Hsa_circ_0043278 functions as competitive endogenous RNA to enhance glioblastoma multiforme progression by sponging miR-638

Circular RNAs have a critical function in the pathogenesis of many diseases and can function as competing endogenous RNA or miRNA sponges to inhibit miRNA and therefore upregulate the expression of target genes. However, little is known about the role of has_circRNA_0043278 (circ_0043278) in glioblastoma multiforme (GBM) and its potential downstream miRNA targets. This work validated that circ_0043278 is highly expressed in GMB cell lines and tissues, while knockdown circ_0043278 inhibited GBM cell migration, proliferation, and invasion in vitro and tumorigenesis in vivo. Dual-luciferase reporter assay determined that circ_0043278 directly sponged miR-638 to upregulate the expression of HOXA9, which can activate downstream Wnt/β-catenin signaling in GBM. Moreover, miR-638 inhibition reversed circ_0043278 silencing-induced impairment of malignant tumor behavior. These results showed that circ-0043278/miRNA-638/ Homeobox A9 (HOXA9) axis had a vital function in promoting GBM progression. Our findings may provide potential new targets for the diagnosis and therapy of GBM.


INTRODUCTION
Glioma is a main class of primary brain tumor that leads to half of malignant brain tumor types [1]. The World Health Organization categorizes gliomas into four grades (from I to IV) according to overall survival (OS) and malignancy. Glioblastoma multiforme (GBM) is a malignant fastgrowing grade IV glioma. GBM is the most typical adult malignant primary brain tumor that results in 60-80% of primary brain tumors [2], with a median OS of ~9-15 months and a five-year survival rate <10%, despite chemotherapy, radiotherapy, and surgical treatments [3][4][5]. Early prognosis and diagnosis are necessary for accurate GBM treatment. Due to poor outcomes after conventional treatments, it is necessary to discover novel therapeutic approaches and treatment policies for GBM, particularly those targeting complicated gene regulation networks.
Circular RNAs (circRNAs) belong to a new family of endogenous RNAs in mammalian cells. These include covalently closed continuous loops without 5ʹ end caps and 3ʹ end poly (A) tails [6,7]. CircRNAs are mostly derived from exons, which are mainly found in the eukaryotic cell cytoplasm [8]. In addition, circRNA expression is diseasespecific and organic [9]. CircRNAs have been formerly erroneously regarded as byproducts of splicing errors that were not associated with any specific mechanisms. Improvements in high-throughput sequencing technology led to investigations that determined that many human circRNAs might be involved in fetal development [10], myocardial infarction [11], and carcinomas [12]. CircRNAs have important functions in many diseases, such as tumors, by upregulating target gene expression via microRNA (miRNA) sponge action [13]. These characteristics make circRNAs potential candidates for tumor treatment and diagnosis, which are indispensable in medicine of tumor transformation. Previous research studies have indicated that circRNAs have the capability to inhibit or facilitate various tumor progression and development, including gliomas [14][15][16].
The present study identified that circ_0043278 is upregulated in GBM cell lines and tissues. Knockdown circ_0043278 significantly inhibited cell migration, proliferation, and invasion. Study results verified that circ_0043278 upregulated Homeobox A9 (HOXA9) expression by absorbing miR-638 in GBM. This study provides a novel direction for clinical therapy of GBM.

Circ_0043278 targets on miR-638
Science that circ-0043278 is predominantly distributed in the cytoplasmic fraction, it was hypothesized that circ-0043278 may act as a competitive endogenous RNA (ceRNA) in biological processes. Based on the online predictions of circ-interactome (https:// circinteractome.nia.nih.gov) and analysis using the circinteractome database, circ_0043278 shares putative binding sites with miR-638 ( Figure 5A). To explore a potential interaction between circ-0043278 and miRNA-638, dual-luciferase reporter plasmids carrying a fragment of the mutant (mut) or wild-type (wt) circ-0043278 sequence and the predicted miR-638 recognition site were constructed. Luciferase activity was markedly decreased in 293 cells co-transfected with circ-0043278-wt and miR-638 mimics. However, no significant differences in luciferase activity were present in cells co-transfected with circ-0043278-mut and miR-638 mimics (F (3,8)=12.84, p=0.002, Figure  5B). These results suggest that circ_0043278 sponges miR-638.

DISCUSSION
CircRNAs belong to a class of single-stranded RNA molecules with a covalently closed loop structure. They have been characterized by high stability, abundance, conservation, and display tissue stage-specific expression. Furthermore, based on the abundance in distinct body fluids or exosomes, circRNAs present novel biomarkers and targets for the diagnosis and prognosis of cancers [25]. CircRNAs have indispensable functions in different biological processes, especially carcinogenesis, cancer metastasis, and progression in lung adenocarcinoma [26], hepatocellular carcinoma [27], and gastric cancer [28]. Recently, several circRNAs have been reported to play important roles in human glioma. CircRNAs such as circSMARCA5 and circCFH, are expressed in a glioma-specific pattern. These circRNAs may be used as tumor biomarkers. CircNFIX and circNT5E served as tumor promoter in glioma, while circFBXW7 and circSHPRH were reported to function as tumor suppressors [29]. Wu et al. found that circTADA2A, also named circ_0043278, target miR-203 to upregulate CREB3 expression, thus leading to cell migration, invasion, and proliferation in osteosarcoma [21]. Zhang et al. used circRNA microarray dataset analysis and found that circ_0043278 was upregulated in pancreatic ductal adenocarcinoma [20]. Nevertheless, circ_0043278 functions in GBM remain unknown. The current study focused on the role of circ_0043278 in GBM mechanisms of cell proliferation, invasion, and migration. Compared to normal human astrocytes, circ-0043278 is highly expressed in the GBM cell lines. Knockdown of circ-0043278 markedly inhibited cell proliferation, migration, and invasion in vitro and tumorigenesis in vivo, indicating that circ_0043278 is a positive tumor promoter in GBM.
As a newly discovered type of noncoding RNA, circRNAs share many functions with traditional noncoding RNAs. Recently, circRNAs have been reported to function as miRNA sponges, protein sponges, coding RNAs or scaffolds for protein complexes [29]. Long noncoding RNAs (lncRNAs) function as ceRNA to upregulate gene expression by targeting miRNAs in GBM [30]. CircRNAs also can function asceRNAsor miRNA sponges to inhibit miRNA and therefore upregulate the expression of target genes via miRNA response elements [31]. MiRNAs are small non-coding RNAs that are 18-24 nucleotides long and function during gene expression regulation [32]. Translation of targeted mRNAs is inhibited post-transcriptionally via miRNA binding to 3'UTR sequences [33,34]. For example, circRNA-5692 binds to miR-328-5p to enhance DAB2IP expression and thus inhibits cell migration, proliferation, and invasion in hepatocellular carcinoma [35]. Circ_0000285 acts as a ceRNA that increases TGFB2 expression by sponging hsa-miRNA-599 in osteosarcoma [36]. In the current study, miR-638 was downregulated in glioma cells and tissues, while a luciferase reporter assay indicated that circ_0043278 directly targets miR-638.
In addition, HOXA9 was reported to activate the Wnt/βcatenin signaling pathway and is directly targeted by miR-638 [24]. The Wnt/β-catenin pathway is a highly conserved signaling cascade that takes part in different cellular processes, such as cell differentiation, proliferation, adhesion, survival, and migration [37]. U87 and U251 cells were stably transfected with NC or cic_0043278-shRNA1# or co-transfected with cic_0043278-shRNA1# and miR-638 inhibitor. Migration and invasion were evaluated using transwell assays. Results represent the mean±SD for three independent experiments (*p<0.05, **p<0.01, as determined by one-way ANOVA with Bonferroni's multiple comparisons test). Scale bar = 50 μm in all panels. AGING Wnt/β-catenin signaling has been recently found to be related to EMT-like states [38], cancer stem cell biology [39], self-renewal [40], and chemo-resistance [41] in some cancers. Sun et al. reported that some circRNAs affect the proliferation and invasion of gliomas through cancer-associated signaling pathways. For example, circ-0000177 activates the wnt/β-catenin pathway through miR-638/FZD7, while circTTBK2, circSHKBP1 and circHIPK3 activate PI3K/AKT and MARK/ERK signaling pathway through miR217/HNF1β, miR544a/ FOXP1, and miR379/FOXP2, respectively [25]. The present study found that circ_0043278 functions as a ceRNA that increases HOXA9 expression and Wnt/βcatenin signaling pathway by sponging miR-638.  implantation with U87 cells stably transfected with NC or cic_0043278-shRNA1# or co-transfected with cic_0043278-shRNA1# and LV-anti-miR-638 subcutaneously. Tumors were dissected and photographed after one month; (B) Tumor weight was calculated after euthanasia (N = 6 in each group, *p<0.05, **p<0.01, as determined by one-way ANOVA with Bonferroni's multiple comparisons test); (C) Tumor volumes were recorded every six days after mice were injected with stable glioma cells (N = 6 in each group, **p<0.01, ***p<0.001, *** p<0.0001, as determined by two-way ANOVA with Dunnett's multiple comparisons test); (D, E) Western blotting was used to detect HOXA9 protein levels and Wnt/β-catenin signaling. Results represent the mean±SD for three independent experiments (*p<0.05, **p<0.01, as determined by oneway ANOVA with Dunnett's multiple comparisons test).
In conclusion, this study demonstrated that circ_ 0043278 functions as a ceRNA to enhance GBM progression by sponging miR-638. The results indicated that circ_0043278 may function as a new biomarker, which provides a novel direction for GBM clinical diagnosis and treatment.

Patient samples
A total of 30 patients were enrolled in the study. The patients had surgeries to resect glioma tumors, as well as adjacent non-tumor tissues at the First Hospital of Lanzhou University between September 2017 and December 2018. Tumor specimens were snap frozen in liquid nitrogen and kept at -80°C until further use. Tissue sections were inspected by three pathologists during diagnosis to identify tissue variants. Ethics Committee at the First Hospital of Lanzhou University approved the study. Consent forms were obtained from all patients.

Cell culture
Human U87, U251, A172, and U373 glioblastoma cell lines were purchased from the ATCC (Manassas, VA, USA). Primary normal human astrocytes (NHAs) were from Sciencell Research Laboratories (Carlsbad, CA, USA). NHAs were cultured in Dulbecco's modified Eagle's medium (DMEM), while all cultured glioma cell lines were cultured in DMEM-F12. Both mediums were supplemented with 10% fetal bovine serum (FBS, Gibco, Grand Island, NY, USA) and incubated at 37°C and 5% CO2. No mycoplasma contamination was detected in all cell lines by the MycAwayTM Plus-Color One-Step Mycoplasma Detection Kit (Yeasen, Shanghai, China).

RNase R treatment, RNA extraction, and quantitative real-time PCR (qRT-PCR)
Total RNA samples were extracted from tissues and cells using TRIzol (Invitrogen, Carlsbad, USA) following standard procedures. For the RNase R treatment, 2 mg of total RNA were incubated for fifteen minutes at 37°C with or without 3 U/mg RNase R (Epicentre Technologies, Madison, USA). SYBR Premix Ex Taq II (TaKaRa) and script RT reagent kit (TaKaRa, Dalian, China) were used for the mRNA and circRNA analyses. Reactions were measured using a Roche LightCycler® 480II PCR instrument (Basel, Switzerland) following standard procedures. β-actin was used as a standard internal control. For the miRNA analyses, samples were treated with DNase I to remove genomic DNA, while cDNA was synthesized by leveraging the Mir-X miR First-Strand synthesis kit (TaKaRa). SYBR Premix Ex Taq II (TaKaRa) was used for qRT-PCR. U6 was employed as an internal standard control. Relative RNA expression levels were analyzed using a 2 -ΔΔCt method as previously described [21].

Western blot (WB) analysis
Cells were lysed in the RIPA lysis buffer that included a protease inhibitor cocktail on ice. An equivalent amount of total protein was separated from various samples using SDS-PAGE gels at 100 V for 1.5 h and then transferred onto 0.22-μm polyvinylidene difluoride membranes at 280 mA for 1.5 h. The membranes were blocked with 5% non-fat milk for 1 h and then incubated overnight at 4°C with primary antibodies. The primary antibodies were: anti-HOXA9 (ab191178, 1:500), anti-CyclinD1 (ab134175, 1:5000), anti-c-Myc (ab185656, 1:1000), and anti-β-catenin (ab16051, 1:1000; all from Abcam, San Francisco, USA). Membranes were washed and then treated with goat anti-rabbit IgG conjugated with horseradish peroxidase (HRP) for 1 h at 37°C. Protein bands were visualized using chemiluminescence HRP substrate, where β-actin served as an internal control. Blot density was then quantified using Image J software.

Cell proliferation assay
Cell proliferation was determined using a cell counting kit-8 (CCK-8; Dojindo, Japan). A total of 2 × 10 3 cells/well were seeded into 96-well plates for overnight culture. Afterwards, 10 µL of the CCK-8 reagent were added for cell proliferation quantification at A450 nm after 24, 48, or 72 h using an Epoch Microplate spectrophotometer (BioTek, Winooski, USA).

Cell migration and invasion assays
Migration and invasion capabilities were determined using a transwell assay (Corning, USA). For cell migration, culture inserts with 8.0-μM pores (Transwell, NY, USA) were placed into 24-well culture plates. Cell invasion assay was performed using a 40 μl 1:4 mixture of Matrigel Basement Membrane Matrix (Corning) according to the manufacturer's protocols. Briefly, 200 μL of serum-free DMEM-F12 medium containing 5×104 treated cells for the migration assay, or 1 ×105 cells/well for the invasion assay were added to the upper chamber. Then, 600 μL of DMEM-F12 containing 10% FBS were added into the lower chamber. After incubation in a humidified atmosphere for 24 h, non-migrated or non-invaded cells were removed by scraping the upper surface of the membranes with a cotton swab, cells attached to the lower membrane were fixed in 4% paraformaldehyde at room temperature for 10 min and stained with 0.1% crystal violet at room temperature for 20 min. Five random fields of view were evaluated under a microscope with magnification of ×200.

Luciferase reporter assay
Wild-type (wt) or mutant (mut) seed circ_0043278 regions were designed to include putative target sites for miR-638 (Shanghai GeneChem, Shanghai, China) and cloned into the pGL3 vector. HEK-293 cells were cotransfected with the reporter plasmid and miR-638 mimics or mimics NC (RiboBio). After transfection for two days, luciferase activity was analyzed with a Dual Luciferase Reporter Assay System (Promega) and normalized using Renilla luciferase activity.

Subcutaneous tumor models
Animal Ethics Committee at the First Hospital of Lanzhou University approved all animal experiments. All protocols were established following provided guidelines. Five-week-old male nude mice were purchased from Shanghai SLAC Laboratory Animal Co. Ltd. (Shanghai, China) and housed in laminar airflow cabinets under pathogen-free conditions. Animals were randomly divided into three groups (NC, circ_0043278-shRNA1# and circ_0043278-shRNA1#+LV-anti-miR-638) (six in each group). Tumor formation was initiated via subcutaneous injection of infected U87 cells (2× 10 6 cells) into nude mouse flanks. Tumor volume was computer using an equation, where volume = length × width 2 /2. Tumor volume was measured every six days. After 30 days, mice were euthanized and tumor grafts were excised, weighed, and harvested. Proteins were extracted from the one-half of tumors, and the expression of HOXA9, CyclinD1, c-Myc and β-catenin were measured by WB. The other half was fixed in 4% paraformaldehyde.

Statistical analysis
Results are presented as the mean ± standard deviation (SD) from three independent experiments performed in triplicate (n=3). Statistical analysis was performed using GraphPad Prism 6 (GraphPad software, USA). Student's t-test (paired or unpaired) was used to compare differences between two groups and one-way or two-way analysis of variance (ANOVA) was used to evaluate differences among more than two groups. p values<0.05 were considered statistically significant.

Editorial note
& This corresponding author has a verified history of publications using a personal email address for correspondence.

CONFLICTS OF INTEREST
The authors declare no conflict of interest.