Abstract

Objective. Provide a reference to elucidate the mechanism of circRNAs regulating osteoarthritis (OA) and the clinical treatment. Methods. Herein, articles about circRNAs (hsa-circ) and osteoarthritis in the recent 5 years have been reviewed and the differential expression and regulatory effect of circRNAs in OA deduced. Based on these conclusions and Protein-Protein Interaction (PPI), Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of the acquired circRNAs, the potential functions and interactions of circRNAs in OA and the involved signaling pathways are discussed. Results. A total of 33 studies meeting the inclusion criteria were included in this study, and 27 circRNAs were upregulated and 8 circRNAs were downregulated in OA. A total of 31 circRNAs were finally included in the PPI, GO, and KEGG analyses. From PPI, 12 map nodes and 7 map edges were interrelated. VWF had the biggest node and edge size. From GO, VWF showed a majority of the functions. From KEGG, circRNAs are enriched in PI3K/AKT, human papillomavirus infection (HPI), and focal adhesion (FA) pathways, and VWF was involved in major pathways. Conclusion. We found that most articles about circRNAs regulating OA in the recent 5 years focused on the mechanism, especially the absorption effect of circ-miRNA as sponges in the recent 2 years, while most of the articles about their functions addressed ECM and PI3K, AKT, and mTOR signaling pathways. Future studies might focus on the functions of circRNAs, and circRNA VWF, with preferable functions, interactions, and involvement, can be used as a biological indicator to detect OA in clinical practice.

1. Introduction

Osteoarthritis (OA) is a common clinical disease that has a long process from early inflammation in the joint to the wear and tear of the cartilage layer and the formation of subchondral osteophytes, eventually leading to the failure of the joint to carry out daily movements and perform daily functions [1, 2]. In clinical practice, OA can only be relieved and improved but cannot be cured fully [3, 4]. The mechanism of OA has not yet been defined in existing studies; however, some studies have shown that circular RNAs (circRNAs) play a role in the occurrence and development of OA, but the functions and mechanism of circRNAs in OA were still not very clear [5, 6]. The present study reviewed the articles about circRNAs and OA in the recent 5 years to provide some reference to elucidate the mechanism of circRNAs in OA. Also, based on PPI, GO, and KEGG analysis of the acquired circRNAs, the potential functions of circRNAs in OA and the involved signaling pathways are also discussed in this article. This review can provide some reference for the fundamental research of the prevention and treatment of OA.

2. Material and Methods

2.1. Data Source

“Circular RNA (circRNA)” and “osteoarthritis (OA)” were used as keywords to search relevant articles from January 1, 2016, to 2021 in China National Knowledge Infrastructure (CNKI), PubMed, and Web of Science. There are no ethics committee approval and informed consent in this article.

2.2. Criteria
2.2.1. Inclusion Criteria

Inclusion criteria are as follows: (1) experimental articles with the keywords in the databases and (2) articles with the circRNA ID starting with hsa (human gene).

2.2.2. Exclusion Criteria

Exclusion criteria are as follows: (1) overviews in the databases, (2) repeated articles, (3) articles of poor quality, (4) articles with the circRNA ID starting with mmu (mouse gene), and (5) articles about rheumatoid arthritis (RA).

2.3. PPI, GO, and KEGG Analyses

Included circRNAs were retrieved from circBase (http://www.circbase.org/) to check the accuracy of information, and the circRNAs that had no relevant information were deleted.

The PPI network (https://string-db.org/) was mapped. Choose multiple proteins-gene symbol, and the minimum required interaction score was 0.400.

GO analysis (DAVID, https://david.ncifcrf.gov/summary.jsp) and KEGG analysis (KOBAS, http://kobas.cbi.pku.edu.cn/kobas3) were performed to discuss the potential functions and the participating signaling pathways (species, Homo sapiens; input type, gene symbol; ). As shown in Figure 1.

2.4. Data Processing and Analysis

The acquired data were analyzed and mapped using Cytoscape (3.7.2) and R language (R x64 4.0.2).

3. Results

3.1. Current Studies on the Correlation between circRNAs and OA

A total of 33 studies meeting the inclusion criteria were included in this study, and 35 circRNAs were sorted out by circRNA ID, gene symbol, regulation, miRNA, target gene/signal pathway, reference, and year (Table 1). As shown in Table 1, 81.82% of the articles about circRNAs in the recent 5 years discussed the circ-miRNA axis. As shown in Table 2, 27 circRNAs had upregulated expression and 8 had downregulated expression. As shown in Figure 2, most articles (61%) about circRNAs were published in 2020, and 15% of the articles were published in 2021.

3.2. PPI, GO, and KEGG Analyses of circRNAs

Repeated circRNAs (SERPINE2, VWF, EPS15, and UNK) and those having no information in circBase (hsa_circ_9119, hsa_circ_7, PSM3, and hsa_circ_100226) were excluded, and a total of 31 circRNAs were finally included in the PPI, GO, and KEGG analyses. The final result showed .

Figure 3 of PPI analysis shows a network of 30 circRNAs (RP11-909M7.3 not found in STRING); of these, 12 map nodes and 7 map edges were interrelated. VWF and DUSP5 had the biggest map node size (degree 2); IQGAP1-VWF-SERPINE2 and PLOD1-COL6A3 had the bigger map edge size (0.906, 0.928, and 0.913).

GO enrichment analysis usually covers molecular function (MF), cellular component (CC), and biological process (BP). The results of this study showed that the functions of 31 circRNAs were mainly focused on MF, including protein kinase activity and glycosaminoglycan binding; CC were proteinaceous ECM, platelet α granule, extracellular matrix (ECM), cellular exosome, and endoplasmic reticulum membrane; BP included peptidyl-serine phosphorylation, ECM organization, and cell adhesion (Figure 4). Among them, VWF showed a majority of the functions (6/10, Table 3).

KEGG signaling pathway analysis showed that circRNAs are enriched in PI3K/AKT, human papillomavirus infection (HPI), focal adhesion (FA), and other seven pathways (Figure 5), and VWF and COL6A3 were involved in 4/7 pathways (Table 4).

4. Discussion

4.1. Brief Information and Functions of circRNAs

circRNAs are a type of noncoding RNAs mainly found in the cytoplasm of mammalian cells. circRNAs consist of the 3- and 5-phosphodiester bonds covalently linked to form a circular structure, which is stable and resistant to RNA exonuclease-mediated degradation, and hence are termed as circRNAs. Three types of circRNAs, exon circRNAs (exon circular RNA (ecircRNAs)), intron circRNAs (intron circular RNA (ciRNAs)), and exon-intron circRNAs (exon-intron circular RNA (EIcircRNAs)) [4244], especially ciRNAs, are conserved across evolution and have a half-life of >48 h, which also confirms their high stability. In addition, circRNAs are highly stable and sensitive in body fluids and used for biochemical tests [4547].

Current studies showed that the functions of circRNAs are as follows. (1) They adsorb microRNAs (miRNAs), bind to miRNAs as sponges, affect the corresponding message RNAs (mRNAs), and eventually regulate the expression of target genes. (2) They regulate the activity of RNA-binding proteins (RBPs) and transport them or act as their scaffold to facilitate the formation of new complexes. Additionally, circRNAs can also interact with proteins, selectively cut or transcribe parent genes (binding RNA polymerases), and encode the proteins [4852], as shown in Figure 6.

4.2. Studies on the Mechanism of circRNAs in OA and circRNAs as Biological Indicators of OA

This study showed that the articles on circRNAs in OA in the recent 5 years mainly focused on the mechanism while they also discussed circRNAs as clinical, biological indicators of OA.

According to the statistical results of this study, 61% of the relevant articles in the recent 5 years were published in 2020 and 15% in 2021. The majority of these articles focused on the mechanism of circ-miRNA with respect to the absorption effect of circRNAs as sponges on miRNAs in OA. Kulcheski et al. [53] proposed that circRNAs are sponges of miRNAs and can serve as the novel type of biomarkers. circRNA 0092516 regulates chondrocyte differentiation and apoptosis via miRNA-337-3p/PTEN (phosphatase and tensin homolog), according to Huang et al. [8], while circRNA UBE2G1 regulates lipopolysaccharide- (LPS-) induced OA chondrocytes via miR-373/hypoxic inducible factor 1 alpha (HIF-1α), according to Chen et al. [9]. Wu et al. [12] demonstrated that lowly expressed circRNA HIPK3 regulates SRY-related high-mobility group box gene 8 (SOX-8), a critical marker of chondrocyte development as the sponge of miR-124, thus promoting the apoptosis of osteoarthritis chondrocytes. Ma et al. [13] found that circRNA VCAN promotes the apoptosis of OA chondrocytes by blocking the NF-κB signaling pathway. Wu et al. [16] showed that circRNA PDE4D protected OA by binding to miR-103a-3p and regulating the fibroblast growth factor 18 (FGF18), and Zhou et al. [17] found that circRNA ANKRD36 regulated Casz1 (miR-599 target gene) and prevented the apoptosis and inflammation of OA chondrocytes by targeting miR-599.

Additionally, some studies also discussed circRNAs as biological indicators to detect and evaluate OA. In the study by Wang et al. [9], patients with Kashin–Beck disease (KBD) and OA were subjected to circRNA sequencing to observe differential expression; the result of which showed that circRNA 0020014 could serve as the potential marker of OA to evaluate the progression of OA. Wang et al. [23] analyzed the gene expression profile, wherein VWF (hsa_circ_0025119) and other three genes served as OA markers. Xiao et al. [26] demonstrated that, on the Illumina HiSeq platform, circRNA 0045714 was expressed differentially in OA. Xiang et al. [54] revealed the expression profile of circRNAs in OA through RNA sequencing and identified 122 circRNAs of differential expression. Based on these studies, VWF (hsa_circ_0025119) had the highest value (Figure 3, Tables 3 and 4), indicating a significant interaction between VWF and other circRNAs; also, additional functions and signaling pathways were detected in the BP. Therefore, we speculated that VWF (hsa_circ_0025119) is more feasible to be used as a biological indicator compared to other circRNAs, to detect OA in clinical practice.

4.3. Studies on the Potential Functions of circRNAs in OA and Involved Signaling Pathways

The current study showed that circRNAs play a critical role in ECM. Shen et al. [7] showed that the overexpression of circRNA SERPINE2 downregulates the miR-1271-ERG (E26 transformation-specific-related gene) pathway to reduce HCS (human chondrocyte) apoptosis and promote ECM anabolism, thus slowing down OA development. Zhu et al. [11] found that circRNA GCN1L1 regulates miR-330-3p and TNF-α to promote OA synovial cells and reduce ECM anabolism. Wu et al. [18] demonstrated that circRNA 0005105 upregulates the expression of NAMPT (miR-26a target gene) and promotes ECM degradation in chondrocytes by absorbing miR-26a as sponges. In addition, circRNA TMBIM6 promotes ECM degradation of OA-induced chondrocytes via the miR-27a/matrix metalloproteinase-13 (MMP-13) axis, according to Bai et al. [21]. circRNA SERPINE2 reduces IL-1β-induced apoptosis and ECM degradation of chondrocytes by regulating the miR-495/transforming growth factor-beta receptor 2 (TGFBR2) axis [32]. Furthermore, the functions of circRNAs also include protein kinase activity, glycosaminoglycan binding, endoplasmic reticulum membrane, and peptidyl-serine phosphorylation, which can be the focus of future studies on the mechanism of OA.

In this study, VWF and COL6A3 are involved in the PI3K/AKT signaling pathway (Table 4). According to Zhou et al. [24], circRNA7 regulates PI3K/AKT/mTOR by absorbing miR-7, thus aggravating OA and indicating that the PI3K/AKT signaling pathway may play a critical role in circRNAs regulating the development of OA. The PI3K/AKT/mTOR signaling pathway functions in cartilage degeneration, subchondral bone dysfunction, and synovial inflammation [5557]. Therefore, in future studies on the mechanism of circRNA-regulated OA chondrocytes and synovial cells, the correlation between the circ-PI3K/AKT/mTOR axes can be observed, and the role of PI3K/AKT/mTOR is discussed. Multiple collagen factors were also detected in the ECM-receptor interaction pathway in Figure 5. Collagen is a vital component of cartilage composition and plays a crucial role in protecting cartilage tissues [5860]. This finding suggested that the ECM-receptor interaction signaling pathway may also play a critical role in the mechanism underlying circRNA-regulated OA (Figure 7).

We also found that the expression of most circRNAs was upregulated, while a few were downregulated in OA. According to Wang et al. [31], circRNA RNF121 aggravated the progression of OA via the miR-665/MYD88 axis (MYD88 is the canonical adaptor for inflammatory pathway), and according to Xiao et al. [33], circRNA CSNK1G1 promotes the progression of OAs by targeting the miR-4428/FUT2 (fucosyltransferase) axis. Jiang et al. [35] demonstrated that circRNA DHRS3 accelerates OA progression via miR-183-5p/GREM1 (Gremlin, the miR-183-5p target gene). Wang et al. [38] found that circRNA 0114876 aggravates OA via the miR-671/TRAF2 (TNF receptor-associated factor 2) axis. Yang et al. [40] found that circRNA RSU1 aggravates OA via the miR-93-5p/MAP3K8 (mitogen-activated protein kinase 8) axis, and Shen et al. [14] showed that circRNA CDK14 protects OA via the sponge tissue miR-125a-5p and enhances the expression of Smad2 (gene of TGF-β family). Moreover, in the study by Chen et al. [22], circRNA 9119 was shown to prevent apoptosis of IL-1β-treated OA chondrocytes by blocking the miR-26a/PTEN axis, and circRNA ADAMTS6 protects OA by absorbing miR-431-5p [36]. Another study showed that circRNA 0045714 exerted a protective effect on OA via the miR-193b/insulin-like growth factor 1 receptor (IGF1R) axis [37]. In summary, 77.78% of the circRNAs were upregulated and 22.23% were downregulated, and the overexpression of the majority of the circRNAs aggravates the occurrence and development of OA.

Herein, the studies on the correlation between circRNAs and OA in the recent 5 years and the circRNAs with differential expression and reliable mechanism of action in OA were reviewed. We found that most articles about circRNAs regulating OA in the recent 5 years focused on the mechanism, especially the absorption effect of circ-miRNA as sponges in the recent 2 years, while most of the articles about their functions addressed ECM and PI3K, AKT, and mTOR signaling pathways. Based on the GO and KEGG analysis results, future studies might focus on the functions of circRNAs, such as protein kinase activity, glycosaminoglycan binding, endoplasmic reticulum membrane, and peptidyl-serine phosphorylation, as well as ECM-receptor interaction-related signaling pathways. circRNA VWF, with preferable functions, interactions, and involvement, can be used as a biological indicator to detect OA in clinical practice. However, although the absorption effect of circ-miRNA as sponges in the mechanism of OA has been under intensive focus in the recent 2 years, studies are still rare. Therefore, further studies would focus on the database of the circ-miRNA axis in OA in order to provide a reference for the clinical treatment based on the mechanism of OA.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

The first author took responsibility for the conception and design of the study, or acquisition of data, or analysis and interpretation of data and drafted the article or revised it critically for important intellectual content. The corresponding author took responsibility for the final approval of the version to be submitted. Wei Wu is the first author.

Acknowledgments

This study was funded by Shanghai Key Laboratory of Human Sport Competence Development and Maintenance (Shanghai University of Sport) (No. 11DZ2261100) and the National Natural Science Foundation of China (81871835).