Circular RNA circCCDC66 promotes glioma proliferation by acting as a ceRNA for miR-320a to regulate FOXM1 expression

Samples

12 pairs of glioma samples and corresponding nearby control of surgical resection of patients ranging from January 2012 to January 2014 were derived from our hospital. All samples were preserved at -80° C until RNA extraction. Our experiments were approved by the Ethics Committee of our hospital, and all the subjects were unanimously signed informed consent.

Cell lines

Normal human astrocyte cell (HEB) and human glioma cells (U87, SW1783 and U373) were derived from ATCC or our lab. HEB U87 and U373 cells were cultured in DMEM medium (Gibco, USA) containing 10% FBS (TransGen Biotech, China). SW1783 cells were cultured in RPMI-1640 (Hyclone, USA) medium supplied with 10% FBS (TransGen Biotech, China). All the indicated cells were under a 37° C incubator containing 5% CO₂.

RNA extraction and quantification

Trizol (Takara, China) was used to extract total RNA from cells and tissues as described in the manual. Reverse transcription system was performed as follows: 10 μL volume included 1mg RNA with TransScript First-Strand cDNA Synthesis SuperMix (TransGen Biotech, China), followed by subjecting to qRT-PCR with indicated primers via TransStart Green qPCR SuperMix (TransGen Biotech) on Applied Biosystems 7500 System (Applied Biosystems, USA). We finally calculated relative RNA expression by the 2^−ΔΔCt method.

Cell transfection

SiRNAs targeting circCCDC66 were synthesized by GenePharma (Shanghai, China). MiRNA mimics, inhibitor, mimics NC, inhibitor NC, and siRNAs were provided by GenePharma. pcDNA3.1 circRNA mini construction comprising circCCDC66 was a highly expressed cassette, and pcDNA3.1 including FOXM1 indicated over-expressed plasmid. The above-mentioned plasmids were separately transfected into indicated cells by Lipofectamine 3000 (Invitrogen) as manual instructions. The sequence was as follows: si-circCCDC66, CAAUUAGAGCAUCAGGAAA; si-NC, UUCUCCGAACGUGUCACGUTT.

CCK-8 assay

Cell proliferation assay was conducted by Cell Counting Kit (Dojindo, Japan) at 24 h post-transfection. 2000 cells were re-seeded in 96-well per well and then added 10ul of CCK-8 solution at the indicated time for 2 h incubation. Cell proliferation was measured at specified days. Absorbance value in 450 nm was detected by the microplate reader (BioTek Instruments, USA) [14].

Transwell assay

8μm pores of transwell chamber (Costar, USA) with 8.0 mm pore size polycarbonate membrane (Corning, USA) and Matrigel film (BD Biosciences) for coating upper chambers were applied to conduct invasion assay. At the same time, the upper chamber was a pre-coated membrane without matrix gel. The cells were placed in the upper chamber for the migration assay. The treatment chamber including 200μl of medium absence serum and lower chamber with 600μl of medium containing 10% FBS had seeded 1×10⁵ cells as indicated at 37° C with 5% CO₂ overnight. On the following day, non-immigrated or non-invasive cells on the top side were eliminated by cotton swabs. The insert was fixed by methanol for 20 minutes and then dyed by DAPI at the concentration of 10ug/mL for 15 minutes. We counted migrated or invade cells to the membrane bottom and captured images by microscope and calculating the number of cells from three independent experiments in triplicate.

Luciferase assay

A circular RNA Interactome database was applied to predict the downstream target of miRNA. The circCCDC66 sequence or the wild-type and mutant binding sites of miR-320a or miR-320a in FOXM1 3'UTR were subcloned into pmirGLO dual-luciferase reporter to construct circCCDC66-Wt/Mut and FOXM1-Wt/Mut, and co-transfect the plasmid with miR-320a mimic or miR-320a mimic into SW1783 or U373 cells. The association between circCCDC66/ FOXM1 3’-UTR or miR-320a was detected by luciferase assay. Relative luciferase activity was determined by dual-luciferase reporter assay (Promega Corporation, USA) [15]. The luciferase reporter analysis required a biological triple repeat.

Statistical analysis

The representative data were shown as mean ± SD. All the data was calculated after three independent experiments, followed by SPSS 16.0 software analysis (IBM, USA). The difference that existed in two comparison groups or multiple groups in indicated experiments was determined by Student's t-test. The differences between tumor and normal tissues were counted by paired samples t-test. Correlation between circCCDC66 and clinical-pathological profile was determined by Fisher's exact test [16–20]. The log-rank test was applied to inspect the difference between highly and lowly expressed circCCDC66 groups. The obvious difference was indicated as the P-value was less than 0.05 [21–23].

Enhanced circCCDC66 in promoted glioma cell proliferation

As far as we know, circCCDC66 has not been reported in gliomas. This study aims to explore whether circCCDC66 affects gliomas and their mechanisms. To begin our research, we used quantitative real-time polymerase chain reaction to detect the expression of circCCDC66. qRT-PCR data demonstrated circCCDC66 level was increased in glioma samples compared to that in the normal group (Figure 1A). A similar result was observed in glioma cell lines. Compare to a normal cell, the most heightened level of circCCDC66 was shown in SW1783, while the lowest level of that was exhibited in U87 (Figure 1B). Our data suggested that circCCDC666 perhaps exhibited considerable function during the cancerous process of glioma. Then, we performed a functional analysis to study circCCDC66 in glioma cells. SiRNAs specifically targeting circCCDC66 were used in this part. Our data illustrated siRNA demonstrated strong efficiency in SW1783 and U373 cells (Figure 1C, 1D). CCK-8 assay displayed suppressed circCCDC66 retarded both SW1783 and U373 cell growth (Figure 1E, 1F).

Cell migration and invasion were promoted by circCCDC66

Furthermore, cell migration and invasion were then detected using transwell migration and invasion assay. Our results showed that the knockdown of circCCDC66 significantly suppressed cell metastasis in both SW1783 and U373 cells. The migrated and invaded SW1783 cells were remarkably reduced by 83% and 47% after silencing circCCDC66 compared to the control group (Figure 2A, 2B). Meanwhile, the migrated and invaded U373 cells were remarkably reduced by 91% and 75% after silencing circCCDC66 compared to the control group (Figure 2C, 2D).

Construction of circCCDC66-miRNA interacting network

To predict the binding miRNAs of circCCDC66 in gliomas, we applied bioinformatics analysis using RegRNA 2.0 database. Eight miRNAs, including hsa-miR-320a, hsa-miR-518c-5p, hsa-miR-320b, hsa-miR-198, hsa-miR-211-5p, hsa-miR-127-3p, hsa-miR-378a-5p, hsa-miR-138-5p, were identified as the potential targets of circCCDC66. In order to analyze the expression pattern of these miRNAs in glioma samples, we used CGGA database. As presented in Figure 3, we found that Compared with normal samples, circCCDC66 in gliomas was significantly up-regulated. However, circCCDC66 had no significant difference in different WHO grades. hsa-miR-198 was down-regulated in WHO III/IV glioma samples compared to WHO II glioma samples. MiR-320a/b was down-regulated in WHO IV glioma samples compared to WHO II/III glioma samples. However, we found hsa-miR-127-3p and hsa-miR-211-5p were up-regulated in advanced stage glioma samples. The present study focused on hsa-miR-320a/b, which had been reported to be a tumor suppressor in cancers.

Bioinformatics analysis of circCCDC66-miR-320a/b axis

Then, we constructed a circCCDC66-miR-320a/b-mRNAs network in gliomas using Starbase datasets, which was showed in Supplementary Figure 1A. A total of 2142 targets of miR-320a and 1882 targets of miR-320b were included in this network. Meanwhile, 1864 targets were identified to be targets of both miRNAs (Supplementary Figure 1B). The interaction among these proteins was also revealed using PPI network analysis (Supplementary Figure 1C). GO analysis showed circCCDC66 was involved in regulating Protein metabolism, Regulation of nucleobase, nucleoside, nucleotide, and nucleic acid metabolism, RNA metabolism, Regulation of gene expression, epigenetic, Regulation of translation, Protein targeting, Mitosis, Morphogenesis (Supplementary Figure 1D). KEGG pathway analysis demonstrated that circCCDC66 was related to the regulation of TRAIL signaling pathway, IGF1 pathway, signaling events mediated by Hepatocyte Growth Factor Receptor (c-Met), PDGF receptor signaling network, signaling events mediated by VEGFR1 and VEGFR2, VEGF and VEGFR signaling network, EGFR-dependent Endothelin signaling events, Sphingosine 1-phosphate (S1P) pathway, Proteoglycan syndecan-mediated signaling events (Supplementary Figure 1E).

Identification of key targets of circCCDC66-miR-320a/b axis

Then, we aimed to identify key targets of the circCCDC66-miR-320a/b axis using Mcode analysis in Cytoscape software. Three key PPI networks were identified.

The network 1 included 68 nodes (Supplementary Figure 2A) (SF3B1, TRIM36, RBMX, NCBP2, TRIM41, SRSF7, UBE2C, PAPOLA, HNRNPF, AREL1,). The network 2 included 40 nodes (Supplementary Figure 3A) (STAG2, RAB5B, TFRC, CLASP1, CLIP1, SH3GL1, SKA2, EPN2, PAFAH1B1, ADRB2) and the network 3 included 93 nodes (Figure 4A) (FOXM1, ASF1B, CLCN6, MRTO4, TRAK1, PPP3CA, EXO1, TRIM24, CCNA2, NAT10).

Bioinformatics analysis showed network 1 was enriched in protein ubiquitination involved in ubiquitin-dependent protein catabolic process, fibroblast growth factor receptor signaling pathway, mRNA cis splicing, DNA-templated transcription, termination (Supplementary Figure 2B). network 2 was enriched in regulation of dendritic spine morphogenesis, Arp2/3 complex-mediated actin nucleation, sister chromatid cohesion (Supplementary Figure 3B). network 3 was involved in regulating calcineurin-NFAT signaling cascade, cellular response to ionizing radiation double-strand break repair, DNA damage checkpoint, DNA duplex unwinding, DNA recombination, double-strand break repair via nonhomologous end joining, ERCC4 double-strand break repair via, homologous recombination, maturation of SSU-rRNA, mitotic cytokinesis, mitotic DNA damage checkpoint regulation of DNA repair, mitotic nuclear envelope disassembly, mRNA export from the nucleus, mRNA transport, nucleus organization, regulation of cellular response to heat, regulation of cytokinesis, regulation of DNA recombination, regulation of protein dephosphorylation, regulation of response to DNA damage stimulus, regulation of telomere maintenance, response to gamma radiation, response to ionizing radiation, response to X-ray, somatic recombination of immunoglobulin gene segments, telomere maintenance via recombination, telomere maintenance via telomere, tRNA export from the nucleus (Figure 4B).

CircCCDC66 could regulate FOXM1 expression by sponging miR-320a

Previous researches have shown five discovered miRNAs could be sponged by circCCDC66. Here, our results revealed that circCCDC66 was a probable interactor with eighteen miRNAs. Based on that, miR-320a was selected as a candidate to explore the interaction with circCCDC66. Figure 5A preliminary displayed the links between circCCDC66 and miR-320a. Luciferase assay demonstrated that overexpression of miR-320a reduced the luciferase activity of circCCDC66-WT vector, however, it did not affect the luciferase activity of circCCDC66-Mut vector (Figure 5B).

Through database analysis, we found that 1864 targets were identified as miR-320a/b targets, among which FOXM1 has a higher degree. We chose FOXM1 for the next analysis. Very interestingly, we found that FOXM1 was a target of CircCCDC66/miR-320a/b and involved in regulating DNA damage response pathway. Thus, we performed bioinformatics analysis to infer probably occurred interaction between FOXM1 and miR-320a. The qRT-PCR analysis revealed miR-320a mimics lowered FOXM1 mRNA expression (Figure 5C), but the effects could be reversed by miR-320a inhibitor (Figure 5D). Luciferase assay demonstrated that overexpression of miR-320a reduced the luciferase activity of the FOXM1-WT vector, however, it did not affect the luciferase activity of FOXM1-Mut vector (Figure 5E).

Further data exhibited FOXM1 mRNA level was down-regulated when ablating circCCDC66. Co-transfection of si-circCCDC66 and miR-320a inhibitor would partly reverse FOXM1 mRNA expression level, respectively in both SW1783 and U373 cells (Figure 5F, 5G). The CCK-8 assay results showed that the inhibitory effect of si-circCCDC66 on the proliferation of SW178 and U373 cells was attenuated by miR-320a inhibitor (Figure 6A, 6B). The results of the Transwell assay showed that the inhibitory effect of si-circCCDC66 on the migration and invasion of SW178 and U373 cells was attenuated by the miR-320a inhibitor (Figure 6C–6F). These data together revealed that circCCDC66 promoted cell proliferation, migration and invasion by negatively regulating the expression of miR-320a.

CircCCDC66 promotes cell progression via FOXM1 and miR-320a pathways in glioma

Figure 7A demonstrated co-transfection of inh-miR-320a or FOXM1 plasmid into U373 cells could rescue the consequence of si-circCCDC66-induced cell growth suppression to some extent. Transwell analysis revealed down-regulation of miR-320a or overexpression of FOXM1 could recover the ability of cell migration and invasion compared with the si-circ group (Figure 7B, 7C). These results suggested that circCCDC66 promotes cell progression via FOXM1 and miR-320a pathways in glioma cells.