Cartilage tissues from the knee joints of patients with OA [7 males and 3 females; age range, 54-66 years; body mass index (BMI), 19.7-23.5 kg/m²] who had undergone total knee arthroplasty, and healthy cartilages from patients with femoral neck fracture (5 males and 5 females; age range, 50-52 years; BMI, 19.9-23.2) from January 2021 to January 2022 were obtained and used for PCR analysis. The study was approved by the Ethics Committee of Zhongda Hospital Southeast University and the signed informed consent was obtained from all subjects (no. ZDL22-21; date: April 23, 2022).

As previously described (9), cartilage tissues from the knee joints of OA patients were collected, and then cut into small sections. After washing five times with PBS (Beyotime Institute of Biotechnology), the specimens were digested in 0.25% trypsin-EDTA (Beyotime Institute of Biotechnology) solution for 30 min and supplemented with Dulbecco's modified Eagle's medium (DMEM; Thermo Fisher Scientific, Inc.) containing collagenase type II for 2 h at 37˚C. The released chondrocytes were seeded in 25 cm² cell flasks. The cells were passaged at a ratio of 1:3, when they reached 80-90% confluency. For stimulation with IL-1β, the isolated chondrocytes were maintained in DMEM (Thermo Fisher Scientific, Inc.) at 37˚C, 5% CO₂ and incubated with IL-1β (10 ng/ml; Abcam) for 24 h.

Overexpression vector for circ0072568 (oe-circ0072568) was provided by Guangzhou BersinBio Biotechnology Co., Ltd.). miR-382-5p mimics and negative control (miR-NC) were obtained from RiboBio (Guangzhou RiboBio Co., Ltd.) and MyBioSource, respectively, without sequence information. miR-382-5p inhibitor (inhib-miR-382-5p) and negative control (inhib-NC) were obtained from Shanghai GenePharma Co., Ltd. Circ0072568 siRNA (si-circ0072568) and its corresponding control (si-nc), as well as TOP1 siRNA (si-TOP1) and its corresponding control (si-NC), were designed and synthesized by Shanghai GenePharma Co., Ltd. Primary chondrocytes, isolated from normal articular cartilage tissues, were cultured in 96-well plates at 37˚C for 24 h and then transfected with 200 pmol/l oe-circ0072568 [or 200 pmol/l negative control (oe-NC)], 50 nM miR-382-5p mimics (or 50 nM miR-NC), 100 nM inhib-miR-382-5p (or 100 nM inhib-NC), or 200 pmol/l si-TOP1 (or 200 pmol/l si-NC) using Lipofectamine 3000^® (Invitrogen; Thermo Fisher Scientific, Inc.). The aforementioned oligonucleotides or plasmids were transfected into chondrocytes using Lipofectamine™ 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's instructions. Briefly, chondrocytes were cultured for 24 h, before the constructs were transfected into the cells using 500 µl Opti-MEM containing 20 µl Lipofectamine 2000 (1 mg/ml) in a humidified atmosphere with 5% CO₂ at 37˚C. After incubating for 6 h, culture medium was replaced by fresh medium. Subsequently, at 48 h following transfection, the cells were harvested for use in subsequent experiments. The sequences of the siRNAs, mimics and inhibitors used were as follows: miR-382-5p mimics forward, 5'-GAAGUUGUUCGUGGUGGAUUGG-3' and reverse, 5'-AAUCCACCACGAACAACUUCUU-3'; miR-NC forward, 5'-CTCGCTTCGGCAGCACA-3' and reverse, 5'-AACGCTTCACGAATTTGCGT-3'; miR-382-5p inhibitor, 5'-CGAAUCCACCACGAACAACUUC-3'; inhib-NC, 5'-CAGUACUUUUGUGUAGUACAA-3'; si-TOP1, 5'-GCAUAAAGACAAACAUAAAGA-3'; si-NC, 5'-UUCUCCGAACGUGUCACGUTT-3'. si-circ0072568, 5'-AACAGUUUUGAUGUGGACAAU-3'; si-nc, 5'-UUCUCCGAACGUGUCACGU-3'.

Total RNA extracted from tissues and cells was incubated in the presence or absence of RNase R (3 U/µg, Epicentre) at 37˚C for 20 min. After TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and the miRNA isolation kit (Thermo Fisher Scientific, Inc.) to extract the total RNA, reverse transcription was performed using a miRNA cDNA kit (Invitrogen; Thermo Fisher Scientific, Inc.), TaqMan™ MicroRNA Assays (Applied Biosystems; Thermo Fisher Scientific, Inc.) and PrimeScript RT Master Mix (Takara Bio, Inc.). The PCR thermal cycling parameters were as follows: 95˚C (5 min), 45 cycles of 94˚C (30 sec) and 60˚C (30 sec). Gene expression was quantified using the 2^-ΔΔCq method (10), using GAPDH and U6 as internal genes for circ0072568 and TOP1, as well as miR-382-5p and miR-545-3p, respectively. The primer sequences used are listed in Table I.

Chondrocytes (1x10⁴ cells/well) were cultured in the 96-well plates. 10 µl MTT stock solution (Beijing Solarbio Science & Technology Co., Ltd.) was added to the culture medium according to the manufacturer's instructions, and the plate was incubated for 4 h at 37˚C. Following the addition of dimethyl sulfoxide (DMSO; Beijing Solarbio Science & Technology Co., Ltd.), the absorbance of the cells was read at 570 nm using a Multiska FC microplate reader (Thermo Fisher Scientific, Inc.).

The apoptotic rate of the chondrocytes (2x10⁵ cells/well) at 80% confluency was analyzed using an apoptosis detection kit [Annexin V-FITC/propidium iodide (PI); Beyotime Institute of Biotechnology] and a BD FACS flow cytometer (BD Biosciences).

The secretion of pro-inflammation factors was detected using corresponding IL-6 (cat. no. PI325), TNF-α (cat. no. PT518) and monocyte chemoattractant protein-1 (MCP-1; cat. no. PC130) ELISA kits (Beyotime Institute of Biotechnology).

After using the bicinchoninic acid (BCA) method (Beyotime Institute of Biotechnology) to quantify total proteins, which were extracted from the cell using radio immunoprecipitation assay buffer (cat. no. BP0013B; Beyotime Institute of Biotechnology), the proteins were separated by 10% SDS-PAGE, transferred onto PVDF membranes. Thereafter, the membranes were blocked with 5% skimmed milk at room temperature for 2 h and then incubated with primary antibodies at 4˚C overnight. Subsequently, the membranes were incubated with secondary antibody IgG (1:2,000; cat. no. ab205718; Abcam) at room temperature for 2 h. The primary antibodies used were as follows: Anti-collagen II (1:1,000; cat. no. ab188570; Abcam), anti-MMP-13 (1:1,000; cat. no. ab51072; Abcam), anti-TOP1 (1:1,000; cat. no. ab131166; Abcam) and anti-GAPDH (1:1,000; cat. no. ab8245; Abcam). Blots were developed using an enhanced chemiluminescence system (Beyotime Institute of Biotechnology). The density of each protein blot was compared with that of GAPDH using ImageJ software (version 1.46r; National Institutes of Health) and is presented as a ratio to the endogenous control.

The PARIS kit (cat. no. AM1921; Invitrogen; Thermo Fisher Scientific, Inc.) was used for nuclear-cytoplasmic fractionation. Briefly, the nuclei were lysed using cell disruption buffer and the supernatant of nuclear and mitochondrial DNA extracted form chondrocytes were mixed with a 2X lysis binding solution, followed by eluting the cytoplasm and nucleus RNA using the corresponding kit solutions.

The downstream target miRNAs of circ0072568 were predicted using the online website StarBase v2.0 (http://starbase.sysu.edu.cn/index.php) and circBank (http://www.circbank.cn/searchCirc.html). Direct targets of miR-382-5p were predicted using the online databases StarBase, miRTarBase (http://mirtarbase.mbc.nctu.edu.tw/index.html), TargetScan (https://www.targetscan.org/vert_80/) and miRDB (https://mirdb.org/). Primary chondrocytes (1x10⁴) were cultured to 70% confluency in 24-well plates. Either wild-type or mutant circ0072568 (circ0072568-wt or circ0072568-mut) and TOP1 (TOP1-wt or TOP1-mut) fragments were inserted into the pGL3-firefly luciferase vectors (Shanghai Genechem Co. Ltd.). Primary chondrocytes were cultured for 24 h before being co-transfected with 600 ng circ0072568-wt or 600 ng TOP1 3'UTR-wt luciferase reporter gene plasmid and 20 nmol miR-382-5p mimics or 20 nmol miR-NC using Lipofectamine 3000 reagent for 48 h at 37˚C, and the dual-luciferase System (Promega Corporation) was used to determine the luciferase activity. The results were normalized to Renilla lucifierase activity.

The Ago-RIP test (MilliporeSigma) was performed using the the Magna RIP RNA-Binding Protein Immunoprecipitation kit and RNeasy MinElute Cleanup kit (Qiagen GmbH). The cells were lysed in RIP buffer (Beyotime Institute of Biotechnology) containing a protease inhibitor cocktail and RNase inhibitors, and incubated at 4˚C with anti-Ago2 (cat. no. ab186733; 1:50; Abcam) or anti-IgG antibodies (cat. no. PP64B; 1:20; EMD Millipore) overnight. RNA was quantified by qRT-PCR.

SPSS v22.0 software was used for statistical analyses. An unpaired two-tailed Student's t-test and one-way ANOVA (for ≥3 groups) followed by post hoc (Tukey's or Dunnett's) tests was used. Data are presented as the mean ± SD of three repeats. The correlation between circ0072568 and miR-382-5p expression in OA tissues was assessed using Pearson's correlation analysis. A value of P<0.05 was considered to indicate a statistically significant difference.

As shown in Fig. 1A, circ0072568 expression was decreased in OA-affected cartilage, as well as in IL-1β-stimulated primary human chondrocytes (Fig. 1B). As demonstrated by the results of RT-qPCR, transfection with circ0072568 overexpression vector (oe-circ0072568) successfully increased the circ0072568 level in chondrocytes isolated from normal articular cartilage tissues (Fig. 1C). Additionally, the downregulation of circ0072568 induced by exposure to IL-1β for 24 h was restored by transfection with circ0072568 overexpression vector (Fig. 1D). The decrease in the viability of IL-1β-stimulated primary human chondrocytes was abolished by the overexpression of circ0072568 (Fig. 1E). The increased apoptosis in IL-1β-induced human primary chondrocytes was also abolished by the overexpression of circ0072568 (Fig. 1F). In addition, circ0072568 overexpression led to the decreased secretion of pro-inflammatory factors (Fig. 1G-I), and to decreased MMP-13 expression and enhanced collagen II expression (Fig. 1J and K).

ECM degradation by IL-1β-stimulated chondrocytes was assessed by detecting the expression of proteins related to ECM catabolism, including collagen II and MMP-13. Western blot analysis revealed that IL-1β stimulation decreased the expression of collagen II and increased the expression of MMP-13, indicative of the increased ECM degradation potential (Fig. 1J and K). By contrast, circ0072568 overexpression led to an increase in collagen II expression and a decrease in MMP-13 expression in IL-1β-stimulated chondrocytes. Taken together, these data suggest that circ0072568 inhibits IL-1β-induced inflammation and ECM degradation by chondrocytes.

The regulatory mechanism of circRNAs as miRNA sponges has been widely confirmed. The present study we identified the distribution of circ0072568 in chondrocytes by subcellular fractionation and found that circ0072568 was mainly enriched in the cytoplasm of chondrocytes (Fig. 2A). The analysis of public databases (StarBase v2.0 and circBank) identified miR-545-3p and miR-382-5p as target miRNAs of circ0072568 (Fig. 2B). However, the expression of miR-382-5p and miR-545-3p did not differ significantly between OA tissues and normal tissues, and the stimulation of human chondrocytes with IL-1β did not result in significant changes in their expression levels (both P>0.05; Fig. 2C and D). In addition, miR-382-5p decreased the relative luciferase activity of circ0072568-wt, but not that of circ0072568-mut vectors (Fig. 2E). Furthermore, the enrichment of circ0072568 in beads was found to conjugate to the Ago2 antibody compared with the IgG controls (Fig. 2F). oe-circ0072568 and its negative control (oe-NC) were then transfected into chondrocytes (Fig. S1), and the results revealed that the overexpression of circ0072568 in chondrocytes led to a decreased expression of miR-382-5p (Fig. 2G), abolishing the IL-1β-induced upregulation of miR-382-5p expression (Fig. 2H). Pearson's correlation analysis further confirmed the negative correlation between the expression of circ0072568 and miR-382-5p in the cartilage tissues of patients with OA (Fig. 2I).

TOP1 was found to be one of the target genes of miR-382-5p, as predicted using the StarBase, TargetScan and miRDB databases (Fig. 3A). The relative luciferase activity of TOP1-wt vectors was decreased in the chondrocytes overexpressing miR-382-5p (P<0.05; Fig. 3B). The overexpression of miR-382-5p in human chondrocytes transfected with miR-382-5p mimic was abolished following transfection of the cells with circ0072568 overexpression vector (all P<0.05; Fig. 3C). While the transfection of chondrocytes with miR-382-5p mimic decreased the expression of TOP1 in chondrocytes, this effect was reversed by co-transfection with circ0072568 overexpression vector (all P<0.05; Fig. 3D and E). In addition, transfection of the chondrocytes with miR-382-5p inhibitor increased TOP1 expression in chondrocytes, whereas this effect was reversed by co-transfection with si-circ0007256 (all P<0.05; Fig. 3F and G). IL-1β stimulation markedly decreased the expression of TOP1 in chondrocytes (all P<0.05), and this effect was reversed by the overexpression of circ0072568 (all P<0.05). Additionally, circ0072568 regulated TOP1 expression via miR-382-5p in chondrocytes (Fig. 3H and I).

miR-382-5p overexpression or TOP1 knockdown abolished the anti-inflammatory effects of circ0072568 overexpression on IL-1β-stimulated chondrocytes. miR-382-5p overexpression or TOP1 knockdown abolished the suppressive effects of circ0072568 overexpression on the levels of pro-inflammatory factors (Fig. 4A-E), and reversed the inhibition of MMP-13 expression and the increase in collagen II expression, which was induced by circ0072568 overexpression (Fig. 4F-H).