Hydroxysafflor yellow A attenuatesovariectomy induced bone loss andpromotes osteogenesis via PI3K/AKT activation in an in vitro BMSC model

Abstract

Osteoporosis (OP) is a chronic systemic metabolic bone disorder that requires further exploration of effective treatments. An optimal anti-osteoporosis agent should exert dual effects on both osteoblasts and osteoclasts. Hydroxysafflor yellow A (HSYA), a flavonoid from the traditional Chinese medicinal plant safflower, is recognized as a potential candidate. This study shows that HSYA supports bone formation by stimulating osteoblast differentiation and increasing calcium nodule deposition. HSYA activates the PI3K/AKT signaling pathway and upregulates Runx2 expression in an in vitro osteoblast differentiation model using BMSCs. Concurrently, HSYA inhibit osteoclast differentiation, thereby mitigating the excessive bone resorption associated with osteoporosis. These findings indicate that HSYA mitigates bone loss in OP in vivo, while enhancing osteoblast activity and activating the PI3K/AKT pathway in an in vitro osteoblast differentiation model using BMSCs, and inhibiting osteoclast-driven bone resorption.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-51480-4.

Introduction

Osteoporosis (OP) is a chronic systemic metabolic bone disorder marked by reduced bone mass and deteriorating bone microarchitecture, leading to increased fragility and fracture risk¹. The prevalence of osteoporosis has escalated significantly in conjunction with the aging global population, leading to a substantial rise in the incidence of osteoporotic fractures^2,3. This trend poses significant health and economic challenges for individuals and society.

Bone tissue continuously remodels under physiological conditions, regulated by the balance between bone resorption and formation⁴. Bone resorption is facilitated by osteoclasts, while osteoblasts are responsible for bone formation. An imbalance in this dynamic process, bone resorption is mediated by osteoclasts through the RANKL-RANK signaling pathway, which activates transcription factors such as NFATc1, driving the expression of osteoclast-specific genes including MMP9 and CTSK that facilitate bone matrix degradation. Natural compounds like HSYA may modulate these pathways, thereby inhibiting osteoclast differentiation and activity. characterized by excessive bone resorption and inadequate new bone formation, can precipitate the development of osteoporosis^5,6. Therapeutic strategies for osteoporosis primarily focus on promoting bone formation by osteoblasts and reducing bone resorption by osteoclasts. First-line clinical interventions include synthetic parathyroid hormone (PTH) and the RANKL-targeting antagonist denosumab, which specifically target osteoblasts and osteoclasts, respectively^7,8. The active fragment of PTH (PTH 1–34) mimics the physiological pulsatile secretion of PTH, thereby stimulating osteoblast differentiation and collagen synthesis, which directly promotes bone formation⁹. Denosumab binds to RANKL, preventing its interaction with RANK on osteoclast precursors, thereby inhibiting osteoclast formation, activity, and survival, which reduces bone resorption and rapidly increases bone density¹⁰. Both pharmacological agents have adverse effects: prolonged PTH use may increase osteosarcoma risk and cause hypercalcemia, injection site reactions, and cardiovascular issues; denosumab can lead to hypocalcemia and severe side effects like osteonecrosis of the jaw and atypical femoral fractures¹¹. Additionally, both treatments can elicit withdrawal reactions, often necessitating combination or sequential therapy¹². The multifaceted pathogenesis of osteoporosis typically involves multiple targets, indicating that solely targeting osteoblasts or osteoclasts is inadequate for restoring homeostasis within the osteoporotic microenvironment^13,14. Thus, an optimal anti-osteoporosis pharmacotherapy should exert dual effects on both osteoblasts and osteoclasts.

Hydroxysafflor yellow A (HSYA) is a flavonoid natural product derived from the traditional Chinese medicinal herb safflower¹⁵. Previous studies have shown that HSYA possesses diverse pharmacological properties, such as anti-inflammatory, antioxidant, anti-apoptotic, and anti-tumor effects, and has been used to treat conditions like atherosclerosis, ischemia-reperfusion injury, cancer, and diabetes^16–19. Furthermore, HSYA has demonstrated protective effects against glucocorticoid-induced bone loss²⁰. Previous studies reported that HSYA promotes osteogenesis via epigenetic regulation of Wnt/β-catenin and prevents OVX-induced bone loss in rodents²¹, and protects against thioacetamide-induced osteopenia in zebrafish²². Building on these findings, our study demonstrates that HSYA simultaneously enhances osteoblast differentiation and suppresses osteoclastogenesis in vitro, and mitigates OVX-induced bone loss in mice, with PI3K/AKT signaling identified as a key mechanistic pathway. We propose that HSYA, due to its anti-inflammatory and antioxidant properties, may inhibit osteoclast-mediated bone resorption and enhance osteoblast differentiation, positioning it as a potential therapeutic agent for osteoporosis.

The aim of our study is to examine the impact of HSYA on osteoporosis and clarify its underlying mechanisms. Our findings suggest that HSYA stimulates bone marrow stromal cell proliferation, significantly enhances osteoblast differentiation and osteogenic activity, and promotes extracellular matrix mineralization. HSYA promotes osteogenesis by activating the PI3K/AKT signaling pathway and upregulating Runx2 expression in an in vitro osteoblast differentiation model using BMSCs. Simultaneously, HSYA mitigates bone resorption and osteoporotic bone loss by inhibiting osteoclast differentiation. Animal studies showed that HSYA effectively mitigates bone loss post-ovariectomy by enhancing bone formation and suppressing osteoclast-driven resorption. Our research provides new insights into pharmacological strategies for treating osteoporosis.

Materials and methods

Isolation and culture of primary cells

Animal experiments received approval from the Ethics Committee of Taicang TCM Hospital Affiliated to Nanjing University of Chinese Medicine(approval number: 2024026)and were conducted following the Guidelines for the Care and Use of Laboratory Animals. Bone marrow stromal cells (BMSCs) were extracted from the tibiae and femora of female C57BL/6 mouse (8 weeks old, housed under specific pathogen-free (SPF) conditions, with controlled temperature (22 ± 2 °C), humidity (50–60%), and a 12-hour light/dark cycle, with ad libitum access to food and water). The cells were cultured for three passages in Dulbecco’s Modified Eagle’s Medium DMEM, HyClone, Logan, USA, Cat#SH30243.01) with 20% fetal bovine serum and 1% penicillin-streptomycin to obtain a pure BMSC population for further experiments. Bone marrow macrophages (BMMs) were isolated from the femora and tibiae of C57BL/6 mice and cultured for 72 h in a medium with 10% FBS, 1% P/S, and 30 ng/mL macrophage colony-stimulating factor (M-CSF) (R & D Systems, Minneapolis, MN) to prepare them for further experiments.

Cytotoxicity analysis

BMSCs were seeded at 2000 cells per well in a 96-well plate and incubated overnight for adhesion. HSYA (Absin Bioscience Inc. Shanghai, China, CAS 78281-02-4, dissolved in DMSO) was administered at concentrations of 0, 1, 10, 25, 50, 100, 500, and 1000 µM for 24, 48, and 72 h. After discarding the medium and subsequent rinsing with phosphate-buffered saline (PBS), CCK-8 solution (Vazyme) was introduced, and absorbance at 450 nm was measured. All cytotoxicity assays were performed with three independent biological replicates, and each measurement was conducted in triplicate (technical replicates) to ensure reproducibility.

Osteogenic differentiation

BMSCs were seeded in 24-well plates at a density of 2 × 10⁴ cells per well and cultured in complete medium at 37 °C with 5% CO₂. After allowing the cells to attach and reach approximately 70–80% confluence (about 24 h after seeding), the culture medium was substituted with osteogenic induction medium consisting of standard culture medium supplemented with 50 µg/mL ascorbic acid, 10 mM β-glycerophosphate, and 100 nM dexamethasone (Sigma-Aldrich, USA), and subsequently treated with HSYA at concentrations of 10 µM, 50 µM, and 100 µM. Osteogenic differentiation was assessed on day 7 by alkaline phosphatase (ALP) staining. ALP activity was quantified using a Beyotime alkaline phosphatase assay kit (China) according to the manufacturer’s instructions. The measured ALP activity was normalized to total protein concentration determined using a BCA protein assay kit, and results were expressed as ALP activity per µg of total protein, following the manufacturer’s protocol. After 7 days of osteogenic induction, BMSCs were washed with PBS and lysed in RIPA (Beyotime, China) buffer. Cell lysates were centrifuged at 12,000 × g for 10 min at 4°C, and the supernatants were collected. ALP activity was determined by measuring the conversion of p-nitrophenyl phosphate (pNPP) to p-nitrophenol at 405 nm using a microplate reader (Bio-Rad, USA). Calcium nodule formation was evaluated after 21 days of osteogenic induction using Alizarin Red S staining: after 21 days of osteogenic induction, BMSCs were washed three times with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde for 15 min at room temperature. Cells were then stained with 2% Alizarin Red S solution (pH 4.2; Sigma-Aldrich, USA) for 20 min to visualize calcium deposits. Excess dye was removed by washing the cells five times with distilled water. For quantitative analysis, the bound dye was solubilized by adding 10% (w/v) cetylpyridinium chloride (CPC) in 10 mM sodium phosphate (pH 7.0) and incubating for 30 min at room temperature with gentle shaking. The absorbance of the resulting solution was measured at 490 nm using a microplate reader (Bio-Rad, USA), providing a semi-quantitative assessment of mineralized nodule formation. Each osteogenic differentiation assay was independently repeated three times (biological replicates), and all measurements were performed in triplicate (technical replicates).

Gene expression analysis

Total RNA was extracted from BMSCs using TRIzol reagent (Invitrogen, USA, Cat#15596026)) after 5 days of induction in osteogenic medium. RNA concentration and purity were assessed with a NanoDrop-2000 spectrophotometer (Thermo Fisher, USA), followed by reverse transcription into cDNA using HiScript III All-in-one RT SuperMix (Vazyme). The cDNA products were mixed with SYBR Green (Vazyme) and primers for reverse transcription polymerase chain reaction (RT-PCR) (Bio-Rad). The mouse primer gene sequences are shown in Table 1.

Table 1.

Sequences of mouse primers used in qRT-PCR (5′ to 3′).

Primers for qPCR
OPN	F	TTTGTAGGCGGTCTTCAAGC
	R	GTGAGATTCGTCAGATTCATCCG
OCN	F	GAGGGCAATAAGGTAGTGAACAGA
	R	AAGCCATACTGGTTTGATAGCTCG
Col1a1	F	TAAGGGTCCCCAATGGTGAGA
	R	GGGTCCCTCGACTCCTACAT
Osterix	F	GCTGCAAGCTCTCCATAACC
	R	GCCAGAAGCTGTGAAACCTC
ALP	F	CCAACTCTTTTGTGCCAGAGA
	R	GGCTACATTGGTGTTGAGCTTTT
Runx2	F	TTCTCCAACCCACGAATGCAC
	R	CAGGTACGTGTGGTAGTGAGT
MMP9	F	GGACCCGAAGCGGACATTG
	R	CGTCGTCGAAATGGGCATCT
CTSK	F	CTTCCAATACGTGCAGCAGA
	R	TCTTCAGGGCTTTCTCGTTC
ATP6v0d2	F	GACCCTGTGGCACTTTTTGT
	R	GCTTGCATTTGGGGAATCTATC
GAPDH	F	CATCACTGCCACCCAGAAGACTG
	R	ATGCCAGTGAGCTTCCCGTTCAG

Western blot assays

Total proteins were extracted from BMSCs using RIPA lysis buffer after 5 days of induction in osteogenic medium. Protein levels were quantified using a BCA kit, followed by equal protein separation via SDS-PAGE and transfer to polyvinylidene fluoride membranes. After blocking with 5% non-fat milk, the membranes were incubated overnight at 4 °C with rabbit-derived primary antibodies against the indicated target proteins (abcam). After washing, the membranes were incubated with HRP-conjugated goat anti-rabbit IgG secondary antibody, protein bands were visualized using a Monad chemiluminescent imaging system, followed by semi-quantitative analysis with ImageJ. Full-length, uncropped, and unprocessed images of all gels and blots are provided in the Supplementary Information.

Transcriptome sequencing analysis

Total RNA was extracted from BMSCs using TRIzol reagent (Invitrogen, USA, Cat#15596026) according to the manufacturer’s instructions. RNA purity and integrity were evaluated using a NanoDrop 2000 spectrophotometer (Thermo Fisher, USA) and an Agilent 2100 Bioanalyzer (Agilent Technologies, USA). RNA‑seq libraries were prepared using the TruSeq RNA Sample Preparation Kit v2 (Illumina, USA, Cat#RS-122-2002) following the standard Illumina protocol for poly(A) + RNA library construction. Transcriptomic sequencing (RNA-seq) analysis was performed on cells from both the osteogenic induction (OIM) group and the OIM + HSYA group. After the extraction of total RNA, Shanghai OE Biotech Co., Ltd (China) conducted the RNA-seq. We identified the differentially expressed genes (DEGs) as a P value < 0.05 and a fold change ≥ 2. The identified DEGs underwent enrichment analysis utilizing the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases, with Gene Set Enrichment Analysis (GSEA) executed via the clusterProfiler package.

Rescue experiment

In the PI3K inhibition experiment, BMSCs were pretreated with 10 nM Copanlisib prior to induction with osteogenic medium for osteogenic differentiation, with 50 µM HSYA administered. ALP and ARS staining were performed at 7 and 21 days of induction, respectively, to evaluate osteoblast viability and calcium nodule deposition. RT-PCR and Western Blot analyses were performed after 5 days of culture to evaluate the expression changes of genes and proteins linked to osteogenic differentiation. All rescue experiments were conducted in three independent BMSC cultures (biological replicates), with measurements in triplicate (technical replicates).

Immunofluorescence staining

BMSCs and BMMs were washed three times with PBS and fixed with cold paraformaldehyde for 15 min, followed by additional PBS washes. Cells were then blocked in immunofluorescence (IF) blocking solution for 1 h at room temperature. Subsequently, cells were incubated overnight at 4 °C with the following primary antibodies: anti-F4/80 (1:500)(Rat monoclonal, Abcam, UK, Cat#ab6640, RRID: AB_1140045), anti-RUNX2 (1:500)(Rabbit polyclonal, Abcam, UK, Cat#ab236639, RRID: AB_2921286) anti-Col1a1 (1:7000)(Rabbit polyclonal, Abcam, UK, Cat#ab34710, RRID: AB_732879), and anti-MMP9 (1:500)(Rabbit polyclonal, Abcam, UK, Cat#ab76003, RRID: AB_1524058); all antibodies were obtained from Abcam. After incubation, cells were rinsed with PBS for 5 min. For actin staining, cells were treated with DY-554 phalloidin. Corresponding secondary antibodies—goat anti-rat IgG H&L (Alexa Fluor 488)(Alexa Fluor 488, Abcam, UK, Cat#ab150157, RRID: AB_2633057) and goat anti-rabbit IgG H&L (Alexa Fluor 647)(Alexa Fluor 647, Abcam, UK, Cat#ab150079, RRID: AB_2753208)—were applied and incubated at 37 °C for 60 min. Finally, cells were mounted using antifade medium with DAPI(Sigma‑Aldrich, USA, Cat#D9542) and imaged under a confocal laser scanning microscope.

Osteoclast differentiation

Bone marrow macrophages (BMMs) were plated in 24-well plates at a concentration of 100,000 cells per well. Following adherence, the medium was substituted with an osteoclastic medium, which is the standard culture medium enhanced with 30 ng/mL M-CSF and 50 ng/mL RANKL (R&D Systems, USA). Seven days post-osteoclast induction, TRAP staining was executed with a TRAP staining kit (Beyotime, China), followed by semi-quantitative analysis via ImageJ. RT-PCR and Western Blot analyses were conducted after 3 days of osteoclast induction to assess gene and protein expression changes associated with osteoclast differentiation. Each osteoclast differentiation experiment was performed in three independent BMM cultures (biological replicates), with all measurements repeated three times (technical replicates).

Establishment and intervention of OVX-induced osteoporosis model

Female C57BL/6 mouse (8 weeks old, 18–22 g) were housed under specific pathogen-free (SPF) conditions, with controlled temperature (22 ± 2 °C), humidity (50–60%), and a 12-h light/dark cycle, with ad libitum access to food and water. A mouse model of osteoporosis was established using female C57BL/6 mouse (8 weeks old). Bilateral ovariectomy (OVX) was performed under anesthesia to induce osteoporosis, while mice in the sham group underwent ovarian exposure only, followed by muscle and skin suturing. After surgery, animals were placed on a constant-temperature heating pad until recovery. Penicillin was administered intraperitoneally for 3 consecutive days postoperatively to prevent infection. Mice subjected to bilateral ovariectomy, excluding the sham surgery group, were categorized into the OVX group (OVX), low-dose HSYA treatment group (OVX + Low-HSYA 10 mg/kg), and high-dose HSYA treatment group (OVX + High-HSYA 100 mg/kg). HSYA was dissolved in sterile 0.9% normal saline. Mice in the OVX group received an equal volume of sterile 0.9% normal saline as the vehicle control. HSYA was administered orally at a dose of 50/100 mg/kg twice weekly for 4 weeks. This dosing regimen was selected based on the relatively low oral bioavailability and rapid metabolism of HSYA, as reported in previous pharmacokinetic studies. Twice-weekly administration ensures adequate systemic exposure while minimizing handling stress to the animals. The dose was further guided by prior in vivo studies demonstrating bone-protective efficacy without observable toxicity. All treatments were performed at consistent times to reduce variability (Scheme 1).

Scheme 1.

Schematic showing inhibition effect of HSYA on OVX-induced osteoporosis.

Micro-CT

After 8 weeks of establishing the OVX model, the mice were anesthetized and euthanized, and the femur and tibia were dissected. Micro-CT analysis was performed using the SkyScan 1176 system (Bruker, Belgium) following fixation in 4% paraformaldehyde. The scanning parameters were set to 9 μm per layer, with a voltage of 50 kV and a current of 800 µA. Subsequent three-dimensional image reconstruction enabled the measurement of bone morphometric parameters such as bone volume/total volume (BV/TV, %), trabecular thickness (Tb.Th, mm), trabecular separation (Tb.Sp, mm), and bone mineral density (BMD, mg/cm³).

Histological analysis

Collect the tissue and fix it in 10% formalin for 48 h. The bone tissue was decalcified in 10% ethylenediaminetetraacetic acid (EDTA, Sigma Aldrich) for 4 weeks. Then dehydrate the tissue and embed it in paraffin. The tissues were sectioned sagittally into 5 μm-thick sections using a microtome (Leica RM2135, Germany). Place the tissue slices in xylene to dissolve the wax. Then rehydrate the slices in ethanol solution. After cleaning, immerse the slices in water and stain with hematoxylin dye (Leadene, Beijing, China) for 3 min. Then perform color separation using 1% hydrochloric acid in ethanol and ammonia. Then immerse these parts in eosin dye (Leadene), wash and dehydrate with ethanol and xylene. For IHC staining, tissue sections were immersed in 5% catalase for antigen repair and incubated at 37 °C for 10 min, then incubated with primary antibodies against osteogenic and osteoclast related factors, including Col1a1 (1:500), OPN (1:1000), and MMP9 (1:500; all from Abcam, Cambridge, UK), overnight at 4 °C. According to the primary antibody host, the secondary antibody (VECTOR, Burlingame, California, USA) is used to bind with the primary antibody. Then perform dimethylaminophenethylamine (DAB, VECTOR) staining to stain positive cells. Obtain microscopic images using an inverted optical microscope (OLYMPUS, Japan). Use Image J software for quantitative analysis. For quantitative evaluation, five random non-overlapping fields per section were selected under ×200 magnification. The percentage of positively stained area and the integrated optical density (IOD) were quantified using ImageJ software (NIH, USA). The mean value from three independent sections per sample was calculated for statistical analysis. All image acquisition and quantitative analyses were performed in a blinded manner.

For all experiments, including ALP staining and activity assay, Alizarin Red S staining, TRAP staining, and immunohistochemistry, at least three biological replicates were performed, and each measurement was conducted in triplicate as technical replicates to ensure reproducibility. Quantitative analyses were based on these replicates. For in vivo experiments, 6 mice per group were used as biological replicates. Quantitative analyses were conducted on at least three tissue sections per mouse, and each measurement was repeated in triplicate (technical replicates).

Statistical analysis

All data are presented as mean ± standard deviation. Comparisons between two groups were conducted using the Student’s t-test, For multiple group comparisons, one-way ANOVA followed by Tukey’s post-hoc multiple comparisons test was performed using GraphPad Prism 8.1. A significance threshold was established at p < 0.05.

Results

HSYA enhances osteogenesis of BMSCs

The chemical structure of HSYA is illustrated in Fig. 1A. Initially, we assessed the cytotoxicity of HSYA utilizing the CCK-8 assay. The findings from the CCK-8 assay revealed that varying concentrations of HSYA did not demonstrate significant cytotoxic effects on BMSCs. Notably, a moderate concentration of HSYA (50 µM) was found to enhance the proliferation of BMSCs, whereas higher concentrations diminished this effect (Fig. 1B). We examined the impact of HSYA on the osteogenic differentiation of BMSCs. ALP staining results after seven days of osteogenic induction indicated that HSYA augmented both ALP staining and ALP activity in a concentration-dependent manner (Fig. 1C, D). ARS staining and quantitative analysis after 21 days of osteogenic induction showed that HSYA significantly enhanced calcium nodule deposition, with the 50 µM concentration having the greatest impact (Fig. 1E, F).

Fig. 1.

The impact of HSYA on the osteogenic differentiation of bone marrow-derived stem cells (BMSCs). (A) Depicts the chemical structure of HSYA. (B) CCK-8 assay was conducted to evaluate the cytotoxicity of HSYA. (C, E) alkaline phosphatase (ALP) staining and Alizarin red (ARS) results are presented to assess the Osteogenic differentiation of BMSCs in osteogenic induction medium (OIM). (D) Analysis of ALP activity (diethanolamine value). (F) ARS staining results are shown quantitative assessment of calcium nodule formation. All bar graphs are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, n = 3 per group.

We conducted an analysis of gene expression related to osteogenic differentiation after administering HSYA treatment. Consistent with the ALP and ARS staining outcomes, HSYA treatment significantly upregulated the expression of genes associated with osteogenic differentiation, such as Runx2, Osterix, OCN, OPN, Col1a1, and ALP (Fig. 2A–F). We validated these findings by evaluating osteogenic-related protein levels after HSYA treatment through Western Blot analysis. The results indicated that HSYA effectively increased the protein expression of Col1a1, OPN, Runx2, and Osterix, thereby promoting osteogenic differentiation (Fig. 2G–K). Collectively, these findings suggest that HSYA facilitates the osteogenesis of BMSCs and holds potential for the treatment of osteoporosis.

Fig. 2.

HSYA enhances the expression of genes and proteins associated with osteogenic differentiation. (A–F) RT-PCR analysis was performed to assess the relative mRNA expression levels of Runx2, osterix, OCN, OPN, Col1a1, ALP. (G–K) Western blot analysis was utilized to measure the protein expression levels of Col1a1, OPN, Runx2, and Osterix. (total protein was 20 µg) (n = 3). The data are presented as the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant. SD, standard deviation.

To further investigate the influence of HSYA on osteogenesis, we developed a model of osteoblast differentiation from BMSCs induced by osteogenic induction medium (OIM) and administered varying concentrations of HSYA. Immunofluorescence staining was performed to visualize the localization and relative fluorescence intensity of Runx2 and Col1a1. Immunofluorescence staining results showed that compared to the OIM group, the fluorescence intensity of Runx2 and Col1a1 in HSYA-treated osteoblasts was significantly elevated (Fig. 3A, B). Immunofluorescence staining revealed that RUNX2 was predominantly localized in the nucleus, consistent with its role as a transcription factor regulating osteoblast differentiation. In contrast, Col1a1 fluorescence was mainly cytoplasmic and extracellular, reflecting its function as a secreted extracellular matrix protein contributing to bone matrix formation. This nuclear versus cytoplasmic/extracellular distribution provides insight into the functional roles of these proteins during osteogenesis. Importantly, the quantitative changes in Runx2 and Col1a1 expression were confirmed by Western blotting and RT-PCR, which provide robust measurement of protein and mRNA levels. Together, these results suggest that HSYA promotes osteoblast differentiation and extracellular matrix formation, with IF images serving as visual confirmation of protein distribution.

Fig. 3.

HSYA can effectively promotes osteoblast differentiation by promoting Runx2 and Col1a1 expression. (A–D) Immunofluorescence staining of Collagen-I and RUNX2 at day 4 after osteogenic differentiation. Scale bar, 100 μm. N = 3. All data were expressed as the mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001, ns, not significant. SD, standard deviation.

HSYA stimulates the PI3K/AKT signaling pathway in an in vitro osteoblast differentiation model using BMSCs

To further elucidate the mechanism by which HSYA promotes osteogenic differentiation, we performed transcriptome sequencing analysis on the osteogenic induction group (OIM) and the HSYA-treated osteogenic induction group (OIM+HSYA). The volcano plot and heatmap analyses revealed a total of 1932 upregulated differentially expressed genes (DEGs) and 2398 downregulated DEGs between the two groups (Fig. 4A, B). KEGG enrichment analysis^23–25 of the DEGs revealed significant changes in the PI3K/AKT signaling pathway after HSYA intervention (Fig. 4C). Gene set enrichment analysis validated that HSYA activated the PI3K/AKT signaling pathway activation (Fig. 4D). We used Western Blot analysis to confirm transcriptome sequencing results by evaluating the activation of the PI3K/AKT signaling pathway following HSYA treatment. The study found that HSYA treatment notably increased PI3K and AKT phosphorylation, enhancing the PI3K/AKT signaling pathway activation, with the 50 µM concentration showing the greatest impact (Figs. 4E–G).

Fig. 4.

HSYA enhances osteogenic differentiation in BMSCs by activating the PI3K/AKT signaling pathway. (A) A volcano plot illustrating differentially expressed genes (DEGs). (B) Heatmap representation of DEGs. (C) KEGG enrichment analysis results. (D) Gene Set Enrichment Analysis (GSEA) targeting the PI3K/AKT signaling pathway. (E–G) Western blot analysis was conducted to assess the protein expression levels of phosphorylated PI3K (p-PI3K), total PI3K, phosphorylated AKT (p-AKT), and total AKT. (total protein was 20 µg) (n = 3). The data are presented as the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ns, not significant. SD, standard deviation.

HSYA activates the PI3K/AKT signaling pathway to promote osteogenic differentiation of BMSCs

Transcriptome sequencing and Western Blot analyses revealed that HSYA activates the PI3K/AKT signaling pathway. We hypothesized that inhibiting PI3K activation could mitigate the beneficial effects of HSYA on osteogenic differentiation. To further investigate this hypothesis, we conducted rescue experiments. Copanlisib, a selective PI3K inhibitor²⁶, was utilized to effectively inhibit PI3K activation, as illustrated in Fig. 5A, B. We administered Copanlisib concurrently with HSYA during the osteogenic differentiation process. As anticipated, the introduction of Copanlisib partially reversed the promoting effects of HSYA on osteogenic differentiation, as confirmed by ALP staining and ALP activity analysis results (Fig. 5C, D). ARS staining and quantitative analysis revealed that PI3K inhibition notably diminished the impact of HSYA on enhancing osteoblast mineralization and calcium nodule deposition (Fig. 5E, F).

Fig. 5.

Osteogenic differentiation is hindered by the inhibition of the PI3K/AKT signaling pathway. (A, B) Chemical structure of Copanlisib. (C, E) ALP staining and ARS results are presented to assess the Osteogenic differentiation of BMSCs in OIM. (D) Analysis of ALP activity (diethanolamine value). (F) ARS staining results are shown quantitative assessment of calcium nodule formation. All bar graphs are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, n = 3 per group. NS, not statistically significant.

Additionally, we examined the expression changes of osteogenic differentiation-related genes and proteins due to PI3K inhibition through RT-PCR and Western Blot analysis. The RT-PCR results demonstrated that Copanlisib effectively inhibited the promoting effects of HSYA on the gene expression of Runx2, Osterix, OCN, OPN, Col1a1, and ALP (Fig. 6A–F). Interestingly, the combination of HSYA + Copanlisib did not fully reverse the inhibitory effects of Copanlisib, resulting in expression levels that were intermediate between HSYA alone and Copanlisib alone. Concomitantly, protein expression of Col1a1, OPN, Runx2, and Osterix showed a similar trend(Figs. 6G–K). Immunofluorescence staining results further corroborated the effects of HSYA and Copanlisib on osteoblast differentiation, revealing that compared to the OIM group, the fluorescence intensity of Runx2 and Col1a1 in HSYA-treated osteoblasts was significantly elevated, while the fluorescence intensity in the HSYA+Copanlisib group was markedly reduced (Fig. 7A, B). In addition to supporting the quantitative data from Western blotting and RT-PCR, immunofluorescence images revealed the subcellular localization and relative distribution of Runx2 and Col1a1 under different treatments. This visual information highlighted that HSYA promotes nuclear translocation of Runx2 and extracellular deposition of Col1a1, while Copanlisib co-treatment attenuates these effects, providing direct evidence for the role of PI3K/AKT signaling in osteoblast differentiation. These observations are consistent with the quantitative results from Western blotting and RT-PCR (Figs. 8 and 9), confirming that HSYA enhances osteoblast differentiation through activation of the PI3K/AKT pathway. Thus, IF images provide visual support for the protein distribution and relative changes observed in the quantitative assays.

Fig. 6.

Suppressing the PI3K/AKT signaling pathway diminishes the expression of genes and proteins linked to osteogenic differentiation. (A–F) RT-PCR analysis was performed to assess the relative mRNA expression levels of Runx2, osterix, OCN, OPN, Col1a1, ALP. (G–K) Western blot analysis was utilized to measure the protein expression levels of Col1a1, OPN, Runx2, and Osterix (total protein was 20 µg) (n = 3). The data are presented as the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant. SD, standard deviation.

Fig. 7.

Inhibiting the PI3K/AKT signaling pathway reduces the mean optical density values for Runx2 and Col1a1 expression. (A) Immunofluorescence staining results for Runx2. (B) Immunofluorescence staining results for Col1a1. Scale bar, 100 μm. n = 3. All data were expressed as the mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001, ns, not significant.

Fig. 8.

HSYA inhibits RANKL-induced osteoclastogenesis in bone marrow monocytes (BMMs). (A–C) TRAP staining images and quantitative evaluation. (D–G) RT-PCR analysis of relative mRNA expression levels of osteoclastic genes. (H–L) Western blot analysis measuring the protein expression levels of Atp6v0d2, MMP9, NFATc1 and CTSK. (total protein was 20 µg) (n = 3). The data are presented as the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant. SD, standard deviation.

Fig. 9.

HSYA can effectively inhibit osteoclast differentiation by inhibiting MMP9 and NFATc1 expression. (A) Immunofluorescence staining results for MMP9. (B) Immunofluorescence staining results for NFATc1. Scale bar, 100 μm (n = 3). The data are presented as the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant. SD, standard deviation.

HSYA inhibits osteoclast differentiation from bone marrow macrophages

To fruther investigate the influence of HSYA on osteoclastogenesis, we developed a model of osteoclast differentiation from BMMs induced by RANKL and administered varying concentrations of HSYA. TRAP staining results indicated that HSYA dose-dependently inhibited osteoclastogenesis, significantly reducing both the area and number of osteoclasts (Fig. 8A–C). We also assessed the impact of HSYA on osteoclastogenesis-related gene expression using RT-PCR analysis. The study revealed that RANKL induction significantly increased mRNA levels of Atp6v0d2, MMP9, NFATc-1 and CTSK compared to the control group, while HSYA treatment effectively reduced these levels in a dose-dependent manner (Fig. 8D–G). Western blot and semi-quantitative analyses confirmed that HSYA dose-dependently decreased the protein levels of Atp6v0d2, MMP9, NFATc-1 and CTSK, thus inhibiting osteoclast activation (Fig. 8H–L).

Immunofluorescence staining results further corroborated the effects of HSYA on osteoclast differentiation, revealing that compared to the RANKL group, the fluorescence intensity of MMP9 and NFATc1 in HSYA-treated osteoclasts was significantly reduced (Fig. 9A, B). NFATc1 was mainly localized in the nucleus after osteoclast differentiation, indicating activation of osteoclast-specific transcriptional programs. MMP9, a secreted matrix-degrading enzyme, showed predominantly cytoplasmic and extracellular staining, reflecting its role in bone matrix remodeling. The distinct subcellular localizations of NFATc1 and MMP9 highlight their respective functional contributions to osteoclast activity. In conclusion, the findings suggest that HSYA can effectively inhibit osteoclast differentiation.

HSYA enhances collagen deposition and osteogenic marker expression in vivo and prevents bone loss in an ovariectomized mouse model

To further investigate the in vivo therapeutic effects of HSYA on osteoporosis, we established an ovariectomy-induced osteoporosis model and administered HSYA. Micro-CT 3D reconstruction analysis showed a significant decrease in femoral bone mass in the OVX group compared to the Sham group. HSYA treatment increased femoral bone mass, with the high-dose group showing the greatest effect (Fig. 10A). Our analysis of bone structural parameters revealed a significant decrease in BV/TV, Tb.Th, and BMD, alongside a notable increase in Tb.Sp, in the OVX group, confirming bone mass loss in the femoral tissue. The HSYA treatment group showed improvements in BV/TV, Tb.Th, Tb.Sp, and BMD compared to the OVX group (Fig. 10B–G). In conclusion, these results substantiate that HSYA can effectively mitigate bone mass loss in OVX mice.

Fig. 10.

HSYA mitigates bone loss induced by ovariectomy (OVX). (A) micro-CT three-dimensional reconstruction of the coronal images of the knee 2 months after OVX. (B–G) Evaluation of bone structural parameters include volume/total volume (BV/TV), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp, bone mineral density (BMD), number of porous and area of porous (n = 6). The data are presented as the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant.

To further evaluate osteogenic activity in vivo, Masson’s trichrome staining and immunohistochemical analyses were performed on femoral sections. Masson staining demonstrated reduced collagen fiber deposition in OVX mice compared with sham-operated controls, indicating impaired bone matrix formation. HSYA treatment markedly enhanced collagen deposition, with the high-dose group showing the most pronounced improvement (Fig. 11A). Quantitative analysis confirmed a significant reduction in newly formed collagen area in OVX mice, which was partially restored following HSYA administration. HSYA treatment significantly increased the number of OPN and Collagen-I-positive cells. Semi-quantitative analysis supported these findings (Fig. 11B–F). These results indicate that OVX impaired osteogenic activity, while HSYA promoted bone matrix formation in vivo.

Fig. 11.

HSYA reverse the bone loss caused by ovariectomy (OVX) in vivo. (A) Compared with the OVX group, representative Masson staining images of subchondral bone in the knee joint showed darker blue collagen fibers in the low-dose and high-dose groups. (B, C) HSYA improved the trabecular structure and stimulated the expression of col1a1 and OPN in both low-dose and high-dose groups. Scale bar = 50 μm, The average density of antibody positive cells is calculated as the integrated optical density of positive cells per unit area (n = 3). The data is expressed as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant. SD, Standard deviation;

To further investigate whether HSYA influences osteoclast differentiation, we performed TRAP staining and immunohistochemical analysis on femoral sections from the OVX murine model. TRAP staining revealed a higher number of TRAP-positive (TRAP⁺) osteoclasts in the OVX group compared to sham controls. This increase was markedly attenuated in both low- and high-dose HSYA-treated groups (Fig. 12A&C). Quantitative analysis further indicated a significant rise in the number of osteoclasts per bone surface (OCs/BS) in OVX mice, which was effectively reversed by HSYA treatment at both dosage levels (Fig. 12D). HSYA administration substantially reduced the expression of both markers, with the most pronounced suppression observed in the high-dose group (Fig. 12B). Quantitative assessment of positive cell numbers further validated these expression changes (Fig. 12E).

Fig. 12.

HSYA attenuated OVX-induced osteoclastic bone resorption in vivo. (A) TRAP staining. (B) Immunohistochemical staining of MMP-9. (C) Number of TRAP-positive cells. (D) Oc.S/BS (%). (E) Number of MMP-9 positive cells. Scale bar = 50 μm. The average density of antibody positive cells is calculated as the integrated optical density of positive cells per unit area (n = 3). The data is expressed as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant. Trap: tartaric acid phosphatase.

Discussion

Osteoporosis is a significant global public health concern impacting millions^27–29. The main treatment medications for osteoporosis include anti-resorptive drugs (such as bisphosphonates, estrogen, calcitonin) and anabolic agents (such as parathyroid hormone, teriparatide)³⁰. Prolonged use of these medications may cause serious side effects, including gastrointestinal issues and osteonecrosis, which, along with poor patient compliance, impede effective long-term osteoporosis management³¹. Consequently, it is crucial to investigate novel, safe, and effective medications for osteoporosis prevention and treatment. In this study, we discovered a new candidate drug, HSYA, which can be used for the treatment of osteoporosis.

Osteoporosis results from the excessive activation of osteoclasts and the inhibition of osteoblast function, leading to an imbalance in bone remodeling. Due to the interrelationship between osteoclasts and osteoblasts, existing treatment methods often simultaneously inhibit or promote both osteoclasts and osteoblasts^32,33. Therefore, there is a need to develop dual-action drugs that promote new bone formation while inhibiting bone resorption³⁴. Natural products are a valuable resource for the treatment of osteoporosis, and HSYA can exert anti-osteoporotic effects by inhibiting osteoclastogenesis and promoting osteoblast differentiation. First, we found that 50 µM of HSYA can promote the proliferation of BMSCs and increase cell viability. Despite higher concentrations diminishing the effect, HSYA demonstrated no significant cytotoxicity relative to the control group, indicating its safety for both in vitro and in vivo applications. Subsequently, we further confirmed the promoting effect of HSYA on osteoblast differentiation. It enhanced osteoblast viability, promoted calcium nodule deposition in the extracellular matrix, and significantly upregulated Col1a1, OPN, Runx2, and Osterix gene and protein expression during osteoblast differentiation. Notably, 50 µM of HSYA exhibited the best therapeutic effect. At the same time, we also evaluated the effect of HSYA on osteoclast differentiation. The RANKL-RANK interaction is essential for osteoclastogenesis, prompting bone marrow-derived mononuclear cells to differentiate into osteoclasts, thereby playing a key role in systemic bone mass maintenance^35,36. HSYA effectively inhibited RANKL-induced osteoclastogenesis, reducing both the number and area of osteoclasts, and significantly lowering the expression of osteoclast differentiation markers Atp6v0d2, MMP9, and CTSK. Furthermore, in the in vivo OVX model, HSYA significantly alleviated bone loss.

In recent years, numerous natural products have been reported to exert anti-osteoporotic effects through regulation of osteoblast and osteoclast activity. For example, icariin has been shown to promote osteogenic differentiation via activation of the PI3K/AKT and Wnt/β-catenin pathways, while suppressing osteoclast formation^37,38. Similarly, resveratrol enhances osteoblast differentiation and inhibits osteoclastogenesis through modulation of oxidative stress and inflammatory signaling pathways³⁹. Other natural compounds, such as quercetin and naringin, have also demonstrated dual regulatory effects on bone remodeling by targeting key signaling cascades including NF-κB, MAPK, and PI3K/AKT pathways^40,41. Compared with these previously reported phytochemicals, our study provides evidence that HSYA not only promotes osteogenic differentiation through activation of the PI3K/AKT–Runx2 axis but also suppresses RANKL-induced osteoclastogenesis in vitro and mitigates OVX-induced bone loss in vivo. Importantly, transcriptomic sequencing combined with pharmacological inhibition experiments further validated the involvement of the PI3K/AKT pathway, strengthening the mechanistic basis of HSYA’s action. These findings position HSYA as a promising dual-functional natural compound with potential translational value for osteoporosis management.

We performed transcriptome sequencing to investigate how HSYA facilitates osteogenic differentiation, revealing that it activates the PI3K/AKT pathway. The PI3K/AKT pathway, activated by various membrane receptors and growth factors, is crucial for promoting osteoblast differentiation^42,43. Prior research has established that activating the PI3K/AKT pathway facilitates osteoprogenitor cell differentiation into osteoblasts and enhances osteoblast differentiation, proliferation, and mineralization via the transcription factor Runx2, thus promoting bone formation^44,45. HSYA enhanced PI3K and AKT phosphorylation, activating the PI3K/AKT signaling pathway, and significantly elevated Runx2 gene and protein expression. To validate the impact of HSYA on the PI3K/AKT signaling pathway, a PI3K inhibitor was administered concurrently with HSYA treatment. We found that this not only significantly weakened the beneficial effects of HSYA on osteoblast viability and calcium nodule deposition but also reduced the expression of Runx2. Notably, while Copanlisib alone significantly inhibited the expression of osteogenic related markers, the combination of HSYA and Copanlisib did not completely reverse Copanlisib’s inhibitory effects. Specifically, for OPN and Runx2, the HSYA + Copanlisib group showed intermediate levels between HSYA alone and Copanlisib alone. This may be due to the potent inhibition of PI3K signaling by Copanlisib, which cannot be fully overcome by HSYA, suggesting that HSYA partially relies on PI3K/AKT activation for promoting osteoblast differentiation. These findings highlight the mechanistic dependence of HSYA on PI3K/AKT signaling while also indicating potential compensatory pathways that may mediate HSYA’s residual effects. While our in vitro experiments in BMSCs clearly demonstrate that HSYA activates the PI3K/AKT signaling pathway to promote osteoblast differentiation via Runx2, we did not directly assess PI3K/AKT activation in the OVX mouse model. Nevertheless, HSYA treatment in OVX mice significantly enhanced bone formation and suppressed osteoclast activity, consistent with the in vitro mechanistic findings. These results suggest that PI3K/AKT activation may contribute to the osteogenic effects of HSYA in vivo, although further studies are needed to confirm its direct involvement in the bone microenvironment of OVX mice. Runx2 is a crucial transcription factor that regulates the differentiation of mesenchymal stem cells into osteoblasts⁴⁶. It regulates early osteoblast differentiation genes, including alkaline phosphatase, facilitating the transition of osteoprogenitor cells to pre-osteoblasts^46,47. Additionally, Runx2 promotes the differentiation of osteoblast precursors into mature osteoblasts by regulating the expression of Osterix⁴⁸. HSYA enhances Runx2 and Osterix expression, whereas PI3K/AKT pathway inhibition reduces their expression levels. HSYA stimulates the PI3K/AKT signaling pathway, enhancing osteogenic differentiation in an in vitro osteoblast differentiation model using BMSCs.

This study also has certain limitations. This study established that HSYA enhances Runx2 expression and transcriptional activity via the PI3K/AKT pathway to support osteogenic differentiation. However, further investigation is required to determine if HSYA directly influences Runx2’s transcriptional activity and to elucidate the mechanisms involved in Runx2 upregulation. Secondly, our research demonstrated that HSYA mitigates bone loss post-OVX by suppressing osteoclast activation, yet the precise mechanisms remain unexplored. Oxidative stress and chronic inflammation are significant factors influencing bone loss in osteoporosis, directly leading to enhanced osteoclast bone resorption and suppressed osteogenic function. HSYA has been reported to regulate oxidative stress, but whether it inhibit osteoclast differentiation by affecting the oxidative stress response of osteoclasts remains to be explored. Although Masson’s trichrome staining demonstrated increased collagen deposition, this method primarily reflects extracellular matrix accumulation rather than dynamic bone formation activity. It does not provide direct evidence of bone formation rate. In the absence of dynamic histomorphometric analysis, such as double calcein labeling to assess mineral apposition rate (MAR) and bone formation rate (BFR/BS), the conclusion that HSYA promotes bone formation in vivo should be interpreted with caution. Nevertheless, the micro-CT analysis showing improvements in BV/TV, Tb.Th, and BMD, together with increased expression of osteogenic related markers (OPN and Col1a1) in immunohistochemical staining, collectively support the bone-protective effects of HSYA. Future studies incorporating dynamic bone labeling techniques are warranted to further validate its role in promoting bone formation in vivo. Besides, a limitation of our study is that PI3K/AKT signaling was not directly assessed in vivo. Future studies are warranted to confirm whether HSYA activates this pathway in osteoblasts and osteoclasts within the OVX bone microenvironment. Future studies using double calcein labeling are required to confirm changes in bone formation rate. We will discuss these related topics in more detail in the future.

Conclusion

Our study demonstrates that HSYA enhances osteogenic differentiation and inhibits osteoclastogenesis in vitro, and mitigates bone loss in an ovariectomized mouse model in vivo. Mechanistically, HSYA activates the PI3K/AKT signaling pathway to promote osteogenic differentiation in an in vitro osteoblast differentiation model using BMSCs. However, as dynamic bone histomorphometric analyses such as calcein labeling were not performed, whether HSYA directly promotes bone formation in vivo requires further investigation, and the involvement of this pathway in vivo requires further investigation. These findings suggest that HSYA may function as a promising bone-protective agent for osteoporosis treatment.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

OP

Osteoporosis

HSYA

Hydroxysafflor yellow A

BMSCs

Bone Marrow Stromal Cells

BMMs

Bone Marrow Macrophages

PTH

Parathyroid Hormone

OIM

Osteogenic Induction Medium

ALP

Alkaline Phosphatase

ARS

Alizarin Red S

PBS

Phosphate Buffered Saline

TRAP

Tartrate-Resistant Acid Phosphatase

RANKL

Receptor Activator of Nuclear Factor κB Ligand

M-CSF

Macrophage Colony-Stimulating Factor

RIPA

Radio-Immunoprecipitation Assay buffer

BCA

Bicinchoninic Acid

IF

Immunofluorescence

DAPI

4’,6-diamidino-2-phenylindole

PI3K

Phosphatidylinositol 3-Kinase

AKT

Protein Kinase B

DEGs

Differentially Expressed Genes

GSEA

Gene Set Enrichment Analysis

KEGG

Kyoto Encyclopedia of Genes and Genomes

GO

Gene Ontology

OVX

Ovariectomy

BV/TV

Bone Volume / Total Volume

Tb.Th

Trabecular Thickness

Tb.Sp

Trabecular Separation

BMD

Bone Mineral Density

Author contributions

Mingzhou Wu: Conceptualization, Methodology, Software. Wei Wang: Data curation, WritingOriginal draft preparation. Zhongwei Ji: Visualization, Investigation. Kai Zheng: Data curation, Writing-Original draft preparation. Weicheng Zhang: Data curation, Writing-Original draft preparation. Gaoran Ge: Software. Chunyang Fan: Data curationDechun Geng: Supervision. Yaozeng Xu: Software, Validation.**Jun Zhou: Writing- Reviewing and Editing.**.

Funding

This research was supported by the Special Project of Diagnosis and Treatment Technology for Key Clinical Diseases in Suzhou (LCZX202302); the Key Project of “Strengthening Health through Science and Education” in Suzhou (ZDXM2024001); the Natural Science Foundation of Nanjing University of Traditional Chinese Medicine (XZR2024069); the Basic Research Project of Medical Innovative Application in Suzhou (SKYD2023071); the Basic Research Program of Medical Application in Suzhou (SKYD2023238); Suzhou Applied Basic Research (Medical and Health) Youth Project (SYWD2024188); the Basic Research on Medical and Health Application in Taicang (TC2024JCYL14); the National Natural Science Foundation of China (82502879); Natural Science Foundation of Fujian Province (2025J08306); Program of Xiamen Health Commission (2024GZL-QN016).

Data availability

Data availability statement: The data are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics statement

The animal study was reviewed and approved by the Ethics Committee of Taicang TCM Hospital Affiliated to Nanjing University of Chinese Medicine.