Dual-engineered cartilage-targeting extracellular vesicles derived from mesenchymal stem cells enhance osteoarthritis treatment via miR-223/NLRP3/pyroptosis axis: Toward a precision therapy

2.1. Identification and characterization of hUCMSCs and hUC-EVs

hUCMSCs displayed good proliferation capability and typical spindle-like morphology (Fig. 1A). For the tri-lineage differentiation potential, hUCMSCs were able to successfully differentiate into the osteogenic (Fig. 1B, left panel, Alizarin Red staining), adipogenic (Fig. 1B, middle panel, Oil Red O staining), and chondrogenic lineages in respective medium (Fig. 1B, right panel, Alcian Blue staining). Expressions of the MSC surface markers were analyzed using flow cytometry, showing that hUCMSCs were positive for CD73 and CD90, and negative for CD34 and CD45 (Fig. 1C).

Transmission electron microscopy (TEM) revealed that hUC-EVs exhibited a typical spherical bilayer membrane structure (Fig. 1D). Nanoparticle Tracking Analysis (NTA) of these vesicles indicated an average diameter of 142.9 nm, and the main peak of particle size was 131.3 nm (Fig. 1E). In addition, the concentration was about 2.2 × 10¹⁰ particles/ml. Further, western blotting revealed that exosomal specific markers (CD81, Tsg101, Alix) were positive, while Calnexin was negative (Fig. 1F). The above results indicated that EVs were successfully isolated from hUCMSCs.

2.2. hUC-EVs promote cell survival and anabolism of IL-1β-induced chondrocytes

We then investigated the effect of hUC-EVs on chondrocytes in vitro, and the experimental schematic is shown in Fig. 2A. The cellular uptake of DiO-stained hUC-EVs was observed under confocal fluorescence microscope after co-incubation for 12 h with chondrocytes (Fig. 2B, negative control in Fig. S1), which showed the DiO-stained hUC-EVs entered chondrocytes successfully. IL-1β (10 ng/mL, all the same below) was used to imitate the inflammatory microenvironment of OA [44]. IL-1β-induced chondrocytes were treated with hUC-EVs (5 × 10⁸ particles/ml, hUC-EVs group) or normal medium (IL-1β group) for 24 h. The results of cell counting kit-8 (CCK-8) assay showed that hUC-EVs could partially reverse IL-1β-decreased chondrocytes viability (Fig. 2C). Meanwhile, the protein levels of cartilage anabolic factors (Collagen II and SOX9) and catabolic factor (matrix metalloproteinase 13, MMP13), which were highly related to the pathology of OA [45,46], were measured by western blotting. Consistent with former studies [47,48], IL-1β treatment significantly reduced Collagen II and SOX9 expression, while upregulated the level of MMP13. The supplement of hUC-EVs reversed the effect of IL-1β on chondrocytes (Fig. 2D and E). The results of immunofluorescence staining (Collagen II and Aggrecan) also supported the findings (Fig. 2F–H). In summary, hUC-EVs could promote cell survival and anabolism of IL-1β-treated chondrocytes.

2.3. hUC-EVs alleviate the inflammation and cartilage degeneration in OA rat model

To evaluate the role of hUC-EVs on articular cartilage protection, the monosodium iodoacetate (MIA)-induced OA model of rat was carried out. The schematic diagram of animal experiments is shown in Fig. 3A. The OA rat knees were treated with 10 μL negative control (PBS group) or 10 μL hUC-EVs (10¹⁰ particles/ml, diluted in PBS) (hUC-EVs group) once a week for 4 times 1 week after the MIA injection. The knee joints were harvested and examined at 5 weeks. Pro-inflammatory factors in the cartilage, which play key roles in OA progression, were measured by ELISA, including IL-1β, MMP2, tumor necrosis factor α (TNF-α), prostaglandins E2 (PGE2), etc [49]. The results suggested that hUC-EVs markedly inhibited the generation of IL-1β, MMP2, TNF-α, and PGE2 compared to the PBS group (Fig. 3B).

The cartilage degradation, characterized by gradual loss of collagen II, aggrecan, and proteoglycan in extracellular matrix (ECM), is a typical pathological manifestation of OA [50]. The mRNA levels of anabolic factors (Collagen II, SOX9), and catabolic factor (MMP13) were measured using qRT-PCR, which showed that hUC-EVs could reverse the inhibition of cartilage anabolism induced by MIA (Fig. 3C). Histological analyses, including Hematoxylin-eosin (H&E) staining, Safranin O/Fast Green staining (aggrecan and proteoglycan), and immunohistochemical staining (collagen II) of knee joint sections were performed. As the representative images shown in Fig. 3D, the control group showed normal structures of the articular cartilage, with smooth surfaces, and aggrecan, proteoglycan, and collagen II were uniformly distributed. In the PBS group, articular cartilage was obviously degenerated after MIA injection, with ECM loss, representing the progression of OA. The changes were partly rescued in the EVs-treated group, in which the aggrecan, proteoglycan, and collagen II were obviously upregulated compared to the PBS group. The maximal and summed scores of the Osteoarthritis Research Society International (OARSI) scoring system (Fig. 3E), as well the Mankin's score (Fig. 3F), were evaluated to quantify the severity of cartilage degeneration. Moreover, Collagen II protein levels in articular cartilage were assessed using the immunohistochemical staining results (Fig. 3G). Collectively, these results revealed that hUC-EVs injection could significantly inhibit inflammation, and alleviate the cartilage degradation in MIA-induced OA models.

2.4. miR-223 mediates the inhibitory effect of hUC-EVs on NLRP3 inflammasome

To further investigate the underlying mechanism of hUC-EVs promoting chondrocyte survival and cartilage repair, we performed high-throughput miRNA sequencing (service provided by BGI, Shenzhen) of cartilage tissues in the PBS group and hUC-EVs group. As the volcano map demonstrates, 12 miRNAs were upregulated and 30 downregulated in the hUC-EVs group (Fig. 4A), and the details of the miRNA differential expression are presented in the heat map (Fig. 4B). In recent years, cell pyroptosis and NLRP3 inflammasome has attracted increasing attention for their roles in promoting OA progression and cartilage degeneration [36,37,51]. The assembling of NLRP3 inflammasome activates Caspase-1, which then leads to the release of pro-inflammatory cytokines IL-18 and IL-1β, as well as gasdermin D (GSDMD)-triggered cell pyroptosis [28,52]. We hypothesized that hUC-EVs may protect chondrocytes and articular cartilage by inhibiting NLRP3-mediated cell pyroptosis and inflammatory cascade in OA rat models. In order to determine potential miRNA targets NLRP3, we used the website http://www.targetscan.org. And miR-223, which could directly target the 3′UTR of NLRP3 mRNA, decrease NLRP3 expression [53,54], was found to be significantly upregulated in cartilage treated with hUC-EVs (Fig. 4B). Then we examined the expression of miR-223 among chondrocytes treated with hUC-EVs or PBS using qRT-PCR, which confirmed that miR-223 level was significantly elevated in the EVs-treated chondrocytes (Fig. 4C). The probable binding site between NLRP3 mRNA and miR-223 was predicted (Fig. 4D). To verify the direct interaction between NLRP3 and miR-223, we built luciferase reporting vectors containing the wild type (WT) and mutant type (MUT) of 3′UTR sequence in NLRP3 mRNA and transfected the chondrocytes. As has been expected, the relative luciferase activity of the WT in the miR-223 mimics group was significantly reduced, while for the MUT there was no obvious change (Fig. 4E). Finally, the effect of hUC-EVs on NLRP3 protein expression levels was reconfirmed using western blotting. The protein level of NLRP3 and GSDMD distinctly decreased in chondrocytes cultured with hUC-EVs, which was partially reversed by miR-223 inhibitor, indicating the miR-223 dependence of above effects by hUC-EVs (Fig. 4F, quantification analysis is presented in Fig. S2). In conclusion, miR-223 was upregulated both in chondrocytes and cartilage tissues after hUC-EVs treatment, and it could directly target NLRP3 to mediate the inhibitory effect on NLRP3 expression.

2.5. CTP- and mir-engineering of hUC-EVs

Based on above findings, to further enhance the hUC-EVs’ effect on anabolism promoting and inflammation inhibiting of chondrocyte and cartilage, and provide a promising therapeutic regimen for osteoarthritis treatment, two engineering methods were carried out to generate three kinds of engineered EVs (Fig. 5A). Three plasmids, encoding the Collagen II-targeting peptide (CTP)-lysosome-associated membrane glycoprotein 2 b (Lamp2b), enhanced green fluorescent protein (EGFP)-Lamp2b, or CTP-EGFP-Lamp2b were constructed and packaged with lentiviral vectors. The lentiviruses were then applied to transfect hUCMSCs to generate surface-modified EVs (Fig. 5B). The efficiency of the transfection in hUCMSCs was evaluated by fluorescence imaging, and more than 80% hUCMSCs transfected with lentivirus vectors are positive for EGFP, indicating that hUCMSCs with EGFP-Lamp2b or CTP-EGFP-Lamp2b expression were successfully established (Fig. 5C). Then, the CTP-EGFP-Lamp2b-EVs and EGFP-Lamp2b-EVs were injected into joint cavity of rat knees, and after 24 h, the fluorescence imaging indicated that CTP-EGFP- Lamp2b-EVs were mainly distributed in the articular cartilage, while a large proportion of EGFP-Lamp2b-EVs were absorbed by synovium, showing good cartilage-targeting ability of CTP-modified EVs (Fig. 5D, the quantification data is presented in Fig. S4A as the ratio of MFI in cartilage and synovium). After 48 h, fluorescence imaging confirmed that CTP-EGFP-Lamp2b-EVs entered knee chondrocytes more effectively than EGFP-Lamp2b-EVs (Figs. S3 and S4B). Then, miR-223 was loaded into the ordinary hUC-EVs and primarily-engineered CTP-EVs by electroporation, generating the miR-223-abundant Mir-EVs and dual-engineered CTP/Mir-EVs, respectively. The morphology of EVs remained intact after miR-223 electroporation, which was observed by TEM and dynamic light scattering (Fig. S5). We measured the miRNA loading efficiency, by comparing the free miR-223 left in the solution to the total input, and it turned out to be around 60% in either hUC-EVs or CTP-EVs (Fig. S6).

2.6. CTP- and mir-engineering both amplify the effect of hUC-EVs on promoting survival and improving anabolism of chondrocytes via mir-223/NLRP3/pyroptosis axis

We next observed the therapeutic effects of the various engineered EVs in vitro and validated our conjecture that the miR-223/NLRP3 axis mediated the beneficial effects of hUC-EVs and subsequent engineered EVs on chondrocyte protection. Briefly, we treated IL-1β-induced chondrocytes with aforesaid different EVs (Mir-EVs, CTP-EVs and CTP/Mir-EVs) at the concentration of 5 × 10⁸ particles/ml for 24 h. In reverse validation experiment group, for activating NLRP3, the cells were treated with a NLRP3 agonist -- nigericin sodium salt (NSS) (1 μM). The schematic is shown in Fig. 6A and experimental grouping in Fig. 6B.

To assess the effect of engineered EVs on cell viability and pyroptosis, CCK-8 assay (Fig. 6C), and LDH activity analysis of supernatant (Fig. 6D), which is used as a measure of pyroptosis were carried out. Mir-EVs and CTP-EVs could markedly reverse IL-1β-induced viability decrease of chondrocytes, along with the LDH release, and no significant difference was observed between the two groups. The dual-engineered CTP/Mir-EVs improved the viability and downregulated the LDH level further, which showed significant differences compared with Mir-EVs and CTP-EVs. Interestingly, the gain effects of all engineered EVs were reversed by NSS, which proved that engineered EVs exert their protective roles via miR-223/NLRP3/pyroptosis axis. The protein levels of Collagen II, SOX9, MMP13, NLRP3 inflammasome and pyroptosis-related molecules (NLRP3, cleaved caspase-1, GSDMD-N) was measured by western blotting (Fig. 6E and F). Accordingly, Mir-EVs and CTP-EVs treatment significantly increased Collagen II and SOX9 expression, while downregulated the protein level of MMP13, NLRP3, cleaved caspase-1, and GSDMD-N. CTP/Mir-EVs further enhanced these effects and NSS reversed the above. The results of immunofluorescence staining also supported the findings (Fig. 6G-L). In summary, the above results showed that CTP- and Mir-engineering both amplify the effect of hUC-EVs on promoting survival and improving anabolism of chondrocytes via mir-223/NLRP/pyroptosis axis, and the dual-engineered CTP/Mir-EVs achieved better therapeutic results than single modified Mir-EVs or CTP-EVs.

2.7. Dual-engineered CTP/Mir-EVs further promote cartilage repair, inhibit inflammation and ameliorate the severity of OA

Finally, we validated that the newly engineered EVs could further promote cartilage repair and ameliorate the severity of OA in rat model and provide as a novel and rewarding therapeutic choice for clinical osteoarthritis treatment.

The MIA-induced OA rat were treated with the newly engineered EVs (diluted in PBS) or negative control (PBS) once a week for 4 times 1 week after the MIA injection. The knee joints were harvested and examined at 5 weeks. In order to observe the activation of inflammasomes more directly and objectively, IF staining of NLRP3 and GSDMD-N in articular cartilage sections was performed (Fig. 7A), and the semiquantitative analysis was in Fig. S7. The results validated that hUC-EVs decreased the number of chondrocytes positive for NLRP3 and GSDMD-N in the cartilage, inhibiting the NLRP3 inflammasome activation and pyroptosis in OA knees, and engineered EVs amplified this effect, among which the CTP/Mir-EVs performed the best. To verify the anti-inflammatory effects of the engineered EVs, pro-inflammatory factors in cartilage samples were measured by ELISA, which suggested that CTP/Mir-EVs markedly reduced the generation of MMP2, IL-1β, TNF-α, and PGE2 in vivo compared with Mir-EVs and CTP-EVs (Fig. 7B). Consistent with the above results, H&E, Safranin O/Fast Green, and immunohistochemical staining (collagen II, collagen I) of knee joint sections showed that Mir-EVs and CTP-Mirs both partly reversed the changes of OA, while CTP/Mir-EVs further restored nearly normal articular cartilage structure (Fig. 7C, Fig. S8A). The CTP/Mir-EVs group had the lowest maximal and summed scores of the OARSI (Fig. 7D), along with the Mankin's score (Fig. 7E). Moreover, collagen II and collagen I protein levels in articular cartilage were also the highest and the lowest in the CTP/Mir-EVs group, respectively (Fig. 7F, Fig. S8B). Micro-CT was performed to investigate the progressive cartilage damage and subchondral bone degeneration, which usually shows joint space narrowing and bone volume/total volume (BV/TV) decreasing [55]. The results revealed that the joint space widths (JSWs) significantly decreased in the PBS group compared with the control group, the EVs, especially the CTP/Mir-EVs treatment significantly increased the JSW (Fig. 8 A, C, and D). Also, the subchondral bone was obviously destroyed, with the BV/TV declined in the PBS group, and CTP/Mir-EVs significantly suppressed this effect, indicating that CTP/Mir-EVs can alleviate joint damage effectively (Fig. 8B, E). Collectively, the above results showed that CTP/Mir-EVs could further inhibit NLRP3 inflammasome activation, chondrocyte pyroptosis and cartilage degradation, which amplified the therapeutic effect for MIA-induced OA.

A schematic overview of the proposed underlying mechanism of our dual-engineered CTP/Mir-EVs in promotion of cartilage repair was concluded and presented as Scheme 1.

4.1. Cell preparation and identification

hUCMSCs were bought from Oricell™ (Cyagen Biosciences, China) and cultured in α-MEM medium (Hyclone, USA), supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 100 units/mL penicillin/streptomycin (1% P/S; Servicebio, China) in an incubator with 5% CO₂ at 37 °C. Passage 4 hUCMSCs were seeded in six-well plates switched to osteogenic, adipogenic, or chondrogenic differentiation medium (Oricell™, Cyagen Biosciences) for 3 weeks. Alcian Blue, Oil Red O, and Alizarin Red S staining were performed according to standard protocols using Alcian Blue 8GX (Sigma), Oil Red O (Sigma), and Alizarin Red S (Sigma), respectively, to confirm their multiple differentiation potential. Surface protein markers of hUCMSCs were detected using flow cytometry with monoclonal antibodies (ThermoFisher, USA).

For extracting chondrocytes, the articular cartilage of Sprague-Dawley rats (male, 5 days old) was processed as previously described [56]. In brief, the cartilage was diced and digested with 0.2% type II collagenase (Sigma, USA) in DMEM/F12 (Invitrogen, USA) solution at 37 °C for 6 h. Then, the cell suspension was filtered, and chondrocytes were obtained. The chondrocytes were cultured with DMEM/F12 medium supplemented with 10% FBS and 1% P/S which was exchanged every other day. The chondrocytes were not passed over three generations to prevent phenotypic loss and dedifferentiation.

4.2. Isolation and characterization of hUC-EVs

Cells were washed twice with PBS after reaching 80% confluency, and then cultured with medium supplemented with 10% exosome-depleted FBS (SBI, USA) and 1% P/S. We collected conditioned medium from hUCMSCs after two days to isolate EVs. EVs were extracted by ultracentrifugation, as previously described [57]. Briefly, the conditioned medium was centrifuged at 300 g for 10 min, 2000 g for 20 min and then at 10000 g for 30 min at 4 °C, which allowed the remaining cells and debris to be clearly removed. The supernatant was then ultracentrifuged at 100,000 g for 70 min twice at 4 °C, the pellet (EVs) was resuspended in 50 μL PBS. We stored hUC-EVs at −80 °C until further use. We used a NanoSight LM10 instrument (Malvern, UK) to measure EV concentration and size distribution. To determine the morphology of hUC-EVs, transmission electron microscopy (TEM) was used (Hitachi H7500, Japan). The surface markers of hUC-EVs were analyzed by western blotting (antibodies purchased from Proteintech, China).

4.3. Lentivirus construction and transfection

The CTP-(EGFP-)Lamp2b expressing lentiviral vector was custom designed on VectorBuilder (Chicago, IL, USA) and obtained from Yunzhou Co, LTD (Guangzhou, China). hUCMSCs were seeded in six-well culture plates at a density of 3 × 10⁵ cells per well and allowed to adhere for 24 h in a humidified incubator with 5% CO₂ before transduction. Following the manufacturer's instructions, cells were transduced with 1 mL of media containing the appropriate virus at a MOI of 10. We generated stable over-expression hUCMSC lines after a 3-day antibiotic selection with puromycin. A fluorescence microscope was used to visualize and count EGFP-positive cells during selection to confirm transduction efficiency. To collect engineered CTP-EVs, stable cell lines were seeded in a 10 cm dish at a density of 5 × 10⁶.

4.4. miRNA loading by electroporation

An electroporation buffer containing 1.15 mM potassium phosphate, pH = 7.2, 25 mM potassium chloride, and 21% Optiprep was used to mix 0.5 μmol miR-223 mimics with 10 μg of total EVs. Using a Gene Pulser II system (BioRad Laboratories, CA), the mixture was electroporated in a 4 mm electroporation cuvette for 30 min at 350 V and 150 μF, and then incubated at 37 °C for 30 min to assure the membrane recovery of the EVs. Ultracentrifugation at 120,000×g and 4 °C for 70 min removed free unloaded miR-223, and the final pellet (EVs) was resuspended in PBS and stored at −80 °C. Using a fluorimeter that excited at 532 nm and emitted at 580 nm, the loading efficiency was calculated as follows: miRNA loading efficiency = (encapsulated miRNA/initial input of miRNA) *100%.

4.5. Chondrocyte treatment

To estimate the effect of EVs on chondrocyte viability and metabolism, chondrocytes were treated with the various EVs (5 × 10⁸ particles/ml) in the presence or absence of IL-1β (10 ng/mL) for 24 h, in which IL-1β was used to simulate the microenvironment of OA in vitro [44].

4.6. Cell viability test

CCK-8 (Dojindo, Kumamoto, Japan) was used to evaluate chondrocyte viability. We cultured the chondrocytes for 12 h in a 96-well plate (2000 per well). A complete medium containing different EVs was added to each well. Following further incubation for 24 h, each well was washed three times with complete medium. Each well was then added 100 μL of complete medium containing 10% CCK-8 reagent and the absorbance at 490 nm was measured by a microplate reader (Bio-Rad, US) after 4 h of incubation.

4.7. Western blot analysis

A BCA protein assay kit (Beyotime, China) was used to measure protein concentrations in cell lysates harvested using RIPA (Beyotime, China). The Western blot analysis was performed as described previously [58]. Anti-collagen II (Abcam, ab34712), anti-Sox9 (Abcam, ab185966), anti-MMP13 (Abcam, ab39012), anti-NLRP3 (Abclonal, A5652), anti-caspase-1 (Abclonal, A0964), anti-GSDMD-N (Abclonal, A20197) were used as primary antibodies.

4.8. Quantitative RT-PCR analysis

TRIzol Reagent (Invitrogen, America) was used to extract total RNA for mRNA detection. PrimeScript™ RT reagent Kit with gDNA Eraser (Takara, Japan) was used to generate complementary DNA templates. MiRNAs were isolated using the miRNA Isolation Kit (BioFlux, Japan) for detection. RT-PCR was performed using SYBR Premix Ex Taq™ II Kit (Takara, Japan) on Mx3000 P system (Stratagene, America), with each sample prepared in triplicate as recommended by the manufacturer. An All-in-One™ miRNA Fist-Strand cDNA Synthesis Kit and an All-in-One™ miRNA qPCR Kit (GeneCopoeia, America) were used for RT-PCR. RNA expression levels were calculated using the 2^−ΔΔCT method. Table S1 lists the primers used for RT-PCR.

4.9. Immunofluorescent staining in vitro

To evaluate the expression of collagen II, aggrecan, MMP13, NLRP3, and GSDMD, the chondrocytes were sequentially stained with anti-collagen II (Abcam, ab34712), anti-aggrecan (Abcam, ab3778), anti-MMP13 (Abcam, ab39012), anti-NLRP3 (Abclonal, A5652), and anti-GSDMD-N (Abclonal, A20197), then secondary antibodies, respectively. Nuclei were stained with DAPI (Thermo Fisher Scientific).

4.10. Induction of osteoarthritis and Intra‐articular injection of EVs

Study design and animal experiments were approved by the Ethics Committee of Shanghai Sixth People's Hospital (Animal Welfare Ethics No: DWLL2022-0594). The OA model was established using SD rats (8 weeks, 200–250 g) purchased from Shanghai Sixth People's Hospital. Sodium pentobarbital at 0.5% was injected intraperitoneally (0.9 mL/100 g) for general anesthesia. Through the infrapatellar ligament, 50 μL MIA (Sigma-Aldrich) at 20 mg/mL was injected intraarticularly to induce OA. Intra-articular injections of 50 μL PBS or 50 μL EV (10¹⁰ particles/ml) for 4 weeks (once a week) 1 week after the MIA injection were administered to rats.

4.11. Specimen preparation and histopathology analysis

Rats were sacrificed 5 weeks after MIA injection by anesthetic overdose. We fixed the knee joints in 4% paraformaldehyde, decalcified them in 20% formic acid, and embedded them in paraffin. At 50-mm intervals, serial sagittal sections were taken including the entire joint. Safranin O/Fast green, HE, immunohistochemical, and immunofluorescent stainings were performed on the sections. Matrix proteoglycan and overall joint morphology were assessed by Safranin O/Fast green-staining. Osteoarthritis Research Society International (OARSI) scoring system [59], along with the Mankin's score [60] were used to evaluate the grading of OA progress and cartilage degeneration of every sample in each group. Anti-collagen II antibody (Abcam, ab34712), anti-NLRP3 (Abclonal, A5652), and anti-GSDMD-N (Abclonal, A20197), were used as primary antibodies for IHC and IF, respectively.

4.12. ELISA analysis

The supernatant of chondrocytes in different groups was harvested for ELISA to evaluate the level of LDH (Cloud-Clone, USA). The cartilage tissues were also harvested and homogenized in PBS (1 mL). After centrifugation at 10,000×g for 15 min at 4 °C, the supernatants were used for detecting the expression of IL-1β (Enzolife, ADI-900-131 A), MMP2 (Raybiotech, ELR-MMP2-1), TNF-α (Enzolife, ADI-900-086 A), and PGE-2 (Cayman, 514,010) by the corresponding ELISA kits. Microplate readers (Bio-Rad, USA) were used to measure the absorbance at 450 nm.

4.13. Micro-computed tomography imaging

A high-resolution micro-CT scanner Skyscan 1176 (Bruker, Kontich, Belgium) was used to examine the knee joints of each rat.

4.14. MiRNA sequencing and analysis

Total RNA from cartilage was extracted for miRNA sequencing (miRNA-seq). Library preparation and miRNA-seq were performed by BGI (Shenzhen, China). To prepare the library, small RNAs ranging from 18 to 30 nucleotides (nt) were fractionated from the total cartilage RNA. A HiSeq 2500 platform (Illumina, San Diego, CA, USA) was used to sequence the PCR products.

4.15. Dual-luciferase reporter assay

A dual-luciferase miRNA target expression vector (GP-miRGLO, GenePharma, Shanghai, China) was cloned from the 3′UTR region of NLRP3 mRNA containing the miR-223 binding site, wild or mutant (AACUGAC mutated to GGAGUCU). MiR-223 mimics and mimic NC were co-transfected into HEK293T cells with wild and mutant vectors. A Dual-luciferase Reporter Assay (Promega, Madison, WI, USA) was performed to measure the Firefly and Renilla luciferase activities.

4.16. Statistical analysis

Numerical data are presented as mean ± standard deviation (SD). All statistical analyses in this study were performed using SPSS 26.0 software (IBM Corp., Armonk, NY, USA). In the figure legend, the sample size (n) for each statistical analysis is indicated. One-way ANOVA with Tukey post hoc analysis was used to compare the mean values among groups. The level of significance was set at α = 0.05 (two-tailed).