A Novel Prenylflavonoid Icariside I Ameliorates Estrogen Deficiency-Induced Osteoporosis via Simultaneous Regulation of Osteoblast and Osteoclast Differentiation

Abstract

Regulation of osteoblast-mediated bone formation and osteoclast-mediated bone resorption is crucial for bone health. Currently, most clinical drugs for osteoporosis treatment such as bisphosphonates are commonly used to inhibit bone resorption but unable to promote bone formation. In this study, we discovered for the first time that icariside I (GH01), a novel prenylflavonoid isolated from Epimedium, can effectively ameliorate estrogen deficiency-induced osteoporosis with enhancement of trabecular and cortical bone in an ovariectomy (OVX) mouse model. Mechanistically, our in vitro results showed that GH01 repressed osteoclast differentiation and resorption through inhibition of RANKL-induced TRAF6-MAPK-p38-NFATc1 cascade. Simultaneously, we also found that GH01 dose-dependently promoted osteoblast differentiation and formation by inhibiting adipogenesis and accelerating energy metabolism of osteoblasts. In addition, both in vitro and in vivo studies also suggested that GH01 is not only a non-toxic natural small molecule but also beneficial for restoration of liver injury in OVX mice. These results demonstrated that GH01 has great potential for osteoporosis treatment by simultaneous regulation of osteoblast and osteoclast differentiation.

Introduction

It is well known that maintaining bone health depends on the balance between osteoblast-mediated bone formation and osteoclast-mediated bone resorption.^1,2 Osteoporosis is an aging-related skeletal degenerative disease accompanied by bone loss and microarchitectural deterioration.³⁻⁵ Typically, postmenopausal osteoporosis in women aged above 55 is orchestrated by estrogen deficiency and an imbalance of bone turnover favoring bone resorption by osteoclasts over bone formation by osteoblasts, leading to bone loss and fractures.⁶ Mature multinucleated osteoclasts are originated from hematopoietic lineage and fuse induced by the macrophage colony-stimulating factor (M-CSF) and receptor activator of the NF-κB ligand (RANKL).⁷⁻⁹ When the RANKL binds to the receptor activator of NF-κB (RANK), it induces osteoclastogenesis processes that facilitate the formation of osteoclasts.¹⁰ Then, RANK recruits the tumor necrosis factor (TNF) receptor-associated factor (TRAF6) and activates mitogen-activated protein kinase (MAPK) signaling that controls osteoclast formation including c-Fos, nuclear factor of activated T cells (NFATc1), and activator protein-1 (AP-1).¹¹⁻¹³ NFATc1 is the key transcription factor that regulates numerous osteoclast-specific genes including cathepsin K (Ctsk),¹⁴ tartrate-resistant acid phosphatase (Trap),¹⁵ and matrix metallopeptidase-9 (Mmp9) in vitro and in vivo.¹⁶ Meanwhile, mesenchymal stromal cells and pre-osteoblasts are involved in the differentiation process of osteoblasts,¹⁷ commonly characterized by upregulation expression of osteoblast-specific transcription factors such as Runt-related transcription factor 2 (Runx2) and Osterix (Osx) and secretion of protein markers like alkaline phosphatase (ALP), osteocalcin (OCN), and osteopontin (OPN).¹¹ Mature osteoblasts synthesize the bone matrix and coordinate the mineralization of the skeleton.¹In vivo, the N-terminal propeptide of type I procollagen (PINP) and C-telopeptide of type I collagen (CTX-I) are biomarkers of bone formation and resorption, respectively, which are the bone turnover indicators that are recommended for clinical use.²

An imbalance induced by the number of osteoclast elevation and/or osteoblast reduction causes bone loss; therefore, pharmacological interventions have received considerable attention for treatment of pathological bone diseases such as osteoporosis and arthritis. To date, bisphosphonates including alendronate, risedronate, zoledronate, and ibandronate are the most widely used clinical drugs for osteoporosis treatment because they can effectively lower the risk of nonspine fractures.¹⁸ Clinical trials show that peptides as valid therapeutic agents can be applied to repair bone defects.¹⁹ For example, the parathyroid hormone 1–34 peptide (teriparatide) was reported to repress bone resorption by inhibiting NF-κB activity and promote the proliferation and differentiation of osteoblasts to increase bone formation.²⁰ However, taking bisphosphonates orally or subcutaneous injection of teriparatide with high dosage may induce undesirable side-effects such as stomach upset¹⁸ and increased risk of osteosarcoma.²¹ The development of new therapeutic agents is in high demand for prevention and treatment of osteoporosis by regulating the differentiation of osteoblasts and osteoclasts.

Previous studies showed that Epimedium-derived prenylflavonoids can be used to treat osteoporosis, cardiovascular diseases, sexual dysfunction, and menstrual irregularities.²²⁻²⁴ Icariin (Figure 1A), a main component of Epimedium, has been officially allowed to treat osteoporosis by enhancing osteogenic differentiation and mineralization of human MSC via activation of c-Jun N-terminal kinase (JNK) and P38 pathways.^25,26 However, its clinical use is daily requirement with a high dosage for a long-term clinical administration period, leading to liver and digestion problems.^27,28 Furthermore, its clinical application is limited by the low oral bioavailability, attributing to the poor absorption in the intestine.²⁷ Increasing evidence suggests that icariside I (GH01) (Figure 1B) and icartin (Figure 1C), two derivatives of icariin with a similar molecular structure, also exhibit biological activity of “bone-strengthening”.²⁹ Especially, they were found to enhance host systemic immunity for cancer therapy with no toxicity and safe for mouses.³⁰ However, less work has been performed on the molecular mechanisms responsible for biological activities against anti-osteoporosis probably due to their extremely low content in plants.

Figure 1.

Chemical structures of icariin and its derivatives isolated from Herba Epimedii. (A) Icariin, (B) icariside I, and (C) icaritin.

In this study, we discovered a novel prenylflavonoid GH01 that effectively ameliorates estrogen deficiency-induced osteoporosis in vivo. We further revealed the mechanisms by which GH01 inhibits osteoclast differentiation and resorption and concomitantly promotes osteoblast differentiation and formation in vitro. These results demonstrated that GH01 has strong anti-osteoporosis activity by maintaining the balance of bone turnover and may be developed as a potential drug for treatment of osteoporosis.

Materials and Methods

Reagents

GH01 was provided by Guangdong Golden Health Biotechnology Co., Ltd (Guangdong, China), and in vivo safety evaluation of GH01 (95%) was previously finished. For cell culture, GH01 was dissolved in dimethyl sulfoxide (DMSO) (Sigma, D2650) to make 50 mM stock solution. The corresponding concentration of GH01 was diluted by cell culture medium, and DMSO concentration was below 0.1%. For the animal experiment, GH01 was dissolved in corn oil (vehicle), providing final dosages of 5 and 50 mg/kg body weight for low (L) and high (H), respectively.

OVX-Induced Osteoporosis Mouse Model

All animal experiments were reviewed and approved by the animal ethics committee of the Innovation Academy for Precision Measurement Science and Technology (APM No: APM20029, China). A total of 32 C57BL/6 12 week-old female mice (Charles River Co. Ltd, Beijing, China, around 20 g) were housed under specific pathogen-free conditions with a 12 h light cycle and provided sterilized food and autoclaved water ad libitum at the Wuhan Institute of Virology (Hubei, China). All mice were randomly divided into four groups (n = 8): sham, ovariectomy (OVX), OVX and low dose of GH01 (L), and OVX and high dose of GH01 (H) groups. After acclimation for a week, bilateral OVX or sham operation was used to obtain bone loss in the mouse model under the condition of tribromoethanol anesthesia. After OVX for 8 weeks, OVX mice were given GH01 (5 and 50 mg/kg body weight) and vehicle (corn oil) for 6 days per week by gavage. Following a 4 week treatment of GH01, mice were sacrificed under isoflurane anesthesia after 8 h of fasting. We collected tissue samples immediately and stored them at −80 °C for the following experiments. Serum was used for detecting the biomarker of PINP and CTX-1 (n = 6) by ELISA kits. A part of liver tissues was fixed in 4% paraformaldehyde for 48 h and then dehydrated and embedded in paraffin wax. Then, the specimens were sectioned at 4 μm and stained with H&E and oil-red-O stain.

Cell Culture

Cells were cultured in a humidified incubator containing 5% CO₂ at 37 °C. Primary bone marrow macrophages (BMMs) were flushed from femurs and tibias of 8 week-old C57BL/6 mice and cultured in DMEM medium (8121258; Gibco) with 10% fetal bovine serum (FBS) (10099141; Gibco) and 1% penicillin–streptomycin (15140-122; Gibco). Suspending cells were collected overnight and treated with 25 ng/mL M-CSF (511112-MNAH-20; Sino Biological) for 3 days. For induction of osteoclastogenesis, osteoclast precursors were treated with 50 ng/mL mouse RANKL (462-TEC-010; R&D systems) for up to 7 days. Osteoclasts with RANKL stimulation can be divided into three stages, including early (1–2 days), intermediate (3–4 days), and late (5–6 days) stages.

Primary osteoblastic cells were isolated from newborn C57BL/6 mouse calvarias.^31,32 Briefly, calvarias were dissected aseptically, adherent soft connective tissue was removed, and they were immersed in cold phosphate-buffered saline (PBS) (20012027; Gibco) and washed twice. Then calvarias were cut into pieces in a 1.5 mL EP tube with trypsin–EDTA (2520007; Gibco) for 20 min. After centrifuging for 5 min at 1000 rpm, the obtained supernatants were discarded. Sequentially, calvarias were digested in α-MEM containing 125 unit/mg collagenase II (C8150; Solarbio) for 60 min. After centrifuging for 5 min at 1000 rpm, the precipitate was resuspended by the growth medium. The growth medium was provided with MEM Alpha basic (1×) (41500-034; Gibco) medium and 10% FBS and 1% penicillin–streptomycin. For osteoblastogenesis experiments, the growth medium was added with ascorbic acid (A4403; Sigma) and β-glycerophosphate (G9422; Sigma).

Cell Viability Assay

The viability assay was performed with Cell Counting Kit-8 (CK-04; Dojindo). The primary BMMs and primary osteoblastic cells (5 × 10³ cells/well) were seeded in 96-well plates overnight. The cells were then replaced with fresh medium containing the appropriate concentration of GH01 (0, 0.1, 10, 100, and 1000 nM), and the medium was changed every other day. At the endpoint, we added 10 μL of the CCK-8 reagent to each well, and 96-well plates were incubated at 37 °C for 2 h. A microplate reader (K3; Labserv) was used to read the absorbance at 450 nm after shaking for 10 s.

TRAP Staining for Osteoclasts

Cells were fixed in 4% paraformaldehyde solution and stained with a TRAP kit (G1492; Solarbio) followed by the manufacturers’ instructions. Images were taken using an Olympus BX51 microscope and analyzed using Image J software (1.52v; USA). TRAP⁺ multinucleated cells (>3 nuclei) were counted as osteoclasts.

Resorption Pit Assay

For assessment of osteoclast function, specifically for pit formation, primary BMMs were cultured with 50 ng/mL RANKL combined with 25 ng/mL M-CSF in six-well plates. As osteoclasts formed, cells were detached and seeded as 8000 cells/well on an Osteo Assay Surface plate (3988; Corning) with or without GH01 treatment. Following incubation for 7 days with changing differentiation solution every 3 days, osteoclasts were gently bleached and discarded. The plates were allowed to air-dry completely at room temperature before being captured on the microscope. The area of pits was measured using Image J software (1.52v; USA).

RNA Isolation and Quantitative Real-Time PCR Analysis of Gene Expression

Total RNA from osteoclasts and osteoblasts was extracted using the RNAiso Plus reagent (9109; TaKaRa) following the standard procedure. cDNA was generated from RNA using the PrimeScriptRT Master Mix (RR036A; TaKaRa) or reverse transcription kit (AT341-02; TransGen). Then, we performed real-time PCR by applying PowerUp SYBR Green Master Mix (A25742; Thermo Fisher) on the ABI StepOne real-time PCR system (Applied Biosystems). The PCR cycling parameters were set by the instrument’s own system. The relative expression was measured by the comparative 2^–ΔΔCT method. The primers of genes used for PCR are shown in Supporting Information, Table S1.

Western Blotting

RIPA buffer (Beyotime Biotechnology) containing protease inhibitors was used to obtain total protein at the corresponding time. A BCA (bicinchoninic acid) protein assay kit (C603021; Sangon Biotech) was used to detect protein concentration. We then separated lysates on the appropriate percentage of sodium dodecyl sulfate polyacrylamide gel and transferred them to the polyvinylidene fluoride membrane (ISEQ00010; Millipore) by wet transfer systems (1645050; Bio-Rad). Then, membranes were blocked with QuickBlock blocking buffer (P0252; Beyotime) for 15 min and then immersed in primary antibodies overnight. The membrane was then incubated with a suitable secondary antibody the following day. A ChemiDoc imager (Bio-Rad) was used for visualization of the membranes with exposure by the enhanced-chemiluminescence HRP substrate (P90719; Millipore). Primary antibodies used were as follows: phospho-JNK (1:1000; 4668; cell signaling), JNK (1:1000; 9252; cell signaling), phospho-p38 (1:1000; 4511; cell signaling), p38 (1:1000; 8690; cell signaling), phospho-extracellular signal-regulated kinase (Erk) (1:2000; 4370; cell signaling), Erk (1:1000; 4695; cell signaling), NFATc1 (1:200; sc-7294; Santa Cruz), TRAP (1:1000; ab185716), c-FOS (1:2000; 66590-1-Ig; Proteintech), RUNX2 (1:1000; ab23981; Abcam), OCN (1:1000; A18241; ABclonal),OPN (1:1000; ab8448; Abcam), TRAF6 (1:5000; 66498-1-Ig; Proteintech), β-Tubulin (1:1000; 10094-1-AP; Proteintech), GAPDH (1:50000; 60004-1-Ig; Proteintech), β-ACTIN (1:50000; AC026; ABclonal), and secondary anti-mouse/rabbit HRP-conjugated antibodies (SA00001-1 and SA00001-2; Proteintech) were subsequently applied.

Immunofluorescence Staining

For immunofluorescence staining, primary BMMs were seeded in 35 mm glass bottom cell culture dishes (801002; Nest) at a concentration of 8 × 10³/well, and the osteoclast precursors were treated with 50 ng/mL RANKL (462-TEC-010; R&D systems) and 25 ng/mL M-CSF (511112-MNAH-20; Sino Biological) in the presence or absence of GH01. After 5 days of stimulation, cells were fixed with 4% paraformaldehyde (G1101; Servicebio) for 20 min. Next, cells were permeabilized with 0.5% Triton X-100 in PBS, blocked with 5% goat serum, and then incubated with NFATc1 (1:200; sc-7294; Santa Cruz) and FITC-F-actin (1:100; CA1620; Solarbio) overnight at 4 °C. Samples were incubated with appropriate species-specific Alexa Fluor 488 antibodies (1:400; A-10680; ThermoFisher). Samples were imaged with FluoView FV1000 (Olympus) after adding DAPI (C0065; Solarbio).

ALP Activity ELISA

ALP activity was measured according to standard techniques by the ELISA ALP assay kit (A059-2-2; Nanjing Jiancheng Bioengineering Institute). Osteoblast cells were collected after osteoblastogenesis differentiation for 7 days. Then, cells were extracted by ultrasonically aided method.

ALP Staining and Alizarin Red Staining

ALP staining and Alizarin red staining were performed on primary osteoblastic cells cultured in osteoblastogenesis medium for 7 and 14 days, respectively. After being fixed with 4% paraformaldehyde solution and washed with PBS, cell staining was carried out according to the BCIP/NBT ALP color development kit (C3206; Beyotime) and Alizarin Red S solution (G1452; Solarbio). The cells were then rinsed with distilled water to stop the reaction. Images were obtained using the scanner (LIDE220; Canon) and microscope (BX51; Olympus).

¹H NMR-Based Cellular Metabolomics

Primary osteoblastic cells were cultured for 7 days in osteoblastogenesis medium and collected (around 100 mg). Metabolic profiles of osteoblast extraction were obtained from ¹H NMR spectra, and metabolites were assigned with a series of two-dimensional (2D) nuclear magnetic resonance (NMR) experiments. Comparative O-PLS-DA was performed between normalized NMR data from the GH01-treated (0.1 and 100 nM) and control cells at matched time points for osteoblast extraction. In the model, the values of R² and Q² represent the quality of fit and predictability, respectively, and were all qualified. The O-PLS-DA models were further validated by CV-ANOVA values (p < 0.05), indicating a significant difference between GH01-treated groups and controls in the ¹H NMR profiles of osteoblast extraction. The sample preparation, NMR spectral acquisition (Bruker AVANCE III 600 MHz spectrometer) and processing, and multivariate data analysis were performed as described previously.³³

μCT Analysis

Following mouse sacrifice, left femurs were obtained, fixed with 4% paraformaldehyde for 48 h, and stored in ethanol (75%). μCT analysis was conducted on a Skyscan (1276; Bruker) at 6 μm resolution to obtain trabecular and cortical bone morphometry. Scanning was carried out by applying 55 kV voltage and 200 μA source current with an angular increase of 0.3°. Pixel size was 6 μm. In the femoral trabecular region, we began with 50 slices below the distal growth plate and extended for 100 slices proximally. The bone volume/tissue volume (BV/TV, %), bone mineral density (BMD) (mg HA/mm³), trabecular number (Tb.N, mm^–1), trabecular thickness (Tb.Th, mm), and trabecular separation (Tb.Sp, mm) were measured by the CT analyzer program (Bruker). For the cortical trabecular region, we chose 100 slices in the middle of the whole femur. Total cross-sectional area (Tt.Ar, mm²), cortical bone area (Ct.Ar, mm²), Ct.Ar fraction (Ct.Ar/Tt.Ar, %), and cortical thickness (Ct.Th, mm) were measured. CTvol and Data-viewer were used to obtain images in 2D and 3D (Bruker Belgium), respectively.

Statistical Analysis

All data and statistical analysis were shown as mean ± S.D. using Graph-Pad Prism 8.0 software. The P values were determined by one-way ANOVA with multiple comparisons, and two-tailed Student’s t-test was performed between two groups. p < 0.05 was considered as statistical significance.

Results

GH01 Represses Osteoclast Differentiation and Resorption In Vitro by Suppressing MAPK-p38-NFATc1 Cascade

To explore the effect of GH01 on osteoclastogenesis, we assessed osteoclast differentiation using primary BMMs with the M-CSF and RANKL. We found that GH01 treatments at different doses (0.1, 1, 10, and 100 nM) significantly inhibited osteoclast differentiation (Figure 2A), shown with marked reduction of the number of TRAP-positive multinucleated osteoclasts (>3 nuclei; corresponding p value for 0.0044, 0.0023, 0.0746, and 0.0588) (Figure 2B). TRAP staining further showed that GH01 exposure at a dose of 100 nM strongly inhibited osteoclastogenesis, manifested by marked reduction of the number and size of TRAP-positive multinucleated osteoclasts with RANKL stimulation in all stages (p < 0.001) including early (1–2 days), intermediate (3–4 days), and late (5–6 days) stages (Figure 2C,D). In addition, no significant cytotoxicity was observed in RANKL-induced BMMs exposed to GH01 at different doses (Figure S1A).

Figure 2.

GH01 inhibits RANKL-induced osteoclast differentiation in vitro. (A) Primary BM macrophages were incubated with the M-CSF (25 ng/mL) and RANKL (50 ng/mL) and treated with GH01 at different doses (0.1, 1.0, 10, and 100 nM). Scale bar = 200 μm. (B, D) Number of TRAP-positive (purple, cytoplasm) multinucleated osteoclasts in each well. The findings showed three independent experiments. (C) Primary BM macrophages were cultured with 100 nM GH01 on day 1–2, 2–4, or 4–6. Mature osteoclasts were assessed via TRAP staining (>3 nuclei). Data are shown as mean ± SD. P values were obtained by one-way ANOVA with multiple comparisons, *p < 0.05, **p < 0.01, and ***p < 0.001.

To determine the effects of GH01 on resorption capability of mature osteoclasts, osteoclast precursors were incubated on calcium hydroxyapatite-coated 96-well plates with and without GH01 treatment at dosages of 0.1 and 100 nM. We found that GH01 significantly repressed RANKL-induced resorption capability of osteoclasts in a dose-dependent manner with p value below 0.05 (Figures 3A and S1B). Administration of GH01 significantly decreased RANKL-induced TRAF6 levels (Figure 3D) and the RANK adaptor protein, thereby inhibiting the RANKL/RANK-mediated MAPK signaling pathway. Given that osteoclast differentiation is mediated by the phosphorylation of JNK, p38, and ERK downstream of MAPK and TRAF6 pathways, we further evaluated the time-dependent effects of GH01 on the MAPK signaling pathway. Continuous decreases were observed in the protein level of phosphorylation of p38 (p-p38) from 5 to 15 min (Figure 3E), whereas the proteins levels of JNK, P-JNK, ERK 1/2, and P-ERK 1/2 remained virtually unchanged at the corresponding time points (Figure 3E). Inhibition of osteoclast formation by GH01 was further confirmed by markedly decreased mRNA levels of corresponding osteoclast specific genes such as Ctsk, Mmp9, and Nfatc1 (p < 0.05) in RANKL-induced BMMs (Figure 3C). Consistently, osteoclast specific proteins such as NFATc1, c-FOS, and TRAP were also substantially decreased in RANKL-stimulated BMMs after GH01 treatment (Figure 3B,F).

Figure 3.

GH01 represses osteoclast differentiation and resorption in vitro by suppressing MAPK-p38-NAFTc1 cascade. (A) Primary BMMs differentiated on hydroxyapatite-coated wells (in white). Scale bar = 200 μm. (B,D) Western blotting of NFATc1, c-FOS, TRAP, and TRAF6 in osteoclasts treated with GH01 at 0.1 and 100 nM. (C) mRNA expression level of NFATc1, Mmp9, and Ctsk in osteoclasts treated with GH01 at 0.1 and 100 nM. (E) Western blotting of RANKL-induced phosphorylation of p38 involved in the MAPK pathway of osteoclasts treated with GH01 for 0, 5, 15, and 30 min. (F) Immunofluorescence staining for NFATc1 (red), F-actin (green), and DAPI (blue). Scale bar = 50 μm. Data are shown as mean ± SD. P values were obtained by one-way ANOVA with multiple comparisons, *p < 0.05, **p < 0.01, and ***p < 0.001.

GH01 Promotes Osteoblast Differentiation and Formation In Vitro by Accelerating Cell Metabolism

New bone formation and remodeling play pivotal roles in bone homeostasis that is mediated by osteoblast differentiation and formation. First, no significant cytotoxicity was observed in osteoblasts after GH01 exposure for 3 and 5 days at different dosages (0.1, 1, 10, 100, and 1000 nmol/L) (Figure 4A). ALP staining and ALP activity measurement showed that GH01 (0.1, 1, 10, 100, and 1000 nM) markedly promoted differentiation and formation of osteoblasts (p value 0.0585, <0.01, and <0.001) (Figure 4B,D), which were further verified by Alizarin red staining, suggesting significant formation of mineralized nodules after GH01 treatment (Figure 4D). Concomitantly, GH01 (0.1 and 100 nM) markedly upregulated downstream signal factors such as RUNX2 (p < 0.01) and OCN (p < 0.05) at both mRNA and protein levels involved in osteoblastogenesis (Figure 4C,E).

Figure 4.

GH01 promotes osteoblast differentiation and formation in vitro. (A) Cell viability of primary osteoblasts incubated for 3 and 5 days with different dosages of GH01. (B) ALP activity of osteoblasts assessed with different dosages of GH01 for 7 days. (C) mRNA expression level of Runx2 and Ocn in osteoblasts treated with GH01 at 0.1 and 100 nM. (D) ALP staining and Alizarin red staining for the primary osteoblasts treated with 0.1, 1, 10, 100, and 1000 nM GH01 for 7 and 14 days (scale bar = 500 μm). (E) Western blotting of RUNX2, OPN, and OCN in osteoblasts treated with GH01 at 0.1 and 100 nM at day 14. Data are shown as mean ± SD. P values were obtained by one-way ANOVA with multiple comparisons, *p < 0.05, **p < 0.01, and ***p < 0.001.

Supportive evidence of osteoblastogenesis promoted by GH01 treatment could also be found in the accelerated energy metabolism of primary calvarial osteoblasts with GH01 administration (0.1 and 100 nM) (Figure 5). Compared with the control group, GH01 (0.1 nM) significantly upregulated the levels of choline, glutamate, and glutamine and downregulated the levels of lactate and uridine (Figure 5A). A relatively high dose of GH01 (100 nM) administration induced significant elevation in the levels of glutamate, glutamine, succinate, choline, adenosine monophosphate (AMP), adenosine diphosphate (ADP), inosine, adenosine, hypoxanthine, and UDP-GlcNAc and a decrease in the levels of some amino acids including valine, isoleucine, leucine, alanine, lysine, phenylalanine, and tyrosine in primary calvarial osteoblast extraction (Figure 5B).

Figure 5.

GH01 accelerates osteoblastic metabolism. (A,B) OPLS-DA scores (left) and coefficient loading plots (right) from ¹H NMR spectra of osteoblasts in the control, GH01 (0.1 nM), and GH01 (100 nM) groups. Abbreviations: valine (Val); isoleucine (Isoleu); leucine (Leu); alanine (Ala); lysine (Lys); glutamate (Glu); glutamine (Gln); choline (Cho); guanosine triphosphate (GTP); UDP-N-acetylglucosamine (UDP-GlcNAc); adenosine monophosphate (AMP); tyrosine (Tyr); phenylalanine (Phe); and adenosine diphosphate (ADP).

GH01 Ameliorates Estrogen Deficiency-Induced Osteoporosis In Vivo Without Significant Hepatotoxicity

Given that GH01 can effectively regulate osteoblast and osteoclast differentiation in vitro, we then examined whether GH01 could ameliorate estrogen deficiency-induced osteoporosis in vivo using an OVX-induced osteoporosis mouse model. As shown in Figure 6A, mice were OVX or sham operated for 8 weeks and then treated with GH01 (5 and 50 mg/kg body weight) and vehicle (corn oil) by gavage daily for 6 days per week over 4 weeks. OVX surgery induced a significant increase in the average body weight of mice, which were not markedly affected by GH01 treatments (Figure 6B, p < 0.001). Furthermore, OVX mice exhibited significant liver injury with inflammatory infiltration, vacuolated hepatocytes, and lipid droplets, shown with histopathological H&E staining (Figure 6D) and oil-red-O staining (Figure 6E). However, there was no significant change seen in the weight of livers (Figure 6C), and GH01 treatments for 4 weeks at different doses markedly improved such OVX-induced liver injury (Figure 6D,E).

Figure 6.

GH01 ameliorates estrogen deficiency-induced osteoporosis in vivo without significant hepatotoxicity. (A) Experimental design for treatment in the study. (B) Body weight changes during the experiment. (C) Liver weight of mice (n = 6). (D,E) Histopathological assessment of H&E-stained and oil-red-O-stained in liver sections, respectively (scale bars = 50 μm). Data are shown as mean ± SD. P values were obtained by one-way ANOVA with multiple comparisons, *p < 0.05, **p < 0.01, and ***p < 0.001.

Micro-computed tomographic (μCT) images and three-dimensional (3D) reconstruction of trabecular and cortical bone showed that OVX induced striking reduction of BMD (p < 0.05), BV/TV ratio (p < 0.05), Tb.N (p = 0.088), cortical thickness (Ct.Th, p < 0.01), and cortical bone fraction (Ct.Ar/Tt.Ar, p < 0.01) accompanied by significant upregulation of Tb.Sp (p = 0.141) in mice (Figure 7A,B). In addition, significant elevation of P1NP (Figure S2A) and CTX (Figure S2B, p < 0.05), respective markers of bone formation and resorption, was also found in serum of OVX mice. Here, oral administration of GH01 at different doses (5 and 50 mg/kg body weight) for 4 weeks ameliorated OVX-induced osteoporosis to some extent, manifested by marked restoration of the abovementioned parameters including elevation of BMD (p = 0.1548 and p = 0.067), BV/TV (p = 0.1247 and p = 0.1365), Tb.N (p = 0.1733 and p = 0.1565), Ct.Th (p = 0.0548), Ct.Ar/Tt.Ar (p < 0.01), and PINP (p < 0.05) together with downregulation of Tb.Sp (p < 0.01) and CTX (p < 0.05, p < 0.01) in OVX mice (Figures 7 and S2).

Figure 7.

GH01 ameliorates OVX-induced bone loss in vivo. (A) μCT imaging and 3D reconstruction of trabecular and cortical bone. Scale bar = 200 μm. (B) Quantitative analyses of 3D parameters for trabecular and cortical bone microarchitecture, including BMD, BV/TV, Tb.N, Tb.Sp, Tb.Th, Ct.Th, and Ct.Ar/Tt. Ar (n = 6). Data are shown as mean ± SD. P values were obtained by one-way ANOVA with multiple comparisons, *p < 0.05, **p < 0.01, and ***p < 0.001.

Discussion

The process of osteoblast formation and osteoclast resorption maintains the normal architecture of bones.¹ Osteoblasts originate from mesenchymal cells, synthesize the bone matrix, and make new bone formation, whereas osteoclasts stem from macrophages and are responsible for bone resorption and releasing the mineral matrix.¹⁷ The delicate balance between osteoclasts and osteoblasts maintains the bone homeostasis. However, as the age increased, especially menopause in women, osteoclasts become more active and gradually lead to microarchitectural deterioration of bone.³⁴ Nowadays, bisphosphonate, peptides, and oestrogen have been comprehensively used for osteoporosis treatment clinically. Unfortunately, they more or less can cause severe adverse effects like stomach upset and even breast cancer after high dosage oral administration for a long term.^21,35 In this study, we discovered and demonstrated that GH01, a novel non-toxic prenylflavonoid, could inhibit osteoclastogenesis and promote osteoblasts formation in vitro simultaneously and thus effectively ameliorate estrogen deficiency-induced osteoporosis in vivo.

One of the most striking findings in this study was the significant inhibition of osteoclast formation and bone resorption by GH01 administration. Upon RANKL binding with RANK, downstream NF-kB, MAPK/AP-1, and ROS signaling pathways are subsequently activated in vitro. Within the MAPK family, there are three main subgroups including ERK, p38, and JNK signaling pathways.³⁶ Phosphorylated JNK and p38 have been shown to highly associate with osteoclastogenesis, while ERK is critical to osteoclast function and survival.³⁷⁻³⁹ Interestingly, our prenylflavonoid, GH01, selectively suppressed the MAPK-p38 pathway without affecting JNK and ERK signaling, suggesting that GH01 attenuated osteoclast differentiation but not survival. Following RANKL stimulation, GH01 treatment consequently repressed NFATc1 expression at both mRNA and protein levels, the master transcription factor for osteoclast differentiation and formation located downstream of the MAPK-p38 signaling pathway. Eventually, downstream osteoclastogenic genes such as Ctsk and Mmp9, which are all mediated by NFATc1 directly, were also markedly inhibited in osteoclasts by GH01. In addition to GH01, some prenylflavonoid compounds such as icaritin,²⁹ icarisoside A,⁴⁰ naringenin,⁴¹ and xanthohumol⁴² present in various plants have also been shown to prevent or treat osteoporosis and osteoclastogenesis closely related to the MAPK-NFATc1 pathway. Actually, NFATc1 deficiency results in the failure to form osteoclasts from primary BMMs, ultimately leading to increased bone mass and inhibited osteoclastogenesis in a rodent model.^43,44 Mechanistically, these results suggested that GH01 dose-dependently repressed osteoclast differentiation and resorption through inhibition of RANKL-induced TRAF6-MAPK-p38-NFATc1 cascade.

Another prominent finding in this study was the marked promotion of osteoblast differentiation and formation by GH01 treatment. At the cellular level, GH01 dose-dependently increased activity of ALP, the biomarker of the early stage of osteogenic differentiation. Concomitantly, significant upregulation of osteogenic factors such as RUNX2, OCN, and OPN at both mRNA and protein levels together with calcium deposition was observed in the process of osteoblastogenesis, further suggesting that GH01 treatment promoted osteoblast differentiation and formation in vitro. At the molecular level, here global ¹H NMR-based cellular metabolomics was employed to extract the differential metabolites mainly contributing to osteoblast differentiation and formation with and without GH01 treatment. Significant upregulation in the levels of intracellular glutamate and glutamine by GH01 at both doses suggested that GH01 treatment markedly regulated osteoblast differentiation, since glutamate is secreted by osteoblasts and plays a physiological role in regulating the maturation of osteoblasts and osteoclastogenesis.⁴⁵ It has been reported that activation of both glutamate receptors for N-methyl-d-aspartate and α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid can stimulate osteoclastogenesis and significantly increase the mineralized deposition.⁴⁵ Another study suggested that glucocorticoids have potential effects on glutamine synthetase expression in human osteoblastic cells, which is a key enzyme catalyzing the conversion of glutamate and ammonia to glutamine, ultimately resulting in osteoporosis.⁴⁶ The significant reduction of cellular branched-chain amino acids (BCAAs) in GH01-treated pre-osteoblasts suggested that GH01 inhibited the differentiation process from pre-osteoblasts into adipocytes due to the high contribution of BCAAs to adipocyte metabolism in vivo and in vitro through protein catabolism.⁴⁷ In addition, marked up-regulation of nucleoside and nucleotide metabolites such as ADP, AMP, GTP, inosine, and adenosine in pre-osteoblasts treated by GH01 showed that GH01 accelerated energy metabolism of osteoblasts, which was further verified with up-regulation in the levels of succinate, an intermediate product of the tricarboxylic acid cycle involved in cell energy metabolism, in GH01-treated osteoblasts at high dosage. Collectively, these observations suggested that GH01 dose-dependently promoted osteoblast differentiation and formation through inhibition of adipogenesis and acceleration of energy metabolism of osteoblasts.

In summary, our in vitro and in vivo findings demonstrated that GH01 is a non-toxic natural small molecule that simultaneously inhibits bone resorption and promotes bone formation, maintaining the balance of bone turnover. Although the clinical application of GH01 is limited due to its relatively low bioavailability, it may be developed as a potential therapeutic candidate against osteoporosis. Some possible influencing factors such as the age of the animals and cell sources for primary osteoporosis in vitro experiments may also be the limitations of this study, which warrants further investigation to address the challenges to clinical application of GH01.

Glossary

Abbreviations

OVX

ovariectomy

GH01

icariside I

NF-κβ

nuclear factor kappa-B

RANKL

activator of NF-κβ ligand

TNF

tumor necrosis factor

M-CSF

macrophage colony-stimulating factor

TRAF6

TNF-receptor-associated factor 6

NFATc1

nuclear factor of activated T-cells

c-Fos

protooncogene C-Fos

MAPK

mitogen-activated protein kinase

ERK

extracellular signal-regulated kinase

JNK

c-Jun N-terminal kinase

CTSK

cathepsin K

TRAP

tartrate-resistant acid phosphatase

MMP9

matrix metallopeptidase-9

ALP

alkaline phosphatase

RUNX2

runt-related transcription factor 2

OCN

osteocalcin

OPN

osteopontin

BM

bone marrow

μCT

micro-computed tomography

BMD

bone mineral density

BV/TV

bone volume per tissue volume

Tb.N

trabecular number

Tb.Th

trabecular thickness

Tb.Sp

trabecular separation

Ct.Ar

cortical bone area

Ct.Ar/Tt.Ar

cortical bone area fraction

IF

immunofluorescence

DAPI

4,6-diamidino-2-phenylindole

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.2c00192.

• Primer sequences of RT-PCR, cell viability of BMM, area of erosion pit, and ELISA of serum PINP and CTX (PDF)

Author Contributions

Conceptualization: L.Z. and D.H.X.; methodology: L.Z., D.H.X., and C.C.; investigation: C.C. M.W., H.L., Z.C., F.W., Y.C.S., C.Z., M.Q, C.Z. R.D., L.Z., and Y.J.L.; writing original draft: C.C., L.Z., and D.H.X.; writing, review, and editing: C.C., J.Z., Y.J.L., L.Z., and D.H.X.; funding acquisition: L.Z., and D.H.X.; resources: L.Z., and D.H.X.; and supervision: L.Z., and D.H.X.

The authors declare no competing financial interest.