Polygonatum sibiricum polysaccharide attenuates BMSC senescence by inhibiting NCOA4-mediated ferritinophagy

Abstract

Context

BMSC senescence is associated with osteoporosis progression. Polygonatum sibiricum polysaccharide (PSP) shows potential anti-senescence activity, but its effects on bone marrow mesenchymal stem cells (BMSCs) remain unclear.

Objective

To investigate the protective effects and molecular mechanisms of PSP against tert-butyl hydroperoxide (TBH)-induced senescence in BMSCs.

Materials and methods

Senescence was induced in primary rat BMSCs with TBH, followed by PSP intervention. Viability was assessed using the Cell Counting Kit-8 assay, while senescence-associated β-galactosidase, alkaline phosphatase, and Oil Red O staining assessed senescence and osteogenic/adipogenic differentiation. Ferritinophagy-associated proteins (NCOA4, FTH1, and LC3) and intracellular Fe²⁺ were quantified using western blot, immunofluorescence, and FerroOrange staining. Mitochondrial function was evaluated using MitoSOX and JC-1 assays and ATP quantification. Cytosolic mtDNA accumulation and the cGAS–STING pathway were assessed using immunofluorescence colocalization, quantitative polymerase chain reaction, and Western blotting. NCOA4 involvement was examined using an NCOA4-overexpressing lentiviral vector.

Results

PSP concentration-dependently attenuated TBH-induced senescence phenotypes in BMSCs, partially restored osteogenic differentiation, and reduced adipogenic differentiation. PSP downregulated NCOA4 expression, attenuating ferritinophagy-associated changes and intracellular Fe²⁺ accumulation. Reduced Fe²⁺ accumulation preserved mitochondrial membrane potential, reduced mitochondrial reactive oxygen species generation, and reduced cytosolic mtDNA accumulation, alongside attenuated cGAS–STING signaling. Lentiviral overexpression of NCOA4 attenuated the protective effects of PSP on mitochondrial homeostasis and cGAS–STING pathway suppression.

Discussion and conclusions

PSP may attenuate BMSC senescence by suppressing NCOA4-mediated ferritinophagy, thereby preserving mitochondrial homeostasis and inhibiting cGAS–STING signaling. These findings provide a rationale for further investigation of PSP as a potential therapeutic strategy for osteoporosis-associated BMSC dysfunction.

Introduction

Osteoporosis (OP) is a progressive systemic skeletal disorder characterized by diminished bone mass and microarchitectural deterioration, thereby increasing bone fragility and fracture risk (Compston et al. 2019; Föger-Samwald et al. 2022). Bone marrow mesenchymal stem cells (BMSCs) play essential roles in bone homeostasis, tissue repair, and regeneration. Pathological senescence is increasingly recognized as an important contributor to the pathogenesis of degenerative skeletal conditions, including OP and impaired fracture healing (Jang et al. 2025; Li and Li et al., 2025; Wang et al., 2025). Under pathological microenvironmental stress, BMSCs undergo senescence-associated phenotypic and functional changes. Crucially, their differentiation balance shifts from osteogenesis toward adipogenesis, contributing to progressive bone loss and increased marrow adiposity, thereby accelerating OP progression (Chen et al. 2025; Zhu et al. 2025; Liu et al. 2026). Consequently, elucidating the molecular mechanisms governing BMSC senescence is essential to identify viable pharmacological interventions and develop targeted therapeutic strategies for OP.

Dysregulation of intracellular iron homeostasis contributes to BMSC senescence. Nuclear receptor coactivator 4 (NCOA4)-mediated ferritinophagy, a selective autophagic pathway, regulates intracellular iron turnover by targeting ferritin heavy chain 1 (FTH1) for lysosomal degradation, thereby releasing free iron (Bengson et al. 2023; Li et al. 2024). Although basal ferritinophagy maintains metabolic equilibrium, pathological stress hyperactivates this process, resulting in excessive intracellular labile ferrous ion (Fe²⁺) accumulation (Wang et al., 2025). Excess Fe²⁺ promotes reactive oxygen species (ROS) generation via the Fenton reaction, leading to mitochondrial membrane depolarization and structural damage (Santana-Codina et al. 2022; Gu et al. 2025; Xiao et al. 2025). Consequently, mitochondrial damage can promote the release of mitochondrial DNA (mtDNA) into the cytosol, serving as a damage-associated molecular pattern (DAMP). Cytosolic mtDNA is recognized by cyclic GMP-AMP synthase (cGAS), activating the stimulator of interferon genes (STING) inflammatory cascade, thereby promoting inflammatory signaling and the senescence-associated secretory phenotype (SASP)(Decout et al. 2021; Luo et al. 2024; Guo et al. 2025; Wu et al. 2025).

Polygonatum sibiricum polysaccharide (PSP) is a bioactive macromolecule derived from the medicinal plant Polygonatum sibiricum. Pharmacological evaluations indicate that PSP exhibits antioxidant, immunomodulatory, and mitochondrial protective properties (Huang et al. 2025; Li and Liu et al. 2025; Li and Yao et al.2025). Previous studies have demonstrated that PSP attenuates bone microarchitectural deterioration in postmenopausal OP by modulating canonical osteogenic pathways and reducing cellular senescence by restoring mitochondrial calcium homeostasis (Du et al. 2016; Zheng 2020; Chen et al. 2024; Gao et al. 2025). However, whether PSP regulates intracellular iron homeostasis, particularly through NCOA4-mediated ferritinophagy, remains unclear. Specifically, whether PSP can limit ferritinophagy-associated iron release, inhibit mtDNA-associated cGAS–STING signaling, and preserve BMSC homeostasis remains unknown.

Accordingly, using a tert-butyl hydroperoxide (TBH)-induced senescence model in primary rat BMSCs, we evaluated the effects of PSP on senescence-associated phenotypes and osteogenic/adipogenic lineage commitment. Furthermore, we investigated the role of NCOA4-mediated ferritinophagy in BMSC senescence and examined its effects on mitochondrial homeostasis and the cGAS–STING signaling pathway.

Methods

Cell culture and treatment

Primary rat BMSCs (VYR00055, Weisai Biological) were thawed and maintained in BMSC-specific complete medium (VYP00006, Weisai Biological) at 37 °C in a humidified 5% CO₂ incubator. When cells reached approximately 90% confluence, they were dissociated with 0.25% trypsin-EDTA (T8150, Solarbio). Trypsinization was terminated by adding complete medium, and cells were collected by centrifugation at 150 × g for 3 min and subcultured. TBH (B802372, Macklin) was diluted in a serum-free medium to prepare a 10 mM stock solution and was used within one week. The stock solution was diluted with culture medium immediately before treatment. To determine the optimal TBH concentration for establishing the senescence model, BMSCs were treated with different concentrations of TBH (25, 50, 100, 200, and 400 μM) for 48 h. Cell viability and morphological changes were evaluated, and 100 μM TBH was selected as the optimal concentration for subsequent experiments. PSP (S27804, Yuanye Bio-Technology) was dissolved directly in complete medium to prepare working solutions at concentrations of 50, 100, and 200 μg/mL. PSP solutions were freshly prepared before each experiment and added directly to the culture medium to treat cells. For pharmacological intervention, the cells were treated with PSP at 50, 100, or 200 μg/mL for 24 h. For the ferritinophagy rescue experiment, BMSCs were treated with 100 μM TBH for 48 h and simultaneously treated with 200 μM ferric ammonium citrate (FAC) and 200 μg/mL PSP for 24 h. For mechanistic validation, adherent BMSCs were transduced with a lentiviral vector overexpressing rat NCOA4 or a FLAG-tagged empty vector (negative control; KH74525, APE) for 48 h. After transduction, the indicated cell groups were treated with 100 μM TBH (48 h), followed by 200 μg/mL PSP (24 h).

Cell viability and cytotoxicity assay

The cytotoxicity of PSP on BMSCs was evaluated using a Cell Counting Kit-8 (CCK-8) assay (BS350A; Biosharp), according to the manufacturer’s protocol. BMSCs were seeded into 96-well plates at 5 × 10³ cells/well, with five technical replicates per condition. After cell attachment, the cells were incubated with varying concentrations of PSP (0, 50, 100, 200, 400, 800, and 1600 μg/mL) for 24 h. Subsequently, 10 μL of CCK-8 reagent was added to each well, and the plates were incubated for 2 h at 37 °C. Optical density (OD) was measured at 450 nm using a microplate reader (800TS; BioTek). Relative cell viability was normalized to the OD of the untreated control group (0 μg/mL) to establish the optimal nontoxic concentration range for subsequent pharmacological interventions.

Intracellular ATP quantification

Intracellular ATP levels were quantified using an enhanced ATP assay kit (MA0440, MeilunBio) according to the manufacturer’s protocol. The washed cells were lysed and centrifuged (10,000 × g, 5 min, 4 °C) to isolate the supernatant. Supernatant aliquots (10 μL) were combined with 100 μL of ATP detection reagent in microplates. Chemiluminescence, expressed as relative light units, was measured using a Synergy H1 microplate reader (BioTek, United States). ATP concentrations were calculated from a standard curve generated using ATP standards ranging from 0.01 to 10 μmol/L ATP. To normalize ATP levels to cellular protein content, the total protein concentration of the corresponding samples was determined using a bicinchoninic acid (BCA) protein assay kit (WLA004, Wanleibio). The final ATP content was normalized to the protein concentration and expressed as nmol/mg protein.

Intracellular Fe²⁺ and mitochondrial membrane potential (MMP) assessment

Intracellular Fe²⁺ levels were assessed using a FerroOrange probe (MX4559; MKBio). Treated BMSCs were washed with phosphate-buffered saline (PBS) and incubated with the FerroOrange probe for 30 min at 37 °C in the dark to detect intracellular Fe²⁺ levels. FerroOrange staining was performed according to the manufacturer’s protocol. The cells were washed and incubated with JC-1 staining solution (BL711A, Biosharp) for 20 min at 37 °C, according to the manufacturer’s protocol. After staining, all samples were washed with designated buffers. Fluorescence images were acquired using an IX53 inverted fluorescence microscope (Olympus, Japan). MMP was quantified by calculating the ratio of red (J-aggregates) to green (J-monomers) fluorescence intensities.

Flow cytometric analysis of mitochondrial ROS (mtROS)

The mtROS levels were measured using the MitoSOX probe (B8808, APE), according to the manufacturer’s protocol. Following treatment, BMSCs were washed with PBS and incubated with 500 nM MitoSOX probe (B8808, APE) for 30 min at 37 °C in the dark. After staining and washing, the cells were dissociated using 0.25% trypsin-EDTA, neutralized, and pelleted by centrifugation (150 × g, 3 min). The cell pellets were resuspended in ice-cold PBS to generate single-cell suspensions for immediate fluorescence acquisition using a NovoCyte flow cytometer (Agilent, United States). Data were analyzed using NovoExpress software, with the mean fluorescence intensity (MFI) serving as a quantitative metric for mtROS accumulation.

Osteogenic and adipogenic lineage induction and phenotypic evaluation

BMSCs were cultured in lineage-specific induction media supplemented with designated PSP concentrations for 7 days. For osteogenesis, cells were maintained in osteogenic induction medium (PD-008, Procell), and the medium was refreshed every 3 days. The cells were cultured using alternating 24-h cycles of adipogenic induction media A and B (PD-010, Procell). After 7 days of induction, osteogenic and adipogenic differentiation were evaluated by histochemical staining. To assess adipogenesis, cells were fixed for 20 min at room temperature and stained with Oil Red O (G1262, Solarbio) for 20 min to visualize intracellular lipid droplets. To assess osteogenesis, cells were fixed in situ for 15 min and incubated with alkaline phosphatase (ALP) staining solution (C3206, Beyotime) for 30 min in the dark. All samples were thoroughly washed with PBS to remove unbound staining solution, and bright-field images were acquired using an IX53 inverted microscope (Olympus, Japan).

Western blot analysis

Total cellular proteins were extracted using RIPA lysis buffer (WLA019, Wanleibio) supplemented with 1% PMSF. Following a 5 min incubation on ice, lysates were centrifuged (12,000 × g, 10 min, 4 °C) to isolate the supernatant. Protein concentrations were determined using a BCA protein assay kit (WLA004, Wanleibio) and an ELX-800 microplate reader (BioTek, USA). Denatured protein samples (40 μg/lane) and prestained protein marker(26616, Thermofisher) were resolved via 10%-15% SDS-PAGE (80 V, 2.5 h) and subsequently transferred onto PVDF membranes (IPVH00010, Millipore) at a constant voltage of 80 V for 1.5 h. Membranes were washed with TBST (WLA025, Wanleibio), blocked with blocking buffer (WLA066, Wanleibio) for 1 h at room temperature, and incubated overnight at 4 °C with the following primary antibodies: cGAS (1:1000, HA500023, HUABIO), STING (1:1000, HA722832, HUABIO), NCOA4 (1:1000, A5695, ABclonal), FTH1 (1:500, WL05360, Wanleibio), LC3 (1:500, WL01506, Wanleibio), Lamin B1(1:750, R24826, zenbio), GAPDH(1:5000, bsm-33033M, Bioss), TOMM20(1:3000, A19403, abclonal), and β-actin (1:2000, WL01372, Wanleibio). Following TBST washes, membranes were incubated with an HRP-conjugated goat anti-rabbit IgG secondary antibody (1:5000, WLA023, Wanleibio) for 45 min at 37 °C. Protein bands were visualized using an ECL substrate (WLA006, Wanleibio), and densitometric quantification was performed using Tanon Image software.

Immunofluorescence analysis and cytosolic mtDNA quantification

Post-treatment BMSCs cultured on coverslips were fixed with 4% paraformaldehyde (80096618, Sinopharm) for 15 min at room temperature and permeabilized with 0.1% Triton X-100 (ST795, Beyotime) for 30 min. Standard PBS washes were performed between all sequential steps. Cells were blocked with 1% BSA (A602440, Sangon Biotech) for 15 min at room temperature. To visualize mitochondrial morphology and the spatial distribution of cellular DNA, coverslips were co-incubated overnight at 4 °C with primary antibodies against TOMM20 (1:100, WL03626, Wanleibio) and DNA (1:50, Sc-58749, Santa Cruz). Following PBS washes, samples were incubated with species-matched secondary antibodies—Cy3-conjugated goat anti-rabbit IgG (1:200, A27039, Invitrogen) and FITC-conjugated goat anti-mouse IgG (1:200, ab6785, Abcam)—for 60 min at room temperature in the dark. Nuclei were counterstained with DAPI (D106471, Aladdin), and coverslips were mounted using an anti-fade mounting medium (S2100, Solarbio). Fluorescence images were acquired utilizing a BX53 microscope (Olympus, Japan) to facilitate subsequent colocalization analysis.

For quantitative analysis of cytosolic mtDNA release, mtDNA was isolated from whole-cell and cytosolic fractions and quantified by qPCR as previously described (Bryant et al. 2022). Briefly, BMSCs were harvested by trypsinization and divided into two aliquots for DNA isolation from whole-cell extracts(WCE) and cytosolic fractions, respectively. For whole-cell extracts, cell pellets were lysed in 10 mM Tris (pH 8.0) containing 1% SDS at 95 °C for 5 min. For cytosolic fractionation, cell pellets were resuspended in Buffer A containing 20 μg/mL digitonin (20 mM HEPES, pH 7.4/KOH, 200 mM sucrose, 50 mM KCl, 10 mM NaCl, 1 mM EGTA, 2 mM MgCl₂, and 1× Halt^™ Protease Inhibitor Cocktail) and incubated on a rotating mixer at 4 °C for 10 min. The supernatant was collected and subsequently centrifuged at maximum speed for 15 min to remove residual organelles, and the resulting supernatant was retained as the cytosolic fraction. Fraction purity was assessed by immunoblotting with antibodies against GAPDH (cytosolic marker), TOMM20 (mitochondrial marker), and Lamin B1 (nuclear marker). Before qPCR analysis, DNA was extracted from both whole-cell and cytosolic fractions. Each extract was treated with 50 μg/mL RNase A at 37 °C for 2 h, followed by treatment with 200 μg/mL proteinase K at 55 °C for 1 h, and DNA was subsequently purified by phenol/chloroform extraction and precipitation. qPCR was performed using Hieff^™ qPCR SYBR^® Green Master Mix (Shanghai Yeasen BioTechnologies, 11201ES03) with primers targeting the mitochondrial D-loop region and the nuclear TERT gene. Potential nuclear DNA contamination in the cytosolic fractions was assessed by qPCR for TERT, with Cq < 30 considered indicative of nuclear DNA contamination. For whole-cell extracts, mtDNA was first normalized to nuclear DNA as ΔCq(whole cell) = Cq(D-loop) − Cq(TERT). Cytosolic mtDNA was then normalized to the corresponding whole-cell value as ΔCq(cytosol) = Cq(D-loop, cytosol) − ΔCq(whole cell). ΔΔCq was calculated relative to the median cytosolic ΔCq of the control group, and relative cytosolic mtDNA levels were expressed as 2^(−ΔΔCq). The primer sequences were as follows: D-loop, forward 5′-GGTTCTTACTTCAGGGCCATCA-3′ and reverse 5′-GATTAGACCCGTTACCATCGAGAT-3′; TERT, forward 5′-GGTCTTCCGCAGCTTGGTTG-3′ and reverse 5′-CAGCAGGTAGAGCGCACAGT-3′.

Statistical analysis

Statistical analyses were conducted using the Statistical Package for the Social Sciences software (version 26.0). Data are presented as mean ± standard deviation (SD). Normality and homogeneity of variance were assessed using the Shapiro–Wilk and Levene’s tests, respectively. For normally distributed data with equal variances, statistical comparisons between two groups were performed using an unpaired Student’s t-test. Multiple group comparisons were evaluated using one-way analysis of variance, followed by Tukey’s HSD or Bonferroni post hoc tests. Statistical significance was defined as p < 0.05 (*p < 0.05, **p < 0.01, ***p < 0.001).

Results

PSP attenuates TBH-induced senescent phenotypes in BMSCs

The CCK-8 assay indicated that PSP concentrations of 50–200 μg/mL did not significantly reduce BMSC viability, whereas 400 μg/mL significantly reduced cell viability. Consequently, 50, 100, and 200 μg/mL were established as the selected concentration range for subsequent experiments (Figure 1A). To establish an appropriate cellular senescence model, BMSCs were exposed to various concentrations of TBH. The CCK-8 assay demonstrated that 100 μM TBH significantly reduced cell viability while preserving sufficient viable cells for subsequent senescence analyses. However, 200 μM and higher caused extensive cell death; therefore, 100 μM was selected for senescence induction (Figure 1B). Senescence-associated beta-galactosidase (SA-β-gal) staining revealed a significant increase in the proportion of SA-β-gal-positive cells. Conversely, PSP treatment dose-dependently reduced the proportion of SA-β-gal-positive cells (Figures 1C and D). Furthermore, lineage induction assays demonstrated that TBH-induced senescence was associated with reduced osteogenic capacity and increased adipogenic commitment. PSP treatment attenuated this shift in a dose-dependent manner, concurrently preserving the balance between osteogenic and adipogenic differentiation. These data indicate that PSP attenuates senescent phenotypes, in part, by preserving the osteogenic-adipogenic lineage balance (Figure 1C, E, and F).

Figure 1.

Effects of PSP on BMSC viability and cellular phenotypes. (A) Cell viability following 24 h of PSP treatment, evaluated via CCK-8 assay (n = 5). (B) Cell viability following 48 h of TBH treatment, evaluated via CCK-8 assay (n = 4). (C) Representative images of SA-β-gal, ALP, and Oil Red O staining across all experimental conditions. (D) Quantification of SA-β-gal staining. (E) Quantification of ALP staining. (F) Quantification of Oil Red O staining.

PSP mitigates iron metabolism dysregulation in senescent BMSCs by inhibiting ferritinophagy

FerroOrange fluorescence imaging demonstrated that TBH exposure significantly increased intracellular Fe²⁺ accumulation, a pathology significantly attenuated by PSP intervention (Figures 2A,B). Further analysis of the ferritinophagy pathway by Western blot revealed that TBH treatment significantly activated ferritinophagy, characterized by an increased LC3-II/LC3-I ratio, elevated NCOA4 expression, and reduced FTH1 levels, consistent with increased ferritin turnover. PSP treatment significantly attenuated these alterations (Figures 2C–F). To further examine the involvement of iron metabolism in the protective effects of PSP, an iron-repletion experiment was conducted using FAC, increasing the proportion of senescent cells and partially reversing the PSP-associated preservation of osteogenic differentiation (ALP) and suppression of adipogenesis (Oil Red O) (Figures 2G–K). These results indicate that PSP attenuates intracellular Fe²⁺ accumulation and alters ferritinophagy-associated markers in senescent BMSCs.

Figure 2.

PSP mitigates iron metabolism dysregulation in senescent BMSCs by inhibiting ferritinophagy. (A, B) Representative fluorescence images and quantification of FerroOrange staining. (C–F) Western blot analysis and densitometric quantification of ferritinophagy-associated proteins (FTH1, NCOA4, and LC3). (G, H) representative images and quantification of SA-β-gal staining. (I) Representative images of ALP and Oil Red O staining. (J) Quantification of ALP staining. (K) Quantification of Oil Red O staining.

PSP mitigates BMSC senescence through the suppression of NCOA4-mediated ferritinophagy

To assess the involvement of NCOA4 in PSP-mediated regulation of ferritinophagy, BMSCs were transduced with a lentiviral vector to overexpress NCOA4 (LV-NCOA4). Functional rescue experiments revealed that lentiviral NCOA4 overexpression attenuated the inhibitory effects of PSP on ferritinophagy. Specifically, relative to the PSP+LV-NC control, the PSP+LV-NCOA4 group exhibited a significant re-accumulation of intracellular Fe²⁺ (Figure 3A, B). Concomitantly, NCOA4 overexpression elevated the LC3-II/LC3-I ratio and re-initiated the degradation of FTH1 (Figure 3C–E). These findings support the involvement of NCOA4 downregulation in PSP-mediated preservation of intracellular iron homeostasis and suppression of excessive ferritinophagy.

Figure 3.

PSP mitigates BMSC senescence by suppressing NCOA4-mediated ferritinophagy. (A, B) Representative fluorescence images and quantification of intracellular Fe²⁺ via FerroOrange staining. (C–E) Western blot analysis and densitometric quantification of the LC3-II/LC3-I ratio and FTH1 protein levels.

PSP restores mitochondrial functional homeostasis in senescent BMSCs

Flow cytometric analysis of MitoSOX fluorescence revealed that TBH exposure significantly increased MitoSOX fluorescence, indicating increased mtROS accumulation, which was effectively attenuated by PSP (Figure 4A). Furthermore, JC-1 assays demonstrated that TBH-induced mitochondrial membrane depolarization, characterized by a decreased red-to-green fluorescence ratio, was effectively attenuated by PSP (Figure 4B,C). ATP quantification indicated that PSP partially restored intracellular ATP synthesis observed in the senescent model (Figure 4D). Immunofluorescence colocalization analysis of TOMM20 and DNA revealed that TBH treatment reduced TOMM20/DNA colocalization, whereas PSP partially restored their colocalization (Figure 4E,F). Quantitative qPCR confirmed that TBH increased cytosolic mtDNA, while PSP attenuated this effect (Figure 4G–I). Collectively, these data demonstrate that PSP preserves mitochondrial functional homeostasis in senescent BMSCs. Given that cytosolic mtDNA leakage serves as a primary trigger for the cGAS-STING cascade, we subsequently evaluated the activation status of this inflammatory pathway.

Figure 4.

PSP restores mitochondrial functional homeostasis in senescent BMSCs. (A) Quantification of mtROS in BMSCs via MitoSOX staining. The values shown in the flow histogram indicate MitoSOX fluorescence intensity. (B, C) Representative fluorescence images and quantification of mitochondrial membrane potential (MMP) via JC-1 staining. (D) Quantification of intracellular ATP content. (E, F) Immunofluorescence analysis of mitochondrial morphology (TOMM20, red) and DNA localization (DNA, green). (G-H) representative Western blots and densitometric quantification of whole cell extract (WCE) and cytosolic fraction from normal rat BMSCs showing strong GAPDH enrichment in the cytosol with negligible TOMM20 and Lamin B1 contamination, confirming cytosolic fraction purity. (I) The qPCR analysis of cytosolic mtDNA levels in BMSCs cells treated with PSP.

PSP inhibits the cGAS-STING signaling pathway by preventing cytosolic mtDNA leakage

Cytosolic mtDNA can activate the cGAS–STING axis, thereby promoting inflammatory signaling and potentially contributing to BMSC senescence. After observing that PSP reduced cytosolic mtDNA translocation, the activation status of the cGAS–STING pathway was evaluated. Western blot analysis demonstrated that TBH exposure significantly increased the protein levels of cGAS and STING. PSP intervention effectively attenuated the pathological overexpression of these proteins (Figures 5A–C). These findings indicate that PSP may attenuate cGAS–STING signaling alongside reduced cytosolic mtDNA accumulation.

Figure 5.

PSP inhibits the cGAS-STING signaling pathway by preventing cytosolic mtDNA leakage. (A-C) Representative Western blot bands and densitometric quantification of cGAS and STING protein expression levels in BMSCs across all experimental groups.

PSP maintains mitochondrial homeostasis and inhibits the cGAS-STING axis through regulation of NCOA4-mediated ferritinophagy

To evaluate the involvement of NCOA4 in the observed signaling cascade, functional rescue experiments were performed by overexpressing NCOA4 (LV-NCOA4) alongside PSP intervention. Sequential assessments of mitochondrial function and downstream signaling were conducted. Relative to the PSP+LV-NC control, NCOA4 overexpression (PSP+LV-NCOA4) triggered a resurgence in mtROS levels (Figure 6A), secondary mitochondrial membrane depolarization (Figure 6B,C), and a significant decline in ATP biosynthetic capacity (Figure 6D). Immunofluorescence colocalization analysis of TOMM20 and DNA demonstrated that NCOA4 overexpression decreased the overlap between mitochondrial markers and DNA signals compared with the LV-NC group (Figure 6E,F). Consistently, qPCR quantification of cytoplasmic fractions confirmed that LV-NCOA4 significantly increased cytosolic mtDNA levels compared to the LV-NC control (Figure 6G). Consistent with these mitochondrial abnormalities, Western blot analysis demonstrated that cGAS and STING protein levels were significantly increased in the PSP + LV-NCOA4 group (Figure 6H–J).

Figure 6.

PSP maintains mitochondrial homeostasis and inhibits the cGAS-STING axis through regulation of NCOA4-mediated ferritinophagy. (A) Quantification of mtROS in BMSCs via MitoSOX staining. The values shown in the flow histogram indicate MitoSOX fluorescence intensity. (B) Representative images of JC-1 staining in BMSCs across all experimental groups. (C) Quantification of mitochondrial membrane potential (MMP) via JC-1 staining. (D) Quantification of intracellular ATP content. (E-F) Quantitative analysis of mitochondrial (TOMM20) and DNA colocalization to assess changes in the spatial association between mitochondria and cellular DNA. (G) The qPCR analysis of cytosolic mtDNA levels in BMSCs cells treated with PSP. (H-J) Representative Western blot bands and densitometric quantification of cGAS and STING protein expression levels.

Discussion

PSP, a bioactive polysaccharide derived from Polygonatum sibiricum, has attracted increasing attention for its antioxidant, anti-inflammatory, and potential anti-senescence properties. Previous studies have demonstrated that PSP can alleviate oxidative damage by reducing ROS accumulation, suppressing lipid peroxidation, and improving mitochondrial function in senescence-related models (Ma et al. 2021; Li and Liu et al., 2025). These findings suggest that PSP may protect cells against stress-induced dysfunction by regulating oxidative homeostasis. However, the molecular mechanisms underlying the protective effects of PSP against stem cell senescence, particularly in OP-associated BMSC dysfunction, remain largely unknown. In this study, we established a TBH-induced senescence model in rat BMSCs and demonstrated that PSP effectively alleviated senescent phenotypes, preserved the balance between osteogenic and adipogenic differentiation, and improved mitochondrial homeostasis. Mechanistically, these findings identified NCOA4-mediated ferritinophagy as a potential regulatory pathway involved in PSP-mediated protection. PSP attenuated ferritinophagy-associated changes, thereby limiting intracellular Fe²⁺ accumulation, reducing mitochondrial damage, and inhibiting mtDNA-triggered cGAS–STING signaling.

Maintaining intracellular iron homeostasis is important for cellular function and may contribute to the maintenance of BMSC homeostasis under stress. NCOA4-mediated ferritinophagy plays a critical role in the metabolic reprogramming associated with stem cell aging; however, its overactivation has been associated with pathological iron accumulation (Wang and Zhang 2022). In the TBH-induced senescence model, oxidative stress was associated with increased NCOA4 expression. Functioning as a specific cargo receptor, NCOA4 functions as a ferritin cargo receptor that mediates the delivery of ferritin, including FTH1, to lysosomes for degradation. This process can increase the labile intracellular iron pool and contribute to Fe²⁺ accumulation under pathological conditions (Zhang et al. 2023). This mechanism aligns with the pathogenic role of ferritinophagy previously reported in ovariectomized-induced osteoporosis and premature ovarian failure (Ma et al. 2024). Evidence suggests that, in specific microenvironments such as diabetic wound healing (Liu et al. 2024), activation of NCOA4-mediated ferritinophagy may paradoxically promote apoptosis and facilitate the clearance of senescent cells, thereby facilitating tissue repair. This divergent phenotypic outcome highlights the context dependence of ferritinophagy across distinct cell types and pathological stages. Although transient ferritinophagy activation may benefit processes requiring rapid cellular turnover, such as wound healing, chronic or excessive activation of this pathway may adversely affect BMSC function under pathological conditions, potentially contributing to impaired BMSC function and regenerative capacity. The pharmacological intervention of PSP aligns with the principles of systems biology; rather than fully blocking physiological iron metabolism, PSP modulates it by downregulating NCOA4 expression to restore iron homeostasis from a pathologically imbalanced state.

As a key target of iron-mediated oxidative damage, mitochondrial dysfunction serves as a critical link between dysregulated iron metabolism and inflammatory responses (Santana-Codina et al. 2022). Excess intracellular labile iron facilitates ROS production via the Fenton reaction, inducing lipid peroxidation that impairs mitochondrial oxidative phosphorylation (Fujimaki et al. 2019). These morphological and functional data demonstrate that senescent BMSCs exhibit significant structural compromise and a bioenergetic crisis, characterized by mtROS bursts, profound loss of MMP, and decreased intercellular ATP levels. Moreover, senescent BMSCs exhibit abnormal lineage commitment with enhanced adipogenic differentiation. This shift may disrupt metabolic homeostasis, increase oxidative stress, and further aggravate mitochondrial dysfunction and bone marrow microenvironment deterioration (Lin et al. 2025). This mitochondrial dysfunction was accompanied by increased cytosolic mtDNA accumulation. As a potent DAMP, cytosolic mtDNA is recognized by cGAS, which catalyzes cGAMP synthesis and activates the STING-dependent sterile inflammatory signaling pathway (Sui et al. 2025). Recent evidence suggests that aberrant activation of this axis is implicated in osteoclast differentiation, exacerbating OP (Salminen et al. 2025; Wang et al. 2026), and functions as a conserved sensing pathway for environmental toxin-induced stem cell senescence (Izquierdo 2025; Xiao et al. 2025). Persistent cGAS–STING signaling has been reported to prompt SASP-associated inflammatory responses, comprising pro-inflammatory cytokines and matrix metalloproteinases. These factors may influence neighboring cells within the bone marrow microenvironment, suppressing adjacent osteoblast activity. The protective effect of PSP is associated with suppression of excessive upstream ferritinophagy. By attenuating the accumulation of labile intercellular iron, PSP may reduce iron-dependent oxidative damage to mitochondria, attenuate cytosolic mtDNA accumulation, and suppress the cGAS–STING-associated inflammatory signaling pathway. These findings provide a potential mechanistic link among the antioxidant, mitochondrial-protective, and anti-inflammatory effects of PSP through ferritinophagy, mitochondrial dysfunction, and cGAS–STING signaling.

The ferritinophagy–mitochondrial dysfunction–cGAS–STING axis identified in this study follows a defined cascade. NCOA4 overexpression provided functional evidence supporting the involvement of NCOA4 in this pathway. Exogenous NCOA4 overexpression attenuated the inhibitory effects of PSP on ferritinophagy, resulting in the resurgence of the LC3-II/LC3-I ratio, renewed FTH1 degradation, and intracellular Fe²⁺ re-accumulation. This secondary disruption of iron homeostasis confirms that pathological upregulation of NCOA4 drivers dysregulated ferritinophagic flux, aligning with observations in retinal pigment epithelial cell senescence models (Li et al. 2023). Concurrent with the disruption of iron homeostasis, the mitochondrial network underwent a secondary bioenergetic crisis. Although regulated iron mobilization is essential for cellular iron utilization and mitochondrial function, increased labile iron availability under senescent stress may contribute to mitochondrial dysfunction (Fujimaki et al. 2019; Sui et al. 2025). Crucially, mitochondrial damage can promote cytosolic mtDNA accumulation. Cytosolic mtDNA can act as a DAMP and activate cGAS–STING signaling, potentially contributing to the maintenance of the senescence-associated phenotype. Recent cross-disease models in neuroinflammation (Sui et al. 2025) and autoimmune disorders (Wang et al. 2026) have corroborated that the ferritinophagy–mtDNA accumulation–cGAS–STING axis has been implicated as a signaling link between iron dysregulation and inflammatory responses. These findings support NCOA4-mediated ferritinophagy as an important regulatory pathway involved in the protective effects of PSP. The axis constitutes a comprehensive pathological cascade in which modulation of NCOA4 is associated with downstream changes in iron homeostasis, mitochondrial function, and inflammatory signaling. This study provides a systemic understanding of the mechanisms underlying BMSC senescence and highlights the regulatory involvement of NCOA4-mediated ferritinophagy in PSP-mediated protection.

Although this study provides a comprehensive evidentiary chain at the cellular and molecular levels, several limitations remain that require consideration. First, PSP is a structurally complex natural polysaccharide (Li and Yao et al., 2025). The specific bioactive components and the molecular mechanism by which PSP modulates NCOA4 remain unidentified. Future investigations using chromatographic separation and mass spectrometry are required to isolate the specific bioactive fractions associated with NCOA4 regulation. Second, these findings are derived exclusively from in vitro cellular models. Whether PSP can effectively attenuate BMSC senescence and improve bone microstructure in osteoporotic animal models through a similar mechanism remains to be validated. NCOA4 conditional knockout could help validate the role of NCOA4 in vivo. Nevertheless, a study by Chen et al.(Chen et al. 2025) demonstrated that PSP improves diabetic OP in zebrafish models, supporting the feasibility of such in vivo research. Finally, the regulatory crosstalk between ferritinophagy, mitochondrial function, and the cGAS–STING axis is highly complex. The potential existence of additional feedback loops within this signaling network warrants further exploration to fully delineate the mechanisms of action of PSP.

Conclusions

This study provides mechanistic evidence that PSP attenuates BMSC senescence by integrating metabolic and immunological regulation. These results suggest that PSP attenuates NCOA4-mediated ferritinophagy and is associated with improved intracellular iron homeostasis and attenuating mitochondrial oxidative stress. This attenuation of iron dysregulation further attenuates mitochondrial dysfunction and attenuates cytosolic mtDNA-associated cGAS-STING signaling, thereby improving the senescent phenotype and altered osteogenic and adipogenic differentiation in BMSCs. These findings provide additional insight into the potential role of ferritinophagy-associated iron dysregulation in BMSC dysfunction relevant to OP and highlight PSP as a bioactive compound with potential effects on both metabolic homeostasis and inflammatory signaling. Further in vivo validation and clinical investigations are required. However, this research provides a theoretical foundation and identifies a potential therapeutic candidate for developing novel anti-osteoporotic strategies centered on regulating iron metabolism (Figure 7).

Figure 7.

PSP attenuates BMSC senescence by inhibiting NCOA4-mediated ferritinophagy.

Funding Statement

This work was supported by the National Natural Science Foundation of China (Grant No. 82575113, SY), the Project Supported by Scientific Research Fund of Hunan Provincial Education Department (Grant No. 25A0288, ML), and the Natural Science Foundation of Hunan Province of China (Grant No.2025JJ90072, XL).

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.