CircRNA_0058063 functions as a ceRNA in bladder cancer progression via targeting miR-486-3p/FOXP4 axis

Bladder cancer (BC) was ranked as the main factors that are responsible for the cancer-related deaths worldwide, with roughly 74,000 developed cases in the United States in 2015 [1,2]. The risk of BC increases with age, peaking between 50 and 70 years old, and men are three times higher than women [3]. The recurrence rate is as high as over 70% within 5 years, and bladder cancer is required for lifelong cystoscopy [4].

Circular RNAs (circRNAs) represent a sort of non-coding RNA, which are widespread presented in mammals [5]. circRNAs are formed by RNA splicing, which occurs on a so called ´head-to-tail’ splicing joint where the receptor splicing site at exon 5′ ends is linked to the donor site at downstream exon 3′ ends. [6]. Most circRNAs (∼85%) resided with known protein-coding genes with structures have intact exons, suggesting that RNA polymerase II (RNA pol II) transcribes them and their biogenesis may be mediated by the splice [7–9]. In the past two decades, they have been considered non-functional due to splicing errors. To date, numerous circRNAs have been continuously explored in various cell lines and tissue species [10]. However, their biological processes and potential functions are poorly understood. Several circRNAs have been reported to regulate the progression of BC. For example, circUBXN7 were identified to suppress BC cancer viability and metastasis [11]. circSLC8A1 acts as a ceRNA of miR-130b/miR-494 to regulate PTEN expression and thus inhibits BC progression [12]. It has been reported that circFAM114A2 represses BC carcinogenesis through regulating ∆NP63 by sponging miR-762 (Liu et al.). Another study also reported that by binding to microRNA-589-5p circ_0091017 suppresses BC cell growth, migration, as well as invasion (Zhang et al.). However, the mechanistic function of major circRNAs in BC is merely understood. In hepatocellular carcinoma, circTRIM33–12 has thus far been verified to induce tumorigenesis via its action as a buffering molecule for miR-191 [13]. Specifically, in lung cancer circRNA circPRKCI stimulates tumor cell growth through a ceRNA mechanism [14]. In pancreatic cancer, circ-LDLRAD3 was found to be significantly associated with lymphatic/venous invasion and metastasis [15]. Taken together, all these explorations potentiate circRNA's regulatory roles as ceRNA in cancer progression.

In this research, we reanalyzed the previous microarray datasets and found that circRNA_0058063 was aberrantly expressed in both different BC cell lines and BC carcinoma tissues. Silencing of circRNA_0058063 signaling blocked cell proliferation but promoted cell apoptosis. Furthermore, circRNA_0058063 could bind to miR-486-3p and regulate the expression of FOXP4. Therefore, circRNA_0058063 would be a potent candidate for future diagnostic and therapeutic purpose of BC.

All of 94 cases of BC cancerous tissues and the non-cancerous normal tissues were retrospectively analyzed from the First Affiliated Hospital of Wenzhou Medical University during April 2013 to Oct 2014. BC cell lines (BIU-87, 5637 and RT-112) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, U.S.A.). 293T cell line was donated by Dr Teng of Shanghai Eighth People’s Hospital (Shanghai, China). Cells were cultured in DMEM/F12 supplemented with 10% FBS (Gibco, Grand Island, NY, U.S.A.), penicillin (100 µg/ml, Gibco) and streptomycin (100 µg/ml, Gibco) at 37°C in 95% humidified air and 5% CO₂. All studies were conducted by following the protocol provided by the manufacturer. Studies are performed in accordance with the Declaration of Helsinki and adhere to local ethical regulations. We have acquired the written informed consent from each participant before the initiation of this experimental study. The protocol of the current investigation was authorized by the Institutional Review Board of the First Affiliated Hospital of Wenzhou Medical University.

Full-length FOXP4 was inserted in pcDNA3.1 plasmid to construct the overexpression plasmid. Transfection was performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s introductions.

We used microarray datasets for identification of candidate circRNAs at platform GPL19978 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE92675). Four pairs of BC carcinoma tissues and paired para-carcinoma tissues were selected for detection of the expression and potential circRNAs functions.

Total RNA was extracted using TRIzol reagent (Invitrogen, U.S.A.). Then, cDNA was reverse-transcribed by the Prime Script RT reagent Kit (New England Biolabs, MA, U.S.A.) and subjected to RT-PCR measurement using PowerUp SYBR Green Master Mix (Takara Bio, Dalian, China). The relative expressions of each gene were acquired by using the 2-ΔΔCt method. The primers were listed below:

• 5′-TATGATCCTGTTTGGTGGTCGGCA-3′ (forward), 5′-TGGACCAAGATGGGTAGCTTGTGA-3′ (reverse) for circRNA_0058063

• 5′-CAATGTACGTTCGCTATCGGC-3′ (forward), 5′-CTCTCACGCACTTAATGCGAT-3′ (reverse) for miR-486-3p

• 5′-CGGCACAGCAGCTGAACTTA-3′ (forward), 5′-GCAACACCTGAAAAAGTGTGA-3′ (reverse) for FOXP4

• 5′-CTCGCTACACCTCAATACATCG-3′ (forward), 5′-GCGCCATAAGTCTAGTATTGAGA-3′ (reverse) for GAPDH

siRNA for circRNA_0058063 or scramble siRNA (si-NC) was synthesized by Invitrogen. The miR-486-3p mimic and the negative control (miR-NC) were synthesized by GenePharma (Shanghai, China). Lipofectamine 2000 (Invitrogen, Carlsbad, CA) system was used to conduct cell transfection assay by following the manufacturer’s protocol.

As previously described [16], 2 ×10³ cells were split and cultured in 96-well plates at 24 h before its use for CCK-8 assay. After 48 h of transfection, 10 μl of CCK-8 solution was then pipetted into the medium. Absorbance was detected at 450 nM on a Spectra Max 250 spectrophotometer (Molecular Devices, U.S.A.). After cells were fixed for 15 min in 4% of paraformaldehyde, the colonies were stained by 0.05% Crystal Violet (20 min) followed by visualized under a microscope.

The rate of apoptosis of the regulatory cells was examined. Both flow cytometry and annexin V/propidium iodide staining were applied to detect cell death rate by using the Annexin V-fluorescein isothiocynate Apoptosis Detection Kit (KeyGen, Nanjing, China).

At the beginning 20% fetal calf serum was put into the lower chamber, and the conditioned cells were suspended in a medium without FBS and inoculated onto an upper chamber. Forty-eight hours after adding the cells, get rid of the Matrigel in the upper chamber, at the same time the attached cells of invading ability were fixed, and Crystal Violet was applied to stain the cells, and then it was subjected for imaging with the aid of a microscope.

With the aid of Lipofectamine 2000 transfection system, 1 × 10⁷ cells were incubated in the RIP buffer transfected with MS2bs-circRNA_0058063mt or MS2bs-circRNA_0058063, and control MS2bs-Rluc. After 48 h of transfection, RIP was conducted by using the Magna RIP RNA-binding protein immunoprecipitation kit (Millipore). For the RIP assay of Ago2, anti-Ago2 antibody (Millipore) was applied to RIP assay after the transfection for 48 h.

First, proteins were subjected to electrophoresis with 10% SDS-PAGE and then were transferred onto the PVDF membrane (Millipore Corp., MA, U.S.A.) for 1 h. Then, the membranes were blocked with 5% non-fat milk prepared by washing buffer for 1 h. Next, the membranes were prepared to be incubated with corresponding antibodies against FOXP4 (1:500, Abcam, U.S.A.) to detect the relative protein expression of targets. Anti-β-actin antibody (1:1000, Affinity, U.S.A.) was applied as loading control. After incubating with primary antibody at 4°C overnight, the membranes were washed with tris-buffered saline for 4 times and each time for 5 min. At last, specific corresponding secondary antibodies (1:2000, CST, U.S.A.) were further incubated with the membrane at room temperature. Protein bands were visualized and analyzed using Quantity One 4.6.2 software (Bio-Rad, Hercules, CA, U.S.A.).

SPSS 21.0 software for Windows (SPSS Inc.) was utilized to conduct statistical analysis. Data sets were expressed as mean ± S.E.M. The differences between two groups or more were compared by two-tail t test and χ2 test, respectively. Kaplan–Meier survival curves were analyzed using log-rank tests. High/low circRNA_0058063 expression groups were divided according to the medium expression values. Statistical differences were taken when P < 0.05. Triplicate experiments were performed in any of the assay unless otherwise stated.

A total of 312 circRNAs were differentially expressed (Fold change > 2, P < 0.05) with the analysis of circRNA microarray from GEO database. About 195 circRNAs were up-regulated, and 117 circRNAs were significantly down-regulated in four bladder cancer samples. The top dysregulated circRNAs were presented in the heat map (Figure 1A). Our results indicated that circRNA_0058063’s expression level was augmented in both BC cell lines (Figure 1B) and the tissues (Figure 1C). Kaplan–Meier analysis showed 94 cases of BC patients with the higher expression level of circRNA_0058063 had a relatively low 5-year OS, indicating that circRNA_0058063 expression was negatively correlated with carcinogenesis in individual BC patients (Figure 1D). And it was further confirmed that higher circRNA_0058063 expression contributes to advanced stage and metastasis of the disease (Table 1). Taken together, it potentiates circRNA_0058063’s critical role in the regulation of BC.

Since the expression of circRNA_0058063 was induced in BC, the expression of circRNA_0058063 was silenced in BC to determine the regulatory events. As shown in Figure 2A, silencing of circRNA_0058063 was successfully achieved. Moreover, silencing of circRNA_0058063 inhibited the proliferation rate of 5637 and BIU-87 cell lines (Figure 2B). Transwell assay showed that inhibition of cell invasion by circRNA_0058063 (Figure 2C). Apoptosis assays indicated that inhibition of circRNA_0058063 contributed to elevate cell apoptosis (Figure 2D). Therefore, it would be interesting to conclude that silencing of circRNA_0058063 inhibits both 5637 and BIU-87 cell proliferation and cell invasion, and promotes apoptosis of 5637 and BIU-87 cell lines during BC progress in vitro.

The intracellular residence of circRNA_0058063 was further explored, and it was revealed that circRNA_0058063 was mainly distributed in the cytoplasm (Figure 3A), suggesting that circRNA_0058063 may act as a miRNA sponge to induce miRNAs. Based on the results, the Circular RNA Interactome (https://circinteractome.nia.nih.gov/index.html) was applied to study the underlying mechanism of circRNA/miRNA interactions and we found the complementary site of miR-486-3p in the circRNA_0058063 sequence. (Figure 3B). miR-486-3p was chosen for further studies as it has been shown to a biomarker for prediction and prognosis of BC (Jiang et al.). Furthermore, miR-486-3p expression was mitigated in 5637 and BIU-87 cell lines and BC tissues (Figure 3C,D). Next, the direct binding relationship between miR-486-3p and circRNA_0058063 was confirmed with a luciferase reporter assay. When luciferase reporter gene and the miR-486-3p mimic were co-transfected into the 5637 and BIU-87 cell lines, the luciferase intensities were greatly reduced (Figure 3E). To elucidate whether the binding of circRNA_0058063 and miR-486-3p was direct or indirect, a RIP assay was conducted. It was indicated that miR-486-3p was predominantly resided at the site of MS2bs-circRNA_0058063, potentiating a direct binding of circRNA_0058063 and miR-486-3p (Figure 3F). In summary, our data confirm that circRNA_0058063 can directly inhibit the miR-486-3p expression by directly combining to the binding site.

To investigate whether circRNA_0058063 modifies miR-486-3p to regulate target gene expression, the potential targets of miR-486-3p were predicted by using TargetScan and FOXP4 was selected to be a potential candidate (Figure 4A). FOXP4 has been demonstrated to play an important role in the progression of multiple cancers, such liver cancer and oral squamous cell carcinoma, and breast cancer (Qi et al.; Xu et al.; Ma et al.); however, its function in bladder cancer remains unknown. We detected FOXP4 expression level and found that the expression of FOXP4 was significantly increased in the BC cell lines (Figure 4B) and tissues (Figure 4C). Overexpression of miR-486-3p inhibited the expression of FOXP4 and reversed the up-regulation of FOXP4 induced by transfection with FOXP4-overexpressing plasmid (Figure 4D). Next, the luciferase reporter assay suggested that the luciferase activity of FOXP4 was significantly decreased by the miR-486-3p mimic (Figure 4E). Next, the RIP assay for Ago2 showed circRNA_0058063, FOXP4 and miR-486-3p were mainly enriched to Ago2 (Figure 4F). This result indicated that circRNA_0058063 and FOXP4 were subjected to Ago2-related RISC, and then it promoted their interaction with miR-486-3p. Furthermore, inhibition of circRNA_0058063 lowered the recruitment of Ago2 to circRNA_0058063, but elevated the recruitment of Ago2 enrichment to FOXP4 (Figure 4G), which revealed that circRNA_0058063 can act as a ceRNA and antagonize with FOXP4 to bind with miRNA. In addition, the expression of FOXP4 was detected when circRNA_0058063 was silenced and it was found that FOXP4 expression was significantly reduced, but it was significantly reversed by miR-486-3p LNA (Figure 4H). Taken together, circRNA_0058063 tightly regulated the expression of FOXP4 as a ceRNA by inducing miR-486-3p (Figure 5).

An abundant of researches have explored circRNA’s crucial functions in the occurrence and progression of diverse cancers [17,18]. In lung cancer, the regulatory roles of circRNA have been uncovered in various cell biological processes [19]. During the progression of glioma, circPTN blocked cancer growth and metastasis [20]. Several studies were performed to investigate the function of circRNAs in bladder cancer (Liu et al.; Lu et al.; Liu et al.; Zhang et al.). However, the exploration of circRNAs in BC is still rare. Therefore, we utilized the circRNA expression microarray assay in BC and indicated that circRNA_0058063 was up-regulated in BC. Hsa_circ_0058063 is located at the domain of chr2: 216177220-216213972 according to the human reference genome (GRCh38.p13) and is thought to originate from the gene ATIC, which residues at chromosome 2q35. Therefore, hsa_circ_0058063 was named as ‘circRNA_0058063’. Further investigations demonstrated that silencing of circRNA_0058063 blocked cell proliferation and cell invasion, and promoted cell death, suggesting that circRNA_0058063 can modulate BC progression. Therefore, circRNA_0058063 potentiates its potential as a target for pathological detection and treatment.

The reported researches have indicated that circRNA regulates gene expression through cavernous miRNAs. In colorectal cancer, CCDC66 promotes cancer growth and metastasis through targeting miRNA-33b and miR-93 [21]. Bi et al. proved that circRNA_102171 regulates cell proliferation and apoptosis, and causes malignant transformation by blocking CTNNBIP1 in papillary thyroid carcinoma [22]. miR-486-3p has been reported to have a critical role in several cancers, such as lung, liver, breast cancers (ElKhouly et al.). It has also been reported that lncRNA-PVT1 promotes the proliferation of cervical cancer cells by increasing ECM1 expression via sponging miR-486-3p (Wang et al.). In oral cancer, overexpression of miR-486-3p inhibits oral cancer cell growth and induces apoptosis by down-regulating of DDR1 (Chou et al.). Here, we found that circRNA_0058063 directly bound with miR-486-3p and worked as a miRNA sponge.

It was reported that microtubule-associated proteins (MAPs) mainly reside in the cytoplasmic fraction and bind with tubulin and facilitate their aggregation [23]. Others have suggested MAP’s involvement in the control of mitochondrial balance or apoptosis [24]. Previous studies have reported that FOXP4 overexpresses and regulates tumor cell growth in non-small cell lung cancer [25]. FOXP4 has been certified in the progression of prostate cancer [26]. It has been also reported that FOXP4 enhances the migration and invasion of breast cancer cells via promoting EMT (2019). In the present study, the abundance of FOXP4 was found to be elevated in BC. Moreover, our data demonstrated that FOXP4 was a novel target gene of miR-486-3p. Further exploration suggested circRNA_0058063 acts as a bait for miR-486-3p to modulate FOXP4 expression. Inhibition of circRNA_0058063 decreased FOXP4’s expression, which was overturned by silencing of miR-486-3p.

To sum up, the expression of circRNA_0058063 was elevated in BC. Silencing of circRNA_0058063 blockeds cell proliferation and cell invasion, however promoted cell death. Mechanistically, circRNA_0058063 regulates the expression of FOXP4 by the sponge miR-486-3p. circRNA_0058063 potentiates as a valuable candidate for the early detection and prognosis of BC.

The authors declare that there are no competing interests associated with the manuscript.

The authors declare that there are no sources of funding to be acknowledged.

Conception and intellectual input: Y.T.Y.; Designed and performance of experimentation: L.H.T.; Data collection and manuscript drafting: H.H., L.Y.P. Statistical analyses and data interpretation: L.Y.Y. All authors read and approved the final manuscript.

The protocol of the current investigation was authorized by the Institutional Review Board of the First Affiliated Hospital of Wenzhou Medical University. Written informed consent was obtained from all the participants in the study.

The authors declare that the data supporting the findings of this study are available within the article.

BC

bladder cancer

circRNA

circular RNA

RIP

RNA immunoprecipitation

RNA pol II

RNA polymerase II