2.1 Patients and controls

No public gene expression database contains the expression of Circ_HECW2 in OA patients, possibly because that Circ_HECW2 is a recently identified circRNA. To analyze the differential expression of Circ_HECW2 in OA, this study enrolled a total of 64 OA patients (60.5 ± 6.3 years, 38 cases of knee OA and 26 cases of hip OA, 40 females and 24 males) and 64 healthy controls (60.4 ± 6.5 years, 40 females and 24 males) at Peking University Shenzhen Hospital between May 2018 and May 2020. To diagnose OA, OA-related bone changes and damages were revealed by X-ray. Joint aspiration was also performed on all OA patients to exclude other bone disorders. This study excluded OA patients with preadmission treatment and recurrent OA. The Ethics Committee approved the study before the admission of patients. All patients signed the written informed consent.

2.2 Extraction of synovial fluid

All OA patients were subjected to joint aspiration to collect synovial fluid (1.0 ml, affected sites). Synovial fluid (1.0 ml) was also extracted from the corresponding site of healthy controls (38 cases of the knee and 26 cases of the hip) to serve as the control. All synovial fluid samples were stored in liquid nitrogen.

2.3 Chondrocytes (OA) and cell culture

Human OA-affected chondrocytes cells were from Sigma-Aldrich (402OA-05A) and cultured in Dulbecco's modified Eagle's medium (11330032; Gibco, Thermo Fisher Scientific) containing 10% fetal bovine serum (10270-106; Gibco) and 1% penicillin/streptomycin (15140148; Thermo Fisher Scientific) at 37°C under 5% CO₂. To induce inflammatory injury, chondrocytes were stimulated with LPS (916374, LPS; Sigma-Aldrich) at various concentrations (0, 1, 5, 10, and 20 μg/ml) for 48 h.

2.4 Cell transfections

Circ_HECW2 sequence was designed to insert into pcDNA3.1 plasmid to construct a Circ_HECW2 expression plasmid (pcDNA3.1-Circ_HECW2). Both pcDNA3.1-Circ_HECW2 and miR-93 mimics were purchased from Invitrogen. OA cells were seeded in six-well plates at a density of 1 × 10⁵ cells per well and incubated overnight to achieve monolayer confluence and transfected with indicated vectors or microRNAs (miRNAs) using Neon Transfection System (MPK5000; Thermo Fisher Scientific) according to the manufacturer's instruction. The cells were transfected for 48 h at 37°C under 5% CO₂ and then subjected to subsequent assay.

2.5 Preparation of RNA samples

Trizol reagent (15596026; Invitrogen) was used to isolate total RNA from synovial fluid and chondrocytes following the manufacturer's protocols. DNase I (18047019; Invitrogen) was added to all RNA samples and incubated at 37°C until an OD260/280 ratio close to 2.0 (indicating pure RNA) was reached. Polyacrylamide gel electrophoresis gel (5%) electrophoresis was performed to analyze RNA integrity. Subsequent assays were performed using high-quality RNA samples.

2.6 Reverse-transcription polymerase chain reaction analysis

Total RNA was used for reverse transcription (RT) to prepare complementazry DNA (cDNA) samples using SSRT-IV (18090010, Invitrogen) system. The synthesized cDNA was used as the template to evaluate Circ_HECW2 expression by quantitative reverse-transcription polymerase chain reaction (RT-qPCR) analysis using SYBR Green Master Mix (1725270; Bio-Rad) with 18S rRNA as the internal control. Poly(A) was added to mature miRNAs using all-in-one miRNA qRT-PCR Detection Kit (QP115; GeneCopoeia). The same kit was used to perform miRNA RTs and quantitative PCRs (qPCRs) to measure mature miR-93 expression levels. The PCR reaction was carried out with Applied Biosystems AB7500 Real-Time PCR system (Applied Biosystems) at the thermal cycle conditions of 95°C for 10 min and 40 cycles of 95°C for 15 s and 60°C for 1 min. Relative gene expression levels were calculated using the 2-∆∆Ct method. The primers used for PCR were Circ_HECW2 forward 5′-CCCACCACTTTGAACGCTAC-3′ and reverse 5′-GGCTGTCAATGCGTGCCT-3′ and miR-93 forward 5′-AGGCCCAAAGT GCTGTTCGT-3′ and reverse 5′-GTGCAGGGTCCGAGGT-3′.

2.7 Methylation-specific PCR (MSP)

Genomic DNA was extracted from chondrocytes using the conventional method. EZ DNA Methylation-Gold Kit (D5006; ZYMO Research) was used to convert DNA samples, followed by performing both methylation-specific PCRs (MSPs) and routine PCRs to determine the expression of the methylated and unmethylated miR-93 gene, respectively, using Tag 2X master mix (M0270L, NEB). The primers used in MSP were methylation forward: 5′-CTGGGGGCTCCAAAGTGCTGTTCG-3′ and 5′-CCGGGGGCTCGGGAAGTG CTAG-3′ and unmethylation forward 5′-CTGGGGGCTCCAAAGTGCTGTTCG-3′ and reverse 5′-CCGGGGGCTCGGGAAGTGCTAG-3′.

2.8 Analysis of cell apoptosis

At 48 h after transfections, chondrocytes cells were seeded into 96-well plates with 3000 cells in 0.1 ml medium per well. Cells were treated with 10 μg/ml LPS for 48 h in culture media. Then cells were washed with fresh medium, digested by trypsin, and collected by centrifugation. Cell apoptosis was detected using Annexin V-FITC Apoptosis Detection Kit (15342-54; nacalai tesque). In brief, cells were resuspended in 195 μl Annexin V-fluorescein isothiocyanate (FITC) binding solution, followed by the addition of 5 μl Annexin V-FITC and 10 μl propidium iodide. The mixture was incubated at room temperature in the dark for 30 min. After that, cell apoptosis was analyzed using flow cytometry.

2.9 Statistical analysis

Quantitative data were presented as the mean ± SD of experimental replicates and processed using GraphPad Prism 6.0. Two and multiple independent groups were compared by unpaired t test and analysis of variance Tukey's test, respectively. p < .05 was considered statistically significant.

3.1 OA patients exhibited altered expression of Circ_HECW2 and miR-93

To evaluate the expression level of Circ_HECW2 and miR-93 in OA patient, synovial fluid samples were collected from 64 OA patients and 64 healthy controls. The Total RNAs from synovial fluid samples were extracted, and qPCRs were performed. The results showed that the OA group exhibited increased Circ_HECW2 expression (Figure 1A, p < .05) and decreased miR-93 expression (Figure 1B, p < .05) compared with the healthy control group.

3.2 LPS treatment altered the expression of Circ_HECW2 and miR-93

A previous study has shown that LPS-induced inflammatory injury of chondrocytes can be served as an in vitro OA model. To investigate whether Circ_HECW2 and miR-93 are involved in LPS-induced chondrocyte injury, we assessed changes in the expression of Circ_HECW2 and miR-93 in response to LPS treatment in chondrocytes in vitro. The RT-qPCR analysis demonstrated that Circ_HECW2 expression was significantly upregulated by LPS exposure in chondrocytes (Figure 2A, p < .05). Moreover, LPS exposure also resulted in markedly downregulated miR-93 expression (Figure 2B, p < .05). These data indicated that Circ_HECW2 and miR-93 expression in OA are altered by LPS, which plays a critical role in OA.

3.3 Circ_HECW2 and miR-93 are closed correlated across OA samples

To analyze the correlation between Circ_HECW2 and miR-93 in OA samples and healthy controls, the correlation analysis was performed. The results showed that Circ_HECW2 and miR-93 were closed correlated across OA samples (Figure 3A), but not across healthy control samples (Figure 3B). This data suggested that the crosstalk between Circ_HECW2 and miR-93 is only significant in disease conditions.

3.4 Circ_HECW2 overexpression decreases miR-93 expression through methylation in chondrocytes

To further investigate the crosstalk between Circ_HECW2 and miR-93, chondrocytes were transfected with Circ_HECW2 expression vector or miR-93 mimic to overexpress Circ_HECW2 and miR-93, respectively. The overexpression was monitored and confirmed from 24 to 96 h after transfection (Figure 4A, p < .05). In addition, Circ_HECW2 overexpression significantly suppressed miR-93 expression (Figure 4B, p < .05). However, miR-93 overexpression did not affect Circ_HECW2 expression (Figure 4C, p < .05), suggesting that it was the increased Circ_HECW2 level that led to the low miR-93 level in OA patients and LPS-induced chondrocytes.

Because DNA methylation is the most common epigenetic modification to regulate gene expression level, we performed MSP analysis to evaluate the methylation level followed by Circ_HECW2 overexpression. The results demonstrated that miR-93 methylation was upregulated upon Circ_HECW2 overexpression (Figure 4D). Together, these data indicated that Circ_HECW2 overexpression decreased miR-93 expression via methylation in chondrocytes.

3.5 Circ_HECW2 overexpression increases LPS-induced chondrocyte apoptosis via miR-93

To explore the biological function of Circ_HECW2 in regulating LPS-induced chondrocyte injury, we performed overexpression experiments by transfecting Circ_HECW2 expression vector and miR-93 mimic into LPS-treated chondrocytes and then analyzed their effects on cell apoptosis. The results showed that Circ_HECW2 overexpression significantly increased apoptosis of LPS-treated chondrocytes, while miR-93 overexpression reversed the effect of Circ_HECW2 on cell apoptosis (Figure 5, p < .05). This data suggested that Circ_HECW2 overexpression increased apoptosis of LPS-treated chondrocytes via inhibiting miR-93 expression.