ZM-306416 prevents ovariectomy-induced bone loss by promoting osteoblastogenesis and inhibiting osteoclastogenesis

Abstract

Background

Osteoporosis affects millions of people worldwide, and current medications such as bisphosphonates and denosumab are not effective enough to reverse bone loss. Moreover, these treatments have drawbacks, including jaw osteonecrosis and skin eczema. Hence, there is an urgent need for new drugs to treat osteoporosis.

Methods

Drug library screening was performed via alkaline phosphatase (ALP) staining in osteoblasts to identify potential candidates for osteoporosis treatment. qPCR, Western blotting, ALP staining, alizarin red staining, and tartrate-resistant acid phosphatase (TRAP) staining were conducted to assess the impact of ZM-306416 (ZM) on osteoblast and osteoclast differentiation in vitro. Additionally, RNA sequencing and pathway analysis were carried out to explore the underlying molecular mechanisms involved. Micro-CT scanning and immunostaining were used to determine bone phenotypes in vivo.

Results

Drug library screening revealed that ZM enhances ALP activity in osteoblasts, indicating its potential as a pro-osteogenic agent. ZM exerts dual effects by promoting osteoblast differentiation through the Wnt/β-catenin signaling pathway and simultaneously inhibiting osteoclast differentiation through the NF-κB and MAPK signaling pathways. In an OVX mouse model, ZM effectively prevents bone loss by stimulating osteoblast formation and inhibiting osteoclast development.

Conclusions

Our study revealed that ZM has a dual anti-osteoporosis effect by promoting osteoblastogenesis and inhibiting osteoclastogenesis, which is mediated by activation of the Wnt/β-catenin signaling pathway and suppression of the NF-κB/MAPK cascades. These findings suggest that ZM could be a promising therapeutic agent for alleviating osteoporosis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-025-06654-7.

Introduction

Adult bone undergoes continuous remodeling to maintain bone homeostasis and strength, which involves osteoblast-mediated bone formation and osteoclast-mediated bone resorption [1–3]. Osteoblasts, the chief bone-making cells, arise from bone marrow mesenchymal stromal cells and secrete specific extracellular proteins (e.g., osteocalcin, alkaline phosphatase, and type I collagen) to facilitate mineralization of the extracellular matrix [4]. In contrast, osteoclasts, specialized multinucleated cells derived from monocytes-macrophages under stimulation by macrophage colony-stimulating factor (M-CSF) and receptor activator of NF-κB ligand (RANKL), absorb old bone [5].

With advancing age and estrogen deficiency, an imbalance in bone remodeling occurs, leading to increased bone resorption and insufficient bone formation, culminating in osteoporosis—a metabolic bone disease characterized by reduced bone mass and impaired microarchitecture, resulting in increased fracture risk [6]. The prevalence of osteoporosis has been steadily increasing due to the aging population, with approximately 200 million individuals affected worldwide, thus presenting a significant healthcare challenge [7, 8].

Early screening is key for osteoporosis management: in 1598 women ≥ 50 years old, weight < 57.4 kg has been identified as the most effective predictor of osteoporosis (outperforming BMI and age), providing practical cutoff values for clinical screening [9].

Common treatments for osteoporosis include bisphosphonates, RANKL antibodies, monoclonal sclerostin antibodies, and parathyroid hormone receptor agonists [10]. These agents have variable efficacy: denosumab shows superior efficacy in increasing hip and spine BMD, romosozumab excels in vertebral fracture reduction, and ibandronate excels in hip fracture prevention [11, 12]. Mechanistically, parathyroid hormone receptor agonists increase osteoblast activity and inhibit osteoclast recruitment, whereas other treatments target osteoclast resorption. Nonetheless, these treatments have limitations and side effects [13–18]. For example, long-term bisphosphonate use has been associated with the development of atypical femur fractures and osteonecrosis of the jaw [18]. RANKL antibodies also have side effects, such as skin eczema, flatulence, cellulite inflammation, and osteonecrosis of the jaw [19]. No single drug addresses all clinical needs, underscoring the urgent need for novel therapeutic interventions.

Beyond pharmacological therapies, surgical interventions like percutaneous vertebroplasty (PVP) are used for painful OVCFs, but their efficacy evidence remains inconsistent. A double-blind randomized sham-controlled trial verified PVP’s pain-relief effect in OVCFs with MRI-confirmed edema [20], highlighting the need for evidence-based non-pharmacological options alongside novel drugs. Similar to PVP, percutaneous kyphoplasty (PKP) also has postoperative risks: low BMD, preoperative thoracolumbar fascia injury, massive bone cement distribution, and low cement filling ratio have been identified as independent risk factors for short-term residual low back pain after PKP, indicating existing surgical protocols need optimization [21]. Non-pharmacological approaches like pulsed electromagnetic fields (PEMFs) can promote osteoblast proliferation and differentiation [22], but their application is limited by equipment dependence and individual variability—this limitation is particularly critical for frail elderly populations with fragility fractures. Fragility fractures in the elderly pose severe socioeconomic burdens: diagnosis relies on tools like DXA and FRAX, and management requires ortho-geriatric comanagement, timely surgery, and early rehabilitation [23]. Unfortunately, delayed diagnosis and suboptimal treatment still lead to poor outcomes, emphasizing the need for both targeted pharmacotherapies and optimized clinical protocols tailored to frail individuals.

Special subtypes of osteoporosis further demand personalized strategies: glucocorticoid-induced osteoporosis (CIO), the most common secondary osteoporosis, responds well to pharmacological interventions, with alendronate outperforming other agents in increasing BMD, reducing fractures, and minimizing adverse events [24].Even for rare subtypes like transient bone osteoporosis (TBO), conservative treatment is effective, and bisphosphonates can alleviate pain and accelerate recovery [25]. Both further validate that targeting bone metabolism pathways is a viable strategy across diverse osteoporotic subtypes. Pharmacogenetics and biomarkers support precision medicine: the rs2228570 polymorphism in the vitamin D receptor (VDR) gene modulates postmenopausal women’s response to bisphosphonates or strontium ranelate, with TT genotype carriers showing superior BMD improvement [26]. Bone turnover markers (BTMs) such as bone alkaline phosphatase (bALP), N-telopeptides of type I collagen (NTx), and procollagen type I N propeptide (PINP) are reliable tools for monitoring treatment response and predicting fracture risk—higher bALP correlates with lower femoral BMD, and NTx is associated with non-vertebral fractures [27, 28]. Integrating genetic testing and BTM monitoring enables tailored anti-osteoporotic therapy.

Given the limitations of current treatments and the need for personalized options, high-throughput screening of bioactive compound libraries has proven effective for identifying bone metabolism regulators. A study screened 512 small molecules and identified higenamine, which promotes osteogenesis via the IQGAP1/SMAD4 signaling pathway and prevents bone loss in murine models of osteoporosis. This work directly validates library screening as a robust tool for discovering bone-anabolic agents [29]. Another study developed an aptamer-based competitive screening assay to evaluate 96 compounds (including FDA-approved drugs) and identified 6 sclerostin inhibitors that enhance osteoblastic activity. This work further confirms the reliability of library-based screening for targeting key bone regulatory molecules [30]. Epidermal growth factor receptor (EGFR) is closely associated with bone metabolism, as its signaling pathway regulates the function of osteoblasts, osteoclasts, and chondrocytes [31]. Moreover, mesenchymal stem cells (MSCs) are critical for bone regeneration, and regulating their osteogenic differentiation is a key strategy for osteoporosis treatment. Targeting molecules involved in this process (e.g., EGFR) provides a promising direction for developing anti-osteoporosis agents [32, 33].

In this study, we performed ALP staining screening on a bioactive compound library (L1022P, APExBIO) using osteoblasts. Our analysis revealed that ZM-306416 (referred to as ZM), an EGFR inhibitor, is a potential enhancer of osteoblast differentiation. Therefore, our objective was to investigate the effects of ZM on both osteoblast and osteoclast differentiation and explore its therapeutic potential for osteoporosis.

Materials and methods

Mice and ovariectomy murine model

C57BL/6 mice, aged 10 to 12 weeks, were obtained from the Laboratory Animal Research Centre of Southern Medical University and housed at a temperature of 20–22 °C and a humidity of 30–70%, with four or five mice per cage. All animal care protocols and experiments were reviewed and approved by the Southern Medical University Laboratory Animal Ethics Committee, and this study was compliant with all relevant ethical regulations regarding animal research. The mice were divided into three groups: the OVX group (n = 6), the OVX + ZM group (0.5 mg/kg, n = 6), and the sham group (n = 6). For anaesthesia during bilateral OVX, tribromoethanol anaesthetic (1.25% tribromoethanol solution) was used. It was prepared by weighing 2.5 g of tribromoethanol powder, dissolving it in 5 mL of tert-amyl alcohol, and diluting to 200 mL with deionized water. The solution was magnetically stirred until homogeneous, protected from light with tin foil, and stored at 4 °C. The injection dosage was 0.3 ml/20 g body weight via intraperitoneal injection, ensuring adequate anaesthesia while minimizing toxicity. The OVX group and the OVX + ZM group underwent bilateral OVX after being anaesthetized with this tribromoethanol anaesthetic, whereas the sham group only had fat tissue near the ovary removed. The ZM group was injected with 0.5 mg/kg ZM peritoneally, whereas the Sham group and the OVX group were injected with corn oil twice a week for 8 weeks. In the 9th week, all the mice were euthanized by cervical dislocation (a humane and rapid method that minimally interferes with tissue analysis, consistent with ethical animal research guidelines), and the femurs were isolated after excess tissue was removed for histological and morphological evaluation, micro-CT, and microscopic examination.

Primary osteoblast isolation and culture

Primary osteoblasts were harvested from the calvaria of 3-day-old C57BL/6 mice as described previously [34]. Specifically, the excess soft tissue was removed, after which the calvaria was cut into small bone fragments via sterile scissors. The bone fragments were fully digested with 0.25% trypsin for 10 min and 0.2% type II collagenase for 90 min. Primary osteoblasts were then collected by gently pipetting the digested fragments. The cells were then suspended in complete α-MEM (containing 10% fetal bovine serum) and incubated at 37 °C in a 5% CO2 incubator.

BMM isolation and culture

BMMs were generated from the bone marrow of 6-week-old C57BL/6 male mice as described previously [35]. The bone marrow cells were then obtained by flushing the bones with a 1 mL syringe. The cells were subsequently suspended in complete α-MEM supplemented with 10% fetal bovine serum and 30 ng/mL M-CSF. The cells were cultured at 37 °C in a 5% CO2 incubator. After three days, the adherent cells were identified as bone marrow-derived macrophages.

CCK-8 assay

The cells were seeded into 96-well plates with a liquid volume of 200 µL per well and counted via a cell counting plate, resulting in approximately 8,000 cells per well. Four replicate wells were set up. The 96-well plates were then incubated overnight at 37 °C in a 5% CO2 incubator to allow the cells to adhere. Subsequently, 20 µL of CCK-8 solution was added to each well. A blank control group was also established, where the same volumes of cell culture medium and CCK-8 solution were added, but no cells were present. The 96-well plates were further incubated, and the absorbance was measured at 450 nm after 0.5, 1, 2, and 4 h via a microplate reader (BIO-TEK, Synergy HTX). On the basis of pre-experimental data, the 2-hour time point was selected for final analysis, as it yielded absorbance values within the linear range (0.8–1.2 OD) for all groups, ensuring accurate quantification of viable cells without signal saturation or underdetection. The OD values for each group at this time point were recorded to calculate cell viability.

ALP staining and Alizarin red staining

Osteoblasts were cultured in osteogenic induction medium (OIM; 50 µg/mL ascorbic acid and 5 mM sodium β-glycerophosphate) for approximately 7 days, and the OIM was changed every 2 days to maintain stable concentrations of osteogenic factors and minimize the accumulation of metabolic byproducts. The cells were subsequently fixed with 4% paraformaldehyde for 20 min. A BCIP/NBT staining working solution was prepared following the provided instructions. ALP staining was performed at room temperature to stain the cells for 10–30 min. Finally, the 6-well plate was scanned via an EPSON Perfection V800 scanner (EPSON, Tokyo, Japan), and the resulting images were stored for subsequent analysis. After a 21-day induction period with the osteogenic induction solution, the cells were fixed with 4% paraformaldehyde for 20 min and incubated with a 1% alizarin red staining solution, which was prepared according to the instructions. The cells were stained with alizarin red solution for approximately 5–10 min at 37 °C. The 6-well plate was then scanned via a scanner, and the resulting images were stored for analysis.

TRAP staining

Bone marrow-derived macrophages were induced with complete α-MEM (containing 10% FBS) supplemented with 30 ng/mL M-CSF and 100 ng/mL RANKL. After 7 days of induction, multinucleated osteoclasts were observed under a microscope. To perform TRAP staining, the staining solution was prepared following the provided instructions. The cells were then incubated with TRAP staining solution at 37 °C for 30–60 min in the dark. The cells were subsequently washed 1–2 times with deionized water. Finally, the cells were observed, and pictures were taken under a microscope. Osteoclasts were identified as TRAP-positive cells with more than 3 nuclei. The ability to form osteoclasts was determined by analysing the number of TRAP-positive cells in each group.

Western blotting (WB)

The cells were seeded in 6-well plates at a density of 4 × 10⁵ cells per well and cultured overnight. After treatment, the cells were lysed on ice for 10 min in RIPA lysis buffer containing protease inhibitors to extract total protein. The total protein was separated by SDS‒PAGE and transferred to polyvinylidene fluoride (PVDF) membranes. Following blocking with skim milk for 1 h, the membranes were incubated overnight at 4 °C with the following primary antibodies: anti-OSX (1:1000, BOSTER, China); anti-RUNX2, anti-CTSK, and anti-NFATC1 (1:200, Santa Cruz, USA); anti-Tubulin (1:5000, Affinity, China); anti-ERK, anti-p-ERK, anti-P38, anti-P-P38, anti-P65, anti-P-P65, anti-JNK, and anti-P-JNK (1:1000, Cell Signaling, USA); anti-β-actin (1:20000, Proteintech, China); and anti-β-catenin (1:1000, Abmart, China). The membranes were subsequently washed three times with TBST (5 min each) and incubated with specific secondary antibodies: goat anti-mouse IgG H&L (HRP) pAb (1:20000, PTM Bio, China) and goat anti-rabbit IgG H&L (HRP) pAb (1:10000, PTM Bio, China) for 1 h at room temperature. After three additional washes with TBST, the membranes were incubated with chemiluminescent substrate (Abkine, Cat. No. BMP3010) and visualized via the GeneSys capture system.

Real-time quantitative PCR (qPCR)

Total RNA was isolated from the cells via TRIzol Reagent (Cat. No. 9109; Takara Biotechnology) following the manufacturer’s protocol. The reverse transcription reaction system (Cat. RR036A, Takara Biotechnology) was then prepared according to the manufacturer’s instructions. The system was subsequently placed into the reverse transcription PCR instrument to generate complementary DNA (cDNA). The obtained cDNA was used for qPCR with ChamQ SYBR qPCR Master Mix (Cat. No. Q311-02, Vazyme Biotech) in a LightCycler96 instrument (Roche Molecular Biochemicals). The relative quantification of gene expression was performed via the comparative threshold method. Changes in mRNA expression levels were calculated after normalization of the values to those of the GAPDH housekeeping gene. The primers used are listed in Table S1.

Immunohistochemistry

The femurs were separated from the mice and fixed in 4% paraformaldehyde for 24 h. After decalcification in 10% EDTA (pH 8.0) for 21 days, the decalcified tissues were embedded in paraffin and prepared as 4 μm sagittal-oriented sections for histological analysis. The paraffin sections were then incubated at 65 °C for 2 h. After the paraffin sections were dewaxed and hydrated, antigen retrieval was performed. Next, the paraffin sections were incubated overnight at 4 °C with a primary antibody against osteocalcin (OCN), followed by incubation with secondary antibodies for 1 h at room temperature. Finally, diaminobenzidine (DAB) was used to visualize the positive cells, and the samples were observed and photographed via an Olympus BX51 microscope. Immunohistochemical staining was evaluated by the number of positive cells per bone perimeter (B.Pm).

Micro-CT analysis

Micro-CT analysis was conducted using a CT40 scanner (Scanco Medical AG, Bassersorf, Switzerland) to examine the femoral structure. The scanning conditions were set to 55 kV, 145 µA, and a voxel size of 10 μm (optimal for trabecular bone resolution in murine femurs). Image segmentation was performed using a threshold range of 220–1000 HU, which effectively distinguishes mineralized bone tissue from soft tissue and marrow in mouse femurs. For the analysis of bone trabeculae, 200 section planes were scanned from the distal femoral growth plate towards the proximal end. The analysis of cortical bone involved scanning from the middle of the femur towards the proximal end. The parameters analysed included the trabecular bone density (Tb. BMD, mg HA/ccm), bone trabecular volume fraction (BV/TV, %), bone trabecular number (Tb. N, mm⁻¹), bone trabecular thickness (Tb. Th, mm), bone trabecular separation/porosity (Tb. Sp, mm), cortical bone density (Ct. BMD, mg HA/ccm), and cortical bone thickness (Ct. Th, mm).

ELISA

Blood was collected from the mice via eyeball extraction and left at room temperature for 2 h. After centrifugation at 1000 × g for 20 min, the supernatant was collected for serum preparation. The serum was stored at -80 °C. The serum sample and the CTX-1 ELISA kit were subsequently brought to room temperature for equilibration. Next, the procedures to measure the levels of CTX-1 in the serum of each group were carried out according to the instructions provided with the ELISA kits.

Hematoxylin and Eosin (HE) staining

The paraffin-embedded sections were stained with HE. The sections were observed and photographed using an Olympus BX51 microscope. Subsequently, histological analysis was conducted via ImageJ software.

RNA sequencing and pathway analysis

Primary osteoblasts were treated with or without 10 µM ZM in osteogenic induction medium for 6 days, and the cells were divided into a control group and an intervention group. Concomitantly, bone marrow-derived macrophages were treated with or without 10 µM ZM in complete α-MEM (containing 10% FBS) supplemented with 30 ng/mL M-CSF and 100 ng/mL RANKL for 2 days, and the cells were also divided into a control group and an intervention group. Total RNA was extracted via TRIzol Reagent (Cat. No. 9109; Takara Biotechnology) following the manufacturer’s instructions, and three samples per group were used. The RNA concentration and purity were quantified via a UV spectrophotometer (NanoDropOne, Thermo Fisher Scientific, USA). After RNA extraction, the raw sequencing reads were filtered via Fastp to remove low-quality and adapter-containing reads. Clean reads were aligned to the mouse reference genome (mm10) with HISAT2. Gene read counts were generated by FeatureCounts. Differentially expressed genes (DEGs) were identified via DESeq2, with the criteria of |log₂-fold change (FC)| > 1 and adjusted P value < 0.05. The Kyoto Encyclopedia of Genes and Genomes (KEGG; http://www.genome.ad.jp/kegg) was employed for pathway analysis via enriched R packages. The terms or pathways were mapped and visualized by selecting the top 20 according to the number of genes for which they were enriched. The original data were uploaded to the GEO database (https://www.ncbi.nlm.nih.gov/, GSE307058, GSE307904).

Key reagents

ZM-306416 (ZM) was purchased from APExBIO Technology LLC (Houston, USA; Catalogue No. A8684; CAS No. 690206-97-4). The compound has a purity ≥ 95% (verified by HPLC, as confirmed by the batch-specific certificate of analysis). ZM was dissolved in DMSO to prepare a 10 mM stock solution (solubility ≥ 33.4 mg/mL in DMSO, insoluble in water) and stored at -20 °C according to the manufacturer’s recommendations. ICG-001 was purchased from Selleck Chemicals (Catalogue No. S2662). It was dissolved in DMSO to prepare a 10 mM stock solution and stored at -20 °C. For in vitro experiments, a working concentration of 10 µM was used, as determined on the basis of references indicating that this concentration effectively inhibits Wnt/β-catenin signaling [36].

Drug library screening

Primary osteoblasts were isolated from the calvaria of 3-day-old C57BL/6 mice and seeded in 24-well plates at a density of 5 × 10⁴ cells per well. Following 24 h of adherence in α-MEM supplemented with 10% fetal bovine serum (FBS), the cells were treated with compounds from the DiscoveryProbe™ Bioactive Compound Library Plus (Cat. No. L1022P, APExBIO, USA). The compounds were diluted in osteogenic induction medium (OIM) to a final concentration of 10 µM. Each compound was administered in triplicate (n = 3). The cells were cultured in OIM for 7 days, and the medium was changed every 48 h. Alkaline phosphatase (ALP) staining was performed via BCIP/NBT kits according to the manufacturer’s instructions. The staining intensity was semiquantified via ImageJ software, and the compounds were evaluated relative to 0.1% dimethyl sulfoxide (DMSO) vehicle controls.

Statistical analysis

The experiments were conducted a minimum of three times, and the results are presented as the mean ± SEM. Data analysis was performed via GraphPad Prism 10.1.2. Student’s t- test was used to compare two groups, whereas one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test was used to compare differences between three or more groups. A p value of less than 0.05 was considered statistically significant for all the experiments.

Results

ZM promotes osteoblast differentiation

In a drug screening assay using a bioactive compound library (L1022P, APExBIO), our results demonstrated that ZM significantly enhanced alkaline phosphatase (ALP) activity (Fig. 1A). To verify the role of ZM in osteoblast differentiation, primary mouse osteoblasts were treated with varying concentrations of ZM, and the differentiation phenotypes were analysed. Initially, a CCK8 assay was conducted to detect the effect of ZM on osteoblast proliferation. The findings indicated that concentrations ranging from 1.25 to 10 µM had no significant effect on osteoblast proliferation (Fig. 1B). A concentration of 10 µM was chosen for subsequent cell experiments. Following 7 days of osteogenic induction, the expression of the osteoblast differentiation markers Runx2, Alpl, and Bglap was assessed. We found that ZM markedly upregulated these osteogenic differentiation marker genes (Fig. 1C). Consistent with the mRNA findings, the protein levels of OSX and RUNX2 were also significantly upregulated (Fig. 1D). On the 7th day of osteogenic induction, ALP staining was subsequently performed, followed by alizarin red staining on the 14th day. The results demonstrated that ZM increased alkaline phosphatase activity and calcium deposition in primary osteoblasts (Fig. 1E and F). Overall, these findings highlight that ZM effectively promotes osteoblast differentiation in vitro without affecting cell growth.

Fig. 1.

ZM promotes osteoblast differentiation. (A) ALP staining of candidate drugs from the plate 1022P-17. The ZM and control groups are labelled in red. (B) CCK8 assay of ZM on primary mouse osteoblasts. (C) Primary mouse osteoblasts were cultured with ZM, and the mRNA levels of osteoblast differentiation markers were detected. (D) Primary mouse osteoblasts were cultured with ZM, and the protein levels of osteoblast differentiation markers were detected. (E) Primary mouse osteoblasts were cultured with ZM, and ALP staining was performed. (F) Primary mouse osteoblasts were cultured with ZM, and alizarin red staining was performed. NS, not significant; **, p < 0.01; ***, p < 0.001

ZM inhibits osteoclast differentiation

Considering the coupling role of osteoblasts and osteoclasts in bone remodeling and the development of osteoporosis, we aimed to examine whether ZM could inhibit osteoclast differentiation. We initially assessed the effect of ZM on BMM proliferation and found no significant effect (Fig. 2A). Following a 3-day period of M-CSF and RANKL stimulation, we evaluated the mRNA and protein levels of osteoclast differentiation markers. Notably, ZM inhibited both the mRNA and protein levels of these markers in a concentration-dependent manner (Fig. 2B and C). M-CSF and RANKL induced the formation of numerous TRAP-positive multinucleated osteoclasts from BMMs (Fig. 2D). However, ZM at a concentration of 10 µM inhibited M-CSF and RANKL-induced osteoclastogenesis almost completely (Fig. 2D). Collectively, these findings demonstrate the in vitro inhibitory effect of ZM on osteoclast differentiation.

Fig. 2.

ZM inhibits osteoclast differentiation. (A) CCK8 assay of ZM on BMMs. (B) BMMs were cultured with ZM, and the mRNA levels of osteoclast differentiation markers were detected. (C) BMMs were cultured with ZM, and the protein levels of osteoclast differentiation markers were detected. (D) BMMs were cultured with ZM, and TRAP staining was performed. NS, not significant; ***, p < 0.001; ****, p < 0.0001

ZM prevents OVX-induced bone loss in vivo

To investigate the potential of ZM in restoring bone loss in vivo, we conducted a study to determine whether it could prevent the established bone loss caused by OVX in mice. We observed that ovariectomy-induced estrogen deprivation resulted in a decrease in the size and weight of the mouse uterus (Fig. 3B). Since estrogen deficiency causes an imbalance in bone formation and bone resorption in postmenopausal women, leading to bone loss and osteoporosis, and our results indicate that ZM stimulates osteogenesis and inhibits osteoclastogenesis, we hypothesized that ZM treatment could prevent estrogen deficiency-induced bone loss. For this purpose, OVX mice were intraperitoneally injected with ZM (Fig. 3A). Micro-CT analysis of the distal femoral metaphysis and diaphysis revealed that OVX mice presented noticeable loss of trabecular and cortical bone (Fig. 3C and D). We found that bone mineral density (BMD), bone volume per tissue volume (BV/TV), the trabecular number (Tb.N), the trabecular thickness (Tb.Th), and the trabecular bone area were significantly reduced in mice 8 weeks after OVX (Fig. 3C and D, and 4A). Conversely, trabecular spacing (Tb.Sp) and the number of TRAP-positive osteoclasts were significantly increased, indicating severe bone loss in the femur as a result of OVX (Figs. 3C and 4B). The serum levels of C-terminal telopeptide of collagen type I (CTX-1), a marker for bone resorption, were significantly greater in OVX mice than in control mice (Fig. 3E). These results indicated severe bone loss in the female mice as a result of OVX. To assess the therapeutic potential of ZM, we administered either vehicle or ZM (0.5 mg/kg) intraperitoneally to OVX mice twice a week for 8 weeks. We subsequently conducted micro-CT analysis and histological staining of femurs from these mice. Compared with vehicle-treated OVX mice, OVX mice treated with ZM presented substantial bone restoration and significantly greater BMD, BV/TV, Tb.N, Tb.Th and trabecular bone area (Fig. 3C and D, and 4A). Histomorphometric analysis further revealed that the number of osteoblasts (N.Ocn/B. Pm) was increased in ZM-treated OVX mice, whereas the osteoclast surface was significantly smaller and the number of osteoclasts (N.Oc/B. Pm) was lower than that in vehicle-treated OVX mice (Fig. 4B). Taken together, our results suggest that ZM treatment is a potential therapeutic approach for postmenopausal bone loss.

Fig. 3.

ZM prevents OVX-induced bone loss in vivo. (A) Schematic representation of the experimental design involving the intraperitoneal injection of ZM into OVX mice. (B) Representative image and weights of uteri from sham and OVX mice. (C) Representative micro-CT images and histomorphometric analysis of the metaphysis of femurs from sham and OVX mice. (D) Representative micro-CT images and histomorphometric analysis of the diaphysis of femurs from sham and OVX mice. (E) Serum levels of CTX-1 in sham and OVX mice. NS, not significant; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001

Fig. 4.

ZM treatment promoted osteoblastogenesis and inhibited osteoclastogenesis in OVX mice. (A) HE staining and quantitative analysis of the metaphysis of femurs from sham and OVX mice. (B) Ocn and TRAP immunostaining and quantitative analysis of the metaphysis of femurs from sham and OVX mice. *, p < 0.05; **, p < 0.01

ZM regulates osteoblast and osteoclast differentiation through the Wnt/β-catenin signaling pathway and NF-κB/MAPK signaling, respectively

To elucidate the molecular mechanisms by which ZM modulates osteoblast and osteoclast differentiation, we performed transcriptomic and biochemical analyses. KEGG pathway enrichment analysis of differentially expressed genes (DEGs) in osteoblasts revealed significant enrichment of the Wnt signaling pathway, indicating that this pathway plays a critical role in ZM-mediated regulation of osteoblastogenesis (Fig. 5A). Numerous studies have confirmed that canonical activation of the Wnt/β-catenin pathway is essential for osteoblastogenesis and bone formation [37]. To test the hypothesis that ZM promotes osteogenesis by activating the Wnt/β-catenin pathway, we examined the protein levels of β-catenin (a key mediator of canonical Wnt signaling) and RUNX2 (a downstream transcription factor indicative of osteogenic commitment) in osteoblasts treated with ZM [38, 39]. Western blot analysis revealed significant upregulation of both β-catenin and RUNX2 in ZM-treated osteoblasts compared with OIM-treated controls (Fig. 5B). To further validate the involvement of the Wnt/β-catenin signaling pathway, we co-treated primary osteoblasts with ZM and ICG-001, a specific inhibitor of the canonical Wnt/β-catenin pathway, in OIM, ensuring that the experiments were conducted under conditions that mimic the osteogenic microenvironment [40, 41]. As expected, ICG-001 significantly attenuated the ZM-induced upregulation of β-catenin and RUNX2 (Fig. 5B), confirming that the osteogenic effect of ZM is mediated through canonical Wnt/β-catenin activation. This effect was further validated by ALP staining, where ZM-treated osteoblasts presented stronger ALP activity (darker staining) than OIM-treated controls did, whereas co-treatment with ICG-001 diminished this enhancement (lighter staining), confirming that the osteogenic effect of ZM relies on canonical Wnt/β-catenin activation (Fig. 5C).

Fig. 5.

ZM regulates osteoblast and osteoclast differentiation through the Wnt/β-catenin signaling pathway and NF-κB/MAPK signaling, respectively. (A) KEGG pathway enrichment analysis of differentially expressed genes in osteoblasts (B) Primary mouse osteoblasts were cultured with OIM, ZM and ICG-001, and the protein levels of osteoblast differentiation markers were detected. (C) Primary mouse osteoblasts were cultured with OIM, ZM and ICG-001, and ALP staining was performed. (D) KEGG pathway enrichment analysis of differentially expressed genes in osteoclasts. (E) Phosphorylation levels of NF-κB (p65) and MAPK (p38, JNK, ERK) proteins in BMMs treated with M-CSF/RANKL ± ZM. NS, not significant; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001

For osteoclasts, we performed parallel KEGG pathway analysis to identify signaling pathways involved in ZM-mediated regulation. This analysis revealed significant enrichment in the “osteoclast differentiation” pathway (Fig. 5D), a pathway well-established to be tightly regulated by the NF-κB and MAPK signaling cascades. These cascades are recognized as core mediators of osteoclastogenesis downstream of RANKL-RANK activation [42, 43], which aligns with our focus on deciphering how ZM inhibits RANKL-induced osteoclast differentiation and prompts us to investigate the involvement of these two signaling cascades. Thus, we examined the phosphorylation levels of NF-κB (p65) and MAPK family members (p38, JNK, and ERK), which are crucial and fundamental downstream of RANKL–RANK signaling [43]. As expected, co-stimulation with M-CSF and RANKL triggered rapid phosphorylation of NF-κB (p65) and MAPKs (Fig. 5E). Notably, treatment of BMMs with ZM significantly suppressed M-CSF and RANKL-induced phosphorylation of p65, p38, ERK, and JNK (Fig. 5E), indicating that ZM targets these signaling cascades to exert its inhibitory effect on osteoclast differentiation. The results of WB demonstrate that ZM-306416 exerts time-specific inhibition on RANKL-induced NF-κB/MAPK pathway activation, which aligns with the cascade’s physiological timeline: at 5 min (early response), it reduced p-P65 (NF-κB) phosphorylation to block initial RANKL signal initiation; at 10–20 min (peak activation), it attenuated the peak phosphorylation of p-P38, p-ERK1/2, and p-JNK (MAPK) to impair signal amplification for osteoclast precursor differentiation; and at 30 min (early resolution), it sustained this inhibition to limit residual signal effects. This pattern targets the 5–30 min “signal-sensitive window” of RANKL-induced osteoclastogenesis—blocking signal initiation, weakening amplification, and restricting residual signals—and this synergistic, time-specific suppression constitutes a key mechanism of ZM-306416’s anti-osteoclastogenic effect, which aligns with our goal of targeting early signaling events to suppress osteoclast lineage commitment.

Taken together, these findings demonstrate that ZM promotes osteoblast differentiation by activating the canonical Wnt/β-catenin pathway and inhibits osteoclastogenesis by suppressing NF-κB/MAPK signaling, supporting our conclusion that ZM regulates bone remodeling via these two distinct pathways.

Discussion

In this study, we performed high-throughput drug library screening and identified ZM as a potential candidate that promotes osteoblast differentiation. Notably, ZM has dual effects: it not only enhances osteoblast differentiation but also inhibits osteoclast differentiation. Consistent with these in vitro findings, ZM was further shown to prevent OVX-induced bone loss in vivo, highlighting its potential to regulate bone remodeling bidirectionally. This identification of ZM relied on high-throughput drug library screening, which laid the foundation for our subsequent mechanistic studies.

High-throughput drug library screening is a valuable approach in drug discovery and development. It offers several advantages, including rapid screening of a large number of compounds, which significantly reduces the time compared with traditional methods. This enables researchers to rapidly identify potential drug candidates, explore new targets, and discover drugs for previously unexplored therapeutic areas. However, high-throughput screening requires substantial manpower and material resources. Despite these limitations, it remains a valuable tool in the early stages of drug discovery when used in conjunction with other approaches. After confirming the initial activity of ZM by screening, we further explored its molecular mechanisms through in vitro experiments.

Through in vitro experiments, we demonstrated that ZM activates the Wnt/β-catenin signaling pathway, as evidenced by KEGG enrichment of Wnt signaling-related genes in ZM-treated osteoblasts, alongside upregulation of β-catenin (a central mediator of canonical Wnt signaling) and RUNX2 (a master transcription factor driving osteogenic commitment). Furthermore, the use of ICG-001, a specific inhibitor of the Wnt/β-catenin signaling pathway, inhibited osteoblast differentiation, confirming that activation of the Wnt/β-catenin signaling pathway is indispensable for regulating osteoblast differentiation. Functional validation via ALP staining further confirmed that ZM enhances osteogenic commitment through this cascade. Moreover, ZM inhibits osteoclast differentiation by targeting nuclear factor κB (NF-κB) and mitogen-activated protein kinase (MAPK) signaling. KEGG analysis revealed significant enrichment in the “osteoclast differentiation” pathway in ZM-treated osteoclast precursors, a pathway tightly regulated by the NF-κB and MAPK cascades downstream of RANKL-RANK activation. We confirmed that ZM suppresses the RANKL-induced phosphorylation of p65 (NF-κB) and MAPKs (p38, JNK, and ERK), which aligns with their known role as core mediators of osteoclastogenesis [44, 45].

To contextualize the dual effects of ZM, we compared it with those of clinically approved dual-acting anti-osteoporosis agents. For example, romosozumab, a sclerostin antibody, promotes bone formation via Wnt pathway activation and inhibits resorption through downstream effects. In contrast, ZM modulates both the Wnt/β-catenin pathway (to increase osteogenesis) and the NF-κB/MAPK pathway (to suppress osteoclastogenesis) directly, suggesting a distinct mechanistic profile that may confer unique therapeutic advantages. This distinct mechanism of action of ZM addresses a longstanding gap in current osteoporosis treatment, as most existing drugs lack dual effects on bone formation and resorption. Unlike Asperuloside, which acts via autophagy/Nrf2 to only promote osteoblast function [46], ZM bidirectionally regulates bone metabolism by activating Wnt/β-catenin (osteoblasts) and inhibiting NF-κB/MAPK (osteoclasts), reinforcing its translational potential.

For several years, the medical community has been conducting research on therapies that aim to increase bone mass by stimulating new bone formation while also inhibiting bone resorption. However, most drugs currently used to treat osteoporosis focus only on inhibiting bone resorption and lack the dual effects of stimulating bone formation [6, 10, 47]. Our findings present a potential promising candidate for the development of anti-osteoporosis drugs. The results of our study suggest that ZM may be an effective therapeutic agent for osteoporosis by simultaneously inhibiting bone resorption and promoting bone formation.

However, this study has limitations that should be acknowledged. First, we did not assess the long-term toxicity of ZM, which is critical for evaluating its translational potential. Second, we only tested ZM in an OVX mouse model and did not evaluate its efficacy in aged mouse models or models of glucocorticoid-induced osteoporosis, which are scenarios that are highly relevant to clinical osteoporosis presentations. Additionally, we did not explore the impact of genetic polymorphisms on ZM’s efficacy—an important gap as genetic background affects postmenopausal osteoporosis (PMOP) susceptibility. GLP-1R gene SNPs (e.g., rs3765468) have been linked to PMOP risk in Chinese Han women [48], suggesting personalized strategies may be needed.

Conclusion

Our study revealed that ZM has a dual anti-osteoporosis effect by promoting osteoblastogenesis and inhibiting osteoclastogenesis via the activation of Wnt/β-catenin signaling and suppression of NF-κB/MAPK cascades. These findings suggest that ZM could be a promising therapeutic agent for alleviating osteoporosis.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

YL: Data curation, Formal analysis, Investigation, Writing – original draft. SS: Data curation, Formal analysis, Resources, Writing – original draft. DY: Formal analysis, Writing – review and editing. LZ: Investigation, Writing – review and editing. TC: Funding acquisition, Writing – review and editing. KX: Software, Methodology, Writing – original draft. WL: Data curation, Writing – original draft. JC: Funding acquisition, Project administration, Resources, Writing –resources, writing, review and editing. JC: Conceptualization, Funding acquisition, Project administration, Writing – review and editing.

Funding

This work was supported by the general project of the Anhui Province Outstanding Young Talents Support Program for Universities (Grant No. gxyq2022009); the Anhui Institute of Translational Medicine (Grant No. 2022zhyx-C90); the Medical Scientific Research Foundation of Guangdong Province, China (Grant No. B2025168); and the Longhua District Medical and Health Institution 2025 District-Level Scientific Research Project (Grant No. 2025036).

Data availability

The raw RNA-seq data were deposited in the NCBI Gene Expression Omnibus (GEO) database under accession numbers GSE307058 and GSE307904. Further inquiries regarding the data can be directed to the corresponding author.

Declarations

Ethics approval and consent to participate

All animal care protocols and experiments were reviewed and approved by the Southern Medical University Laboratory Animal Ethics Committee (Approval No. LAEC-21-123), and this study was compliant with all relevant ethical regulations regarding animal research.

Consent for publication

Not applicable.

Competing Interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.