Circular RNA circZNF652 is overexpressed in osteoarthritis and positively regulates LPS-induced apoptosis of chondrocytes by upregulating PTEN

Abstract Circular RNA circZNF652 promotes LPS-induced inflammation, contributing to the development of osteoarthritis (OA), indicating the potential involvement of circZNF652 in OA. This study was carried to explore the involvement of circZNF652 in OA. RT-qPCR was performed to analyse the expression of circZNF652 and PTEN mRNA in synovial fluid samples from 60 OA patients and 60 healthy controls. Correlations between circZNF652 and PTEN mRNA were analysed by Pearson’s correlation coefficient. Overexpression and siRNA silencing of circZNF652 were achieved in chondrocytes, followed by performing RT-qPCR and Western blot to analyse the expression of PTEN. The role of circZNF652 and PTEN in regulating the apoptosis of chondrocytes induced by LPS was analysed by cell apoptosis assay. We found that circZNF652 was overexpressed in OA and positively correlated with PTEN, MMP13, and NF-KB mRNA. In chondrocytes, circZNF652 overexpression increased the expression of PTEN, MMP13, and NF-KB; circZNF652 siRNA silencing decreased the expression of PTEN, MMP13, and NF-KB. Moreover, circZNF652 and PTEN positively regulated the apoptosis of chondrocytes induced by LPS. PTEN overexpression reversed the inhibitory effects of circZNF652 siRNA silencing on cell apoptosis. Therefore, circZNF652 is overexpressed in OA and positively regulates LPS-induced apoptosis of chondrocytes by upregulating PTEN.


Introduction
Osteoarthritis (OA), as the primary cause of disability and chronic pain, is the main subtype of arthritis that affects about 13% females and 10% males worldwide [1]. OA occurs when the protective cartilage wears down during ageing or after injuries [2]. Theoretically, OA can cause irreversible damages to all joints, while the most affected sites are knees, hips, hands, and spine [2,3]. At present, OA treatment mainly focuses on the relief of symptoms, such as the use of acetaminophen to reduce moderate or mild pain, while currently available medications may not work in severe cases [4,5]. In addition, full recovery is still rare in clinical practices [6]. Therefore, novel treatment approaches are needed.
Ageing is the major risk factor for OA while ageing itself is not sufficient for the occurrence and development of OA. OA development and progression require multiple molecular signalling pathways, such as Notch, NF-jB, and PTEN [7,8]. Some key molecular players have been proven potential targets for targeted OA therapy in clinical practices to treat OA by regulating gene expression networks [9,10]. Despite lacking protein-coding capacity, circular RNAs (circRNAs) regulate gene expression to participate in human diseases, including OA, by regulating the expression of key genes [11][12][13]. Therefore, circRNAs are promising targets for targeted OA therapy. However, the function of cirRNAs in OA remains hardly known.
The serum concentration of lipopolysaccharide (LPS), an endotoxin secreted by gram-negative bacteria, is known to positively correlate with the pathogenesis process of OA [14]. Gram-negative bacteria infection could increase LPS levels in the circulation system [15]. In addition, increased LPS also negatively affects the physiological processes sensitive to inflammatory cytokines [16]. LPS administration to joints has been shown to induce synovitis in horses [17]. It has been reported that circZNF652 promotes LPS-induced inflammatory responses [18], thus contributing to OA development [19]. Our preliminary microarray analysis revealed altered circZNF652 expression in OA and its close correlation with PTEN, a key player in OA [8]. This study was performed to further explore the potential crosstalk between circZNF652 and PTEN in OA.

OA patients and controls
Sixty newly diagnosed OA patients (40 males and 20 females) and 60 healthy controls (40 males and 20 females) who were admitted to Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology from May 2018 to May 2020 were enrolled in the study. All patients and controls provided informed consent. The age of OA patients was 41-60 years old with a median of 51, and that of the healthy controls was 41-60 years old with a median of 52. Patients were excluded if they (1) had initial therapy within 3 months before admission, recurrent OA, and other severe clinical disorders. Among the 60 OA patients, hip was affected in 31 cases, and knee was affected in 29 cases. Healthy controls had normal physiological functions after systemic physiological exams. The study was approved by the Ethics Committee of our hospital.

Synovial fluid and chondrocytes
Synovial fluid (2 ml) was extracted using a syringe from the affected sites of patients prior to therapy and from the corresponding sites of healthy controls (31 cases of hip and 29 cases of knee) and stored in liquid nitrogen prior to the subsequent assays.
Primary chondrocytes from an adult with OA were purchased from Sigma-Aldrich (Cat# 402OA-05A, USA) and included in this study as the cell model of OA. Cell culture was performed following the manufacturer's instructions. For LPS treatment, cells were cultured in media supplemented with 0, 2, 4, 6, 8, and 10 mg/ml LPS for 48 h prior to the subsequent experiments.

Enzyme-linked immunosorbent assay (ELISA)
The LPS-binding protein (LPB) level in synovial fluid was determined using ELISA with a Human LBP ELISA kit (ab279407) following the manufacturer's instructions. The absorbance at 450 nm of each well was measured using a microplate reader (Thermo, Massachusetts, USA). All samples were measured in triplicate. The inter-and intra-assay coefficients were derived from six and eight samples, respectively.

Vectors, siRNAs, and transfections
The expression vectors of circZNF652 and PTEN were constructed with pcDNA3.1 vector (Invitrogen) and pcDNA3.1(þ) CircRNA Mini Vector (Addgene) as the backbone, respectively. SiRNA negative control (NC) and circZNF652 siRNA were purchased from Invitrogen. Expression vector (1 lg) or NC miRNA (40 nM) and their respective controls were transfected into chondrocytes (10 8 ) using Lipofectamine 2000 (Invitrogen). Cells were further cultured in fresh media for 48 h prior to the subsequent assays. Untransfected cells were used as the control (C) in all transfections.

RNA extraction
Isolation of RNA from synovial fluid and chondrocytes was performed using RNAzol (Sigma-Aldrich) followed by genomic DNA removal with DNase I digestion at 37 C for 2 h. RNA integrity was analysed by 5% urea-PAGE gel electrophoresis.

RT-qPCR
RNA purity was analysed by OD260nm/280 nm ratios. RNA samples with an OD 260/280 ratio close to 2.0 (pure RNA) were subjected to reverse transcriptions using the SS-IV-RT system (Invitrogen). SYBRV R Green Quantitative RT-qPCR Kit (Sigma-Aldrich) was used to perform all qPCRs with GAPDH as an internal control to determine the expression of circZNF652, MMP-13, NF-KB, and PTEN mRNA. Three technical replicates were included in each reaction. Ct value normalizations were performed using the 2 -DDCt method.

Cell apoptosis analysis
Chondrocytes with transfections were subjected to cell apoptosis assay. Chondrocytes (20,000 cells per well) were cultured in 6-well plates in media containing 10 mg/ml LPS for 48 h at 37 C. Following 0.25% trypsin digestion, cells were stained with propidium iodide (PI, Dojindo, Japan) and Annexin V-FITC (Dojindo, Japan) followed by flow cytometry to analyse cell apoptosis.

Statistical analysis
Expression levels of circZNF652 and PTEN mRNA in synovial fluid samples from both OA patients (n ¼ 60) and healthy controls (n ¼ 60) were expressed as the average values of three technical replicates and analysed using unpaired t test. Data of multiple cell transfection groups and LPS treatment groups were expressed as mean ± SD values of three biological replicates and compared using ANOVA Tukey's test. Correlations were analysed by Pearson's correlation coefficient. p < .05 was deemed statistically significant.

CircZNF652 and PTEN mRNA were overexpressed in clinical OA samples and closely correlated in vivo
CircZNF652 and PTEN mRNA expression in synovial fluid samples from both OA patients (n ¼ 60) and healthy controls (n ¼ 60) was analysed by RT-qPCR. Unpaired t-test analysis revealed that CircZNF652 (Figure 1(A)) and PTEN mRNA (Figure 1(B)) were significantly overexpressed in OA samples compared to the control samples (p < .01). Pearson's correlation coefficient was performed to analyse the correlations between CircZNF652 and PTEN mRNA across both OA and control samples. It was observed that CircZNF652 and PTEN mRNA were positively and significantly correlated with each other across OA samples ( Figure  1(C)) but not control samples (Figure 1(D)), suggesting the existence of the specific crosstalk between them in OA.

LPS treatment increased the expression of both circZNF652 and PTEN mRNA in chondrocytes in a dosedependent manner in vitro
To explore the correlation between circZNF652 expression and LPS in OA patients, circZNF652 level and LPB in synovial fluid samples from OA patients were measured using qRT-PCR and ELISA assays, respectively. Pearson's correlation showed that circZNF652 and LPB levels in synovial fluid samples from OA patients were significantly correlated (Figure 2(A)). Moreover, circZNF652 and PTEN mRNA levels in chondrocytes treated with 0, 2, 4, 6, 8, and 10 mg/ml LPS for 48 h were determined by RT-qPCR. It was observed that circZNF652 (Figure 2(A)) and PTEN mRNA levels (Figure 2(B)) were increased by LPS treatment in a dosedependent manner. Therefore, circZNF652 and PTEN mRNA might participate in OA through the LPS-dependent pathway.

CircZNF652 positively regulates O a progress in chondrocytes in vitro
To explore the crosstalk between circZNF652 and PTEN, chondrocytes were transfected with either circZNF652 or PTEN expression vector or circZNF652 siRNA, followed by the confirmation of transfections by RT-qPCR ( Figure 3(A), p < .05). CircZNF652 overexpression increased PTEN expression (Figure 3(B), p < .05), and circZNF652 siRNA silencing decreased PTEN expression (Figure 3(C), p < .05). In contrast, PTEN overexpression failed to significantly alter circZNF652 expression (Figure 3(D)). Furthermore, circZNF652 overexpression or silencing also affected the

CircZNF652 positively regulated the apoptosis of chondrocytes induced by LPS through PTEN in vitro
Chondrocytes were cultured in media containing 10 mg/ml LPS for 48 h at 37 C, followed by cell apoptosis assays. It was observed that circZNF652 and PTEN overexpression increased cell apoptosis, while circZNF652 siRNA silencing decreased cell apoptosis. Moreover, PTEN overexpression reversed the inhibitory effects of circZNF652 siRNA silencing on cell apoptosis (Figure 4, p < .05).

Discussion
The differential expression of circZNF652 in OA and its role in regulating cell apoptosis were explored in this study. We found that circZNF652 was overexpressed in OA and could positively regulate PTEN to promote the apoptosis of chondrocytes induced by LPS.
OA is essentially a type of inflammatory disease in which LPS-induced inflammation plays an important role [20]. In a recent study, Liu et al. reported that LPS treatment increased circZNF652 expression in WI-38 cells, fibroblasts derived from lung tissue, and circZNF652 knockdown suppressed inflammatory damages caused by LPS treatment [18]. Consistently, our study showed that LPS treatment increased circZNF652 expression in chondrocytes derived from an adult with OA. Therefore, LPS might induce circZNF652 expression in different cells.
In this study, we showed that circZNF652 was significantly overexpressed in OA. Chondrocytes, which are the only mature cells observed in healthy cartilage, produce collagen and proteoglycans to maintain the cartilaginous matrix [19]. In OA, increased chondrocyte apoptosis contributes to the development of OA [21]. In addition, cell apoptosis analysis showed that circZNF652 positively regulated the apoptosis of chondrocytes induced by LPS treatment. Moreover, circZNF652 overexpression enhanced the expression of MMP13 and NF-KB, both of which contribute to the change of extracellular matrix and OA development [22]. Therefore, circZNF652 might promote OA by increasing chondrocyte apoptosis, and circZNF652 silencing might serve as a potential therapeutic target for OA.  The main role of PTEN is to suppress PI3K/Akt pathway, which is the main cell survival pathway, to induce cell apoptosis [23]. In OA, PTEN was overexpressed to promote disease development by increasing cell apoptosis [8].
Consistently, our study revealed the upregulation of PTEN in OA and its enhancing effects on the apoptosis of chondrocytes induced by LPS. It has been well established that the main role of circRNAs is to regulate gene expression at transcriptional and/or translational levels [11]. In this study, we showed that circZNF652 overexpression increased PTEN expression at both mRNA and protein levels. Therefore, circZNF652 overexpression might promote PTEN transcription. However, circZNF652 and PTEN were only positively correlated only across OA samples, not control samples. Therefore, certain pathological factors may mediate the interaction between them.

Conclusion
CircZNF652 is overexpressed in OA and might promote the apoptosis of chondrocytes induced by LPS by positively regulating PTEN expression.

Ethical approval
Informed consent was obtained from all individual participants included in the study. All procedures were approved by the Ethics Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology and operated in compliance with the standards set out in the Announcement of Helsinki and the Laboratory Guidelines of Research in China.

Disclosure statement
No potential conflict of interest was reported by the author(s). We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.

Author contributions
Xuefeng Yuan, Chengla Yi: study concepts, literature research, clinical studies, data analysis, experimental studies, manuscript writing and review; Yingchi Zhang: study design, literature research, experimental studies, and manuscript editing; Cong Cai: definition of intellectual content, clinical studies, data acquisition, and statistical analysis; Chaoxu Liu: data acquisition, manuscript preparation and data analysis; Jie Xie: data acquisition and statistical analysis.

Data availability statement
The data that support the findings of this study are available on request from the corresponding author, Chengla Yi, Department of Traumatic Surgery, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, No. 1095 Jiefang Avenue, Wuhan City, Hubei Province, 430030, P. R. China. Email: ChenglaYiSurgery@163.com. Tel: 18062607945.
The data are not publicly available due to their containing information that could compromise the privacy of research participants.