Alpha minimum essential medium (α-MEM), penicillin/streptomycin, and fetal bovine serum (FBS) were purchased from Excell (Excell Bio, Shanghai, China). Cell culture plate was purchased from In Vitro Scientific (Hangzhou Xinyou Biotechnology Co., Ltd, China). T75 flask was purchased from Jet Biofil (Guangzhou, China). JSH-23 (purity = 99.07%, batch No. S735101, CAS No. 749886–87–1) was purchased from Selleck (Houston, TX, United States). A Cell Counting Kit-8 (CCK-8) assay kit was purchased from MultiSciences (Hangzhou, China). Primary antibodies against NFATc1 (ab177464), c-Fos (ab222699), p-p65 (ab76302), p65 (ab32536), p-Akt (ab192623), Akt (ab179463), p-ERK (ab201015), ERK (ab184699), p-JNK (ab124956), JNK (ab179461), p-p38 (ab178867), p38 (ab170099), TRAF6 (ab33915), Rac1 (ab180683), NOX1 (ab131088), HO-1 (ab68477), Catalase (ab209211), NQO1 (ab80588), GSR (ab124995), Bcl-2 (ab182858), Bax (ab182733), Runx2 (ab192256), BMP2 (ab214821), and GAPDH (ab8245) were purchased from Abcam (Cambridge, MA, United States). Primary antibody against OCN (AB10911) was purchased from Sigma-Aldrich (St. Louis, MO). Primary antibody against p-IκBα (#2859), IκBα (#4814), Cleaved Caspase9 (#9507), Cleaved Caspase3 (#9664), and Nrf2 (#12721) were purchased from Cell Signaling Technology (Danvers, MA, United States). Recombinant mouse M-CSF and mouse RANKL were purchased from R&D Systems (Minneapolis, MN, United States). RNA interference sequences were purchased from GeneChem (Shanghai, China).

Bone marrow macrophages (BMMs) were used as osteoclast precursors and isolated from the bone marrow of four-week-old C57BL/6 mice that were euthanized according to procedures approved by the Fourth Military Medical University. The BMMs were grown in α-MEM supplemented with 10% FBS, 1% penicillin-streptomycin solution, 30 ng/ml M-CSF and incubated in an atmosphere of 5% CO₂ at 37°C. The culture medium was changed every 2 days.

MC3T3-E1 cells were used in our experiments as osteoblasts. The cells were cultured in α-MEM supplemented with 10% FBS and 1% penicillin-streptomycin solution in an atmosphere of 5% CO₂ at 37°C. The culture medium was changed every 2 days.

BMMs were seeded into 96-well plates (8 × 10³ cells/well) and incubated at 37°C. After 24 h, the cells were treated with various concentrations of JSH-23 (range 1–200 μM) and cultured for 48 or 96 h. Ten microliters of CCK-8 buffer was added to each well, the cells were incubated at 37°C for 1 h, and the absorbance was measured at 450 nm on an ELX800 microplate reader (Bio-Tek, Vermont).

BMMs were seeded at a density of 8 × 10³ cells/well in 96-well plates and treated in complete medium containing 30 ng/ml M-CSF and 100 ng/ml RANKL with or without different concentrations of JSH-23 (10, 20 and 40 μM). After 5 days, the cultured cells were fixed with 0.25% glutaraldehyde for 20 min at room temperature and then washed three times with PBS. The tartrate-resistant acid phosphatase (TRAP) enzymatic activity of the cells was detected using a leukocyte acid phosphatase staining kit according to the manufacturer’s procedures. TRAP-positive cells (>3 nuclei) were regarded as osteoclasts and counted under a light microscope (Nikon Corporation, Tokyo, Japan). After TRAP staining, the cells were washed three times with PBS and stained with Actistain 488 Fluorescent Phalloidin (Cytoskeleton, Denver, CO, United States) at room temperature for 30 min in the dark. Then BMMs were washed three times with PBS, and the nuclei were visualized with 1 mg/ml DAPI. Images of osteoclastic rings were obtained using an immunofluorescence microscope.

After RANKL stimulation for 3 days, equal numbers of BMM-derived preosteoclasts were seeded onto bovine bone slices and treated with JSH-23 (10, 20 and 40 μM) for another 48 h. Cells that had adhered to the bone slices were then removed by mechanical agitation and sonication. Resorption pits were visualized under an SEM (FEI Quanta 250), and the bone resorption area was quantified using ImageJ software.

Cells from different treatment groups were washed with cold PBS and lysed with RNAiso Plus (Takara Bio, Otsu, Japan) to obtain RNA according to the manufacturer’s protocol. The total RNA was then reverse-transcribed to obtain cDNA. The cDNA was used for quantitative real-time PCR, which was performed with a CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules, CA, United States). The values were normalized to the levels of GAPDH. The primer sequences (Sino Biological Inc., Beijing, China) are listed in Supplementary Table S1.

Cultured cells from different treatment groups were washed with cold PBS three times for 5 min. RIPA lysis buffer (Sangon Biotech) was used to lyse the cells for 20 min at 4 C. Then, the obtained cell lysates were centrifuged at 10,000 × g for 10 min. The obtained supernatants were collected and dissolved in loading buffer. A 10 μL mixture was separated on 10% SDS-PAGE gels, and the separated proteins were then transferred to PVDF membranes (Bio-Rad, Hercules, CA, United States). The membranes were washed with Tris-buffered saline-Tween 20 (TBST) twice for 10 min and blocked with 5% nonfat dry milk at room temperature for 1 h. Next, the membranes were incubated with primary antibodies overnight at 4°C and with secondary antibodies for 1 h at room temperature. The protein bands were detected using electrochemical luminescence reagent (Millipore, Billerica, MA, United States). The grayscale values of the bands were quantified using ImageJ software (National Institutes of Health, Bethesda, MD, United States).

2′,7′-Dichlorodihydrofluorescein diacetate (DCFH-DA) staining was performed to measure ROS production in vitro. Cells were fixed in 4% formaldehyde for 15 min and then washed with PBS. DCFH-DA (10 μM) was added to the wells, and the cells were incubated for 30 min at 37°C. The cells were washed twice with PBS, and fluorescence images were obtained. The fluorescence intensity was quantified.

In vivo ROS production was assessed by injection of dihydroethidium (DHE) into the calvaria 24 h prior to killing, according to a protocol previously reported (Ni et al., 2020).

Cells were washed with PBS three times, fixed in 4% paraformaldehyde for 30 min at room temperature, blocked with 5% (w/v) BSA in PBST, and immunostained with primary antibodies overnight at 4°C followed by secondary antibodies. After being washed with PBS three times, the cells were stained with DAPI and observed under a fluorescence microscope.

BMMs or MC3T3-E1 cells were transfected with siRNA using Lipofectamine 3,000 (Invitrogen) according to the manufacturer’s protocol. Briefly, cells were seeded in 6-well plates at a density of 2 × 10⁴ cells/well and transfected with 20 nM siRNA. After 6 h, the medium was replaced with complete α-MEM medium, and the cells were cultured for another 48 h. The efficiency of transfection was observed under a fluorescence microscope. Cells were analyzed by western blotting to assess the expression of nuclear factor E2-related factor 2 (Nrf2) and heme oxygenase 1 (HO-1).

After different treatments, MC3T3-E1 cells were washed with PBS and then labeled with FITC-Annexin V and PI in binding buffer according to the manufacturer’s instructions. Ten thousand events were recorded and analyzed to obtain the percentage of apoptotic cells using a FACScan flow cytometer (BD Bioscience, San Jose, CA, United States).

ALP activity is expressed as a percentage of enzyme activity relative to the control value. Alizarin red (AR) staining was performed to measure the degree of mineralization. MC3T3-E1 cells were pretreated with JSH-23 for 24 h before induction of differentiation for seven or 21 days and were then exposed to 400 μM H₂O₂ for 4 h for detection of ALP and AR staining, respectively. The immobilized cells were treated with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (BCIP/NBT, 7 days) or with 2% AR solution (21 days) for 30 min at room temperature. Then, the cells were washed three times with PBS and observed under a light microscope.

The animal experiments were approved by The Fourth Military Medical University. Twenty-four six-week-old female C57BL/6 mice were equally divided into four groups: the sham group (treated with PBS), the vehicle group (treated with LPS), the low group (treated with LPS and 1 mg/kg JSH-23), and the high group (treated with LPS and 3 mg/kg JSH-23). The doses of JSH-23 used in the animal experiments was determined based on previous reports (Kumar et al., 2011; Wang et al., 2018). The mice were subcutaneously injected at the calvarial center with PBS or JSH-23 24 h before injection of LPS. Seven days after LPS injection, all mice were sacrificed, and the calvaria were dissected and fixed in 4% paraformaldehyde for micro-computed tomography (CT) and histological analysis.

Three-dimensional images of the whole calvaria were reconstructed using a high-resolution micro-CT scanner (Skyscan 1,176; Skyscan; Aartselaar, Belgium). Images were acquired using a 50 kV X-ray tube voltage, an 800 μA current, an isotropic pixel size of 14.4 μM (1,024 × 1,024 pixel image matrix) and a 0.75 mm thick aluminum filter for beam hardening. A square region of interest around the midline suture was chosen for further qualitative and quantitative analysis after reconstruction.

The specimens were decalcified with 10% EDTA solution for 3 weeks and then embedded in paraffin. Histological sections were prepared and stained for hematoxylin and eosin (H&E) staining and TRAP activity assessment. The sections were then photographed under a microscope. Bone histomorphometric analyses were performed using BIOQUANT OSTEO histomorphometry software (BIOQUANT Image Analysis Corporation, Nashville, TN).

Immunofluorescence was performed to analyze Runx2, OCN, Nrf2 and HO-1 expression in vivo. Briefly, paraffin-embedded sections were deparaffinized and incubated in 1 mM EDTA (pH = 8.0) at 80°C for 15 min to retrieve the antigens. Then, the samples were blocked with 10% normal goat serum and incubated with primary antibodies. Secondary antibodies and DAPI were further applied, and the samples were observed under a fluorescence microscope. The fluorescence signal intensity was analyzed using ImageJ.

All experimental data are presented as the mean ± SD obtained from three experiments, and statistical significance was determined by Student’s t-test and one-way ANOVA. p < 0.05 was considered to indicate significance.

The chemical structure of JSH-23 is shown in Figure 1A. To assess the cytotoxicity of the compound, BMMs were exposed to varying doses of JSH-23 for 48 and 96 h. The CCK-8 assay showed that JSH-23 exerted no detectable toxic effects at concentrations under 50 μM (Figure 1B). Therefore, we chose concentrations of 10, 20 and 40 μM JSH-23 for further study. TRAP staining showed that JSH-23 suppressed the total number of cells undergoing osteoclastogenesis and the area of RANKL-induced osteoclastogenesis in a dose-dependent manner (Figures 1C,D). Next, we investigated whether JSH-23 suppressed F-actin formation. F-actin forms loop structures on the bone surface for bone resorption. As shown in Figures 1E,F, phalloidin-Alexa Fluor 488 staining revealed characteristic F-actin rings in osteoclasts without JSH-23 treatment. In contrast, under JSH-23 treatment, the areas of F-actin rings were reduced, and more pleomorphic F-actin rings appeared; these effects were dose-dependent.

Bone resorption is the most important function of osteoclasts and is also one of the critical causes of bone loss. Therefore, we further evaluated whether JSH-23 inhibits osteoclast bone resorption. As shown in Figure 1G, compared to the control treatment, treatment with JSH-23 (10, 20 and 40 μM) reduced the resorption area in a dose-dependent manner (Figure 1H). Collectively, these data confirm that JSH-23 inhibits osteoclastogenesis, F-actin formation and bone resorption with negligible cytotoxicity.

In order to further clarify the role of JSH-23 in osteoclast differentiation and function, we utilized qPCR to examine the mRNA expression levels of osteoclast-specific genes, including NFATc1, c-Fos, TRAP, CTSK and DC-STAMP. Compared with the control group, the JSH-23-treated groups exhibited strong suppression of RANKL-induced osteoclast-specific gene expression, and this effect of JSH-23 was dose- and time-dependent (Figures 2A,B). Additionally, western blotting assays indicated that this compound inhibited the protein expression of NfATc1 (days 1, 3 and 5) and c-Fos (days 3 and 5) (Figures 2C,D).

During RANKL-induced osteoclastogenesis, NFATc1 self-amplification and nuclear translocation are critical for the expression of osteoclast-specific genes. As expected, NFATc1 nuclear translocation was significantly promoted by RANKL treatment but disrupted by JSH-23 (Figures 2E,F).

These data further demonstrate the inhibitory effects of JSH-23 on osteoclasts.

To elucidate the molecular mechanisms underlying the effects of JSH-23 on osteoclastogenesis, signaling pathways involved in osteoclastogenesis were investigated. As shown in Figures 3A,B, JSH-23 significantly attenuated p65 and IκBα phosphorylation. However, the phosphorylation levels of Akt, ERK, JNK and p38 were unchanged after JSH-23 treatment. Further immunofluorescence analysis showed that JSH-23 inhibited p65 nuclear translocation (Figure 3C), which was confirmed by quantitative analysis of the number of cells in which p65 was translocated to the nucleus (Figure 3D).

ROS have been shown to be extremely important in regulating RANKL-dependent osteoclast differentiation and are therefore considered therapeutic targets for the treatment of osteolysis (Chen et al., 2019). As shown in Figure 4A, the intensity of DCF fluorescence was markedly higher in the RANKL-stimulated group than in the control group, but the effect of RANKL stimulation was significantly attenuated in the presence of JSH-23 at concentrations of 20 and 40 μM (Figure 4B). During osteoclastogenesis, RANKL binds to its receptor RANK and generates intracellular ROS through activation of tumor necrosis factor receptor-associated factor 6 (TRAF6), Ras-related C3 botulinum toxin substrate 1 (Rac1), and NADPH oxidase 1 (NOX1). As shown in Figures 4C,F, TRAF6, Rac1 and NOX1 expression was significantly upregulated by RANKL stimulation but dose-dependently inhibited by JSH-23 treatment.

Cells have several protective mechanisms against oxidative stress. Nrf2 is a transcription factor that controls the gene expression of many antioxidant enzymes that combat oxidative stress, such as HO-1, catalase, NAD(P)H:quinone oxidoreductase 1 (NQO1), and glutathione reductase (GSR) (Kleszczynski et al., 2016). We demonstrated that JSH-23 activated Nrf2 expression during RANKL-induced osteoclastogenesis (Figures 4D,G). HO-1, one of the main oxidative stress markers induced by Nrf2 activation, is involved in defense against various oxidative stress-inducing agents. The expression of HO-1 was reduced by RANKL stimulation but was recovered and enhanced dose-dependently by JSH-23 treatment. Similarly, JSH-23 enhanced the expression levels of catalase, NQO1 and GSR (Figures 4E,H).

Nrf2 nuclear translocation is a prerequisite for activation of downstream antioxidative target genes. Immunofluorescence staining for Nrf2 and nuclei (DAPI) in BMMs revealed that Nrf2 was localized evenly in the cytoplasm of control cells, while the intensity of green staining was markedly increased in nuclei or concentrated on the edge of nuclei in cells treated with JSH-23 (Figure 4I). In addition, JSH-23 increased HO-1 fluorescence intensity (Figure 4J), consistent with our western blotting results.

Next, to confirm that the Nrf2/HO-1 antioxidant pathway is involved in the inhibition of osteoclastogenesis by JSH-23, we silenced Nrf2 or HO-1 expression using small interfering RNA (siRNA), and the silencing effect was confirmed by western blotting analysis (Figures 4K–N). Interestingly, silencing Nrf2 reduced HO-1 expression, further indicating that HO-1 is downstream of Nrf2 (Figure 4K). Moreover, downregulation of Nrf2 or HO-1 attenuated the inhibitory effect of JSH-23 (Figure 4O), reversed the impairment of osteoclastogenesis, and caused mature osteoclasts to appear (Figure 4P). Collectively, these results indicate that JSH-23 reduces RANKL-induced ROS production in osteoclastogenesis by downregulating the TRAF6/Rac1/NOX1 signaling pathway and enhancing the expression of Nrf2/HO-1.

Since osteoclasts and osteoblasts are both important in bone homeostasis, the effects of JSH-23 on osteoblasts were studied. Although ROS maintain osteoclast differentiation and survival, oxidative stress induces osteoblast apoptosis, reduces osteogenic bone formation, decreases bone mass and triggers bone loss (Tao et al., 2020). Here, we found that the viability of osteoblasts decreased in a dose- and time-dependent manner when the cells were exposed to different concentrations of H₂O₂ (range of 0–800 μM) (Figure 5A). In addition, flow cytometric analysis demonstrated that osteoblast apoptosis increased when cells were exposed to H₂O₂ (Figures 5B,C). According to the results, 400 μM H₂O₂ for 4 h was chosen as the condition for oxidative stress model establishment in MC3T3-E1 cells for further research. Western blotting analysis showed that H₂O₂ increased cleaved caspase-9 and cleaved caspase-3 expression and decreased the Bcl-2/Bax ratio in a dose-dependent manner (Figures 5D,E). These data confirmed that H₂O₂ induced apoptosis in MC3T3-E1 cells. Notably, treatment with JSH-23 decreased the percentage of apoptotic cells under H₂O₂ treatment (Figures 5F,G). Further research showed that JSH-23 exerted its antiapoptotic effect by decreasing cleaved caspase-9 and cleaved caspase-3 expression and increasing the Bcl-2/Bax ratio (Figures 5H,I).

We further evaluated the effect of JSH-23 on H₂O₂-induced changes in osteoblast function during early differentiation and late-stage mineralization in vitro. We found that H₂O₂ markedly reduced osteoblast differentiation, as measured by analysis of ALP staining at 7 days and AR staining at 21 days (Figures 6A,B). However, pretreatment with JSH-23 for 24 h markedly alleviated the subsequent H₂O₂-mediated downregulation of ALP expression and reduction in mineralization (Figures 6A,B).

To investigate the effects of JSH-23 on H₂O₂-induced osteogenic dysfunction, we investigated the mRNA and protein expression levels of osteoblast markers by qPCR and western blotting analysis, respectively. H₂O₂ markedly downregulated the mRNA levels of Runx2, OCN, BMP2 and OSX, while pretreatment with 20 and 40 μM JSH-23 increased the mRNA levels of these genes (Figure 6C). Similar trends were observed at the protein level: JSH-23 pretreatment prevented H₂O₂-induced marked reductions in Runx2, OCN and BMP2 levels (Figures 6D,E). Collectively, these results suggest that JSH-23 pretreatment prevents H₂O₂-induced decreases in osteogenic differentiation.

Treatment of MC3T3-E1 cells with H₂O₂ alone clearly increased the levels of ROS. However, pretreatment with JSH-23 significantly prevented this effect (Figures 7A,B). Western blotting analysis demonstrated that 20 and 40 μM JSH-23 treatment significantly increased Nrf2 and HO-1 protein levels (Figures 7C,D). Similar to the results showing that JSH-23 promoted the Nrf2/HO-1 pathway in osteoclasts, the immunofluorescence analysis results showed that JSH-23 promoted Nrf2 nuclear translocation and increased HO-1 fluorescence intensity in MC3T3-E1 cells (Figures 7E,F).

Next, we silenced Nrf2 and HO-1 expression using siRNA in MC3T3-E1 cells, and the silencing effect was confirmed by western blotting (Figures 7G–J). After downregulation of Nrf2 or HO-1, the protective effect of JSH-23 against the H₂O₂-induced reductions in the levels of osteoblast markers (Runx2, OCN and BMP2) was attenuated (Figures 7K,L). These results suggest that JSH-23 prevents H₂O₂-induced impairment of osteoblast differentiation by scavenging intracellular ROS and increasing activation of the Nrf2/HO-1 pathway.

Having determined that JSH-23 inhibits osteoclastogenesis and promotes osteoblastogenesis in vitro, we next explored the potential protective effect of JSH-23 under the pathologic setting of LPS-induced osteolysis in vivo. As shown in the three-dimensional reconstruction images in Figure 8A, compared to the sham group, the vehicle group exhibited extensive erosion on the calvarial bone surface. However, the severity of LPS-induced osteolysis was significantly inhibited by JSH-23 treatment (Figure 8B). Consistent with the micro-CT results, H&E staining showed that JSH-23 significantly reduced the area of LPS-induced bone erosion (Figure 8C). TRAP staining further revealed that JSH-23 treatment dose-dependently decreased the numbers of TRAP-positive osteoclasts and osteoclasts lining the bone surface (OC.S/BS) (Figures 8C,D).

The anabolic effect of JSH-23 on bone formation in vivo was also examined. Calvarial bone sections were subjected to immunofluorescence double staining for the osteoblast-specific markers Runx2 and OCN. As shown in Figure 8E, the JSH-23-treated group exhibited higher expression of Runx2 and OCN on the calvarial bone surface than the vehicle group (Figure 8F). Taken together, these results provide evidence that JSH-23 is an effective agent for the treatment of pathologic bone loss conditions, such as LPS-induced osteolysis.

As JSH-23 exhibited intracellular antioxidant activity, we next assessed in vivo ROS levels on the bone surface using the ROS probe DHE. Consistent with the in vitro results, ROS fluorescence intensity in bone tissue was much higher in the vehicle group than in the sham group, but JSH-23 dramatically reversed ROS production in vivo (Figures 9A,B). To investigate Nrf2 and HO-1 expression in vivo, total protein was extracted from each group and detected using western blotting. Enhancement of Nrf2 and HO-1 expression was observed in the JSH-23-treated group (Figures 9C,D). Furthermore, to confirm our western blotting results, immunofluorescence double staining for Nrf2 and HO-1 were performed. The results showed that Nrf2 and HO-1 staining were more intense in the JSH-23-treated group than in the vehicle group (Figures 9E,F). A proposed scheme of the mechanism by which JSH-23 restrains bone loss by scavenging ROS production and activating Nrf2/HO-1 signaling is shown in Figure 10. Collectively, these results demonstrate that JSH-23 prevents LPS-induced osteolysis in vivo by reducing ROS production and enhancing Nrf2/HO-1 expression.