MiR-455-3p regulates glioma cell proliferation by targeting PAX6

Purpose: To investigate the role of miR-455-3p in gliomas. Method: Quantitative real-time polymerase chain reaction was used to measure miR-455-3p and paired box 6 (PAX6) levels in glioma cell lines. Western blot analysis was used to determine the expression of cell cycle regulators. In addition to over-expression, silencing of miR-455-3p or PAX6 was performed to study the functions of miR-455-3p in gliomas. Results: The levels of miR-455-3p were significantly up-regulated in glioma cell lines (p < 0.05), while miR-455-3p over-expression increased glioma cell proliferation and interfered with the progress of the cell cycle (p < 0.01). Furthermore, endogenous miR-455-3p silencing prevented glioma cell proliferation by regulating cell cycle progression (p < 0.05).The results also showed that PAX6 controlled the cell cycle while PAX6 silencing selectively regulated p21 expression (p < 0.01). Furthermore, miR-455-3p and PAX6 influenced p53 expression. Re-introduction of PAX6 expressing vector into glioma cells rescued the pro-tumoral effect of miR-455-3p overexpression. Conclusion: These findings demonstrate the role of miR-455-3p as a tumour oncogene in gliomas via regulation of the cell cycle, indicating that miR-455-3p might act as a new treatment strategy for glioma cell tumours and a predictor of survival in glioma patients.


INTRODUCTION
Glioma is a type of malignant tumour of the central nervous system (CNS) that includes four clinical grades that are based on histology and prognosis [1,2]. Due to the inefficiency of current therapeutic approaches such as surgery, chemotherapy, and radiation [3], patients with gliomas have a median survival time of only 15 months and a low average survival rate of 5 years [4,5]. Although the effects of treatment on symptoms have improved in recent years, traditional treatment approaches for gliomas are usually invasive and have poor outcomes. Thus, it is necessary to find new intervention and treatment choices for glioma patients. Recent efforts have focused on new therapy strategies to modify the expression of tumourigenic genes or biochemical signalling pathways [6].
MicroRNAs (miRNAs) comprise a large family of endogenous noncoding RNA molecules with approximately 20 nucleotides [2]. These molecules cause translation inhibition or mRNA degradation, which result in biological effects in the body at the transcriptional and posttranscriptional levels via binding to the 3'untranslated region (3'-UTR) of target mRNAs [7]. Owing to their ability to modulate different target genes, miRNAs have crucial roles in diverse cellular behaviours in tumours, including proliferation, apoptosis, invasion, migration, and even glucose uptake [7]. Abnormal expression of miRNAs is reportedly related to tumour progression, and up-regulation or downregulation of miRNA leads to oncogenic and suppressive effects in tumours including lung cancer, gastric cancer, breast cancer, colorectal cancer, and even glioblastoma [8][9][10]. Moreover, miRNAs are a kind of diagnostic and prognostic index of disease progression [1].
Previous studies found that the abnormal expression of miR-455-3p may be closely associated with temozolomide (TMZ) resistance in the treatment of glioblastoma multiforme (GBM) [11,12]. A miRNA profiling analysis report shows that the expression of 30 miRNAs changes in the presence of GBM [13]. Seventeen miRNAs are down-regulated in migrating tumour cells. Among the most deregulated miRNAs, miR-455-3p is markedly up-regulated [13]. The expression pattern, biological functions, and potential molecular mechanism of miR-455-3p in gliomas are still unknown.

EXPERIMENTAL Cell lines
Human glioma cell lines (SW1783, A172, U87, U251, and H4) and normal human astrocytes (NHA) cell lines were procured from the American Type Culture Collection (ATCC). The cell lines were cultured in modified medium containing 10 % fetal bovine serum and penicillin-streptomycin (Gibco, Grand Island, NY).

Quantitative real-time PCR (qRT-PCR)
TRIzol Reagent (Tiangen, China) was used to extract RNA. Then, total RNA (1 µg) procured using DNase (Life Technologies, Carlsbad, CA), was reverse-transcribed to cDNA by the Kit (Tiangen, China). The qRT-PCR was performed using the SYBR Green PCR Mix (Tiangen, China). The PCR parameters were: 10 min at 95 °C, followed by 40 cycles of 15 s at 95 °C and 15 s at 60 °C. GAPDH was used as an endogenous control, and the U6 level was used as the miRNA control. The primer sequences are shown in Table 1. Cells were collected and lysed 48h after transfection for the luciferase activity assay. Luciferase activity was measured using a dualluciferase reporter assay system (E1910, Promega, USA) according to the manufacturer's instructions.

Western blotting
Total cell proteins were subjected to 15% SDS-PAGE, and then transferred onto polyvinylidene difluoride (PVDF) membranes. Next, membranes were blocked with 5 % non-fat-dried milk. Primary antibodies were added to all membranes and incubated overnight. After incubating with corresponding secondary antibodies (Life Technology), the membranes were scanned. β-Actin was used as an internal control.

MTT assay
U87 and U251 cells were seeded in 96-well plates and subsequently transfected with 100 nM of NC (no silencing), miR-455-3p, or ASO-miR-455-3p. After 48 h, MTT reagent was added to each well that contained 100 mL of medium that had been cultured for 3 days. Following incubation at 37 °C with a humidified 5 % CO 2 for 2 h, the well absorbance at 490 nm was tested.

Cell cycle assay
Cells (1 × 10 6 /well) in a 6 -well plate were transfected with 100 nM of expressing plasmids.
After 48 h of culture, the cells were collected and washed three times with PBS. After fixing with ice-cold 70 % ethanol, the fixed cells were incubated in a mixture of 0.5 % Triton X-100, 1 mg/mL RNase A, and 50 μg/mL propidium iodide (Becton Dickinson and Company, Franklin Lakes, NJ, USA) in PBS at 37 °C for 1 h. Cell cycle was assessed using flow cytometry.

Statistical analysis
Data are expressed as mean ± SEM. Statistical analysis was carried out with SPSS software (version 20.0, Chicago, IL, USA). Statistical differences among groups were compared using a Student's unpaired t-test following the one-way ANOVA method. Values of p < 0.05 were regarded as statistically significant. Statistical analysis was performed using SPSS 20.0.

MiR-455-3p silencing suppressed cell proliferation
To explore the roles of miR-455-3p in cell proliferation, U87 and U251 cells were transfected with ASO-NC, ASO-miR-455-3p, NC vector, or miR-455-3p overexpressing plasmids. Results of qRT-PCR showed an approximately six-fold increase miR-455-3p expression in cells transfected with overexpression plasmids. The miR-455-3p level was significantly decreased in the cells transfected with ASO-miR-455-3p when compared with those transfected with ASO-NC (Figure 2 A). MTT assays demonstrated that miR-455-3p overexpression markedly induced glioma cell proliferation, and that miR-455-3p suppression reversed this effect (Figure 2 B). These results indicate that overexpression of miR-455-3p might promote the proliferation of glioma cells.

PAX6 is a target of miR-455-3p
It is well known that miRNAs function by regulating their downstream target genes. Therefore, TargetScan was used to search for miR-455-3p targets. As predicted, PAX6 was found to be a target of miR-455-3p (Figure 3 A). Western blot assay showed that PAX6 in glioma cell lines was markedly lower than that in astrocyte cell line (Figure 3 B). To confirm the supposition that PAX6 is the target of miR-455-3p, both mRNA and protein levels of PAX6 were evaluated in the U87 and U251 cells with miR-455-3p suppression or overexpression. The data reveal that PAX6 protein and mRNA levels were significantly increased/decreased in the cells with miR-455-3p suppression/overexpression ( Figure  3 C and D). Dual luciferase reporter assay showed that after miR-455-3p overexpression in U87 and U251, the luciferase activity of the cells co-transfected with wild-type 3'-UTR of PAX6 (which have a miR-455-3p binding site), was decreased. This result was not observed in the mutant 3'-UTR of PAX6 or the ASO-miR-455-3p co-transfected group (Figure 3 E). These results indicate that PAX6 were one target of miR-455-3p. (E) Luciferase activity was performed in cells cotransfected with miR-455-3p and wild-type PAX6 3'-UTR or the mutant PAX6 3'-UTR; ■ = NC; □ = miR-455-3p; *p < 0.05 and **p < 0.01 compared between two groups (B); *p < 0.05, and **p < 0.01 compared to the ASO-NC; ( # p < 0.05 and ## p < 0.01 compared with NC vector (C, D and E)

PAX6 regulates the expression of cell cycleassociated proteins in glioma cells
To determine the function of PAX6 on cell cycle in glioma cell lines, PAX6 was knocked down using chemically synthesized PAX6-targeting siRNA (si-PAX6). As expected, results showed si-PAX6-transfected glioma cells had a much lower PAX6 level (Figure 4 A). When treated with si-PAX6, a marked increase in cell proliferation trend was detected in two glioma cell lines (Figure 4 B).The effects of PAX6 silence on cell cycle circuitry detected in glioma cell lines were due to down-regulation of PAX6. Expression levels of Cyclin D1 (CCND1), Cyclin E2 (CCNE2), cyclin-dependent kinases (CDK2, CDK4, and CDK6), and P21 were evaluated in si-PAX6 transfected cells.
The above molecules have been proven to be key modulators of G1 phase [14]. As observed in

Up-regulation of PAX6 reverse miR-455-3p overexpression effects on glioma
To further confirm that miR-455-3p exerted its function by directly targeting PAX6 in the glioma cell lines, miR-455-3p expressing plasmid was transfected together with a PAX6 expressing vector into the two cell lines, followed by a series of functional assays. This finding suggested that miR-455-3p might suppress the p53 expression. First, qRT-PCR and Western blot assay were performed to evaluate whether wild-type p53 was involved in changes induced by miR-455-3p and PAX6 expressions (Figure 5 A). In addition, reintroduction of PAX6 reversed the miR-455-3p overexpression effect on cell proliferation (

DISCUSSION
Glioma is one of the most malignant brain tumours and is common in adults [15,16]. Gliomas possess a subpopulation of selfrenewing cells and cancer stem cells that reproduce even in a poor tumour microenvironment [17]. Currently, therapies targeting cancer stem cells have become popular in the glioma treatment [18]. A growing number of studies have shown that miRNAs and lncRNAs are involved in the occurrence and progression of tumours. These findings may provide promising therapeutic strategies for gliomas.
The miRNAs up-regulated in gliomas always act as oncogenes that exert their impact on cell proliferation through regulating cell cycle systems [19,20]. Previous studies found up-regulated miR-455-3p levels in migrating GBM cells, which could result in elevated migration in vitro [13]. Another study found that miR-455-3p levels are markedly related to cell cycle and invasion in GBM, and miR-455-3pis involved in TMZ resistance in cancer stem cells (+) of GBM patients, providing a new therapeutic target for GBM treatment [11]. Herein, miR-455-3p and PAX6 levels in glioma cell lines were assessed and found to be deregulated in select glioma cell lines. This result implies that the deregulation of miR-455-3p might affect glioma initiation and development.
PAX6 is a member of the PAX family, which participates in encoding a group of paired-box transcription factors in various important bioactivities [21]. Paired box 6 plays a key role as a transcription factor in the development of the central nervous system [22]. Paired box 6 was recently demonstrated in multiple tumours, such as brain, pancreatic, and intestinal tumours [23]. The data show that suppression of PAX6 by miR-455-3p led to an obvious increase in glioma cell proliferation. Luciferase assays proved that miR-455-3p changed gene expression posttranscriptionally by binding to the 3'-UTR of PAX6. Down-expression of PAX6 facilitated cell proliferation, implying that PAX6 functions as an anti-oncogene in glioma. In addition, miR-455-3p and PAX6 co-overexpression reversed the promotion effect caused by miR-455-3p overexpression alone, revealing that miR-455-3p promoted glioma malignant behaviours via suppressing PAX6 expression. This result found that PAX6 is a miR-455-3p target in glioma cells. Using two glioma cell lines, this study further demonstrated that enhancement of miR-455-3p increased cell proliferation. In addition, PAX6 suppression elevated the expression of cyclins, such as cyclin D1, CDK2, and CDK4, in glioma cells. It was still unknown whether upregulation of these cyclins was induced by p21 directly regulated by PAX6. However, the regulation of cyclins and their functions in G1/S transition provided a role of PAX6 involvement in regulating cell cycle. It is possible that PAX6 deficiency in glioma cells resulted in hyperactivation of different cyclin/CDK complexes leading to an increase of DNA replication and even cell mitosis.
Previous studies have reported that p53 could regulate the expression of many miRNAs [24]. Moreover, almost all recent miRNAs studies indicate miRNAs are closely associated with p53 expression [15]. It has been reported that the TP53 tumour suppressor involved in human cancer, also function as a mediator in the tumourigenesis of glioma. p53 was reported to be targeted by miR-25 or miR-32, leading to the down-regulation of both protein and mRNA levels of p53 [25]. Research reports support the conclusion that p53 represses miRNAs through the transcriptional depression of E2F1 and other genes to affect the tumourigenesis process [15].
This study provides evidence of the oncogenic potential role of miR-455-3p in inducing cell growth. In addition, a network mechanism involving miR-455-3p directly co-regulating other factors in a p53-dependent manner through two transcriptional factors, E2F1 and MDM2, is described in this study. These results strongly suggest that PAX6 may function as a regulator controlling the expression level of CDK suppresser, CDKs, and cyclins in the G1/S transition. Taken together, these results bring to light the idea that the miR-455-3p-PAX6-p53 signalling plays a crucial role in human glioma.

CONCLUSION
This study shows that miR-455-3p increases cell proliferation in human glioma in vitro, and that microRNA-455-3p may function by directly targeting PAX6 and abnormal actions of PAX6, thereby disturbing homeostasis by dysregulating the cell cycle in glioma. Furthermore, the effect of abnormal regulation of miR-455-3p on cell cycle gene transcription plays a crucial role in the development of glioma. Further identification of the mechanism and function of miR-455-3p may contribute to understanding the effect of miRNAs on potential mitogenic functions in the development of glioma, thus providing new ideas for therapeutic targets.