Analysis of single-cell RNA-sequence data

The single-cell transcriptome analysis utilized original data sourced from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) (Dataset number: GSE176308). The Seurat software package facilitated object generation and the exclusion of low-quality cells. Subsequent steps involved standard data preprocessing algorithms, which included calculating gene count, cell count, and mitochondrial sequencing count percentages. Cells with a gene count ranging from 200 to 6,000 were considered. To normalize library size effects across cells, unique molecular identifier counts were scaled using a factor of 10,000. The processed normalized data were then used for subsequent analyses as prescribed by the Seurat package. Principal component analysis was conducted on the top 20 variable genes, retaining the first six principal components for further uniform manifold approximation and projection (UMAP) visualization and clustering.

Cell culture

Mouse synovial fibroblasts, sourced from Hycyte (PCC-C224, Suzhou, China), were treated with 100 nM SOD3 mimics (GenePharma, Shanghai, China) or with 100 nM control substances. The sequences for SOD3 mimics and the negative control are documented in Supplementary Table 1. Chondrocytes were harvested from the cartilage of 8-week-old mice. The tissue was excised meticulously using a sterile blade and then the cartilage was minced into 1-mm^3 fragments under a microscope. The fragments were subsequently subjected to digestion in DMEM/F12 medium supplemented with 2 mg/mL type II collagenase (Thermo) for a duration of 6 h. The chondrocytes were digested, centrifuged, and then cultured in DMEM/F12 medium with 10% FBS. For subsequent experiments, chondrocytes at passage two were treated with IL-1β (10 ng/mL) and exosomes (EXOs or S-EXOs, 10 µg/mL).

Isolation of exosomes

Exosomes were isolated from mouse synovial fibroblasts at passages 3–6, extracted from the supernatant of 2 × 107 cell cultures. The medium was initially centrifuged at 300 g for 10 min to remove cells, then at 3,000 g for 15 min to discard cell debris. The supernatant was further centrifuged at 120,000 g for 1 h. The pelleted particles were washed in PBS to remove contaminants, followed by a final centrifugation at 120,000 g for 70 min. The purified particles were resuspended in 200 µL of PBS and stored at -80 °C. Transmission electron microscopy (TEM, Leica, Germany) assessed the morphology and size distribution, while nanoparticle tracking analysis (NTA, Malvern, Worcestershire, UK) measured the particle size. Western blotting was performed to detect exosomal markers CD9, CD63, and TSG101, and Calnexin in the cell lysate.

qRT-PCR

Trizol reagent (Beyotime) was applied to extract total RNA. The synthesis of complementary DNA (cDNA) was conducted utilizing a cDNA synthesis kit from Beyotime, with the reaction being initiated using the SYBR Green kit from Bio-Rad, CA. The relative levels of target genes (Sod2, Sod3, Col2a1, Acan, Mmp13, Adamts5) were quantified using Gapdh as the reference, employing the formula χ = 2-ΔΔCT. Primer sequences used in our study were recorded in Supplemental Table 2.

Western blot

Proteins were isolated using RIPA lysis buffer with protease inhibitors (Beyotime). The concentration of the protein samples was tested using the BCA kit (Beyotime). An equivalent quantity of protein underwent electrophoresis on a 10% SDS-PAGE gel and subsequently underwent transfer to a nitrocellulose membrane. After blocking at room temperature for 1 h, the membranes were incubated with primary antibodies overnight at 4 °C. The antibodies used were COL II (Abcam, ab188570), ACAN (Abcam, ab315486), MMP13 (Abcam, ab39012), ADAMTS5 (Abcam, ab41037), β-actin (Abcam, ab8226), CD9 (CST, 98,327 S), CD63 (Immunoway, YT5525), SOD3 (Santa, 271,170), Calnexin (Affinity, AF5362), and TSG101 (Abcam, ab125011). The subsequent day, the membranes were incubated with HRP-conjugated secondary antibodies (Abcam, ab6721 or ab6788) at room temperature for a duration of 1 h. Detection was achieved using a highly sensitive chemiluminescence substrate (Beyotime), and optical density (OD) values were quantitatively analyzed using β-actin as the loading control.

Immunofluorescence staining

Chondrocytes were fixed in 4% polyformaldehyde solution (Aladdin, Shanghai, China) for 30 min and permeabilized with Triton X-100 solution (Beyotime) for 10 min. After a 30-minute blocking, the cells were incubated with primary antibodies at 4 °C overnight. After incubating with secondary antibodies (Abcam, ab150077 or ab150079), cell nuclei were stained with DAPI solution (Beyotime) and examined under a fluorescence microscope (Zeiss).

Detection of antioxidant capacity

Chondrocytes were divided into four groups: CTRL, IL-1β, IL-1β + EXOs, and IL-1β + S-EXOs. The IL-1β group was stimulated with 10 ng/mL of IL-1β, while the IL-1β + EXOs and IL-1β + S-EXOs groups received IL-1β and exosomes (EXOs and S-EXOs, 10 µg/mL) treatment. Intracellular ROS levels were assessed by treating cells with 5 µM DCFH-DA (Beyotime) for 20 min at 37 °C. Mitochondrial ROS were evaluated using MitoSOX solution (5 µM, Beyotime). Fluorescence microscopy (Zeiss) was used to visualize the cells. In the DPPH assay, a mixture of 2 mL cell medium and 2 mL DPPH solution (0.04 mg/mL) was incubated for 30 min, followed by absorbance measurement at 515 nm using UV-visible spectroscopy. For the ABTS assay, a 0.2 mL ABTS (7.4 mM) solution was mixed with 0.2 mL potassium persulfate solution (2.6 mmol/L), incubated in the dark for 24 h to form ABTS+, then diluted 50-fold with PBS and mixed with the medium for a 10-minute reaction before measuring absorbance at 734 nm. In the PTIO experiment, 2 mL of PTIO solution (0.05 mg/mL) was added to each group’s medium and incubated at room temperature for 2 h, with absorbance measured at 557 nm. The hydroxyl radical scavenging assay was conducted following the kit’s instructions (Solarbio, Beijing, China), and absorbance at 550 nm was measured to quantify hydroxyl radical levels.

Preparation and characterization of EXOs-loaded microspheres

A mixture containing 10 wt% GelMA, a photoinitiator, and 5 wt% Span 80 oil was processed in a microfluidic device to create uniformly sized droplets. The crosslinking of these droplets was achieved through UV irradiation. The resultant microspheres were washed with 75% ethanol solution and stored in deionized water. For surface modification, GelMA microspheres (GMs) were soaked in a polydopamine (PDA) solution (2 mg/mL in 10 mM Tris HCl, pH 8.5) for 6 h, forming GM@PDA. The size and morphology of the microspheres were examined using an optical microscope, and the structure of the freeze-dried microspheres was analyzed with a scanning electron microscope (SEM). Elemental mapping was conducted with an exposure time of 180 s.

Exosome adsorption and release assay

The exosome adsorption process was performed in sterile conditions. Five milliliters of extracellular vesicles (1,000 µg/mL) were mixed with 100 mg of microspheres and incubated at 4oC. The adsorption kinetics of the extracellular vesicles were monitored over specified intervals. At each time point, the supernatant was collected for analysis. To evaluate the release rate of extracellular vesicles, a complex comprising 5 mg of microspheres and extracellular vesicles was submerged in 100 µL of PBS and incubated on a horizontal shaker at 37oC and 1,000 rpm. The supernatant was collected at designated time points and refreshed with new PBS. The concentration of EXOs in the supernatant was determined through the BCA kit.

Cells proliferation and viability

Cell proliferation was evaluated using the Cell Count Kit-8 (CCK-8, Beyotime, Haimen, China). Chondrocytes were incubated with various microspheres (GM@PDA, GM@PDA@EXO, or GM@PDA@S-EXO, 2 mg/well) and the CCK-8 solution at 37 °C. After 1 h, the OD was measured with a multi-plate reader (BioTek, Winooski, VT) at 450 nm. To determine cell viability, chondrocytes were co-cultured with different microspheres (GM@PDA, GM@PDA@EXO, or GM@PDA@S-EXO, 8 mg/well), then washed with PBS and stained using a live/dead staining kit (Beyotime) for 20 min on days 1, 3, and 5. Fluorescence microscopy (Zeiss, Oberkochen, Germany) was used to take photos of live and dead cells.

The establishment of OA model in mice

The animal study protocol was approved by the Ethics Committee of Binzhou Medical College (2023 − 303). Before the surgery, mice were anesthetized with a 2.0% isoflurane and 30% oxygen mixture (RWD Life Sciences, Shenzhen, China). A longitudinal incision was made on the medial side of the knee joint to expose the joint capsule, followed by lateral dislocation of the patella to access the articular cavity. The medial meniscus ligament (MML) was then sectioned using fine scissors. For the sham surgery group, only the joint capsule was exposed, preserving the MML’s structural integrity. Starting one-week post-operation, mice were administered intra-articular injections of GM@PDA, GM@PDA@EXOs, and GM@PDA@S-EXOs microspheres into the joint cavity every two weeks, at 10 µL per injection.

Histology and immunofluorescence staining

Eight weeks post-surgery, mouse knee joints were harvested and fixed in formalin for 48 h. After decalcification and paraffin embedding, 5 μm thick sagittal sections were prepared from the samples. Histological analysis was performed using hematoxylin and eosin (H&E) and Safranin O-fast green staining. Cartilage degradation was evaluated using the Osteoarthritis Research Society International (OARSI) scoring system. For immunofluorescence staining, sections were blocked with 1% fetal bovine serum (FBS) and permeabilized with 0.1% Triton X-100. After antigen retrieval in a solution heated to 110 °C for 10 min, the sections were incubated overnight at 4 °C with ACAN and SOD3 antibodies, followed by incubation with a secondary antibody (ab6721). The nuclei were stained using DAPI (Beyotime) and subsequently captured using a fluorescence microscope (Zeiss).

Statistical analysis

Statistical analysis was performed using GraphPad Prism 9.2 software (GraphPad Software Co., Ltd.). The t-test was employed for data that exhibited a normal distribution, with comparisons between two groups conducted using an independent two-tailed Student’s t-test. In instances of multiple group comparisons, one-way analysis of variance (ANOVA) was applied, followed by Tukey’s post hoc test. Statistical significance was determined by a p-value of < 0.05.