Knockdown of hsa_circ_0008922 inhibits the progression of glioma

Background A glioma is a tumor originating from glial cells in the central nervous system. Although significant progress has been made in diagnosis and treatment, most high-grade glioma patients are prone to recurrence. Therefore, molecular targeted therapy may become a new direction for adjuvant therapy in glioma. In recent years, many studies have revealed that circular RNA (circRNA) may play an important role in the occurrence and development of many tumors including gliomas. Our previous study found that the expression of hsa_circ_0008922 was up-regulated in glioma tissues upon RNA sequencing. The biological mechanism of circ_0008922 is still unreported in gliomas. Therefore, in this study, we preliminarily outlined the expression of hsa_circ_0008922 in glioma and explored its biological functions. Methods The expression of hsa_circ_0008922 in forty glioma tissues and four glioma cell lines (A172, U251, SF763 and U87) was detected by quantitative real-time polymerase chain reaction (qRT-PCR). The correlation between hsa_circ_0008922 expression and clinicopathological features of glioma patients was evaluated by Fisher’s exact test. To understand the potential function of hsa_circ_0008922 in glioma, we constructed small interfering RNA (siRNA) to hsa_circ_0008922 to downregulate its expression in glioma cell lines A172 and U251. With these hsa_circ_0008922 downregulated cells, a series of assays were carried out as follows. Cell proliferation was detected by CCK8 assay, migration and invasion were determined by wound healing assay and transwell assay, respectively. Colony formation ability was evaluated by plate clonogenic assay. Moreover, flow cytometry combined with Western blot was performed to analyze apoptosis status and the expression of apoptotic related proteins (caspase 3 and caspase 9). Finally, the possible biological pathways and potential miRNA targets of hsa_circ_0008922 were predicted by bioinformatics. Results We found that the expression of hsa_circ_0008922 in glioma tissues was 3.4 times higher than that in normal tissues. The expression of has_circ_0008922 was correlated with WHO tumor grade. After down-regulating the expression of hsa_circ_0008922, malignant biological behavior of glioma cells was inhibited, such as cell proliferation, colony formation, migration, and invasion. At the same time, it also induced apoptosis of glioma cells. Predicted analysis by bioinformatics demonstrated that hsa_circ_0008922 may be involved in tumor-related pathways by acting as a molecular sponge for multiple miRNAs (hsa-let-7e-5p, hsa-miR-506-5p, hsa-let-7b-5p, hsa-let-7c-5p and hsa-let-7a-5p). Finally, we integrated our observation to build a circRNA-miRNA-mRNA predictive network.


INTRODUCTION
Glioma is one of the most common primary malignant tumors in the central nervous system. Glioblastoma multiforme (GBM) is the highest grade (IV grade) based on the WHO classification and has high invasiveness and lethality (Ng et al., 2020;Sminia et al., 2021). Despite improvements in current treatment modalities (surgery, radiotherapy and chemotherapy) (Nabors et al., 2020), the prognosis of GBM remains poor and there is still a 90% recurrence rate after treatment (Francis et al., 2022;Easaw et al., 2013). This poses a serious threat to the health of patients and brings heavy economic burden to patients and their families. Therefore, it is critical to identify potential molecules for adjuvant therapy option in glioma.
circRNA is a new class of non-coding RNAs with covalently closed-loop structures (Cocquerelle et al., 1993), which is characterized by high stability and high conservation , and is mainly distributed in the cytoplasm. In recent years, a variety of high-throughput sequencing confirmed that many circRNAs were highly expressed in various tissues including tumors (Xu et al., 2019(Xu et al., , 2018Starke et al., 2015). In addition, circRNA can also act as a sponge for microRNAs (miRNA), endogenous competitive RNA to participate in many pathophysiological processes (Sun et al., 2020). Therefore, the expression profile and role of circRNA in glioma can not only provide a new treatment strategy, but also promote the development of precision medicine in clinical diagnosis and treatment.
With RNA sequencing of a panel of glioma tissues and normal brain tissues we found that hsa_circ_0008922 was highly expressed in glioma tissues compared with normal brain tissues. As far as we know, there is only a report for hsa_circ_0008922 showing its abundant expression in hypopharyngeal squamous cell carcinoma and correlation with worse outcome of patients . However, no studies have verified the function of hsa_circ_0008922 in glioma. In this study, we found that hsa_circ_0008922 expression was significantly associated with tumor grade, tumor size and Ki-67 which is a nuclear protein expressed in all proliferating cells (Røge et al., 2021). The down-regulation of hsa_circ_0008922 expression in vitro can weaken the malignant behavior of glioma cells, such as decreasing cell proliferation, migration, invasion and inducing apoptosis. Additionally, bioinformatic analysis demonstrated that hsa_circ_0008922 may exert a molecular sponge role for some miRNAs in glioma. Therefore, these results provide a basis for further study of hsa_circ_0008922 in glioma in future.

Patients and samples
From May 2018 to November 2021, glioma tissues were obtained from 40 patients with glioma who underwent surgical resection in the Department of Neurosurgery, the First Affiliated Hospital of Guangxi Medical University. All patients were confirmed to be glioma by postoperative pathology without preoperative radiotherapy or chemotherapy. At the same time, 10 non-glioma surgical brain tissue samples were collected. The brain tissue was obtained from the brain tissue needed to be removed due to the surgical approach and the focal edema tissue (normal brain tissue), and the histological characteristics were confirmed by the pathologist. All surgically resected specimens were frozen in liquid nitrogen. The specimen collection was examined and authorized by Medical Ethics Committee of Guangxi Medical University (2018087). We received written informed consent from participants of our study.

RNA sequencing
The tissues were collected as previously described in Liu et al. (2021). Specifically, to reduce the impact of individual differences, we set up each group containing five separate samples from five patients. The 10 tissues were divided into two groups: NB (normal brain tissue, n = 5) and GBM (glioblastoma, n = 5). Total RNA was extracted from glioma tissues using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol and sequenced by Aksomics Biology Technology Co. Ltd (Shanghai, China). The RNA library was constructed by using KAPA Stranded RNA-Seq Library Prep Kit (Illumina, San Diego, CA, USA). The raw sequencing data (FASTQ files generated by the Illumina sequencer) is subjected to quality control to assess whether the sequencing data can be used for subsequent analysis. The expression of circRNA was quantified by calculating the Backsplice junction reads through CIRCexplorer2. Differential expression analysis was conducted using Ballgown (https://rdrr.io/bioc/ballgown/). The circRNA with a P-value ≤ 0.05 and fold change ≥ or ≤1.5 was considered significantly differentially expressed. The statistical power of this experimental design, calculated in RNASeqPower using the web page at https://rodrigo-arcoverde.shinyapps.io/rnaseq_power_calc/ is 0.8083.

RNase R exonuclease digestion experiment and electrophoresis
According to the instructions in RNase R endonuclease digestion kit (Geneseed Biotech Co., Ltd., Guangzhou, China), the total RNA was divided into digestion group (RNase R treatment) and control group, with 5 mg RNA (1 mg RNA required 8 U RNase R endonuclease digestion) in each group. the RNase R was not added in the control group. The above groups were incubated at 37 C for 25 min. Then 1 mg RNA was reversed and detected by qRT-PCR. The qRT-PCR products amplified by the divergent primer were transcribed to Sangon Biotech (Shanghai, China) Co., Ltd. for TA cloning sequencing to determine the full length of PCR products. The PCR products of gDNA and cDNA were further detected by 1.3% agarose gel electrophoresis. PCR products were separated by 110 V electrophoresis for 35 min and detected by UV. GL DNA Marker 100 (AGbio, Beijing, China) was used as a marker of DNA size. The sequences of primers were as follows: hsa_circ_0008922 Divergent Primer: (sense) 5′-TCC ATC AGG ACC CCA GAT GTC-3′, (antisense) 5′-ACT GCA CAT GCA GAC TGT CAC-3′; hsa_circ_0008922 convergent Primer: (sense) 5′-GGG CAT CCT TCA CCC ATC TG-3′, (antisense) 5′-ATC TTG GTG TCA CAC AGG GC-3′.

Wound healing assay
After transfection with siRNA to hsa_circ_0008922 for 48 h, 100 mL of 2 × 10 4 cells were lifted and seeded into the single hole of ibidi plug-in and cultured for 22 h. After the cells were full of ibidi plug-in, the plug-in was removed. Then the DMEM complete medium was replaced with appropriate low serum DMEM complete medium (2% FBS). Photographs were taken at 0, 6, 12 h and 0, 12, 24 h, respectively.

CCK8 assay
CCK8 detection kit (Vazyme A311-01; Vazyme, Nanjing, China) was used to detect the proliferation of glioma cells treated with siRNA. A total of 100 mL of 4 × 10 3 cells were inoculated in 96-well plates, five wells in each group. A total of 10 mL of CCK8 solution was added to each well and cells were incubated at 5% CO 2 37 C for 2 h. The optical density (OD) values at 450 nm were measured at 0, 24, 48, 72 and 96 h, respectively.

Transwell assay
The migration and invasion ability of glioma cells were evaluated using a Transwell chamber (Corning Incorporated, Corning, NY, USA) with or without matrix glue (Absin). Forty-eight hours after transfection, a 200 mL cell suspension (2 × 10 4 cells) was inoculated in the upper chamber of transwell chamber, and DMEM (500 mL) containing 20% FBS was added in the lower chamber of transwell chamber. They were incubated at 37 C for 24 h. Then, 4% methanol was applied to fix the cells for 30 min followed by staining with 0.1% Crystal Violet Stain solution (Solarbio, Beijing, China) for 30 min. Stained cells were photographed by inverted microscope (Olympus, Shinjuku City, Japan) and counted manually.

Flow cytometry
Apoptosis was detected by using FITC Annexin V apoptosis detection Kit (BD Biosciences 556547; BD Biosciences, Franklin Lakes, NJ, USA). After glioma cells were transfected, the cells were completely digested with EDTA-free trypsin and washed twice with PBS buffering at 4 C. The cells were dyed with 5 mL FITC Annexin V and 5 mL propidium iodide (PI) for 15 min at room temperature in the dark and then observed by flow cytometry (BD, Franklin Lakes, NJ, USA). FlowjoV 1.8.1 (Becton, Dickinson & Company, Franklin Lakes, NJ, USA) was used to analyze the results.

Western blot
Western Blot was used to detect apoptosis-related proteins in glioma cells. After transfecting siRNA into cells and culturing cells for 48 h, the cells were harvested and lysed with protease inhibitor and phosphatase inhibitor to extract proteins. Then, the protein was separated by SDS-PAGE and transferred to PVDF membrane. 1:1,000 diluted primary antibodies (caspase 3, rabbit, #14220S; caspase 9, mouse, #9508, Cell Signaling Technology and GAPDH, rabbit, #ab181602 Abcam, Cambridge, UK) were incubated overnight at 4 C. After completing incubation of primary antibodies, the secondary antibodies in 1:5,000 dilution were incubated at room temperature (rabbit, #ab6721; mouse, #ab6728 Abcam Cambridge, UK).

Plate cloning assay
The clone formation rate indicated the proliferation capacity and cell population dependency (Liu et al., 2022).
3 × 10 3 of siRNA treated cells were inoculated in the 6-well plates at 37 C in 5 % CO 2 for 11 days. After culture, cells were fixed with 4% formaldehyde for 30 min and stained with 0.1% Crystal Violet Stain solution for 12 min. The photograph was recorded with a stereomicroscope (Olympus, Shinjuku City, Japan).

Statistical analysis
Data analysis was performed using Graphpad and SPSS 22.0 statistical analysis software. Student's t-test was used for comparison between the experimental groups. Fisher's exact test was used for categorical variables. Quantitative data were expressed as mean ± standard deviation. P < 0.05 was considered statistically significant.

Identifying hsa_circ_0008922 in glioma cell
In order to seek circRNA aberrantly expressed in glioma, we performed RNA sequencing of glioma tissues. Here, we selected a novel circRNA, hsa_circ_0008922, from circRNA identified above because there was no report in glioma except only one study in hypopharyngeal squamous cell carcinoma . As shown in Fig. 1A, hsa-circ-0008922 is produced from regions of exons 3-4 of (MATR3, NM199189),with a junction of 161 nt (http://www.circbank.cn/). We designed divergent primers for hsa_circ_0008922 and convergent primers for MATR3 mRNA. As shown in Fig. 1B, hsa_circ_0008922 can be amplified by divergent primers with cDNA as template, whereas there was no amplification detected at similar sizes using gDNA as template. Contrary, using convergent primers RT-PCR could amplify MATR3 mRNA in the sample without RNase R treatment, while it failed to amplify MATR3 mRNA in the sample with RNase R treatment. Moreover, PCR products amplified by divergent primers were used for TA cloning and sequencing to confirm that hsa_circ_0008922 had the reverse splicing site (Fig. 1C). These results indicated that the divergent primers we designed can specifically amplify hsa_circ_0008922 and can be used for detection of its expression.

Up-regulation of hsa_circ_0008922 in glioma tissues
We cannot gain any information for the expression profile of hsa_circ_0008922 in glioma from five public databases of circRNA (ClRCpedia v2, http://yang-laboratory.com/ circpedia/; circRNADb, circatlas.biols.ac.cn; TSCD, http://gb.whu.edu.cn/TSCD/; MiOncoCirc, https://mioncocirc.github.io/; and circRNADisease, http://cgga.org.cn:9091/ circRNADisease/). Therefore, we tested our clinical samples as a primary study for hsa_circ_0008922 expression. As shown in Fig. 2A, the expression of hsa_circ_0008922 in glioma tissues (n = 40) and normal tissues (n = 10) was detected by qRT-PCR. We found that the relative expression of hsa_circ_0008922 was significantly higher in gliomas than in normal tissues (P = 0.027). It was found that in low-grade tumor and high-grade tumor the expression of hsa_circ_0008922 was significantly higher than normal tissues, respectively  2B). Furthermore, we also detected the relative expression of hsa_circ_0008922 in different types of gliomas. We found that in astrocytoma and glioblastoma and oligodendroglioma the expression of hsa_circ_0008922 was higher than normal tissues, respectively (Fig. 2C).

Correlation between hsa_circ_0008922 expression and clinicopathological parameters of glioma patients
As hsa_circ_0008922 was up-regulated in gliomas, we further investigated the correlation between the expression of hsa_circ_0008922 and clinical data such as age, gender, tumor size, WHO grade and GFAP (glial fibrillary acidic protein). As shown in Table 1, the expression of hsa_circ_0008922 was correlated with the WHO grade, tumor size and Ki-67 of patients. Multiple linear regression showed that tumor size and WHO grade played a significant role in explaining hsa_circ_0008922 (Table 2).

Screening highly efficient siRNA for hsa_circ_0008922
Considering up-regulation of hsa_circ_0008922 correlated with some of clinicopathological parameters of glioma patients, we further conducted loss-of-function experiments to explore the biological function of has_circ_0008922 in vitro. We first detected the relative expression of hsa_circ_0008922 in a panel of glioma cells (A172, U251, U87 and SF763 cells) by qRT-PCR. Figure 3A shown all of cells expressing hsa_circ_0008922 with different levels. A172 expressed the highest hsa_circ_0008922, which was three times of SF763. Whereas U251 expressed the second higher level of hsa_circ_000892 that was 1.6 times of SF763. Therefore, A172 and U251 were selected for hsa_circ_0008922 depletion cell models. Subsequently, we determined the siRNA fragment and optimized its transfection time in vitro. Transfecting cells with two siRNA fragments, s1-hsa_circ_0008922 and s2-hsa_circ_0008922, respectively, it was found that s2-hsa_circ_0008922 had the best effect of -hsa_circ_0008922 down-regulation in the time point of forty-eight hours (Figs. 3B and 3C). Therefore, we chose s2-hsa_circ_0008922 for hsa_circ_0008922 depletion experiments and set the 48 h after siRNA transfection to perform a series experiment for biological behavior in glioma cells.

Down-regulation of hsa-circ-0008922 inhibits migration and invasion in glioma cells
Because migration and invasion were closely related to the malignant degree of tumor (van de Merbel et al., 2018), we used wound healing assays to evaluate lateral migration of glioma cells. Through the analysis of time series data and comparison between groups, it was found that the migration rate of s2-hsa_circ _0008922 treatment group in A172 (Figs. 4A and 4C) and U251 (Figs. 4B and 4D) was lower than that of the control. Further, the transwell assays of migration and invasion showed suppression of longitudinal migration and invasion of A172 (Figs. 5A-5C) and U251 (Figs. 5D-5F) following hsa-circ-0008922 silencing. In summary, the down-regulation of hsa_circ_0008922 in glioma cells can significantly inhibit the migration and invasion ability of glioma cells.  Influence of colony formation, proliferation, and apoptosis after hsa_circ_0008922 depletion in glioma cells After down-regulating hsa_circ_0008922 in A172 and U251, the colony formation ability of cells was detected by plate cloning. It was found that compared to the NC group, the number of cell colonies in the s2-hsa_circ_0008922 group was significantly lower than that in the NC group (Figs. 6A and 6B). CCK8 experiment was used to detect the proliferation in A172 and U251. In this experiment, we detected cell viability at five time points (0, 24, 48, 72, 96 h). It was found that the cell viability in the s2-hsa_circ_0008922 group in A172 (Fig. 6C) and U251 (Fig. 6D) was significantly different from that in the NC group. This result indicated that the down-regulation of hsa_circ_0008922 in glioma cells can significantly inhibit the cell growth in glioma cells. Moreover, flow cytometry was used to study whether hsa_circ_0008922 affected apoptosis in glioma cells. We found that in A172 and U251, the apoptosis rate in the s2-hsa_circ_0008922 group was higher than that in the NC group, and there was significant indigenous difference between the two groups (P < 0.05) (Figs. 7A-7C). Furthermore, we also detected apoptosis proteins caspase 3 and caspase 9. It was found that in A172 (Figs. 7D and 7E) and U251 (Figs. 7D and 7F) the expression of caspase 3 and caspase 9 in the s2-hsa_circ_0008922 group was higher than those in the NC group, respectively. Taken together, hsa_circ_0008922 promoted the proliferation of glioma cells by playing an anti-apoptotic role.
Based on the ceRNA hypothesis (Whereas micrornas are known to cause gene silencing by binding mrnas, cernas can competitively bind micrornas to regulate gene expression (Salmena et al., 2011)), namely, mRNAs should be negatively correlated with targeted miRNAs and positively correlated with circRNA. The top 30 of mRNAs fitting this criterion were selected for further functional analysis. GO analysis showed that the functions of these mRNAs were mainly related to multiple signaling pathways, such as mitotic sister chromatid cohesion, stem cell population maintenance, negative regulation of intrinsic apoptotic signaling pathway in response to DNA damage by p53 class mediator, negative regulation of cAMP-mediated signaling, and regulation of transcription  Table S1). KEGG analysis indicated that these mRNAs were involved in signaling pathways regulating pluripotency of stem cells, cell cycle, and microRNAs in cancer and glioma ( Fig. 9B and Table S2).

DISCUSSION
Glioma is a very common and aggressive intracranial tumor, which leads to brain dysfunction in many patients and poses a serious threat to the health of patients. At present, the traditional treatment of glioma is the comprehensive treatment in surgery, radiotherapy, and chemotherapy, but the prognosis is still poor, and the postoperative recurrence rate is as high as 90% (Ostrom et al., 2014;Van Meir et al., 2010;DeAngelis, 2001). However, the pathogenesis and the progression of glioma still remain unclear. Therefore, it is very important to understand the related molecules for their expression features and their mechanisms in the occurrence and development of glioma. In early studies, circRNA was considered as "noise" and had no biological function. However, in recent years, it was found that circRNAs mostly distributed in the brain and had significant biological functions, especially in cancer (Shahzad et al., 2021;Yang et al., 2018;Rybak-Wolf et al., 2015). In 2019, Xia et al. (2019) showed that AKT3-174aa encoded by circ-AKT3 could inhibit the carcinogenicity of glioma mother cells. In 2020, He et al.
(2020) reported that circ-MAPK4 was up-regulated in glioma, but its expression in high-grade glioma was significantly higher than that in low-grade glioma. At the same time,  found that the expression level of hsa_circ_0074026 was significantly increased in the parent cell samples and cells of glioma, and was correlated with clinical parameters such as tumor size, WHO grade and overall survival. All of these reports imply circRNA has potential for diagnostic and therapeutic application in gliomas.  Through RNA sequencing, we found that hsa_circ_0008922 highly expressed in glioma tissues. hsa_circ_0008922 is derived from exon 3 and exon 4 of Matrin3 (MATR3). MATR3 is a host gene of hsa_circ_0008922 and it is derived from human chromosome 5 and highly expresses in the brain (Coelho et al., 2016). MATR3 participates in many cellular processes, including binding and stabilizing mRNA, regulating mRNA nuclear output, chromatin remodeling, RNA processing, transcription, translation, and apoptosis Salem et al., 2021). In addition, MATR3 has been widely studied in amyotrophic lateral sclerosis (You et al., 2022;Cavalli et al., 2021;van Bruggen et al., 2021). However, the research progressed on circRNA from MATR3 in glioma remains unreported. In this study, we first verified that the expression of hsa_circ_0008922 in glioma tissues was higher than that in normal tissues. Statistical analysis showed that the expression of hsa_circ_0008922 was significantly correlated with WHO grade, indicating that hsa_circ_0008922 could be a significant role in the progression of malignant glioma. The expression of hsa_circ_0008922 was significantly correlated with tumor size and Ki-67, which indicated that hsa_circ_0008922 could have a significant function in proliferation of glioma cells. Down-regulation of hsa_circ_0008922 expression in glioma cell lines A172 and U251 can significantly inhibit the proliferation, migration, and invasion of the two cells, indicating that hsa_circ_0008922 may become a new molecule for molecular targeted therapy.
Previous studies have shown that circRNA is often used as a miRNA sponge to regulate the expression of miRNA in various cancers, forming a circRNA-miRNA-mRNA network to regulate the expression of downstream target genes and affect the development of cancer (Hansen et al., 2013;Kristensen et al., 2018). For example, hsa_circ_0003222 promoted stem cell differentiation and progression of non-small cell lung cancer through miR-527 (Li et al., 2021). Therefore, in this study, we predicted miRNAs regulated by hsa_circ_0008922 through bioinformatics analysis, and selected five miRNAs for bioinformatics prediction based on predicted binding fraction and free energy. Previous studies have shown that hsa-let-7b-5p, hsa-let-7e-5p and hsa-let-7c-5p were all tumor suppressors. In 2019, Xi et al. (2019) found that hsa-let-7b-5p could inhibit the migration, invasion, and cell cycle of glioma cells. Another study found that hsa-let-7e-5p as a tumor suppressor could inhibit the progression of head and neck squamous cell carcinoma by targeting CCR7 expression . A study reported by Fu et al in 2017 demonstrated that hsa-let-7c-5p may act as a tumor suppressor of breast cancer by negatively regulating ERCC6, providing a new strategy for breast cancer treatment (Fu et al., 2017). Thus, the interaction between the hsa_circ_0008922 and these three members of the hsa-let-7 miRNA family may broaden our view for their functions in glioma. Figure 9 (continued) x-axis represents the enrichment score. The size of the bubbles indicates the number of mRNAs in the GO term. Bubble colors represent P-values. (B) A bubble plot of the top 30 enriched pathways in mRNAs negatively correlated with miRNAs downstream of hsa_circ_0008922 was drawn. The y-axis represents the enrichment of the top 30 items, and the x-axis represents the enrichment score. The size of the bubbles indicates the number of mRNAs in the pathway. Bubble colors represent P-values.
Full-size  DOI: 10.7717/peerj.14552/ fig-9 It was interesting that a gene called DNA polymerase lambda (POLL) may be regulated by above mentioned three miRNAs (hsa-let-7b-5p, hsa-let-7e-5p and hsa-let-7c-5p), simultaneously, and its expression may be negatively correlated with the expression of these three miRNAs. Now, there is a report that POLL is differentially expressed in IDH (lactate dehydrogenase) mutant and wild type carriers, relating to the survival rate of patients with LGG (Pang et al., 2020). In future, a series of experiments such as RIP (RNA Binding Protein Immunoprecipitation Assay), RNA pull-down assay and luciferase reporter gene assay have to be performed to illustrate the complicate function of hsa_circ_0008922 in this circRNA-miRNA-mRNA network.
We performed GO analysis and KEGG analysis on the target genes negatively correlated with five miRNAs, and the results showed that hsa_circ_0008922 regulates cancer-related functions and signaling pathways by regulating several miRNAs to play a carcinogenic role. GO analysis showed that the enriched functions of mRNA negatively correlated with miRNAs were mainly related to cell cycle, including regulation of transcription involved in G1/S transition of mitotic cell cycle and mitotic sister chromatid cohesion. Enriched functions were demonstrated in relating to apoptosis, including negative regulation of intrinsic apoptotic signaling pathway in response to DNA damage by p53 class mediator. As known to all, p53 pathway includes DNA repair, cell-cycle arrest, senescence and apoptosis (Riley et al., 2008). The 53 tumor suppressor gene is among the best and longest studied genes in glioblastoma. Since the key function of the gene product p53 is to arrest cells in G0/1 or trigger apoptosis in response to genotoxic stress, so restoration of p53 function by various methods has been extensively studied. However, drugs designed to facilitate the refolding of the mutant protein into a wild-type conformation have not been successful (Rhun et al., 2019). In the result of KEGG analysis, enriched pathways of mRNA negatively correlated with miRNAs were mainly related to tumor, e.g., microRNAs in cancer signaling pathway, glioma signaling pathway. It was also included in cell cycle signaling pathway, Ras signaling pathway which is a pathway closely related to glioma, activating RAS mutation to increase gliogenesis (Breunig et al., 2015). Therefore, we assumed that p53 pathway and Ras signaling pathway might be potential signaling pathways, in which hsa_circ_0008922 may play an oncogenic role in growth and progression of glioma. Nevertheless, these hypotheses should be further validated in our future studies.
There are several limitations in our study. First, the clinic data of this study lacked overall and disease-free survival. However, currently most of the specimens were collected in 2021, less than half a year before the start of this study, so we cannot get the accurate survival time. We will continue to follow up this batch of specimens. Additionally, to confirm the roles of hsa_circ_0008922, in vivo experiments should be further conducted in the mouse models those we have established. Therefore, the roles of hsa_circ_0008922 in glioma warrant further investigation in our future studies.

CONCLUSION
Glioma highly expressed hsa_circ_0008922, which could affect the proliferation, migration, and apoptosis of glioma cells. hsa_circ_0008922 may server as a sponge for multiple miRNAs.

ADDITIONAL INFORMATION AND DECLARATIONS Funding
Ping Wang performed the experiments, authored or reviewed drafts of the article, and approved the final draft. Feng Li performed the experiments, authored or reviewed drafts of the article, and approved the final draft. Yingying Ge conceived and designed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft. Qingmei Zhang conceived and designed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft. Xiaoxun Xie conceived and designed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft. Xisheng Li conceived and designed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft. Bin Luo conceived and designed the experiments, analyzed the data, prepared figures and/or tables, and approved the final draft.

Human Ethics
The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers): The specimen collection was examined and authorized by Medical Ethics Committee of Guangxi Medical University (2018087).

DNA Deposition
The following information was supplied regarding the deposition of DNA sequences: The group II intron/IEP sequences are available at GenBank: OP541861.

Data Availability
The following information was supplied regarding data availability: The Figs. 8 and 9 are available in the Supplemental Files 4 and 5.

Supplemental Information
Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.14552#supplemental-information.