Microarray data
Three expression profile datasets [GSE90604 (GPL17692 platform), GSE65626 (GPL17586 platform), and GSE116520 (GPL10558 platform)] and two microRNA expression profile datasets [GSE65626 (GPL19117 platform) and GSE90604 (GPL21572 platform)] were downloaded from the National Center of Biotechnology Information Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/) database. The lncRNA transcriptome data for GBM were downloaded from The Cancer Genome Atlas (TCGA; https://cancergenome.nih.gov/) database; these data included 5 cases of para-cancerous tissues and 156 cases of tumor tissues. The GSE90604 dataset included 16 GBM tissue samples and 7 healthy brain tissue samples; GSE65626 included 6 GBM and 6 healthy brain tissue samples; and GSE116520 included 17 GBM tumor core tissue samples, 17 GBM peritumoral brain zone tissue samples, and 17 healthy brain tissue samples. Two GEO datasets (GSE65626, GSE90604) with miRNA and mRNA simultaneously detected by Affymetrix microarray were included to reduce variability between samples. Data were normalized using limma software and the edgeR package[17].
Identification of differentially expressed genes (DEGs)
The three expression profiling datasets used for analysis included CEL format files. These datasets of differential RNA expression in GBM were compared with data from normal brain tissue using the limma software package in R. To establish statistical significance, the GSE90604 (miRNA and mRNA) and TCGA datasets were filtered using false discovery rate (FDR) < 0.01 and |fold-change (FC)| > 1 as cutoffs. GSE65626 (miRNA and mRNA) was filtered using cutoffs of p < 0.05 and |FC| > 1. GSE116520 (mRNA) was filtered using cutoffs of FDR < 0.01 and |FC| > 1.5. A Venn diagram was constructed using Venny 2.1.
Gene ontology (GO) and pathway enrichment analysis
GO is widely used in bioinformatics to mainly study three aspects of biology: biological processes (BP), molecular functions (MF), and cellular components (CC). The Kyoto Encyclopedia of Genes and Genomes (KEGG) reflects biological molecular interactions and chemical reactions. We used the Database for Annotation, Visualization, and Integrated Discovery (DAVID) to analyze the DEGs. Values of p < 0.05 and gene counts ≥8 were considered as statistically significant.
Protein-protein interaction (PPI) network and gene module analysis
First, the Search Tool for the Retrieval of Interacting Genes (STRING) database was used to obtain DEG-encoded protein and PPI information. Second, PPI pairs with a combined score >0.4 were downloaded and analyzed. The PPI networks were constructed using Cytoscape software. The plug-in Molecular Complex Detection (MCODE) was used to screen modules or clusters in large PPI networks. The following parameters were used for DEG clustering and scoring: MCODE score > 5, degree cutoff = 2, node score cutoff = 0.2, max depth = 100, and k-score = 2.
Survival analysis of DEGs
Gene Expression Profiling Interactive Analysis (GEPIA), a web server for cancer and normal gene expression profiling and interactive analysis, was used to perform survival analysis of the DEGs, confirm their expression, and identify the median expression among tumor and normal samples in BodyMap[18].
Prediction of target lncRNA-miRNA genes and initial lncRNA-miRNA-mRNA network construction
The initial lncRNA-miRNA-mRNA network was constructed according to the ceRNA hypothesis. miRcode, the miRTarBase database, and TargetScan were used to predict target differentially expressed (DE) mRNAs, and only miRNA-mRNAs that were aligned across the three databases were incorporated into the ceRNA network. Next, DE lncRNA-DE miRNA interactions were predicted based on the miRcode database. Cytoscape (Version 3.7.1) was used to construct the preliminary ceRNA network.
Gene expression and final ceRNA network construction
One-way analysis of variance was used to analyze the DEGs and DE lncRNAs. The cutoffs for the gene expression boxplot were set to |Log2FC| = 1 and p = 0.01. lncBace2.0 and GEPIA (http://gepia.cancer-pku.cn/index.html) were used to validate the DE lncRNAs and screen for genes showing significant expression, respectively. A Ji mulberry was used to construct the final lncRNA-miRNA-mRNA network.
Patient samples
The 37 primary glioma samples were obtained from patients who underwent surgical treatment at the Xiangya Hospital of Central South University from September 2016 to May 2017 but were not administered chemotherapy or radiotherapy. Six normal brain tissues from patients with cerebral trauma surgery were collected as controls. All samples were snap frozen in liquid nitrogen until total RNA was extracted. This study was approved by the Central South University Xiangya Hospital Medical Ethics Committee (No. 201803806).
Cell culture
The U87 and T98G cell lines were obtained from the American Type Culture Collection (Manassas, VA, USA). The cells were cultured in Dulbecco’s modified Eagle’s medium (HyClone, Logan, UT, USA) growth medium supplemented with 10% fetal bovine serum (Gibco, Grand Island, NY, USA) at 37°C in a humidified 5% CO2 incubator.
Cell transfection
The U87 and T98G cells were seeded into six-well plates and cultured until they reached 50%–60% confluence. They were then transiently transfected with small-interfering RNA (siRNA-C1S) and negative control (NC) siRNA, which were designed by RiboBio (Guangzhou, China), using Lipofectamine® RNAiMAX Transfection Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. This transfection reagent and siRNA were diluted in Opti-MEM (Invitrogen) without antibiotics and used for experiments at 48 h after transfection.
RNA isolation and qRT-PCR
Total RNA was extracted from tissues using TRIzol reagent (TaKaRa, Shiga, Japan) according to the manufacturer’s protocol. cDNA was transcribed using the PrimeScript RT Reagent Kit (TaKaRa, RR047A). A SYBR Green PCR Kit (Takara, RR820A) was used to measure the quantity of isolated RNA according to the manufacturer’s instructions. qRT-PCR was performed using the CFX Connect Real-Time System (Bio-Rad, Hercules, CA, USA) as follows: 95°C for 3 min, and 40 cycles of 95°C for 10 s and 60°C for 30 s. GAPDH was used as an internal control. The 2-ΔΔCt method was used to calculate the relative expression levels of the RNAs. The qRT-PCR primers included the bulge-loop RT primer and qPCR primers specific for has-miR-129-5p that were designed and synthesized by RiboBio. The primers for lncRNA MIR155HG were F: 5′-CCACCCAATGGAGATGGCTCT A-3′; and R: 5′- GCAAAAACCCCTATCACGATTA-3′. The primers for mRNA C1S were F: 5′- TCCAAGTCCCATACAACAAACTC-3′, and R: 5′-CAAACCCCGTAA AACGCTCT-3′. GAPDH: F: 5′-CCCATCACCATCTTCCAGGAG-3′, R: 5′-GTTGTCATGGATGACCTTGGC-3′.
Cell proliferation assay
Cell viability was measured in an MTS assay (Cell Titer 96 aqueous one solution reagent, Promega, Madison, WI, USA) according to the manufacturer’s instructions. Briefly, at 24 h before the experiment, cells (2 × 103 cells) were seeded into 96‐well plates, divided into two groups (NC, si-C1S-3), and cultured for 0, 24, 48, 72, 96 h. At the specified time, 10 μL of CellTiter 96 Aqueous One Solution Reagent was added to each well and incubated at 37°C for 30 min. The quantity of formazan product was determined by measuring the absorbance at 490 nm with a microplate reader (Bio-Rad). All experiments were repeated three times.
Wound healing assay
The migration abilities of U87 and T98G cells were assessed via a wound-healing assay. When cells grown on the plate reached a density of approximately 90%, an artificial wound was created by scraping the confluent monolayer with a 200-µL plastic pipette tip. Cell migration was monitored under a microscope at 0 and 24 h, and images were taken immediately. The average wound closure rate was measured in triplicate and the results were evaluated with ImageJ software (NIH, Bethesda, MD, USA).
Cell migration and invasion assays
Cell migration and invasion were evaluated using 24-well transwell chambers (Corning, Inc., Corning, NY, USA) coated with Matrigel or not. Briefly, transfected U87 and T98G cells (2 × 104) were seeded into the upper chambers containing serum-free medium, whereas medium containing 20% fetal bovine serum was added to the lower chambers. After incubation for 24 h (migration assays) and 48 h (invasion assays) at 37°C, the cells were fixed with 4% formalin and stained with 0.1% crystal violet. The cell invasion assay was performed in the same manner, but the upper chamber was covered with Matrigel. Cells were counted under a light microscope in randomly selected fields and analyzed using ImageJ software.
Statistical analyses
Statistical analyses were conducted using R (version 3.6.1, The R Foundation, Vienna, Austria) and SPSS (version 24.0, SPSS, Inc., Chicago, IL, USA). Statistical analysis involved Student’s t-test, Mann-Whitney U test, analysis of variance, or two-way repeated measures analysis of variance with post-hoc Student’s t-test or paired Student’s t-test. Differences with p values <0.05 were considered as statistically significant.