Animal experiments were conducted according to the approved guidelines for experimental animal ethics and welfare by the Institutional Animal Care and Use Committee of Sun Yat-Sen University (Guangzhou, China) (approval no. SYSU-IACUC-2024-002262). For BMSC isolation, 10 3-week-old male Sprague-Dawley rats weighing about 50 g were purchased from the Laboratory Animal Center of Sun Yat-Sen University. Rats were sacrificed by cervical dislocation under anesthetization with isoflurane inhalation (induction, 3%; maintenance, 2%). Animal death was confirmed by respiratory arrest, cessation of heartbeat and disappearance of reflexes. Rats were immersed in 75% ethanol at room temperature for 20 min to sterilize. Tibiae and femurs were dislocated and both ends were cut off under aseptic conditions. The bone marrow was flushed out using a syringe, cut into small fragments and subsequently collected by centrifugation (250 × g, 5 min, 4°C). The fragments were resuspended in DMEM (Gibco; Thermo Fisher Scientific, Inc.) and seeded into a culture flask. Primary BMSCs were cultured in DMEM containing 10% fetal bovine serum (Moregate BioTech) and 100 μg/ml Primocin^® (InvivoGen) in a humidified incubator containing 5% CO₂ at 37°C. The culture medium was changed every 3 days. Passage three BMSCs were used for subsequent experiments.

Quercetin was purchased from MedChemExpress and dissolved in DMSO (Sigma-Aldrich; Merck KGaA). Poly(I:C)-LMW/LyoVec™ [Poly(I:C)], was purchased from InvivoGen and dissolved following the manufacturer's instructions. Briefly, to prepare a stock solution (50 μg/ml), 25 μg lyophilized Poly(I:C) was added to 500 μl endotoxin-free water and gently mixed for ≥15 min. The products were aliquoted and stored at −20°C.

For osteogenic induction, BMSCs (1.5×10⁵ cells/well) were seeded in 12-well plates. After reaching 70-80% confluence, cells were stimulated with 200 μM H₂O₂ for 1 h or treated with 0.1 μg/ml Poly(I:C) for 24 h at 37°C with 5% CO₂. Cells were washed and cultured with osteogenic induction medium (Cyagen Biosciences Inc.) at 37°C with 5% CO₂. Quercetin at concentration of 1 μM was added to osteogenic induction medium. The culture medium was refreshed every 3 days. Alkaline phosphatase (ALP) staining was conducted after 7 days using a BCIP/NBT ALP Color Development kit (Beyotime Institute of Biotechnology) according to the manufacturer's instructions. Alizarin red staining was performed 14-21 days post-induction. Briefly, BMSCs were washed with PBS and fixed with 4% paraformaldehyde at room temperature for 30 min. Cells were stained with alizarin red for 10 min at room temperature. After that, the images were captured by an inverted light microscope (Zeiss Axio; magnification, ×100). The stained area was calculated by Image J 1.48v software (National Institutes of Health).

To assess the geroprotective effect of quercetin treatment and RIG-I knockdown, 2×10⁴ BMSCs were seeded in 24-well plates and cultured at 37°C with 5% CO₂ in complete DMEM for 24 h. Next, cells were treated with 200 μM H₂O₂ for 1 h to induce senescence. For quercetin treatment, cells were washed with phosphate-buffered saline (PBS) and cultured with fresh complete DMEM containing quercetin (0.0, 0.1, 1.0, 10.0 and 100.0 μM) at 37°C with 5% CO₂ for 3 days. For RIG-I knockdown, cells were washed and subjected to short-interfering RNA (siRNA) transfection using Lipofectamine^® RNAiMAX Reagent (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 13778-150) according to the manufacturer's instructions. Briefly, cells were transfected with 50 nM siRNA at 37°C with 5% CO₂ for 6 h. The si-RIG-I (5′-CAUUGAAACCAAGAAAUUACC-3′) were synthesized by GENEray Biotech (Shanghai) Co., Ltd. The si-negative control (NC, 5′-UUCUCCGAACGUGUCACGUTT-3′) was set as the control. To evaluate knockdown efficiency, cells were harvested for qPCR 48-72 h after transfection.

Following treatment with quercetin at 37°C with 5% CO₂ for 3 days, senescent cells were assessed using a SA-β-gal staining kit (Beyotime Institute of Biotechnology) according to manufacturer's instructions. Briefly, cultured cells were fixed at room temperature for 15 min, washed with PBS and incubated with staining solution at 37°C overnight. Images were captured using an inverted light microscope (Zeiss Axio; magnification, ×100) and the ratio of stained cells was quantified using ImageJ 1.48v software (National Institutes of Health).

The optimal concentration of quercetin was selected by assessing its toxic effect using CCK-8 assay. Briefly, BMSCs (3×10³ cells/well) were seeded in 96-well plates and treated with quercetin (0.0, 0.1, 1.0, 10.0, 100.0 and 1,000.0 μM) at 37°C with 5% CO₂. After 3 days, cell viability was measured using CCK-8 (Dojindo Laboratories, Inc.) according to the manufacturer's protocol. Briefly, cells were incubated with CCK-8 solution for 2 h at 37°C. Quercetin concentrations that significantly decreased cell viability compared with the blank control were considered cytotoxic.

The colony formation assay was used to evaluate the self-renewal ability of BMSCs. In brief, 5×10³ BMSCs were seeded in 12-well plates and stimulated with 200 μM H₂O₂ to induce senescence as aforementioned. Then, BMSCs were cultured in complete DMEM (Gibco; Thermo Fisher Scientific, Inc.) at 37°C with addition of quercetin (0.0, 0.1, 1.0, 10.0 and 100.0 μM) for 7-10 days. The number of colonies formed was calculated by crystal violet staining (Beijing Solarbio Science & Technology Co., Ltd.). Briefly, cells were fixed with 4% paraformaldehyde for 15 min, then stained with 0.1% crystal violet for 15 min, both at room temperature. The images were captured and the number of colonies (>50 cells) were manually calculated.

After culturing in DMEM or osteogenic induction medium (Cyagen Biosciences Inc.), total RNA was extracted from cells using the RNA-Quick Purification kit (cat. no. RN001; Yishan Biotechnology Co., Ltd.). RNA was reverse-transcribed into cDNA using HiScript II Q RT SuperMix according to the manufacturer's protocol for qPCR (Vazyme Biotech Co., Ltd.). qPCR was performed using Taq Pro Universal SYBR qPCR Master Mix (Vazyme Biotech Co., Ltd.) on a Light Cycler 480 system (Roche Applied Science). The thermocycling conditions were as follows: Initial denaturation at 95°C for 30 sec, followed by 40 cycles of denaturation at 95°C for 10 sec, annealing and extension at 60°C for 30 sec; and a melting curve between 60 and 95°C. All results were normalized to GAPDH and the quantitative method of relative mRNA expression was used 2^−ΔΔCq (24). The senescence marker gene (p21), SASP genes (IL6, TNF-α, IL1α and IL1β), and osteogenic marker genes [osteoprotegerin (OPG), osteocalcin (OCN), osteopontin (OPN) and type I collagen α 1 (COL1A1)] were detected. The PCR primers are listed in Table I.

BMSCs (2×10⁴ cells/well) were seeded on slides in 24-well plates. Following senescence induction and quercetin treatment, cellular immunofluorescence was used to evaluate expression of γ-H2AX, methylation histone H3 Lys9 (H3K9me3), heterochromatin protein 1α (Hp1α), lamina-associated polypeptide 2 (LAP2), double-stranded (ds)RNA clone rJ2 (rJ2), LINE-1 open reading frame 1 protein (ORF1p) and dsDNA. BMSCs cultured on microscope coverslips at 37°C for 3 days were fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.4% Triton X-100 in PBS for 30 min and blocked with 10% goat serum (BOSTER Biological Technology Co., Ltd.) for 1 h, all at room temperature. Slides were covered with primary antibodies at 4°C overnight followed by the Alexa Fluor 488 dye-(1:300; cat. no. EM35141-01; Beijing Emarbio Science & Technology Co., Ltd.) or Alexa Fluor 594 dye-conjugated secondary antibodies (1:300; cat. no. EM35150-01; Beijing Emarbio Science & Technology Co., Ltd.) for 1 h at room temperature in the dark. Nuclei were stained with DAPI for 10 min at room temperature and slides were mounted using anti-fading mounting medium (Vector Laboratories, Inc.; cat. no. H-1000). Images were acquired using a laser scanning confocal microscope (Carl Zeiss AG; magnification, ×400). The mean immunofluorescence intensity or proportion of positive cells was quantified using ImageJ 1.48v software.

The antibodies used for immunofluorescence staining were as follows: γ-H2AX (1:400; cat. no. 9718; Cell Signaling Technology, Inc.), H3K9me3 (1:400; cat. no. ab8898; Abcam), Hp1α (1:200; cat. no. 2616; Cell Signaling Technology, Inc.), LAP2 (1:100; cat. no. 611000; BD Biosciences), rJ2 (1:100; cat. no. MABE1134; MilliporeSigma), ORF1p (1:200; cat. no. MABC1152; MilliporeSigma) and dsDNA (1:400; cat. no. sc-58749; Santa Cruz Biotechnology, Inc.).

BMSCs (3×10⁵ cells seeded in 6-well plates) were lysed in RIPA buffer (Beyotime Institute of Biotechnology) supplemented with protease and phosphatase inhibitors. Total protein was quantified using BCA Protein Assay kit (Bio-Rad Laboratories, Inc.). Then, 30 μg/lane protein were loaded onto SDS-PAGE in 5-15% Bis-Tris precast gel to separate the different proteins. Subsequently, samples were transferred to a PVDF membrane (MilliporeSigma) and blocked with 5% non-fat milk for 1 h at room temperature. Then, PVDF membranes were incubated with primary antibodies [H3K9me3, Hp1α, LAP2, RIG-I, phosphorylated TANK-binding kinase 1 (p-TBK1), TBK1 or GAPDH] at 4°C overnight followed by secondary antibodies (1:10,000; HRP AffiniPure Goat Anti-Rabbit, cat. no. EM35111-01 or HRP AffiniPure Goat Anti-Mouse, cat. No. EM35110-01; Beijing Emarbio Science & Technology Co., LTD.) for 1 h at room temperature. The signals were detected using the ECL Immobilon Western Chemilum HRP Substrate (cat. no. WBKLS0500; Merck Millipore) and an ultra-high sensitivity chemiluminescence imaging system (Bio-Rad Laboratories, Inc.) and then quantified using ImageJ 1.48v software (National Institutes of Health).

The antibodies were as follows: H3K9me3 (1:1,000; cat. no ab8898; Abcam), Hp1α (1:1,000; cat. no. 2616; Cell Signaling Technology, Inc.), LAP2 (1:500; cat. no. 611000; BD Biosciences), RIG-I (1:1,000; cat. no. A13407; ABclonal Biotech Co., Ltd.), p-TBK1 (1:250; cat. no. AP1418; ABclonal Biotech Co., Ltd.), TBK1 (1:1,000; cat. no. A3458; ABclonal Biotech Co., Ltd.) and GAPDH (1:3,000; cat. no. 60004-1-Ig; Proteintech Group, Inc.).

Following H₂O₂ stimulation and quercetin treatment, RNA-seq analysis of BMSCs was conducted by Shanghai Majorbio Bio-pharm Technology Co., Ltd. Total RNA was obtained using MJZol total RNA extraction kit (cat. no. T01-500; Majorbio). RNA quality was determined by 5300 Bioanalyzer (Agilent) and quantified using ND-2000 (NanoDrop Technologies). The loading concentration of the final library was 10 pM. Agarose gel (Biowest Agarose; Biowest) electrophoresis was used to detect the RNA integrity. mRNA was enriched with Oligo (dT) beads, fragmented into short fragments (300 bp), and reverse-transcribed into cDNA. After being linked to adaptor and purified, cDNA fragments were sequenced on an Illumina NovaSeq Xplus platform (Illumina, Inc.). cDNA library was prepared following Illumina^® Stranded mRNA Prep, Ligation (cat. no. 20040534; Illumina, Inc.). The sequencing library was performed on NovaSeq X plus platform (PE150) using NovaSeq X Series 10B Reagent Kit (300cycles; cat. no. 20085594; Illumina Inc.) and 300 bp paired-end reads were generated. The cleaned reads were mapped to Rattus norvegicus genome (mRatBN7.2; ncbi.nlm.nih.gov/datasets/genome/GCF_015227675.2) using hisat2 (version 2.1.0). For RE analysis, transposable elements were quantified and analyzed using TE transcripts (version 2.2.3). Differentially expressed REs were counted using R package DESeq2 (version 1.38.2) with a cut-off of |log₂(fold change)| >0.3 and P<0.05. The genes significantly affected by H₂O₂ stimulation or quercetin were determined by setting a fold change of ≥1.5 and P<0.05. Venn diagram analysis of differentially expressed genes, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Set Enrichment Analysis (GSEA) were performed on Majorbio Cloud (majorbio.com).

The data are presented as the mean ± standard deviation of three independent experimental repeats. Results were statistically analyzed using GraphPad Prism (version 5.0; Dotmatics). Comparisons were conducted with unpaired Student's t-test or one-way ANOVA followed by Dunnett's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

The accumulation of reactive oxygen species in BMSCs causes oxidative stress-induced senescence and impairs SC properties (25). To induce cellular senescence in the present study, exogenous H₂O₂ was used to stimulate BMSCs. Following treatment with 200 μM H₂O₂, the proportion of SA-β-gal-positive cells significantly increased (Fig. 1A and B), indicating cellular senescence was successfully induced under this condition. To investigate the effect of quercetin on senescent BMSCs, quercetin (0, 0.1, 1, 10 and 100 μM) were used to treat H₂O₂-induced BMSCs for 72 h. The ratio of SA-β-gal-positive cells was reduced with an increase in quercetin concentration, especially when the concentration >1 μM (Fig. 1A and B). Additionally, uninduced BMSCs were treated with quercetin to assess cytotoxicity. Quercetin had no significant effect on cell viability at concentrations of 0.0, 0.1, 1.0 and 10.0 μM. However, it exerted toxic effects on BMSCs when the concentration reached 100 and 1,000 μM (Fig. 1C). To identify the optimal concentration, the restorative effect of quercetin on proliferation of senescent BMSCs was further investigated. The colony formation assay demonstrated the impaired proliferative ability of senescent BMSCs was rescued by quercetin at 1 μM (Fig. 1D and E). Consequently, subsequent experiments were conducted using this concentration. H₂O₂ resulted in upregulation of p21 mRNA expression, which was decreased upon treatment with quercetin (Fig. 1F). Senescent cells do not divide but retain the ability to secrete bioactive molecules known as SASP, which alters the microenvironment for both senescent and surrounding cells (26). SASP-associated IL6, TNF-α, IL1α and IL1β showed increased mRNA expression following H₂O₂ treatment; however, expression decreased following quercetin treatment (Fig. 1G). These data aligned with the SA-β-gal staining and colony formation results showing that quercetin alleviated cellular senescence phenotype induced by H₂O₂.

The instability of the genome caused by persistent DNA damage and accumulating mutations, along with epigenomic instability characterized by loss of heterochromatin structure and RE suppression are mechanisms underlying senescent phenotypes (27). The dsDNA break marker γ-H2AX is a notable molecular indicator for DNA damage and serves as a hallmark of genomic instability (28). Immunofluorescence revealed a significant increase in nuclear γ-H2AX signal in H₂O₂-induced senescent BMSCs. Quercetin resulted in a reduced expression of γ-H2AX, demonstrating that quercetin stabilized the genome of BMSCs (Fig. 2A). In eukaryotic cells, most of the genome is packaged into condensed and transcriptionally suppressed heterochromatin. H3K9me2 and 3 are epigenetic markers of heterochromatin found on transposable elements to ensure their transcriptional suppression (29). There was a noticeable decrease in expression of H3K9me3 in H₂O₂-induced senescent BMSCs; however, this decrease was rescued by quercetin (Fig. 2B). Consistently, western blotting demonstrated a decrease in expression of H3K9me3 in H₂O₂-induced senescent BMSCs, which was rescued by quercetin (Fig. 2E). The HP1 family binds to H3K9me2 and 3 and interacts with other proteins to maintain chromatin condensation (30). Both immunofluorescence staining and western blotting revealed that Hp1α expression decreased in the H₂O₂ group but recovered in the quercetin group (Fig. 2C and E). Additionally, LAP2, involved in the compaction of heterochromatin by anchoring it on the inner nuclear membrane (30), exhibited similar expression patterns as H3K9me3 and Hp1α and was downregulated in the H₂O₂ but rescued in the quercetin group (Fig. 2D and E). These data demonstrated a decrease in the structural integrity of heterochromatin during senescence, with quercetin serving a key role in safeguarding genomic and epigenomic stability, thereby mitigating senescence.

Loss of suppressive epigenetic regulation and heterochromatin structure during cell senescence process may elicit transcriptional activation of REs (9). To elucidate the mechanism by which quercetin alleviates senescence of BMSCs, bulk mRNA sequencing was employed to assess the transcriptional levels of REs. A total of 176 REs was differentially expressed after H₂O₂ treatment. All differentially expressed REs were transcribed at elevated levels in H₂O₂-induced senescent BMSCs. Quercetin resulted in differential expression of 15 repeated sequences in senescent BMSCs, among which 86.67% (13/15) were downregulated (Fig. 3A). To identify REs upregulated by H₂O₂ and rescued by quercetin, Venn diagram analysis was used. The rescued REs primarily included the LINE-1 and ERV families (Fig. 3B). The heatmap revealed augmented expression of these REs in the H₂O₂ group, whereas down-regulation was observed in the quercetin group, suggesting the involvement of LINE-1 and ERVs in quercetin-mediated alleviation of cell senescence (Fig. 3C). Activation of RE transcription results in translocation of endogenous RNA to the cytoplasm and generation of DNA through reverse transcriptase activity (31). Immunofluorescence using rJ2 antibody confirmed the accumulation of cytosolic dsRNA in senescent BMSCs. Quercetin decreased the proportion of cytoplasmic rJ2-positive BMSCs (Fig. 3D and E). In accordance with endogenous expression of dsRNA, there was elevated cytoplasmic protein expression of LINE-1 ORF1p in H₂O₂-induced senescent BMSCs, which was be reversed by administration of quercetin (Fig. 3F and G). RNA derived from retroelements can also be converted to dsDNA in cytoplasm through endogenous reverse transcriptase activity; however, no cytoplasmic dsDNA was detected in any group, indicating that dsDNA was not involved in the mechanism underlying H₂O₂-induced BMSC senescence (Fig. 3H).

Endogenous cytosolic self-nucleic acids can directly trigger innate immune receptor signaling and induce inflammatory responses via recognition of RNA or DNA (15,20). To identify the downstream effect of dsRNA, Venn diagram analysis was employed to identify quercetin-rescued genes that exhibited differential expression in response to H₂O₂ and displayed similar expression patterns as dsRNA transcripts. A total of 20 genes were upregulated by H₂O₂ and differential expression was reversed by quercetin (Fig. 4A). KEGG analysis revealed that these genes were significantly enriched in pathways including 'viral protein interaction with cytokine and cytokine receptor', 'cytosolic DNA-sensing pathway', 'RIG-I-like receptor signaling pathway' and 'toll-like receptor signaling pathway' (Fig. 4B), which are associated with innate immune response. Moreover, GSEA demonstrated activation of innate immune response in H₂O₂-induced senescence of BMSCs (Fig. 4C). RIG-I-like receptor signaling pathway serves a key role in recognizing cytoplasmic dsRNA and initiating an IFN response. To validate the activation of this pathway, western blotting was performed, revealing H₂O₂ treatment upregulated expression of RIG-I and innate immunity-related protein p-TBK1. Conversely, quercetin downregulated these proteins (Fig. 4D and E). To investigate the role of RIG-I-like receptor signaling pathway in senescence induction, siRNA was used to knock down RIG-I expression in rat-derived BMSCs. The effectiveness of knockdown was confirmed by RT-qPCR and western blotting (Fig. 4F). RIG-I knockdown resulted decreased expression of downstream innate immune effector IFN-β which was upregulated in H₂O₂-induced senescence (Fig. 4G). SA-β-gal staining was used to assess cell senescence. Notably, RIG-I knockdown led to a decrease in the proportion of SA-β-gal-positive cells following H₂O₂ treatment, thereby confirming the functional impact of RIG-I on cell senescence (Fig. 4H and I). Overall, the present findings suggested that cytoplasmic dsRNAs induced by REs contributed to activation of RIG-I receptor signaling and subsequent innate immune response. This activation lead to type I IFN response and proinflammatory cytokine production, ultimately promoting senescence.

To evaluate the effect of quercetin on osteogenic differentiation potential of senescent BMSCs, BMSCs were cultured in an osteoblast differentiation medium for 7-21 days. Oxidative stress-induced senescence impaired osteogenic potential of BMSCs, as indicated by a decrease in ALP and alizarin red staining. However, quercetin restored the osteogenic differentiation ability of senescent BMSCs (Fig. 5A and B). The results were further supported by RT-qPCR of osteogenic marker genes, including OPG, OCN, OPN and COL1A1. mRNA levels of osteogenic marker genes (OPG, OCN, OPN, COL1A1) were decreased in senescent BMSCs, whereas they increased after quercetin administration (Fig. 5C). The aforementioned results demonstrated that dsRNA contributes to the mechanism by which quercetin ameliorates BMSC senescence. Furthermore, the role of dsRNA and downstream RIG-I receptor signaling pathway was investigated by adding synthetic dsRNA analogue Poly(I:C), which activates the RIG-I signaling pathway, to the osteoblast differentiation medium. Poly(I:C) resulted in a concentration-dependent decrease in alizarin red staining (Fig. 6A), confirming that dsRNA decreased osteoblast differentiation potential of BMSCs. Moreover, the positive area of alizarin red staining improved with si-RIG-I compared with the si-NC control (Fig. 6B), indicating knockdown of RIG-I partially rescued the impaired osteoblast differentiation ability of BMSCs caused by dsRNA. Similarly, knockdown of RIG-I improved osteogenic differentiation ability of BMSCs impaired by H₂O₂ (Fig. 6C). These data indicated that quercetin restored the osteogenic differentiation ability of senescent BMSCs, potentially due to the inhibition of the dsRNA-triggered RIG-I receptor signaling pathway.