Twenty-four specific pathogen-free (SPF)-grade male C57BL/6 mice (twelve-week-old, 20–30 g) were purchased from Hunan SJA Laboratory Animal Co., Ltd. (Changsha, China). All animals were housed in individually ventilated cages, maintained in a controlled environment with a temperature of 23°C ± 1°C and a 12-h light/dark cycle. They had free access to water and food for 1 week before we performed destabilization of the medial meniscus (DMM) or sham surgery. According to variance analysis requirements, the within-group degrees of freedom (E) should be maintained between 10 and 20 to ensure statistical power [27]. E can be measured by the following formula: E = Total number of animals − Total number of groups. To investigate the effect of low expression of SOX11 in vivo, the mice were randomly divided into four groups, including Sham, OA, OA+sh-NC, and OA+sh-SOX11 (n = 6) groups. Knee OA was induced in mice via DMM as previously described [28]. Briefly, mice were anesthetized, and the right knee joint was aseptically sterilized. Subsequently, the medial joint capsule adjacent to the patellar tendon was carefully incised to expose the intercondylar region. The joint capsule and the overlying skin were sutured closed in layers. For the sham surgery group, the joint capsule was similarly opened but without disruption of the meniscus ligament. One week post-surgery, mice in OA+sh-NC and OA+sh-SOX11 groups were intra-articularly injected 1 × 10⁹ pfu (8 μL) of short hairpin RNA (sh)-SOX11 lentivirus or corresponding negative control empty (sh-NC) lentivirus into the right knee joint, respectively. The mice were anesthetized via intraperitoneal injection of pentobarbital sodium (100 mg/kg) 8 weeks after surgery. Once the animals exhibited cessation of spontaneous breathing for 2–3 min and loss of the blink reflex, knee joint tissues were harvested for histopathologic analysis, Osteoarthritis Research Society International (OARSI) scoring, immunohistochemistry, and immunofluorescence. The Animal Experiment Ethics Committee for Biomedical Research, Hunan Normal University approved all animal experiment procedures (No.: 2025-637). All experimental procedures were conducted in accordance with institutional guidelines for the use of experimental animals.

The mouse knee joints were fixed in 4% paraformaldehyde at 4°C for 48 h, followed by decalcification in 10% EDTA (pH 7.4) for one month. Subsequently, the tissues were embedded in paraffin and serially sectioned at a thickness of 3 μm. The sections were baked at 62°C for 8 h, then subjected to dewaxing and rehydration procedures, and stained with either hematoxylin (Abiowell, AWI0001a, Changsha, China) and eosin (Abiowell, AWI0029a) or stained following Safranin-O/Fast Green kit (Abiowell, AWI0240a). After mounting with neutral balsam, the sections were examined under an optical microscope (MOTIC, BA210T, Xiamen, China).

According to histological analysis, tissue sections were obtained. Following immersion in 0.01 M sodium citrate antigen retrieval buffer (pH 6.0), microwave-mediated antigen retrieval was performed. After cooling at room temperature, the sections were then blocked with 1% periodate solution for 10 min at room temperature. Sections were subsequently washed three times in phosphate-buffered saline (PBS) (5 min per wash). After blocking, sections were incubated overnight at 4°C with the following primary antibodies against SOX11 (Bioss, bsm-60624R, 1:200, Rabbit, Beijing, China) and NOX4 (Proteintech, 14347-1-AP, 1:200, Rabbit, Chicago, IL, USA) or Rabbit IgG Isotype control (Proteintech, 98136-1-RR, 1:200). The samples were washed three times with PBS for 5 min each time. Sections were incubated with a ready-to-use HRP-conjugated polyclonal goat anti-rabbit IgG secondary antibody (Abiowell, AWI0629) at 37°C for 30 min. Following washing three times with PBS for 5 min each time, the sample was added with 50 μL of ready-to-use chromogenic reagent DAB solution (Abiowell, AWB0175) and incubated at room temperature for 30 s. The reaction time was controlled under an optical microscope (MOTIC, BA210T), and distilled water was used for washing to terminate the reaction. After three additional PBS washes (5 min per wash), cell nuclei were counterstained with hematoxylin for 10 min. Sections were observed and analyzed using an optical microscope (MOTIC, BA210T). For each sample, five random fields of view were selected under an optical microscope (MOTIC, BA210T). Positive expression analysis was performed by calculating the ratio of cumulative optical density from the positive expression area to the sample area within each field (Media Cybernetics, Image-Pro-Plus 6.0, Silver Spring, MD, USA). Each experiment is independently repeated three times.

Sections were obtained by referring to the histological analysis methods. Microwave-mediated antigen retrieval was performed. The cells were processed as follows: fixation in 4% paraformaldehyde (30 min), permeabilization with 0.3% Triton X-100 (30 min), and blocking with 5% bovine serum albumin (BSA, Abiowell, AWI0120a) at 37°C for 90 min. Sections were incubated overnight at 4°C with primary antibodies GPX4 (AWA11352, 1:200, Rabbit, Abiowell). Sections were incubated with secondary antibody, CoraLite488-conjugated Affinipure Goat Anti-Rabbit IgG (H+L) (Abiowell, AWS0005a, 1:200), at 37°C for 60 min. DAPI working solution (Abiowell, AWC0293a, 5 μg/mL) was nucleated for 10 min at 37°C. The sections were mounted in buffered glycerin and visualized using a fluorescence microscope (MOTIC, BA210T). For each sample, five random fields of view were selected. Positive expression analysis was performed by calculating the ratio of cumulative optical density from the positive expression area within each field (Media Cybernetics, Image-Pro-Plus 6.0). Each experiment is independently repeated three times.

The tibiae and femora of P3-P5 neonatal mice were carefully dissected and sectioned into fragments approximately 1 mm³ in size. The harvested tissues were initially subjected to digestion using 0.25% trypsin at 37°C for 10 min. Subsequently, they were digested with 0.2% type II collagenase and then placed in an air-oscillating incubator (Thermo Fisher Scientific MaxQ4000). The samples were incubated at a rotation speed of 250 rpm and a temperature of 37°C for 8 h. Following digestion, the primary chondrocytes were isolated and seeded onto culture plates. Mycoplasma contamination testing confirmed the absence of contaminants in the supernatant of primary murine articular chondrocytes. Cells were plated at a density of 5 × 10⁴ cells per well in a six-well plate. These cells were cultured in DMEM/F12 medium (Gibco, 11320033, Shanghai, China), which was supplemented with 1% penicillin-streptomycin solution (Gibco, 15140148) and 10% fetal bovine serum (FBS, Gibco, 10099141). The culture environment was maintained at 37°C with a 5% CO₂ atmosphere. The primary chondrocytes were treated with 10 ng/mL IL-1β (Proteintech, 211-11B) at 37°C for 24 h [29]. Before the IL-1β treatment, cells were cultured in DMEM/F12 medium with a low serum condition (containing 1% FBS) and 1% penicillin-streptomycin solution. To ensure the completeness of the phenotype, we used the third-generation chondrocytes.

Chondrocytes were plated at a density of 5 × 10⁴ cells per well in a six-well plate. Chondrocytes were cultured in DMEM/F12 medium supplemented with 1% penicillin-streptomycin solution and 10% fetal bovine serum at 37°C with a 5% CO₂ atmosphere. Transfection was performed when cells reached 70%–80% confluence. In a 1-mL transfection system, chondrocytes were transfected with either small interfering RNA targeting SOX11 (si-SOX11; 5 μL, 100 nM, HG-SM009234, HonorGene, Changsha, China), NOX4 overexpression plasmid (oe-NOX4; 3 μg, HG-MO015760, HonorGene), or their corresponding negative controls (si-NC/oe-NC) using Lipofectamine 2000 transfection reagent (5 μL) according to the instructions. The transfection groups were designated as: si-NC, si-SOX11, si-SOX11+oe-NC, and si-SOX11+oe-NOX4. The sequences of the targeted siRNAs used in the study are as follows: si-NC: 5^′-TAGGTATCGGATGGAGTTACC-3^′; si-SOX11#1: 5^′-GGCTCTACTACAGCTTCAAGA-3^′; si-SOX11#2: 5^′-GAGTAGTTTTAAACAGATTTT-3^′.

Chondrocytes were seeded in a six-well plate at a density of 5 × 10⁴ cells per well. Cells were lysed in TRIzol reagent (Thermo Fisher, 15596026, Waltham, MA, USA) for total RNA extraction. cDNA was then synthesized using an mRNA reverse transcription kit (CWBIO, CW2569, Beijing, China). The cDNA template was subjected to qRT-PCR amplification with the UltraSYBR Mixture (CWBIO, CW2601). The 30 μL qPCR reaction mixture contained: 5 μL template DNA (or input control), 1 μL Primer R, 1 μL Primer F, 8 μL ddH2O, and 15 μL 2× SYBR GREEN PCR Master Mix. The thermal cycling conditions were: initial denaturation at 95°C for 10 min; followed by 40 cycles of 95°C for 15 s and 60°C for 30 s. mRNA expression levels were calculated by the 2^-ΔΔCt method, with β-actin serving as the internal control. Primer sequences are provided in Table 1.

Total proteins were extracted from either 5 × 10⁴ cells or 25 mg tissues using RIPA lysis buffer (Beyotime, P0013B, Shanghai, China) containing protease inhibitors (583794, Jintai Hongda Biotechnology Co., Ltd., Beijing, China). Following 30 min incubation on ice, lysates were centrifuged at 12,000× g for 15 min at 4°C, and the supernatants were collected for subsequent analysis. The total protein was quantified via the BCA assay kit (Beyotime, P0012S). Proteins (20 μL) were then separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and electrotransferred onto nitrocellulose membranes. The membranes were blocked with 5% (w/v) skimmed milk powder at room temperature for 90 min, followed by overnight incubation at 4°C with primary antibodies or negative control. Subsequently, the membranes were incubated with HRP Goat anti-Mouse IgG (H+L) Secondary Antibody (1:5000, Abiowell, AWS0001) or HRP Goat anti-Rabbit IgG (H+L) Secondary Antibody (1:5000, Abiowell, AWS0002) for 90 min. β-actin served as the internal reference protein. Detailed information regarding the antibodies employed is provided in Table 2. For visualization, the membranes were incubated with ECL chemiluminescence reagent (K-12045-D50, Advansta, Menlo Park, CA, USA) for 1 min, and images were captured using a ChemiScope6100 gel imaging system (CLINX, Shanghai, China). The resulting exposed image bands were subjected to densitometric analysis using Quantity One professional grayscale analysis software (version 4.6.8, BIO-RAD, Hercules, CA, USA).

The ChIP assays followed the protocol of the ChIP Kit (Abcam, ab500). The cells (3 × 10⁶) were cultured in a 10 cm petri dish and crosslinked with 1.1% formaldehyde (25 μL), Buffer A (45 μL), and PBS (0.6 mL) for 10 min at room temperature. The reaction was quenched with 65 μL glycine (1.25 M) at room temperature for 5 min. Cells were harvested by centrifugation at 500× g for 5 min at 4°C. The pellet was washed once with 1 mL of ice-cold PBS and centrifuged again under the same conditions. The washed cell pellet was sequentially resuspended in 1 mL of Buffer B and then 1 mL of Buffer C, with each incubation step performed at room temperature for 10 min. After each resuspension, the mixture was centrifuged at 500× g for 5 min and the supernatant discarded. The resulting nuclear pellet was resuspended in a mixture of 4 μL 25× PI Mix and 96 μL Buffer D. Chromatin was sheared by ultrasonication (XIAO MEI CHAO SHENG, XM-250T, Kunshan, Jiangsu, China) using the following parameters: 4 s ON/6 s OFF pulses, for a total sonication time of 60 s at 20% power output. The sonicated lysate was centrifuged at 14,000× g for 5 min at 4°C. The supernatant, containing the fragmented chromatin, was collected. A 20 μL aliquot of supernatant was mixed with 100 μL of ddH₂O, followed by the addition of 100 μL of DNA purifying solution and thorough mixing. The mixture was incubated at 98°C for 10 min, followed by 20 min at room temperature. After centrifugation at 10,000 × g for 10 s, 1 μL proteinase K was added and vortexed briefly (5 s). Subsequent incubations were performed at 55°C for 30 min and then 98°C for 10 min. The mixture was finally centrifuged at 14,000× g for 1 min at room temperature. The supernatant was transferred to a new 1.5 mL tube. Fragment size distribution was analyzed by electrophoresis on a 1.5% agarose gel using 10 μL of this sample. Each ChIP reaction used 1 μg fragmented chromatin dissolved in 1 mL 1× ChIP buffer and vortexed for 5 s. The ChIP sample was incubated with either 2 μL IgG antibody (Merck Millipore, USA) or SOX11 polyclonal antibody (Thermo Fisher, PA5-47557) at 4°C overnight. The washed agarose beads (40 μL) were resuspended in 1× ChIP buffer (5.3 mL). A 120 μL aliquot of the bead suspension was mixed with 1 mL 1× ChIP buffer, centrifuged at 500× g for 5 min at 4°C, and the supernatant was removed. Meanwhile, the ChIP sample was centrifuged at 14000× g for 5 min at 4°C. The resulting supernatant was transferred to a fresh tube containing the pre-blocked agarose beads and incubated at 4°C for 30–60 min under constant rotation. Following incubation, the mixture was centrifuged at 500× g for 3 min, and the supernatant was discarded. The bead-bound immune complexes were subsequently washed three times with 1× ChIP buffer. The washed agarose beads were resuspended with 100 μL DNA purification slurry at 98°C for 10 min, followed by a 20-min incubation at room temperature, and centrifuged at 10,000× g for 10 s. Proteinase K (1 μL) was added to the obtained sample, vortexed for 5 s, incubated sequentially at 55°C for 30 min and then 98°C for 10 min, and centrifuged at 14000× g for 1 min. Then, 70 μL the clarified supernatant was carefully transferred to a new 1.5 mL microcentrifuge tube. The purified DNA products were tested by qRT-qPCR amplification with the UltraSYBR Mixture. The 30 μL qPCR reaction mixture contained: 5 μL template DNA (or input control), 1 μL Primer R, 1 μL Primer F, 8 μL ddH2O, and 15 μL 2× SYBR GREEN PCR Master Mix. The thermal cycling conditions were: initial denaturation at 95°C for 10 min; followed by 40 cycles of 95°C for 15 s and 60°C for 30 s. Primer sequences are provided in Table 1.

The potential transcriptional binding sites of SOX11 and NOX4 were identified using the Jaspar database (https://jaspar.elixir.no/). The wild-type (WT) promoter region upstream of the NOX4 gene was amplified by PCR, and specific mutations (MUT) were introduced using site-directed mutagenesis techniques. Both the WT and MUT NOX4 promoter fragments were cloned into pGL3-Basic luciferase reporter vectors. HEK 293 T cells (HonorGene, HG-NC071) were maintained in DMEM medium in a six-well plate at a density of 5 × 10⁴ cells per well. For luciferase reporter assays, HEK 293 T cells were transfected using Lipofectamine 2000 Transfection Kit (5 μL/well, Invitrogen, 11668-019, Carlsbad, CA, USA) according to the manufacturer’s instructions. In each experiment, pGL3-Basic-NOX4-WT (2 μg) or pGL3-Basic-NOX4-MUT (2 μg) reporter construct were transfected separately into distinct wells, along with 0.2 μg Renilla luciferase control plasmid (pRL-TK, HonorGene) and either 2 μg si-SOX11 or si-NC plasmid. Following 48 h of incubation at 37°C under 5% CO₂, luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega, GloMax 20/20, Madison, WI, USA) in strict accordance with the manufacturer’s instructions. The firefly luciferase luminescence activity was normalized to the control Renilla luciferase activity. The experiment was independently repeated three times, representing biological replicates.

The concentration of malondialdehyde (MDA), superoxide dismutase (SOD), glutathione (GSH), and Fe²⁺ in cell lysates/tissues was measured using the MDA assay kit (Jiancheng, A003-1-2, Nanjing, Jiangsu, China), SOD kit (Jiancheng, A001-3-2), microreduced GSH assay kit (Jiancheng, A006-2-1), and an iron ion assay kit (Jianglaibio, JL-T1255, Shanghai, China). The determination procedure was based on the kit protocol.

Chondrocytes were seeded in a six-well plate at a density of 5 × 10⁴ cells per well. Following trypsinization, cells were harvested by centrifugation at 1000× g for 5 min at 4°C. Following resuspension in fresh culture medium, the cells were incubated with 5 μM C11-BODIPY (Invitrogen, D3861) at 37°C for 30 min in the dark. Lipid peroxidation was subsequently quantified using flow cytometry (Beckman Coulter, A00-1-1102, Shanghai, China) under 581 nm excitation wavelength, 591 nm emission wavelength.

Chondrocytes were seeded in a six-well plate at a density of 5 × 10⁴ cells per well. Following treatment, cells were detached with 0.25% (w/v) trypsin-EDTA solution (Thermo Fisher) and stained with 10 µM DCFH-DA (Beyotime, S0033S) diluted in serum-free culture medium at 37°C for 30 min. ROS levels were quantified by flow cytometry (Beckman Coulter, A00-1-1102). The maximum excitation wavelength is 488 nm, and the maximum emission wavelength is 525 nm.

Chondrocytes were seeded in a six-well plate at a density of 5 × 10⁴ cells per well. Following incubation with 0.5 mL 1× JC-1 staining solution (Beyotime, #C2006) at 37°C for 10 min, cells were washed three times with JC-1 staining buffer. Flow cytometric analysis was subsequently performed, with the mitochondrial membrane potential (ΔΨm) being quantified through fluorescence ratio analysis. The red/green fluorescence intensity ratio (H1-UL aggregates to H1-UR monomers) served as the indicator of JC-1 polarization status. The maximum excitation wavelength of the JC-1 monomer is 514 nm, and the maximum emission wavelength is 529 nm; the maximum excitation wavelength of the JC-1 aggregates is 585 nm, and the maximum emission wavelength is 590 nm.

Chondrocytes were seeded in a six-well plate at a density of 5 × 10⁴ cells per well. According to the manufacturer’s protocol, cell death was assessed using a live/dead cell staining kit (KeyGEN Bio TECH, KGAF001, Nanjing, China). Briefly, cells were incubated with 2 μM fluorescein AM and 8 μM propidium iodide at room temperature for 30 min. The percentage of cell death was quantified using flow cytometry (Beckman Coulter, A00-1-1102).

All data are presented as the mean ± standard deviation. Statistical analyses were performed using GraphPad Prism 8.0 software (GraphPad Software Inc., San Diego, CA, USA). For comparisons between two groups, an unpaired two-tailed Student’s t-test was utilized. For multiple group comparisons, one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparisons test was employed. A p-value less than 0.05 was considered statistically significant.

To verify the pathological features of the OA mouse model, we employed H&E staining, Safranin-O/Fast Green staining, and molecular detection [matrix metalloproteinase 13 (MMP13) and Collagen II levels]. In the Sham group, the cartilage exhibited a distinct layered structure (surface layer, middle layer, and deep layer) and orderly cell arrangement, characteristic of normal cartilage tissue. In contrast, the OA group displayed obvious surface cartilage damage, indicative of chondrocyte loss and matrix fibrosis. This damage was further quantified by an elevated OARSI score (Fig. 1A). Compared with the Sham group, the thickness of the transparent cartilage layer in the OA group was reduced, while the calcified cartilage thickness increased (Fig. 1B). This change strongly suggested the massive loss of proteoglycans and the occurrence of cartilage calcification. SOX11 and NOX4 expression in the knee joint tissue of the OA group was increased relative to the Sham group (Fig. 1C–E). The immunohistochemical negative control is presented in supplementary Fig. S1A. Additionally, the OA model induced an increase in MMP13 levels and a decrease in Collagen II levels (Fig. 1E). Negative controls of Western blot are shown in supplementary Fig. S1B. Our results indicate that the expressions of SOX11 and NOX4 are upregulated in OA mice, suggesting that they may serve as biomarkers for OA.

Subsequently, the possible regulatory mechanisms of SOX11 and NOX4 in chondrocytes were analyzed in vitro. Compared with the si-NC group, SOX11 expression in the si-SOX11#1 and si-SOX11#2 groups were decreased, confirming the efficiency of si-SOX11 transfection (Fig. 2A,B). SOX11 knockdown reduced NOX4 expression in chondrocytes (Fig. 2A,B). ChIP-qPCR detection revealed that the binding efficiency of SOX11 to the promoter region of NOX4 in the SOX11 knockdown group was lower than that in the si-NC group (Fig. 2C). The dual-luciferase reporter gene system showed that the luciferase activity of the wild-type (WT) NOX4 promoter group decreased after SOX11 knockdown, while the mutant (MUT) promoter group showed no significant change (Fig. 2D). Our results confirm that SOX11 promotes NOX4 transcription in chondrocytes by directly binding to its promoter region.

Emerging evidence establishes SOX11^′s regulatory role in ferroptosis mechanisms [23]. Next, the SOX11 effect on ferroptosis in IL-1β-induced chondrocytes through regulating NOX4 was investigated. IL-1β stimulation upregulated SOX11, NOX4, and MMP13 levels in chondrocytes, while also causing a decrease in Collagen II levels (Fig. 3A,B). Knockdown of SOX11 inhibited NOX4 level in IL-1β-induced chondrocytes. When NOX4 was overexpressed, this inhibitory effect was reversed (Fig. 3C,D), confirming that SOX11 exerts its effect by positively regulating NOX4. IL-1β treatment upregulated ACSL4 level in chondrocytes and downregulated GPX4 and SLC7A11 levels. This phenomenon was reversed by knocking down SOX11. NOX4 overexpression eliminated these protective effects (Fig. 3E,F). Negative controls of Western blot are shown in Fig. S1B. SOX11 knockdown reduced the Fe²⁺ concentration and cell death in chondrocytes subjected to IL-1β, which was counteracted by NOX4 overexpression (Fig. 3G,H). Immunofluorescence results show that SOX11 knockdown increased GPX4 expression in chondrocytes subjected to IL-1β. NOX4 overexpression reversed this phenomenon (Fig. 3I). Our results indicate that knockdown of SOX11 reduces ferroptosis in chondrocytes subjected to IL-1β by down-regulating NOX4.

Then, we further explored the SOX11 role in regulating NOX4 to exert its effect on mitochondrial dysfunction in chondrocytes subjected to IL-1β. Knocking down SOX11 inhibited lipid peroxidation in chondrocytes subjected to IL-1β. NOX4 overexpression reversed the downward trend of lipid peroxidation (Fig. 4A). Knocking down SOX11 increased the mitochondrial membrane potential in chondrocytes subjected to IL-1β, and this effect was counteracted by NOX4 overexpression (Fig. 4B). Additionally, low expression of SOX11 reduced ROS production and MDA levels in chondrocytes subjected to IL-1β, while increasing GSH and SOD levels. When NOX4 was overexpressed, these protective effects brought about by low expression of SOX11 were eliminated (Fig. 4C,D). Our results indicate that knocking down SOX11 alleviates mitochondrial dysfunction in chondrocytes subjected to IL-1β through NOX4.

Finally, in vivo experiments were conducted to investigate SOX11 effects on ferroptosis and mitochondrial dysfunction in OA mice. Immunohistochemical results showed that SOX11 knockdown reduced SOX11 and NOX4 expression in the knee joint tissues of the OA group (Fig. 5A), and negative controls is presented in Fig. S1C. H&E staining results show that the Sham group clearly displayed the layered structure of cartilage (surface layer, middle layer, and deep layer) and cell arrangement, demonstrating the typical characteristics of normal cartilage tissue. However, in the OA group, the surface of the cartilage showed obvious damage, which directly indicated the loss of chondrocytes and matrix fibrosis. At the same time, the OARSI score was increased. Knockdown of SOX11 alleviated the cartilage damage induced by OA and reduced the OARSI score (Fig. 5B). SOX11 knockdown increased the thickness of the transparent cartilage layer in the knee joint tissues of OA model mice and reduced the thickness of calcified cartilage (Fig. 5C). These results indicate that SOX11 knockdown alleviates OA. In addition, SOX11 knockdown increased GPX4 level in the knee joint tissues of OA model mice (Fig. 6A) and mitochondrial membrane potential (Fig. 6B), reduced ROS production, MDA levels, and Fe²⁺ concentration, and increased GSH and SOD levels (Fig. 6C–E). Our results show that SOX11 knockdown alleviates OA through reducing ferroptosis and mitochondrial dysfunction.