Long noncoding RNA ENST00000413528 sponges microRNA‐593‐5p to modulate human glioma growth via polo‐like kinase 1

Summary Aims In this study, we examined the expression of lncRNA ENST00000413528 in glioma and determined its role in glioma development. Methods LncRNA ENST00000413528 was detected in glioma tissues by lncRNA microarray. Then, we performed real‐time PCR, CCK‐8, colony formation assay, flow cytometry, caspase‐3/7 assay and animal experiment to detect the function of ENST00000413528 in glioma after ENST00000413528 knockdown. Subsequent bioinformatics analysis, luciferase reporter assays and RNA immunoprecipitation (RIP) assay western blotting indicated possible downstream regulatory molecules. The expression of PLK1 in glioma tissues was also examined by immunohistochemistry staining. Results Expression of ENST00000413528 was significantly increased in glioma tissues and LN229 and U251 cells. PLK1 protein could not be detected in peritumoral brain edema (PTBE) tissues; however, it showed an increasing number of positively cytoplasmic stained from WHO‐Grade II to Grade III gliomas. Knockdown of ENST00000413528 in glioma cells inhibited cell proliferation and colony formation abilities, induced the G0/G1 arrest of the cell cycle, and promoted apoptosis. The dual reporter assay and RNA immunoprecipitation assay verified the interaction between ENST00000413528 and miR‐593. We also demonstrated that polo‐like kinase 1 (PLK1) was regulated by miR‐593; PLK1 messenger RNA lacking 3’UTR partially reversed the effects caused by ENST00000413528 knockdown or miR‐593 upregulation. Conclusion lncRNA ENST00000413528 is closely related to the development of glioma via the miR‐593‐5p/PLK1 pathway.


| INTRODUC TI ON
Glioma is the most common primary brain tumor and often leads to fatal patient outcomes due to its highly invasive growth pattern and frequent resistance to therapies. 1,2 In recent years, researchers have increasingly focused on the molecular regulatory networks involved in the development of gliomas and on possible new therapeutic targets. The discovery of mutations in specific genes, such as isocitrate dehydrogenase(IDH) gene mutations, 3 BRAF mutations, 4,5 and histone mutations, 6 has changed our understanding of the pathogenesis of many types of gliomas and subsequently driven the biomarker-related classification of gliomas. 7 Notably, the current role of this classification system is not just as a supplement to diagnosis; instead, it goes beyond histological diagnosis.
Human genome sequence data show that most of the transcripts are noncoding ribonucleic acids (RNAs) and that only 2% of the transcripts can encode proteins. 8 Extensive research has been conducted on microRNAs (miRNAs), confirming their role in the regulation of genes and in the function of cell biology in many cancers. 9 Recent studies have also shown that long noncoding RNAs (lncRNAs) play an important role in normal development and numerous diseases, including many kinds of cancers. 10,11 Investigations have also been conducted on lncRNAs and their relevant pathways in glioma. P73 antisense RNA 1T (TP73-AS1) was upregulated in brain glioma tissues and cell lines and was associated with a poorer prognosis in patients with glioma. Additionally, they determined that TP73-AS1 may also act as a competing endogenous RNA (ceRNA) to promote HMGB1 expression by sponging miR-142 to promote the brain glioma growth and invasion. 19 CeRNA is a known mechanism of lncRNA; in this mechanism, ln-cRNA competes for posttranscriptional control by sponging certain relevant miRNAs. 20 Previous surveys have confirmed some lncRNA-miRNA interaction pathways, 21,22 which provide some theoretical basis for our current research.
Polo-like kinase 1 (PLK1) is a serine/threonine protein kinase that is known to be a key regulator of mitosis, with critical roles in cell cycle-associated processes. 25,26 PLK1 is involved in the regulation of the cell cycle and the growth of various tumors, including breast cancer, renal cell cancer, and prostate cancer. 18,27,28 Two studies found some lncRNAs, such as Hox transcript antisense RNA (HOTAIR) and lnc-RI, exerted their regulatory effects through PLK1 31,32 ; however, such lncRNA-PLK1 associated approach has not yet been observed in glioma cells.
In this study, we performed a lncRNA microarray in glioma tissues and found lncRNA ENST00000413528 was overexpressed.
After the qRT-PCR verification in glioma tissues and biological informational analysis, we found a possible ceRNA pathway of ln-cRNA ENST00000413528, which shared a complementary area with miR-593-5p, and associated with the target gene PLK1. We, thus, hypothesized that lncRNA ENST00000413528 may sponge miR-593-5p to regulate PLK1 to exert its effects on glioma cells and designed a series of experiments to verify our hypothesis.

| Ethics statement
The human studies included in this research were conformed to Declaration of Helsinki; both the human and animal studies were approved by the medical research ethics committee of the First Affiliated Hospital of Zhengzhou University in Zhengzhou, China. All patients with glioma signed an informed consent prior to participation.

| Cells and tissues
The glioma cell lines U251, SNB19, U373, SHG44, and LN229 and normal human astrocytes (NHAs) were obtained from the American Collected tissues were immediately placed into liquid nitrogen. The primary antibodies for western blotting or immunohistochemistry against glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (ab8245) and PLK1 (ab17056), as well as secondary antibody (ab6789), were purchased from Abcam (Cambridge, UK).

| Protein extraction and western blotting
Tissues and cells were lysed by radioimmunoprecipitation assay

| Cell Counting Kit-8 assay
Cells were dissociated and seeded in 96-well plates at the density of 2 × 10 4 /well, and siRNA or si-NC was transfected into cells

| Cell colony formation assay
Cells with different treatments were seeded in the 3.5 cm cell culture dish (Corning, Corning, NY, USA) at the density of 150/dish (with 2 mL DMEM with 10% FBS). Then, 200 μL of DMEM with 10% FBS was added into the dish every 2 days to supplement the medium that volatilized into the incubator. Cells were incubated for 10 to 14 days until the colonies were visible to the naked eye on the bottom of the culture dish. After removing the medium, cell colonies were fixed with methanol for 30 minutes and then stained with 0.1% crystal violet (Solarbio, Beijing, China). The number of colonies stained in the dish for analysis was then counted, and the experiment was repeated three times.

| Cell apoptosis assays
The number of apoptotic cells of each group was detected via the

| Cell cycle detection
Cell cycle status of each group was detected by flow cytometry using a cell cycle detection kit (KenGen BioTECH, Jiangsu, China).
Cells (1 × 10 6 ) were dissociated with 0.25% trypsin without EDTA, washed twice with PBS, and fixed with 500 μL of 70% precooled ethanol for 2 hours. After fixation, the cells were washed with PBS once; 500 μL of the working solution (RNaseA: PI = 1:9) was then added to resuspend cells. After incubating cells in the dark for 30 to 60 minutes at room temperature, the red fluorescent signal at the excitation wavelength of 488 nm was detected. The analysis was performed and figures were created using Modfit software.

| Dual luciferase reporter assay
In order to verify the interaction between LncRNA  si-Lnc or si-NC infected U251 cells (5 × 10 6 ) were injected into the left armpit of the mice. The mice were observed weekly to record their weight and tumor volume (volume = 1/2 × length × width 2 ) and sacrificed 4 weeks after implantation.

| RNA immunoprecipitation assay
RNA immunoprecipitation (RIP) assay was performed using the Magna RNA-Binding Protein Immunoprecipitation Kit (Merck KGaA, Darmstadt, Germany) and the Ago2 antibody (ab32381; Abcam, Cambridge, UK) according to the manufacturer's protocols. After the antibody was recovered by protein beads, qRT-PCR was performed to detect the relative expression level of lncRNA ENST00000413528 and miR-593 in the precipitates. 34

| Immunohistochemistry (IHC) and hematoxylin and eosin (H&E) staining
Samples were dehydrated in gradient alcohol, fixed in neutral-buffered formalin and embedded in paraffin. They were prepared into 4-6-μm-thick tissue sections and then dewaxed. For H&E staining, hematoxylin and eosin were stained, rinsed by water, then dehy-

| Statistical analysis
All cell experiments were repeated three times. SPSS 21 software (IBM Corp., Armonk, NY, USA) was used to perform the data analysis. Data were expressed in the format of the mean ± standard deviation (SD). A Student's t test, one-way analysis of variance, Pearson's analysis was carried out to analyze the data. P < 0.05 was considered statistically significant.

| Detection and verification of lncRNA microarray in glioma tissues
The microarray of three human brain glioma tissues and the paired PTBE tissues showed 22 differential expressed lncRNAs, including 13 downregulated and nine upregulated ones ( Figure 1A, Table 1).
LncRNA ENST00000413528 was found to demonstrate a 3.58-fold change upregulation in the glioma specimens. To verify the microarray results, we performed the qRT-PCR of two upregulated and two downregulated lncRNAs in 16 paired human glioma tissues and their corresponding PTBE tissues ( Figure 1B,C,D,E). Our results showed that the relative expression of ENST00000413528 or lnc-NUDT4-4:1 was higher in tumor tissues than that in PTBE tissues (P < 0.05), while NONHSAT207178 and NONHSAT174301 were downregulated in tumor tissues (P < 0.05), all of which were consistent with the microarray results. Thus, we confirmed that the data of the microarray were statistically credible. In addition, the morphological changes in glioma tissues were displayed by H&E staining in Figure 1G. PLK1, the targeted gene in our hypothesis, was also detected by IHC. PLK-1 protein could be mainly cytoplasmic stained in glioma tissues while hardly stained in PTBE tissues. The PLK1 expression showed a significantly and gradually increased tendency from PTBE, low-grade glioma to high-grade glioma tissues. We then detected the relative expression level of ENST00000413528 in different glioma cell lines for further study ( Figure 1F). Expression of ENST00000413528 was significantly increased in glioma cell lines, especially in LN229 and U251 cells, when compared with NHA cells (P < 0.05). Therefore, we selected LN229 and U251 cells to carry out the next experiments.

| Knockdown of ENST00000413528 inhibited cell proliferation and induced apoptosis and cell cycle arrest in glioma cells
LN229 and U251 cells were infected by lentivirus fluid containing si-Lnc or si-NC, and puromycin was used to select cells which may stably knockdown ENST00000413528 expression. We evaluated the relative expression of ENST00000413528 in different groups (Figure 2A), which showed an obvious downregulation in cells F I G U R E 1 A, Differential expressed lncRNAs in glioma tissues detected by the Agilent Human lncRNA V6 chip (Agilent Technologies, Santa Clara, CA, USA). B refers to the brain glioma tissue; C refers to the control tissue (peritumoral brain edema tissue). B, C, D, E, Verification of four differential expressed lncRNAs in 16 pairs of glioma and paired peritumoral brain edema (PTBE) tissues using qRT-PCR with the 2 −ΔΔCT method. F, Relative expression level of lncRNA ENST00000413528 in different glioma cells as detected by qRT-PCR. Each assay was performed three times. G, HE and PLK1 IHC staining in PTBE and grade II, III glioma tissues. The positive staining was majorly localized in the cytoplasm of glioma tissues, grade III glioma expressed a higher level of PLK1. IHC staining showed no detectable PLK1 expression in PTBE tissues Therefore, our results suggest that the downregulation of lncRNA ENST00000413528 may impair glioma cell growth, promote cell apoptosis in LN229 and U251 cells, and the proliferation inhibitory effect was also demonstrated in our xenograft mouse model constructed using U251 cells. Similarly, the upregulation of miR-593 could also lead to a decrease of ENST00000413528 expression ( Figure 3D; P < 0.05). We performed RIP assay to clarify whether ENST00000413528 and  Correlation analysis between the relative expression of miR-593 and ENST00000413528 in glioma tissues demonstrated a negative correlation with R 2 at 0.4147 ( Figure 3G; P < 0.05). Overall, the results of the dual luciferase reporter assay, RIP assay, and correlation analysis suggested that LncRNA ENST00000413528 directly targeted miR-593-5p.

| PLK1 was a target gene of miR-593-5p
MiRNAs regulate gene expression by binding to the three prime untranslated regions (UTRs) of their target genes. We performed the bioinformatics analysis and found a complementary pairing area between the sequences of miR-593-5p and PLK1 ( Figure 4A). To verify whether PLK1 was a target gene of miR-593-5p, we performed the dual luciferase reporter assay. Recombinant plasmids pmirGLO-Wt-PLK1 3′UTR and pmirGLO-Mt-PLK1 3′UTR were cotransfected with miR-NC or miR-593 mimics into U251 and LN299 cells ( Figure 4C).
Our results showed that the luciferase activity of the Wt-PLK1 luciferase reporter vector was notably suppressed in response to the miR-593 mimics' transfection when compared with the NC group (P < 0.05), indicating a direct combination between miR-593 and PLK1 3′UTR. Subsequently, the expression of the PLK1 protein in cells transfected with miR-593, NC control, or blank control was detected by western blotting ( Figure 4B). The upregulation of miR-593 significantly decreased the PLK1 protein expression (P < 0.05).
F I G U R E 3 LncRNA ENST00000413528 regulated miR-593-5p by direct targeting. A, A complementary pairing area was found between ENST00000413528 and miR-593, miR-593, and PLK1. B, pmirGLO-Wt-Lnc or pmirGLO-Mt-Lnc was cotransfected with miR-593 mimics or miR-NC, respectively, and the luciferase activity of each group was detected. C, Relative expression level of miR-593 when ENST00000413528 was knocked down. D, Relative expression level of ENST00000413528 when miR-593 was upregulated. E, Immunoprecipitation of Ago2-associated ENST00000413528 was verified by RT-qPCR. Immunoglobulin G (IgG) served as the negative control. F, Immunoprecipitation of Ago2-associated miR-593 was verified by RT-qPCR. IgG served as the negative control. G, Correlation analysis between the relative expression levels of ENST00000413528 and miR-593 in 16 pairs of glioma and paired peritumoral brain edema (PTBE) tissues. Each assay was performed three times The results from dual luciferase reporter assay, western blotting, and tissue verification suggested that miR-593-5p negatively regulated PLK1 by targeting PLK1 3′UTR. We inferred that lncRNA ENST00000413528 may function as a ceRNA in miR-593-5p/PLK1 pathway when combining the results 3.3 and 3.4.

| Consistent effects of lncRNA ENST00000413528 downregulation and miR-593 upregulation on the cell growth, cell cycle, and apoptosis in U251 and LN229 cells
We have demonstrated that the downregulation of ENST00000413528 affected the growth and apoptosis of U251 and LN229 cells and that there was a negative correlation between ENST00000413528 and miR-593. We then verified a regulation pathway via ENST00000413528/miR-593/PLK1. Therefore, to further clarify the effects of ENST00000413528 and miR-593, we designed four cell groups (ie, cells with the ENST00000413528 knockdown or miR-593 upregulation and their respective negative controls) to carry out the following experiments. The CCK-8 assay showed similar proliferation inhibitory effects in si-Lnc and miR-593 mimics cell groups when compared with the NC groups ( Figure 5A; P < 0.05). A significant decrease in the number of colonies in U251 and LN229 cells was observed with either downregulating ENST00000413528 or upregulating miR-593 ( Figure 5B; P < 0.05). We also found that the cell cycle status was altered by transfecting with si-Lnc or miR-593 mimics versus that found in the NC groups, with an inhibition occurring in the S phase and a relative percent increase happening in the G0/G1 phase; furthermore, there was no significant difference in the effects of the ENST00000413528 knockdown and miR-593 overexpression on cell cycle regulation in U251 and LN229 cells ( Figure 5C; P < 0.05). For antiapoptosis effects, we evaluated the apoptosis cells ( Figure 5D) and caspase 3/7 activity ( Figure 5E) in the 4 groups. The number of apoptosis cells declined significantly in both the si-Lnc and miR-593 mimic groups (P < 0.05). The caspase 3/7 activity increased in the 2 groups but did not result in a significant difference within the two groups (P < 0.05).

| Reversed effects of PLK1 on ENST00000413528/miR-593/PLK1 signaling
Our previous data demonstrated that ENST00000413528 might act as a ceRNA to promote PLK1 expression by sponging miR-593; we performed a restoration experiment for further verification.
Recombinant plasmid pcDNA3.1-PLK1 lacking 3'UTR, which was not able to bind to miR-593, was constructed and transfected into

| D ISCUSS I ON
Recent evidence about lncRNAs has indicated that abnormal lncRNA expression may not only affect the regulation of eukaryotic genomes but also promote the growth of malignant cells, leading to tumor progression. 35,36 Researchers usually use biological informational analysis, expression level regulation, dual luciferase reporter assay, RNA immunoprecipitation assay, and restore assay to conclude a ceRNA function. 37,38 The current study provides evidence for overexpressed ENST00000413528's role as a tumorpromoting gene in glioma tissues and U251 and LN229 cell lines.
As shown in Figures 3A and 4A, the binding sites between lncRNA F I G U R E 5 Effects comparison between ENST00000413528 downregulation and miR-593 upregulation. Cells were divided into four groups with either the si-Lnc or si-NC lentivirus infection, miR-593 mimics, or miR-NC transfection. A, The CCK-8 assay was performed at 0, 24, 48, and 72 h after transfection in the U251 and LN229 cells, respectively, and the OD value as detected by the microplate reader of each group was recorded. B, The number of cell colonies in the four groups was compared. C, The relative percentages of the 3-cell cycle status in the four groups in the U251 and LN229 cells were compared. D, The number of apoptosis cells was detected by flow cytometry of the four groups in the U251 and LN229 cells. E, Caspase 3/7 activity was compared in the four groups It is not uncommon for lncRNA to play a role through ceRNA pathways in gliomas, either as a tumor-promoted gene or a tumor-inhibitory gene. LncRNA CASC2 was reported to be identified as a glioma suppressor gene via miR-21, while lncRNA XIST is an oncogene that acts as a molecular sponge of miR-429. 39 Taken together, our results reveal a regulatory axis of lncRNA ENST00000413528/miR-593/PLK1 in glioma, to our knowledge, this is the first research project of lncRNA ENST00000413528 expression profile in glioma. These findings could help us enrich the knowledge on the mechanism of glioma carcinogenesis and find new diagnostic or therapeutic targets related to lncRNA F I G U R E 6 Reversed effects of PLK1 on ENST00000413528/miR-593/PLK1 signaling. Cells were divided into seven groups: a blank group without any treatment; an si-Lnc group with an si-Lnc lentivirus fluid stable infection; an miR-593 group with miR-593 mimic transfection; an si-PLK1 group with si-PLK1 transfection; a pcDNA3.1-PLK1 group with pcDNA3.1-PLK1 (lacking 3′UTR) transfection; an si-Lnc + pcDNA3.1-PLK1 group with pcDNA3.1-PLK1 transfected into si-Lnc stably infected cells; and an miR-593 + pcDNA3.1-PLK1 group with miR-593 mimics and pcDNA3.1-PLK1 cotransfection. A, The relative expression levels of the PLK1 protein in the seven groups, with GAPDH serving as an internal reference. B, The number of cell colonies in the seven groups. C, The number of apoptosis cells detected by flow cytometry in the seven groups. D, Correlation analysis between the relative expression levels of PLK1 mRNA and ENST00000413528 in 16 pair of glioma and paired peritumoral brain edema (PTBE) tissues ENST00000413528, miR-593-5p, or PLK1 for glioma. However, the roles of lncRNAs in gliomas are diverse and the molecular pathway networks involved are also very complex, further studies are also necessary to elucidate to fully understand the mechanisms.

| CON CLUS ION
LncRNA ENST00000413528 is overexpressed in both human glioma tissues and U251 and LN229 cells; additionally, the knockdown of ENST00000413528 in glioma cells inhibits cell proliferation and colony formation abilities and induces the G0/G1 arrest of the cell cycle, and promotes apoptosis via a miR-593-5p/PLK1 pathway.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.