The human glioma (LN229, U251, A172 and T98G) and Normal Human Astrocyte (NHA) cell lines were obtained from The Cell Bank of Type Culture Collection of the Chinese Academy of Sciences. Cells were incubated in DMEM (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 670087) supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 16140071) and maintained at 37°C in an atmosphere containing 5% CO₂.

Human GBM specimens and adjacent normal brain tissues (located 0.5–1 cm away from the tumor) were obtained from 35 patients with histologically confirmed GBM who underwent surgery at the Second Affiliated Hospital of Harbin Medical University between May 2018 and December 2019 (Harbin, Heilongjiang, China). All GBM samples and adjacent normal brain tissues (17 women and 18 men; age range, 22–74 years; median age, 47.31 years) were confirmed by two senior pathologists. None of the patients had received any preoperative treatment. The specimens were immediately snap-frozen in liquid nitrogen for further research. Written informed consent was obtained from each patient. All procedures were performed in agreement with the Declaration of Helsinki (25). The study was approved by the Institutional Review Board of the Second Affiliated Hospital of Harbin Medical University (No.2018HMUIRB0113, Harbin, China). The clinical characteristics of the patients are summarized in Table I.

The high-throughput sequencing data of nine glioma tissues and three normal samples were acquired from the Gene Expression Omnibus (GEO) database (no. GSE4290 and GSE104267; http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc) (26). The overall survival data were obtained from GBM patients in The Cancer Genome Atlas (TCGA) GBM database (http://cancergenome.nih.gov) (27). Gene Ontology (GO) term analysis was performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.8 (https://david.ncifcrf.gov/). The edgeR package was conducted to explore significantly abnormally expressed lncRNAs for the normalized gene expression profile data. The DEseq2 package was used to analyze the differentially expressed lncRNA with the threshold set as log2 fold-change (FC) level >2 and false-discovery rate (FDR) <0.01. Furthermore, the Venn diagram was generated by FunRich to visualize the overlapping lncRNAs among TCGA database, GSE4290 and GSE104267 datasets.

Glioma cell cytoplasmic and nuclear RNA were harvested and purified by using the PARIS Kit (cat. no. AM1921; Thermo Fisher Scientific, Inc.) according to the manufacturers' instructions.

Total RNA was collected from clinical specimens and T98G and U251 cells using TRIzol Reagent (Beijing Transgen Biotech Co., Ltd.; cat. no. R1021) and reversely transcribed into cDNA using the PrimeScript™ RT Kit (Takara Biotechnology Co., Ltd.; cat. no. RR014A) according to the manufacturers' protocol. RT-qPCR was conducted three times on an ABI 7500HT Real-Time PCR instrument (Applied Biosystems; Thermo Fisher Scientific, Inc.) according to the manufacturers' protocols. The thermocycling conditions were as follows: Initial denaturing step (94°C, 10 min), followed by 40 cycles of denaturing (94°C, 5 sec), annealing (60°C, 30 sec) and extending (72°C, 45 sec). The sequences of the primers used were as follows: KTN1-AS1, forward 5′-ATGCACACTTCTCGGCTAAGAGTC-3′, reverse, 5′-CTACAATGCCACAAGTGATTCCAG-3′; miR-505-3p, forward 5′-CGCGGATCCCAGACTCCCAGCAATCAC-3′, reverse 5′-CCGGAATTCGCAGTATTCCCACCATTT-3′; MMP-9, forward 5′-AGACCTGGGCAGATTCCAAAC-3′, reverse 5′-CGGCAAGTCTTCCGAGTAGT-3′; U6, forward 5′-GGATATTGTTGCCATCAATGACC-3′, reverse 5′-AGCCTTCTCCATGGTGGTGAAGA-3′; and GAPDH, forward 5′-AAGAAGGTGGTGAAGCAGGC-3′ and reverse 5′-GTCAAAGGTGGAGGAGTGGG-3′. U6 served as the endogenous control for miRNA, and GAPDH served as the endogenous control for lncRNA and MMP-9. To verify the expression of KTN1-AS1 and miR-505-3p, endogenous mRNA was synthesized using a SYBR Green PCR Master Mix Kit (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 4309155). Relative gene expression levels were normalized to endogenous controls and were expressed as 2^−ΔΔCt (28).

KTN1-AS1 siRNA, pcDNA3.1-KTN1-AS1 (OE), negative control-siRNA (si-NC), miR-505-3p mimics, miR-505-3p inhibitor and negative control-miRNA (miR-NC) were obtained from Shanghai GenePharma Co., Ltd. T98G and U251 cells were seeded at the density of 5×10⁵ cells per well in 6-well plates and cultured in DMEM containing 10% FBS at 37°C. When the cell confluence reached 70–80%, cells were transfected with 20 µM of each construct using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 11668030) according to the manufacturers' protocol and maintained at 37°C for 6 h. The culture medium was then replaced by fresh DMEM containing 10% FBS, and subsequent experiments were conducted at 24 h post-transfection. The sequences of the constructs were as follows: KTN1-AS1 siRNA, forward 5′-GACUGUGGAUAGAGAUAGAAA-3′, reverse 5′-UCUAUCUCUAUCCACAGUCUG-3′; si-NC, forward 5′-GGUAAGCAGUGGCUCCUCUAA-3′, reverse 5′-ACGUGACACGUUCGGAGAAUU-3′; miR-505-3p mimics, forward 5′-GGGAGCCAGGAAGUAUUGAUGU-3′, reverse 5′-ACAUCAAUACUUCCUGGCUCUU-3′; miR-505-3p inhibitor, forward 5′-GGGAGCCAGGAAGUAUUGAUGU-3′, reverse 5′-CAGUACUUUUGUGUAGUACAA-3′; and miR-NC, forward 5′-CAGUACUUUUGUGUAGUACAA-3′ and reverse 5′-CAGUACUUUUGUGUAGUACAA-3′.

DIANA tools (http://carolina.imis.athena-innovation.gr/diana_tools) and StarBase V2.0 (http://starbase.sysu.edu.cn/) database were used to identify the potential target miRNAs of KTN1-AS1. Among all the statistically relevant miRNAs, miR-23b-3p, miR-23c-3p and miR-505-3p were obtained from two databases, and were therefore selected for further experiments. The human KTN1-AS1 Luc-reporter was used in the ligation of the KTN1-AS1 3′-untranslated region (UTR) PCR product. The psiCHECK2 vector (GeneChem, Inc.) was conducted to construct KTN1-AS1 3′-UTR containing reporter. T98G cells were seeded in 6-well plates at the density of 1×10⁵ cells per well and the miRNA mimics were co-transfected with psiCHECK2-KTN1-AS1-wild type (WT) or mutant (MUT) (20 µl) and miR-505-3p mimics or miR-NC (20 µM) by Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 11668030). The medium was changed at 4 h post-transfection. After 48 h, the intensity of Firefly luciferase was measured using a Dual-Luciferase Reporter Assay System (cat. no. BA0180; BioVision Inc.) according to the manufacturers' instructions. Renilla luciferase intensity was used as an internal control.

The viability of glioma cells was evaluated using Cell Counting Kit-8 (CCK-8; cat. no. CK04; Dojindo Molecular Technologies, Inc.). T98G and U251 cells were seeded in 96-well plates (100 µl containing 3,000 cells/well). Cells were cultured in DMEM at 37°C under 5% CO₂ conditions for 24, 48 or 72 h. CCK-8 solution (10 µl) was then added to the cells for 4 h and the optical density was detected at 490 nm using a Tecan microplate reader (Infinite F50; Tecan Group, Ltd.).

Transfected T98G and U251 cells (5×10⁴) were seeded into the upper chamber of a Transwell on a Matrigel-coated membrane (cat. no. 3495; Costar; Corning, Inc.) and cultured in serum-free medium for 24 h. DMEM containing 20% FBS was placed in the lower chamber. Subsequently, the upper chamber medium was changed and cells on the upper side of the filter were removed. Then, cells that have invaded the lower side were fixed with 4% paraformaldehyde at room temperature for 10 min and stained with 0.1% crystal violet (cat. no. IC0600; Beijing Solarbio Science & Technology Co., Ltd.) at room temperature for 10 min. Stained cells were counted in five randomly selected fields under a light microscope (Nikon Corporation) at ×100 magnification.

Cell slides were precoated with 0.1% poly-L-lysine at 4°C for 24 h and placed into 24-well plates. Subsequently, T98G and U251 cells (1×10⁵) were transfected with si-KTN1-AS1/si-NC in culture dish. After 24 h, cells were collected and cultured on the cell slides in 24-well plates for 24 h. Then the medium was removed and cell slides were incubated with 4% paraformaldehyde in 24-well plates at room temperature for 20 min and 0.1% Triton X-100 for 5 min. The cell slides were washed in 24-well plates and fixed with 5% BSA (cat. no. 9048-46-8; Sigma Aldrich; Merck KGaA) dissolved in PBS at room temperature for 1 h, incubated with primary antibodies targeting MMP-9 (rabbit polyclonal antibody; cat. no. 13667; 1:50; Cell Signaling Technology, Inc.) at 4°C overnight, and with fluorescence-labeled rabbit secondary antibody [tetramethylrhodamine (TRITC)-conjugated goat anti-rabbit IgG; cat. no. SA00007-2; 1:100; ProteinTech Group, Inc.)] for 1 h at room temperature. Nuclei were stained with DAPI (1 µg/ml; cat. no. 4083s; Cell Signaling Technology, Inc.) for 10 min. Cell slides were subsequently collected and placed on glass slides. The cells were observed under a fluorescence microscope (Nikon Corporation) at ×400 magnification.

Four-week-old female BALB/c nude mice (weight, 15–20 g; n=16) were obtained from Beijing Vital River Laboratory Animal Technology, Co., Ltd. The vehicle of the luciferase lentivirus (GV260-purinomycin) and the corresponding transfection reagent (HitransG Infection enhancer fluid) were purchased from GeneChem and transfection was conducted according to the manufacturers' protocol. Then T98G cells were stable transfected with luciferase lentivirus and transfection reagent for 4 h. Then purinomycin was fixed and incubated for 15 day. The mice were assigned to two groups (si-NC and si-KTN1-AS1; n=8 per group) and placed in an anesthesia induction box. Mice were anesthetized with isoflurane (induced concentration, 3–4%) for 2–3 min and maintained anesthetized using at 1–1.5% isoflurane. Subsequently, si-KTN1-AS1 or si-NC-Luc T98G cells (1×10⁶ cells/mouse in 5 µl) were intracranially injected using a stereotactic instrument. Tumor growth was evaluated by bioluminescence imaging (photons/s/cm²) using a Bruker In-Vivo FX PRO Imaging System (Bruker Corporation). Mice were anesthetized using isoflurane and D-luciferin sodium salt injected into the abdomen of mice. Tumor size in the head of mice was evaluated by bioluminescence imaging (BLI). The fluorescence value is associated with the tumor size. After 40 days, the remaining mice were then sacrificed using CO₂ (28% volume displacement per min) and the overall survival time was recorded. All applicable international, national and/or institutional guidelines for the care and use of animals were followed. The study was approved by the Institutional Review Board of the Second Affiliated Hospital of Harbin Medical University (approval no. 2018HMUIRB0113, Harbin, China).

The data were presented as the means ± standard deviation of the mean of three independent experiments. Statistical analyses were conducted using the statistical software SPSS version 19.0 (IBM Corp.). Pearson's rank correlation test was used to determine the correlation between KTN1-AS1 and miR-505-3p or MMP-9 expression. Comparisons between groups were performed using two-tailed Student's t-test or ANOVA followed by Tukey's post hoc test. The association between KTN1-AS1 level and clinicopathological characteristics of the patients was analyzed using χ² test or Fisher's exact test. Kaplan-Meier curve and log-rank test were conducted for survival analysis using GraphPad Prism v.5.0 (GraphPad Software, Inc.). The aberrantly expressed lncRNAs were explored according to Benjamini-Hochberg method (29). P<0.05 was considered to indicate a statistically significant difference.

To identify whether KTN1-AS1 was involved in the development of glioma, KTN1-AS1 expression was evaluated using microarray data downloaded from GEO (GSE104267). The top 50 aberrantly expressed lncRNAs were identified with cut-off values of log2FC>2 and FDR<0.01, and it was observed that KTN1-AS1 level was overexpressed in GBM tissues compared with normal brain tissues (Fig. 1A and B). In total, five intersecting lncRNAs were identified and investigated according to their log2FC level, as displayed in Fig. 1C. The data from TCGA database indicated that patients with glioma and higher KTN1-AS1 expression had a shorter overall survival (Fig. 1D; n=162, P<0.01). To further elucidate the function of KTN1-AS1, the associated gene expression profiles were detected using Co-lncRNA program from the GO database. The results demonstrated that KTN1-AS1 was associated with cell invasion, focal adhesion and zinc ion binding (Fig. 1E). Since GBM is the most malignant and common primary nervous system tumor worldwide (1,2), it is crucial to identify the clinical characteristics of patients with GBM (grade IV in glioma) according to tissue expression of KTN1-AS1. The results demonstrated that expression level of KTN1-AS1 was significantly associated with sex and mean tumor diameter in the clinical GBM samples (n=35; grade IV; P=0.001 and P=0.004, respectively; Table I). Subsequently, KTN1-AS1 expression level was examined in 35 GBM tissues and adjacent normal brain tissues by RT-qPCR. The results demonstrated that KTN1-AS1 expression level was significantly increased in GBM tissues (P<0.001; Fig. 2A) compared with adjacent normal tissues. These findings suggested that KTN1-AS1 may function as an oncogene in glioma progression.

Effect of miR-505-3p on glioma cell viability and invasion. KTN1-AS1 expression was also investigated in glioma cells compared with NHA cells, and the results demonstrated that KTN1-AS1 was most highly expressed in T98G and U251 cells among the glioma cell lines (P<0.01 and P<0.001). These two cell lines were therefore chosen for subsequent experiments (Fig. 2B). To identify the biological function of KTN1-AS1 in the progression of glioma, T98G and U251 cells were transfected with si-KTN1-AS1/si-NC or pcDNA3.1-KTN1-AS1/NC, and the cell proliferation and invasive ability were determined. The transfection efficiencies were confirmed by RT-qPCR, where KTN1-AS1 expression level was significantly increased or increased compared with corresponding NC groups (Fig. 2C). The results from CCK-8 assay indicated that KTN1-AS1 knockdown significantly decreased T98G and U251 cell proliferation after 24 h (Fig. 2D; P<0.05). Furthermore, silencing KTN1-AS1 could significantly decrease glioma cell invasive ability in vitro (Fig. 2E; T98G cells, P<0.01; U251 cells, P<0.05).

Invasion and metastasis serve crucial roles in tumorigenesis. To further determine the underlying mechanism by which KTN1-AS1 could affect the invasive ability of glioma cells, analysis of the TCGA-GBM database was performed. The results demonstrated that KTN1-AS1 expression was positively correlated with expression of matrix metalloproteinase-9 (MMP-9; r=0.281; P=0.0043; Fig. 3A), which is a marker of cell invasion (30). Then, the expression level and cellular location of MMP-9 were determined in glioma cells following KTN1-AS silencing. The results demonstrated that KTN1-AS1 knockdown significantly decreased the expression of MMP-9 in glioma cells (Fig. 3B and C). Taken together, these findings suggested that KTN1-AS1 knockdown may inhibit proliferation and invasive ability of glioma cells.

To investigate the subcellular localization of KTN1-AS1, the nuclear and cytoplasmic fractions of T98G glioma cells were explored. As presented in Fig. 4A, KTN1-AS1 was predominantly localized in the cytoplasm, indicating that KTN1-AS1 may exert both transcriptional and post-transcriptional regulatory functions on glioma cell lines. By using DIANA tools and starBase V2.0 database blast prediction, it was reported that KTN1-AS1 contained a putative targeting site for miR-23b-3p, miR-23c-3p and miR-505-3p. Subsequently, following overexpression or silencing of KTN1-AS1 in T98G cells, it was observed that miR-505-3p could be downregulated or upregulated at the mRNA level (Fig. 4B and C). In addition, miR-505-3p expression level was significantly decreased in GMB tissues compared with adjacent normal tissues (P<0.001; Fig. 4D). Analysis of TCGA-GBM database indicated that low level of miR-505-3p in patients with GBM was associated with better prognosis (n=162; P<0.05; Fig. 4E). Furthermore, miR-505-3p and KTN1-AS1 expression levels in clinical GBM tissues from patients were negatively correlated (n=35; R=−0.3552; P=0.0363; Fig. 4F). Decreased miR-505-3p expression was also observed in glioma cells compared with NHA cells (P<0.05; F=35.37; Fig. 4G). Subsequently, T98G and U251 cells were transfected with miR-505-3p mimics, miR-505-3p inhibitor and miR-NC. The results demonstrated that miR-505-3p was significantly increased or decreased in transfected cells compared with NC group (Fig. 4H). The luciferase reporter assay was then conducted to confirm the putative miR-505-3p target site. In addition, luciferase reporters carrying either the predicted miR-505-3p WT (KTN1-AS1-WT) or its mutated fragment (KTN1-AS1-MUT) were constructed (Fig. 4I). The results indicated that miR-505-3p significantly inhibited the luciferase activity in cells transfected with KTN1-AS1-WT but not in cells transfected with KTN1-AS1-MUT 3′-UTR (Fig. 4J). These data suggested that KTN1-AS1 may interact with miR-505-3p in glioma cells.

miR-505-3p was determined to be negatively associated with KTN1-AS1 in glioma tissues. The expression of miR-505-3p was significantly elevated in KTN1-AS1-silenced cells compared with si-NC cells (Fig. 5A). Subsequently, rescue experiments were performed in T98G cells. Co-transfection of miR-505-3p inhibitor and si-KTN1-AS1 led to a decreased level of miR-505-3p expression compared with miR-inhibitor NC + si-KTN1-AS1 group (Fig. 5B). In addition, the expression of miR-505-3p was increased when cells were co-transfected with si-KTN1-AS1 and miR-505-3p inhibitor compared with si-NC + miR-505-3p inhibitor group. The results from CCK-8 assay demonstrated that miR-505-3p inhibitor could decrease the proliferation induced by silencing KTN1-AS1 (Fig. 5C). Transwell assay results showed that miR-505-3p inhibitor could inhibit the invasive ability induced by silencing KTN1-AS1 (Fig. 5D). These findings suggested that miR-505-3p may be considered as an essential mediator of KTN1-AS1-regulated proliferation and invasion.

To further determine the effect of KTN1-AS1 silencing in vivo, BALB/c nude mice were intracranially inoculated with T98G-luc cells with si-NC or si-KTN1-AS1. BLI of mice was conducted to identify the role of si-KTN1-AS1 on glioma tumor growth. The data demonstrated that tumor formation in KTN1-AS1 silencing group was significantly lower compared si-NC group (P<0.05; Fig. 6A and B). To further determine the inhibitory role of si-KTN1-AS1 on glioma progression in nude mice, the overall survival of mice (n=8 per group) was determined by Kaplan-Meier analysis. The si-KTN1-AS1-treated group displayed a significant improvement in survival compared with the si-NC group (P=0.0008; Fig. 6C). These findings suggested that KTN1-AS1 may promote the tumorigenicity in vivo.