Icaritin ameliorates RANKL-mediated osteoclastogenesis and ovariectomy-induced osteoporosis

Reagents

Recombinant Murine cytokines (RANKL and M-CSF) were provided by Bio-Techne R&D systems (Minneapolis, MN, USA), which were dissolved in 0.1% BSA yielding stocking solution (100 ng/μl) and kept at −80°C before being used. Icaritin (ICT; C₂₁H₂₀O₆; MW: 368.385) was manufactured by Target Molecule Corp. (Boston, MA, USA), prepared as stocking solution by dissolving in DMSO (Dimethyl sulfoxide). The Cell Counting Kit-8 (CCK-8) was also manufactured by Target Molecule Corp. (Boston, MA, USA). The cell culture reagents including minimum essential medium eagle-alpha modification (α-MEM) medium, penicillin/streptomycin, and fetal bovine serum (FBS) were bought from Gibco (Grand Island, NY, USA). The Actin-Tracker and DAPI were purchased from Beyotime (Shanghai, China). The Tartrate Resistant Acid Phosphatase (TRAP) kits, DMSO, and other common reagents of the highest purity available were gained from Sigma-Aldrich (St. Louis, MO, USA).

BMMs isolation and culture

As reported previously, the primary bone marrow cells were isolated from the long bone marrow cavities of 4–6 weeks old mouse by flush and then culture in non-treated plates with α-MEM for 3 days. Subsequently, the BMMs were washed and harvested for further experiments [19]. In all experiments, M-CSF (30 ng/ml) was used to maintain the proliferation of BMMs.

Cell viability assay

The harvested BMMs were seeded into 96-well plate at the same density of about 5 × 10³ cells/well with 30 ng/mL and cultured for 24 hours. Initially, the BMMs were cultured with higher dose of ICT (1, 5, 10, 20, and 40 μM) for 3 days to determine the IC₅₀ of ICT on BMMs. Then, the BMMs were treated with vehicle (DMSO) and lower dose of ICT (0.01, 0.1, and 1 μM) for 1 day, 3 days, 5 days, and 7 days. The absorbance of 450 nm was measured by FlexStation 3 (Molecular Device, Shanghai, China) according to instruction manual of CCK-8 kit. In addition, the BMMs cultured for 7 days were subjected to LIVE/DEAD assay via protocol of LIVE/DEAD Cell Imaging Kit (R37601) (Thermo Fisher Scientific, Waltham, MA, USA). Briefly, the component A was mixed component B to create working solution, which was further added to plate for 15 mins incubation at darkroom. The live and dead cells were showed green fluorescence and red fluorescence respectively.

Osteoclast differentiation and TRAP staining

The harvested BMMs were seeded into 96-well plate at the same density about 1 × 10⁴ cells/well with 30 ng/mL and cultured for 24 hours. Then, the BMMs were cultured in osteoclastogenic medium containing 10% FBS, 1% penicillin/streptomycin, M-CSF (30 ng/mL), and RANKL (50 ng/mL) with vehicle (DMSO) or indicated dose of ICT (0.01, 0.1, and 1 μM). In time-course assay, the 1 μM ICT were supplemented in medium at either day 1 or day 3. The fresh medium was changed every two days until osteoclasts formed, the TRAP staining was carried out by TRAP kit. The TRAP-positive cells with three or more nuclei were recognized as mature osteoclasts, which were captured by Olympus IX-71 (Olympus, Tokyo, Japan) and quantified by ImageJ software.

F-actin ring formation evaluation

The harvested BMMs were seeded into Corning Osteo Assay Surface plate (Corning Incorporated Life Science, Corning, NY, USA) at the same density about 2 × 10⁴ cells/well with 30 ng/mL and cultured for 24 hours. After that, the BMMs were cultured in osteoclastogenic medium mentioned above. When the formed osteoclasts were observed, cells were fixed with 4% paraformaldehyde (PFA) for 20 mins and permeabilized with 0.1% (v/v) Triton X-100 for 5 mins. After that, we used Actin-Tracker and DAPI to label F-actin and nucleus respectively. The images of F-actin ring were acquired by Olympus IX-71 (Olympus, Tokyo, Japan).

Bone resorption assay

The BMMs were cultured in 6-well plates in density of 2.5 × 10³ cells/cm² and treated with M-CSF (30 ng/mL) and RANKL (50 ng/mL) to perform osteoclast differentiation. The formed osteoclasts were digested and seeded on Corning Osteo Assay Surface plate (Corning Incorporated Life Science, NY, USA) at the same density about 2 × 10³ cells/well with osteoclasts induction medium in the presence of vehicle (DMSO) or indicated dose of ICT (0.01, 0.1, and 1 μM). After 3 days culture, the osteoclasts were removed by 5% sodium hypochlorite treatment and the bone resorption pits images were captured by Olympus IX-71 (Olympus, Tokyo, Japan) and quantified by Image J.

RNA isolation and quantitative RT-PCR

The RNA isolation and Quantitative RT-PCR were performed as described previously [18]. Briefly, the isolated BMMs were seeded on 6-well plate at density of 2.5 × 10³ cells/cm² and cultured with M-CSF (30 ng/mL). Twenty-four hours later, the osteoclastogenic medium containing vehicle (DMSO) or ICT (1 μM) was used to induce osteoclast differentiation. Three days later, the total RNA was extracted using RNA-Quick Purification Kit (YISHAN Bio-Tec, Shanghai, China) according to the instructions. Next, cDNA was synthesized using 1 μg of extracted RNA according to instrument of Prime Script RT reagent kit (TaKaRa Biotechnology, Shiga, Japan). The synthesized cDNA was used as templates to perform qRT-PCR using SYBR (TaKaRa Biotechnology, Shiga, Japan). The GAPDH was selected as the housekeeping gene to normalize the relative expression of target genes. The specific primer sequences are as follows: GAPDH: F 5′-AGGTCGGTGTGAACGGATTTG-3′ and R 3′-TGTAGACCATGTAGTTGAGGTCA-5′; NFATc1: F 5′-GACCCGGAGTTCGACTTCG-3′ and R 3′-TGACACTAGGGGACACATAACTG-5′; c-fos: F 5′-CGGGTTTCAACGCCGACTA-3′ and R 3′-TTGGCACTAGAGACGGACAGA-5′; RANK: F 5′-GGACGGTGTTGCAGCAGAT-3′ and R 3′-GCAGTCTGAGTTCCAGTGGTA-5′; Cathepsin K: 5′-GAAGAAGACTCACCAGAAGCAG-3′ and R 3′-TCCAGGTTATGGGCAGAGATT-5′; MMP9: F 5′-CTGGACAGCCAGACACTAAAG-3′ and R 3′-CTCGCGGCAAGTCTTCAGAG-5′; TRAP: F 5′-CACTCCCACCCTGAGATTTGT-3′ and R 3′-CATCGTCTGCACGGTTCTG-5′.

Western blot analyses

The harvested BMMs were seeded into 6-well plates at the same density about 2.5 × 10³ cells/cm² with M-CSF (30 ng/mL). Followed by overnight culture, the BMMs were cultured in osteoclastogenic medium containing vehicle (DMSO) or ICT (1 μM) for 3 days. According to the process mentioned previously [20], the whole protein was prepared by treating cells with RIPA lysis buffer containing broad spectrum phosphatase inhibitors and PMSF, whose concentration was determined by the BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA), and 10 μg proteins was following electrophoresed on 10% SDS-polyacrylamide gel and transferred to the PVDF membranes (Millipore, Burlington, MA, USA). After blocking with 5% non-fat milk at room temperature for 1 hour, the membranes were incubated with specific primary antibodies at 4°C overnight. Then, membranes were washed with TBS-Tween and were incubated with corresponding HRP-conjugated secondary antibodies for 1 hour at room temperature. Lastly, the protein signals were visualized using electrochemical luminescence reagent (ECL) (Millipore, Burlington, MA, USA) on Bio-Rad system and analyzed using Image J software.

The following antibodies were provided by Abmart (Shanghai, China): NFATc1 (TD6446), c-fos (T56596), Cathepsin K (TD6614), MMP9 (TA5228), Parkin (T56641), DRP1 (TD7037), MFN2 (T56638), FIS1 (TD12005). The other antibodies were provided by Proteintech Group (Rosemont, IL, USA): GAPDH (60004-1-Ig). The secondary antibodies, including HRP conjugated anti-rabbit or anti mouse antibody were provided by Jackson ImmunoResearch Inc. (West Grove, PA, USA).

Mitochondrial membrane potential detection

The harvested BMMs were seeded into 6-well plates at the same density about 2.5 × 10³ cells/cm² with M-CSF (30 ng/mL) and cultured for 24 hours. Then, the BMMs were cultured in osteoclastogenic medium with vehicle (DMSO) or ICT (1 μM). Three days later, the JC-1 staining solution (No. 10009172) (Cayman Chemical, Ann Arbor, MI, USA) were added at rate of 100 μl per ml. After incubated in CO₂ incubator at 37°C for 20 mins, the mitochondrial membrane potential was detected by fluorescence microscope.

Transmission electron microscopy (TEM)

The harvested BMMs were seeded into 6-well plates at the same density about 2.5 × 10³ cells/cm² with M-CSF (30 ng/mL) and cultured for 24 hours. Then, the BMMs were cultured in osteoclastogenic medium with vehicle (DMSO) or ICT (1 μM). Three days later, cells were fixed with 2.5% glutaraldehyde, postfixed with 1% osmium tetroxide, dehydrated in a graded series of ethanol, and embedded in resin. Next, the embedded samples were cut into 60 nm sections by an ultramicrotome (Leica EM UC7). After being stained with uranyl acetate and lead citrate, the sections were finally observed at TEM (HT7800/HT7700, HITACHI).

Ovariectomy (OVX) murine model establishment

The animal experiment was designed and performed following the guide of the NIH “Principles of Laboratory Animal Care” (1996 Revised Version) and under the supervision of the Animal Use and Care Committee of Shanghai University of Medicine and Health Sciences Affiliated to Zhoupu Hospital. The wild-type C57/BL6 female mice (12-week-old) were purchased from Gempharmatech Co., Ltd (Nanjing, China) and kept in the specific-pathogen-free animal care facility. All mice were fed with sterile chow and water and kept in constant condition with controlled temperature (25 ± 2°C), humidity (60 ± 5%), and illumination (12:12 light/dark cycle.). Mice (21–25 g) were randomly arranged into three groups (n = 6), that is sham-operated mice with vehicle treatment (Sham), bilateral ovariectomized mice with vehicle treatment (OVX), and bilateral ovariectomized mice with ICT treatment (ICT). The treatment dose of ICT is 10 mg/kg that was referred to previous study that ICT showed no toxicity for mice [21]. The sham and OVX surgery were performed according to previous description [18, 20, 22] and the vehicle and ICT were given by intraperitoneal injections every two day for 6 weeks. All mice were sacrificed by over dose of pentobarbital, then femurs were collected for the following experiments.

Micro-computed tomography (μ-CT) and histomorphometric analysis

After removing muscle and soft tissue, the fixed distal femur was scanned and analyzed by microcomputed tomography (μ-CT) system by following settings: 100 kV source voltage and 98 mA source current with a 10.5 mm resolution. The protocol of μ-CT and histomorphometric analysis referred to our previous study [22]. Briefly, the range at distance of 0.5 mm below subchondral bone was selected as the region of interest (ROI) for further analysis. Three-dimensional reconstruction of femur and evaluation of bone volume/tissue volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), and trabecular separation (Tb.Sp) were all performed in system software installed in μ-CT. After μ-CT scanning, all femurs were decalcified and embedded in paraffin. The 5 μm bone sections were cut along the coronal plane on a microtome and the TRAP staining were performed. Lastly, the representative pictures of distal femur trabecular area were captured and analyzed by microscopy.

Statistical analysis

Each experiment in this study was independently repeated more than three times and the data were described as mean ± SD. We performed one-way ANOVA with Tukey–Kramer honest significant difference (HSD) test to examine the significant differences where more than two groups existed by GraphPad Prism 7.0 software (GraphPad Software, San Diego, CA, USA) and unpaired t-test was used for two groups comparisons. A value of p < 0.05 was considered to indicate statistically significant differences.

ICT inhibits osteoclast differentiation induced by RANKL

In vitro study, we firstly evaluated the potential toxicity of ICT on BMMs via CCK-8 and LIVE/DEAD assay. Initially, for determining the IC₅₀ of ICT on BMMs, BMMs were cultured with higher dose of ICT (1, 5, 10, 20, and 40 μM) for 3 days, and the results of the cell activity assay demonstrated that the IC₅₀ of ICT on BMMs is 16.33 μM (Supplementary Figure 1). Then BMMs were cultured with lower dose of ICT (0.01, 0.1, and 1 μM), which will be used in following study, and CCK-8 assay showed that indicated dose of ICT has no toxicity on BMMs during continuous culture (Figure 1A). Furthermore, in live/dead assay, no obvious dead cells were observed after 7-days culture (Figure 1B). Subsequently, we investigated the effect of ICT on osteoclast differentiation. As shown in TRAP staining, ICT suppressed RANKL-induced osteoclast differentiation at concentration of 1 μM ICT and about 60% osteoclasts numbers were reduced compared to RANKL group (Figure 2A–2C). Furthermore, ICT treatment (1 μM) in the early stage (day 1–3) or in the late stage (day 3–5) of osteoclast differentiation both resulted in decreased osteoclast formation, and according to the count and size of osteoclast, the effect of ICT on early stage of osteoclast differentiation is superior to late stage of osteoclast differentiation (Figure 2D–2F).

Figure 1. ICT has no toxicity on BMMs. (A) Cytotoxicity of ICT on BMMs at day 1, day 3, day 5, and day 7 by CCK-8 assay. (B) The live/dead staining of BMMs after culture with ICT for 7 days. All data were presented as mean ± SD, n = 3. Abbreviation: ns: not statistical significance.

Figure 2. ICT inhibits osteoclast differentiation induced by RANKL. (A) Representative TRAP staining images of BMMs treated with DMSO or different doses of ICT in the presence of M-CSF and RANKL for 5 days. (B, C) Quantification of the number and area of TRAP-positive cells. (D) Representative TRAP staining images of BMMs treated with ICT in the presence of M-CSF and RANKL for the indicated time. (E, F) Quantification of the number and area of TRAP-positive cells. All data were presented as mean ± SD, n = 3. Abbreviation: ns: not statistical significance; ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

ICT suppresses osteoclast function

Due to the inhibition on osteoclast differentiation, we speculated whether ICT also inhibits osteoclastic activity. When we observed differentiated osteoclast in Corning Osteo Assay Surface plate, we used Actin-Tracker to label F-actin ring formation that is an observed trademark for osteoclast mature and provides a prerequisite for osteoclastic function and found that the formation of F-actin ring was disrupted through observing the reduced nucleation per osteoclast and F-actin count (Figure 3A–3C). Unsurprisingly, in bone resorption assay, the resorption pit corroded by osteoclasts was also decreased when formation osteoclast cultured with 1 μM ICT (Figure 3D, 3E).

Figure 3. ICT suppresses osteoclast function. (A) Representative images of F-actin rings. (B) Quantification of the F-actin rings number. (C) Mean number of nuclei in each cell. (D) Representative images of pit area by ICT-treated osteoclast on Osseo Assay plate. (E) Quantification of pit area. All data were presented as mean ± SD, n = 3. Abbreviation: ns: not statistical significance; ^*P < 0.05; ^****P < 0.0001.

ICT suppresses osteoclast-specific protein and gene expression

For further investigating the expression of osteoclast-specific genes during osteoclast differentiation, the BMMs were cultured in osteoclastogenic medium with DMSO or ICT (1 μM) for 3 days. The specific genes responsible for osteoclast formation and osteoclastic bone resorption activity were selected, including NFATc1, c-fos, RANK, Cathepsin K, MMP9, and TRAP, and the normalized data revealed that ICT markedly attenuated RANKL-stimulated expression of NFATc1, c-fos, RANK, Cathepsin K, MMP9, and TRAP (Figure 4A). In western blot assay, the data demonstrated that the protein expression of NFATc1, c-fos, MMP9, and Cathepsin K were also reduced compared with RANKL stimulation, which is in line with the mRNA results (Figure 4B–4E).

Figure 4. ICT abrogates RANKL-induced NFATc1 and c-fos transcription and following osteoclast-related genes. BMMs were cultured in osteoclastogenic medium containing vehicle (DMSO) or ICT (1 μM) for 3 days. (A) The relative mRNA expression of NFATc1, c-fos, RANK, Cathepsin K, MMP9, TRAP. (B) Representative protein bands of NFATc1 and c-fos. (C) Quantitative analysis of NFATc1 and c-fos. (D) Representative protein bands of Cathepsin K and MMP9. (E) Quantitative analysis of Cathepsin K and MMP9. All data were presented as mean ± SD, n = 3, ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

ICT downregulates mitochondrial mass and function

As described previously, osteoclast differentiation and activation are energy-intensive processes, the amount and activity of mitochondria in cell will be enhanced [23]. The BMMs were cultured in osteoclastogenic medium for 3 days. Through TEM, we found that RANKL stimulation increased mitochondria mass and ICT could reduce the number of mitochondria (Figure 5A). Then, we observed the ultrastructure of mitochondria and found that the cristae in BMMs treated with M-CSF and RANKL was richer than cultured with M-CSF only, and ICT could reduce the density of cristae (Figure 5B). In addition, we detected the mitochondrial membrane potential by JC-1 staining, which is a prerequisite for the production of ATP and reflect the activity of the entire mitochondrial function [24]. When mitochondria membrane potentials are high, the JC-1 will aggregate and emit red fluorescence, otherwise, JC-1 will exist in the form of a monomer and produce green fluorescence. Our results showed that during osteoclast differentiation, the mitochondria membrane potentials was increased, which was significantly reduced after treatment with ICT (Figure 6A, 6B). Mitochondrial fusion and fission are known to regulate the function and morphology of mitochondria [25]. Therefore, we detected the protein level involved in mitochondrial fusion and fission by western-blot. As shown in our results, the level of Parkin, DRP1, MFN2, and FIS1 were upregulated when BMMs were treated with RANKL, and ICT inhibited the elevated level of Parkin, DRP1, MFN2, and FIS1 (Figure 6C, 6D). In other word, ICT inhibited the stimulated mitochondrial fusion and fission.

Figure 5. Representative images of ultrastructural analysis of mitochondria during osteoclast differentiation. (A) Representative image of mitochondria. (B) The quantitative analysis of mitochondria number and cristae density. All data were presented as mean ± SD, n = 3, ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

Figure 6. ICT inhibited the mitochondrial membrane potential and reduced mitochondrial fusion and fission. (A) Representative fluorescence photographs of mitochondrial membrane potential stained by JC-1. (B) Quantitative analysis of membrane potential by measuring the ratio of red and green fluorescence intensity. (C) Representative protein bands of Parkin, DRP1, FIS1, and MFN2. (D) Quantitative analysis of Parkin, DRP1, FIS1, and MFN2. All data were presented as mean ± SD, n = 3, ^*P < 0.05; ^**P < 0.01; ^***P < 0.001; ^****P < 0.0001.

ICT mitigates OVX-induced bone loss

In our vitro study, we demonstrated that ICT administration could regulate osteoclast differentiation and osteoclastic bone resorption activity, thus we wanted to further investigate the potential therapeutic effect of ICT on osteoporosis through in vivo study. Through scanning the femur samples, the dramatic compromised trabecular bone mass was observed in OVX group, which was featured with decreased BV/TV, Tb.N, and elevated Tb.Sp, and ICT treatment reversed the bone loss induced by estrogen deficiency (Figure 7A, 7B). In the bone section, the results of TRAP staining indicated that the number of osteoclasts normalized to the bone surface (N.Oc/BS), and surface area of osteoclast to bone surface area (OcS/BS) were increased in distal femur and the osteoclast formation in vivo was inhibited when OVX mice were administrated with ICT (Figure 8A, 8B). According to the data from in vivo study, we concluded ICT ameliorates OVX-induced bone loss via inhibiting osteoclasts formation.

Figure 7. ICT suppresses OVX-induced bone loss. Twelve-week-old female mice were randomly arranged into three groups, that is Sham (Sham + Vehicle), OVX (OVX + Vehicle), and ICT (OVX + ICT). After being subjected to sham or OVX surgery, mice were treated with vehicle or ICT for 6 weeks. (A) Representative 3D reconstruction micro-CT images of the trabecular bone from each group. (B) The measurement data of BV/TV, Tb.Th, Tb.N, and Tb.Sp from each group through built-in software of CT system. All data were presented as mean ± SD, n = 6. Abbreviation: ns: not statistical significance. ^*P < 0.05; ^**P < 0.01; ^****P < 0.0001.

Figure 8. ICT inhibits osteoclast formation in vivo. (A) Representative TRAP staining images of femurs from each group. (B) The quantitative histomorphometric analysis of TRAP-positive osteoclasts, N.Oc/BS and OcS/BS. All data were presented as mean ± SD, n = 6, ^**P < 0.01; ^****P < 0.0001.