Knockdown of LncRNA SCAMP1 suppressed malignant biological behaviours of glioma cells via modulating miR-499a-5p/LMX1A/NLRC5 pathway

2.1 Clinical specimens

A total of 31 glioma specimens and six normal brain tissues (NBTs) were obtained from the Department of Neurosurgery, Shengjing Hospital of China Medical University. NBTs were the rejected materials obtained from surgeries of severe intracerebral haemorrhage, brain trauma and epilepsy. After surgical resection, all human tissues were immediately frozen in liquid nitrogen for long-term preservation. Glioma tissues were classified into four grades according to the WHO classification of tumours in the CNS, then divided into two groups: low-grade glioma group (Grade I–II, 15) and high-grade glioma group (Group III–IV, 16). This research had obtained approval of Shengjing Hospital Ethics Committee.

2.2 Cell lines and cultures

Normal human astrocyte (NHA) cells were purchased from Sciencell Research Laboratories (Carlsbad, CA) and human glioma U87, U251 and HEK293T cells were purchased from Shanghai Institutes for Biological Sciences Cell Resource Center. NHA cells grown in astrocyte medium and U87, U251, HEK293T cells were cultured in Dulbecco's modified Eagle medium (DMEM)/high glucose mixed with 10% foetal bovine serum (Gibco, Carlsbad, CA). All cells were maintained in a humidified incubator at 37°C with 5% CO2.

2.3 Cell transfection

Short-hairpin RNA (shRNA) against SCAMP1, LMX1A and NLRC5 gene, as well as their non-targeting sequences were constructed in pGPU6/GFP/Neo vector from GenePharama (Shanghai, China). Full-length SCAMP1, LMX1A and NLRC5 gene was constructed in pIRES2-EGFP (GenScript, Piscataway, NJ). miR-499a-5p agomir and antagomir, their respective non-targeting sequence were synthesized from GenePharma. After cells reached 70-80% confluence in a 24-well plate, Lipofectamine 3000 and Opti-MEM I (Life Technologies, Waltham, MA) were used to transfect cells. Then stable transfected cells were selected by G418 (Sigma-Aldrich, St Louis, MO) screening. The transfection efficacy was detected by qRT-PCR or Western blot.

2.4 RNA isolation and quantitative RT-PCR (qRT-PCR)

Total RNA was extracted from the tissues and cells according to Trizol reagent (Life Technologies Corporation, Carlsbad, CA) manufacturer's instructions. SYBR Prime-Script RT-PCR Kit (TakaraBio, Japan) was used to detect the mRNA expression levels of SCAMP1, LMX1A, NLRC5 and GAPDH. TaqMan MicroRNA Reverse Transcription kit and TaqMan Universal Master Mix II (Applied Biosystems, Foster City, CA) were used to detect the expression levels of miR-499a-5p and U6. After normalizing to the endogenous control GAPDH or U6, the relative expression values were calculated to represent fold change in gene expression by relative quantification (2^−ΔΔCt method).

2.5 Western blot analysis

Equal amounts of protein samples were separated using 8% or 10% sodiumdodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and then electrophoretically transferred to a polyvinylidene difluoride membrane (Millipore, Shanghai, China). After blocking 2 hours at room temperature, the membranes were incubated overnight at 4°C with primary antibodies as below: LMX1A (1:1000, abcam, ab31006, EUGENE, USA), NLRC5 (1:1000, abcam, ab105411, EUGENE, USA), β-catenin, c-Myc, cyclin D1, MMP-7 (1:1000, Cell Signaling Technology, Beverly, MA) and GAPDH，β-actin (1:5000, Proteintech, China). After incubated 2 hours at room temperature with HRP-conjugated secondary antibodies, protein bands were visualized using ECL (Beyotime) and detected using ECL Detection Systems (Thermo Scientific, Beijing, China). The relative expression was calculated based on the internal control GAPDH or β-actin.

2.6 Cell proliferation assay

U87 and U251 cells were seeded in 96-well plate (2000 cells per well) and cultured at 37°C for 24 hours. Then 10 μL of Cell Counting Kit-8 (CCK-8, Dojin, Japan) solution was added into each well and incubated at 37°C for another 2 hours. The absorbance was measured and recorded at 450 nm.

2.7 Cell migration and invasion assays

For migration assay, U87 and U251 cells were resuspended in serum-free medium at a density of 2 × 10⁵ cells/mL, then 100 μL cell suspension was seeded into the upper chamber of a 24-well transwell chamber (Costar, Corning, NY), meanwhile 600 μL DMEM/high glucose medium containing 10% FBS was added to the lower chamber. After incubation at 37°C for 24 hours, cells were fixed with methanol and stained with 20% Giemsa solution for 30 minutes at 37°C. Then cell numbers were counted by calculating the average of five random fields under an inverted microscope. For invasion assay, 80 μL of 50 ng/μL Matrigel solution (BD, Franklin Lakes, NJ) was pre-coated on the transwell membrane.

2.8 Apoptosis analysis

Annexin V-PE/7-AAD (BD, Biosciences) was employed to detect cell apoptosis. According to the manufacturer's instruction, U87 and U251 cells were collected and washed with cold phosphate-buffered saline twice, then stained 15 minutes with Annexin V-PE/7AAD at room temperature. The apoptotic fractions were then analysed using flow cytometry (FACScan, BD Biosciences).

2.9 Dual-luciferase reporter assays

The predicted miR-499a-5p binding sequence in SCAMP1 and LMX1A 3' UTR sequence and their corresponding mutant sequence were cloned into pmirGLO Dual-Luciferase Vector to construct luciferase reporter vector (SCAMP1-Wt or -Mut and LMX1A-3' UTR-Wt or -Mut; GenePharma). HEK293T cells were seeded into a 96-well plate and co-transfected with the indicated vectors and miR-499a-5p agomir or its negative control respectively. Dual-luciferase assay was performed 48 hours after transfection, the relative luciferase activity was normalized to Renilla luciferase activity and calculated by Dual-Luciferase Reporter Assay System (Promega, Madison, WI).

2.10 Human miRNA microarray analysis

The sample preparation, microarray hybridization and miRNAs analysis were performed by Kangchen Bio-tech (Shanghai, China).

2.11 RNA immunoprecipitation

EZ-Magna RIP kit (Millipore, Billerica, MA) was applied to conduct RNA immunoprecipitation (RIP) assay according to the manufacturer's protocol. Whole cell lysate was incubated with RIP magnetic beads conjugated with anti-Argonaute2 antibody and normal mouse IgG (Millipore). After incubating with Proteinase K buffer, immunoprecipitated RNA was obtained. Finally, the presence of the binding targets was validated by qRT-PCR.

2.12 Chromatin immunoprecipitation

Simple ChIP Enzymatic Chromatin IP Kit (Cell Signaling Technology) was utilized for Chromatin immunoprecipitation (ChIP) assay according to the manufacturer's protocol. Briefly, cells were crosslinked with EBM-2 containing 1% formaldehyde for 10 minutes and collected in lysis buffer. Then chromatin was digested by micrococcal nuclease. The immunoprecipitation sample was incubated with 3 μg of anti-LMX1A antibody or normal rabbit IgG, then gently rotated overnight at 4°C after treating with Protein G Agarose Beads. DNA crosslinks were reversed by 5 mol/L NaCl and proteinase K and were finally purified. DNA was amplified by PCR with respective primers.

2.13 Tumour xenografts in nude mice

Stable expression glioma cells for in vivo study was established as following steps. Lentivirus encoding miR-499a-5p was generated using pLenti6.3/V5eDEST Gateway Vector Kit (Life Technologies Corporation), while shRNA targeting human SCAMP1 was ligated into the LV3-CMV-GFP-Puro vector (GenePharma). After infection, the stable expressing cells of pre-miR-499a-5p and sh-SCAMP1 were established. Subsequently, the miR-499a-5p lentivirus was transduced in stable expressing cells of sh-SCAMP1 to produce pre-miR-499a-5p + sh-SCAMP1 cells. All research methods were conducted strictly in accordance with the protocol of Care and Use of Laboratory Animals, moreover, approvals from the Administrative Panel on Laboratory Animal Care of the Shengjing Hospital also were achieved. Four-week-old BALB/C athymic nude mice were obtained from the National Laboratory Animal Center (Beijing, China). The nude mice were divided into four groups (randomized to each group by two performers in a blinded manner, n = 10 per group): Control, sh-SCAMP1, pre-miR-499a-5p and sh-SCAMP1 + pre-miR-499a-5p group. For subcutaneous implantation, 3 × 10⁵ cells were subcutaneously injected in the right flanks of the mice. Tumour volume was measured every 4 days using the formula: volume (mm³) = length × width²/2. The tumour-bearing mice were executed and tumours were isolated at 40 days after injection. For survival analysis in orthotopic inoculations, 3 × 10⁵ cells were stereotactically implanted into the right striatum of the mice. The number of survived nude mice was recorded and survival analysis was determined using Kaplan-Meier survival curve.

2.14 Statistical analysis

Experimental data are presented as the mean ± standard deviation (SD) from at least three independent experiments. SPSS 22.0 software was used for statistical analysis with the Student's t test, one-way ANOVA, Pearson chi-square test or Log-rank test. Differences were considered statistically significant when P < 0.05.

3.1 SCAMP1 was up-regulated in glioma and overexpressed SCAMP1 was correlated with poor prognosis of glioma patients

The expression levels of SCAMP1 in NBTs, glioma tissues, NHA and glioma cell lines were detected by qRT-PCR. Results showed that SCAMP1 was significantly up-regulated in glioma tissues and cell lines, moreover, its expression level was positively correlated with the pathological grades of glioma (Figure 1A-C). Furthermore, Kaplan-Meier survival analyses and log-rank test in 31 glioma patients suggested that higher SCAMP1 expression indicated poorer overall survival (Figure 1B). The correlation analyses between SCAMP1 expression levels and clinicopathological features of 31 glioma patients are also displayed in Table S1. For deeply investigating the functions of SCAMP1 in glioma, we up-regulated and down-regulated the expression of SCAMP1 in glioma cells, then the effects on malignant biological behaviours were detected by CCK-8 assay, flow cytometry analysis and transwell assay. As shown in Figure 1D, knockdown of SCAMP1 significantly inhibited the proliferation ability while facilitated apoptosis of glioma cells compared with the sh-NC group (Figure 1E). Moreover, reduction in SCAMP1 exhibited weaker cell migration and invasion ability than sh-NC group (Figure 1F). On the contrary, up-regulation of SCAMP1 remarkably enhanced the proliferation, migration and invasion ability, meanwhile decreased apoptosis of glioma cells (Figure S2A-C). Data above suggested that SCAMP1 acted as an oncogene in glioma cells.

3.2 miR-499a-5p was down-regulated and manifested a tumour suppressor in glioma

miRNAs microarray analysis was employed to identify that miR-499a-5p was remarkably up-regulated in glioma cells when SCAMP1 knockdown, implying miR-499a-5p may participate in SCAMP1-induced modulation on glioma cells (Figure S1). Subsequently, as shown in Figure 2A-C, miR-499a-5p expression was proven to be significantly down-regulated in glioma tissues, U87 and U251 cell lines compared with NBTs and NHA respectively. Similarly, the expression level of miR-499a-5p was negatively correlated with the progression of glioma pathological grade. Moreover, Kaplan-Meier survival analyses demonstrated that lower miR-499a-5p expression levels led to poorer prognosis of glioma patients and were associated with advanced pathological grade in 31 glioma patients (Figure 2B and Table S2). To assess the functions of miR-499a-5p in glioma, U87 and U251 cells were transfected with miR-499a-5p-agomir or antagomir. As shown in Figure 2D-F, up-regulation of miR-499a-5p significantly decreased the proliferation, migration and invasion ability, meanwhile enhanced apoptosis of glioma cells compared with the pre-NC group. These evidence above inferred that miR-499a-5p exerted tumour-suppressive functions in glioma cells.

3.3 miR-499a-5p targeted SCAMP1 and its expression was negatively regulated by SCAMP1

Using bioinformatics database (Starbase), SCAMP1 was identified as a potential target of miR-499a-5p. To clarify whether SCAMP1 could bind to miR-499a-5p through the putative binding site, the expression of miR-499a-5p in U87 and U251 cells transfected with sh-SCAMP1 was firstly detected. Results showed that the miR-499a-5p expression was significantly up-regulated after SCAMP1 knockdown (Figure 3A). Conversely, the SCAMP1 expression in glioma cells transfected with miR-499a-5p agomir was remarkably decreased (Figure 3B). Dual-luciferase assay was further quantified the interaction between SCAMP1 and miR-499a-5p. As shown in Figure 3C-D, the luciferase activity in SCAMP1-Wt + pre-miR-499a-5p group was considerably attenuated when compared with that in SCAMP1-Wt + pre-NC group. In addition, RIP assay showed that the expressions of SCAMP1 and miR-499a-5p were both increased in the anti-Ago2 group compared with that in anti-normal group, suggesting SCAMP1 and miR-499a-5p were in an RNA-induced silencing complex (RISC) (Figure 3E). Therefore, results above indicated that there was a reciprocal repression feedback loop between SCAMP1 and miR-499a-5p.

3.4 miR-499a-5p mediated the tumour-suppressive effects of SCAMP1 knock-down on glioma cells

To investigate whether miR-499a-5p could regulate the effects of SCAMP1 inhibition on glioma cells, U87 and U251 cells were co-transfected with sh-SCAMP1 and miR-499a-5p agomir or antagomir. The results of CCK-8 assay, transwell assay and flow cytometry analysis indicated that sh-SCAMP1 + pre-miR-499a-5p group remarkably led to a decrease in proliferation, migration and invasion of glioma cells, while facilitated cell apoptosis compared with the sh-NC + pre-NC group (Figure 3F-H). Moreover, down-regulation of miR-499a-5p obviously rescued the inhibitory effects of SCAMP1 knockdown on biological behaviours of glioma cells (Figure 3F-H).

3.5 Knockdown of LMX1A restrained cell proliferation, migration and invasion of glioma cells while promoting apoptosis

Bioinformatics database (TargetScan) was used to identify that LMX1A was a potential downstream target of miR-499a-5p. To confirm our hypothesis, the expression levels of LMX1A in glioma tissues and cells were detected. Western blot assay showed that LMX1A was up-regulated in glioma tissues and cell lines compared with NBTs and NHA, moreover, the expression level was positively correlated with pathological grades of glioma (Figure 4A-B). Remarkably, knockdown of LMX1A significantly restrained the biological behaviours of glioma cells compared with LMX1A(-)-NC group (Figure 4C-E). These evidence indicated that LMX1A played a cancerogenic role in malignant progression of glioma cells.

3.6 miR-499a-5p directly targeted LMX1A-3′-UTR and suppressed LMX1A-induced malignant biological behaviour in glioma cells

The previous results uncovered that LMX1A exerted cancerogenic functions in glioma cells, however whether LMX1A was involved in SCAMP1 and miR-499a-5p regulatory malignant progression of glioma cells remains blurry. Bioinformatics database (Targetscan) speculated that miR-499a-5p may target LMX1A-3′-UTR. Hence the LMX1A expression of glioma cells altered SCAMP1 and miR-499a-5p expression was firstly detected. Results showed that SCAMP1 knockdown or miR-499a-5p overexpression obviously inhibited LMX1A mRNA and protein expression (Figure 5A-D). Furthermore, the expression of LMX1A in glioma cells co-transfected with sh-SCAMP1 and miR-499a-5p agomir or antagomir was examined. Results showed that down-regulation of miR-499a-5p expression dramatically reversed the suppressive effect on LMX1A by SCAMP1 knockdown in glioma cells (Figure 5E). To deeply clarify the interaction between miR-499a-5p and LMX1A, dual-luciferase assay was carried out (Figure 5F-G). The results above demonstrated that miR-499a-5p directly targeted the specific binding sequence in the LMX1A-3′-UTR.

To further explore whether LMX1A could mediate the tumour-suppressive effects of miR-499a-5p, down-regulated LMX1A by pre-miR-499a-5p was reversed prior to the assessment of the malignant biological behaviours in glioma cells. The results showed that overexpressed LMX1A significantly rescued the inhibitory effect of overexpressed miR-499a-5p on cell proliferation, migration and invasion, moreover diminished cell apoptotic percentage (Figure 5H-J). These findings above indicated that miR-499a-5p suppressed the malignant progression of glioma cells via down-regulating LMX1A expression.

3.7 NLRC5 was involved in LMX1A-mediated modulation on malignant progression of glioma cells and exerted an oncogenic role via activating the Wnt/β-catenin signalling pathway

NLRC5 was inferred as a downstream gene of LMX1A by the bioinformatics databases (DBTSS HOME and JASPAR). To ascertain our hypothesis, the expression levels of NLRC5 in glioma tissues and cells were firstly detected. As shown in Figure 6A-B, NLRC5 expression was significantly up-regulated in glioma tissues and glioma cell lines compared with NBTs and NHA and the expression level was positively correlated with pathological grades of glioma. In addition, NLRC5 knockdown robustly inhibited the ability of cell proliferation, migration and invasion, while augmented cell apoptosis versus sh-NLRC5-NC group (Figure 6C-E). Conversely, up-regulation of NLRC5 expression remarkably enhanced the malignant biological behaviours of glioma cells (Figure S3A-C). These evidence above suggested that NLRC5 functioned oncogenic role in glioma cells.

In the further studies, the expression of NLRC5 in glioma cells treated with altering miR-499a-5p and LMX1A expression was inspected. As shown in Figure 6F-G, LMX1A knockdown obviously reduced NLRC5 expression. Furthermore, overexpression of LMX1A rescued the tumour-suppressive impacts of overexpressed miR-499a-5p on glioma cells via up-regulating NLRC5 expression (Figure 6H). Then ChIP assay deeply affirmed the interaction between NLRC5 and LMX1A. The promoter sequence of NLRC5 was established according to the bioinformatics database (DBTSS HOME). The potential binding region in the promoter was identified when scanning the DNA sequence in the −2000 to +200 bp region of the transcription start site (TSS). As a negative control, PCR was conducted to amply the region 1000 bp upstream of the putative NLRC5 binding sequence. As ChIP results corroborated, there was a direct interaction between LMX1A and NLRC5, while no association was found between LMX1A and the control region (Figure 6I). Overall, evidence above revealed that miR-499a-5p suppressed the expression of NLRC5 by reducing LMX1A in glioma cells.

Accumulating reports suggested that deregulation of Wnt/β-catenin signalling pathway is considered as an important marker in various human tumours.23 Nevertheless, whether Wnt/β-catenin signalling pathway was involved in NLRC5-induced malignant progression of glioma cells remains elusive. To deeper reveal the molecular mechanism of the NLRC5 oncogenic roles, the activity of Wnt/β-catenin signalling pathway was examined. As shown in Figure 6J, the expression of β-catenin, c-Myc, cyclin D1 and MMP-7 was significantly down-regulated after NLRC5 was knockdown. Furthermore, overexpressed NLRC5 remarkably up-regulated the activity of Wnt/β-catenin signalling pathway (Figure S3D). Consequently, data above determined that NLRC5 accelerated malignant progression of glioma cells via enhancing the Wnt/β-catenin signalling pathway.

3.8 Down-regulation of SCAMP1 combined with overexpression of miR-499a-5p inhibited tumour xenograft growth while exhibited higher survival ability

In vivo study further ascertained the above findings. The results showed that the tumour volumes of sh-SCAMP1, pre-miR-499a-5p and sh-SCAMP1 + pre-miR-499a-5p group were smaller than the control group. Additionally, down-regulation of SCAMP1 combined with overexpression of miR-499a-5p presented the smallest tumour volume among all groups (Figure 7A-B). As shown in Figure 7C, the survival analysis in orthotopic inoculations assay manifested that the mice in the sh-SCAMP1, pre-miR-499a-5p and sh-SCAMP1 + pre-miR-499a-5p groups obtained longer survival time compared with control group, moreover the mice treated with sh-SCAMP1 + pre-miR-499a-5p had the longest survival time.