OVX + E₂, OVX + ALN, OVX + GF-L and OVX + GF-H. The impact of GF on bone loss was evaluated using bone microstructure and micro-CT. As shown in the 3D images in Fig. 6A, the OVX group had a decreased BMD relative to the sham group, while the E₂, ALN and GF treatments substantially increased the BMD of the femur, compared with the OVX group. Compared with the sham group, the OVX group exhibited decreased BV/TV and Tb.Th values in the femur, while the E₂, ALN and GF groups demonstrated significant increases in both values, compared with the OVX group; notably, for Tb.Th, there was no significant difference between the GF-L and OVX groups. Additionally, OVX led to increased Tb.Sp in the femur compared with the sham group, while treatment with E₂, ALN and GF resulted in a decrease in Tb.Sp (Fig. 6B-E).

To investigate the trabecular area, the expression of NFATc1 and CTK using H&E and IHC staining was examined. IHC staining demonstrated that the levels of NFATc1 and CTK were significantly increased in the OVX group compared with the sham group, and reduced in the E₂, ALN and GF groups compared with the OVX group (Fig. 7A-D). Next, to evaluate the effect of GF on the trabecular area in the femur, H&E staining was performed (Fig. 7E and F). The OVX group displayed a reduction in trabecular area compared with the sham group. However, the E₂, ALN and GF groups showed an increase in trabecular area relative to the OVX group.

To assess the impact of GF on osteoblasts and the formation of calcified nodules, alizarin red S staining was performed. According to the results depicted in Fig. 8A, GF did not induce a notable effect on the osteoblast differentiation induced by A.A and B.G.P. Additionally, assessment of the extracted dye showed that GF did not demonstrate a higher absorbance than the control (Fig. 8B). Furthermore, the tested concentration of GF did not exhibit any cytotoxicity (Fig. 8C).

Echinocystic acid and oleanolic acid are active ingredients of GF (20). Chromatographic peaks of standard echinocystic acid and oleanolic acid ingredients were detected (Fig. 9A) for the comparison with GF (Fig. 9B). It was demonstrated that peaks for both chromatograms were detected at the same time, and the contents of echinocystic acid and oleanolic acid in GF were confirmed to be 8.97 and 10.8 mg/g, respectively.