Long Non-coding RNA LINC00114 Facilitates Colorectal Cancer Development Through EZH2/DNMT1-Induced miR-133b Suppression

This study aimed to identify the roles of the long non-coding RNA LINC00114 in colorectal cancer (CRC) development. The expression levels of LINC00114 and miR-133b in CRC were determined by reverse transcription (RT)-polymerase chain reaction (PCR) and the functions of LINC00114 in CRC were evaluated in vitro and in vivo. Methylation-specific PCR assay was performed to detect the miR-133b promoter methylation in CRC cells. Bioinformatics analysis, RNA immunoprecipitation, dual luciferase assay, RNA pull-down, co-immunoprecipitation (IP), and chromatin IP (ChIP) assays were used to elucidate whether LINC00114 could recruit EZH2/DNMT1 and bind to the miR-133b promoter region, leading to dysregulated methylation and the depression of miR-133b. The expression levels of DNA methyltransferases (DNMTs), EZH2, and nucleoporin 214(NUP214) were analyzed by western blotting. Data showed that LINC00114 was highly expressed, whereas miR-133b was downregulated in the CRC tissues and cells. In vitro, silencing LINC00114 inhibited cell proliferation and impeded cell cycle at the G1/S phase by upregulating miR-133b. In vivo, LINC00114 knockdown reduced tumor growth. Further analysis showed that the methylation in miR-133b promoter region was increased in the CRC and silencing LINC00114 increased miR-133b expression through depressing methylation of its promoter region. ChIP-PCR experiments demonstrated that EZH2 and DNMT1 could bind to the miR-133b promoter region and it was abolished by LINC00114 knockdown. sh-EZH2 reversed the overexpression of DNMTs and CRC cell cycle progression induced by the LINC00114 upregulation. LINC00114 could regulate the NUP214 protein expression by sponging miR-133b. These results demonstrated that LINC00114 suppressed miR-133b expression via EZH2/DNMT1-mediated methylation of its promoter region, indicating that LINC00114 might be a potential novel target for CRC diagnosis and treatment.


INTRODUCTION
Colorectal cancer (CRC) is the third leading cause of cancerrelated deaths worldwide (1). Importantly, CRC incidence rates have been increasing rapidly in several countries that have been historically at low risk for CRC (2), and relative survival rates indicate poor 5-year CRC prognosis (1). Although recently developed medical techniques have improved diagnosis and treatment, unfortunately, CRC is usually detected at advanced stages, which significantly impedes its treatment (3). Therefore, elucidating early diagnostic markers of CRC and discovering the pathogenesis of CRC are critical.
Long non-coding RNAs (lncRNAs) are a class of nucleotides, approximately 200 bp in length with no protein-coding capacity. Accumulated evidences indicate that lncRNAs play vital roles in tumor growth and metastasis (4). lncRNAs are categorized into several types and exert their functions through different molecular mechanisms. Notably, lncRNAs were reported to be intimately related to the development and progression of CRC. The lncRNA MAPKAPK5 antisense RNA 1, which is upregulated in CRC, can promote the development of CRC by binding to p21 (5). Moreover, nuclear enriched abundant transcript 1, another lncRNA, facilitates the progression of CRC by binding to DEAD-box helicase 5 (DDX5) via activation of the Wnt/βcatenin signaling pathway (6). In addition, lncRNAs regulate the transcription of downstream molecules by regulating DNA methylation or hydroxymethylation as well as histone acetylation or methylation. Hypermethylation often occurs in the promoter region near the tumor suppressor genes transcription factor 21 (TCF21) and TCF21 antisense RNA inducing promoter demethylation (TARID), resulting in gene silencing. The LncRNA Chaer (cardiac hypertrophy associated epigenetic regulator) can directly bind to the catalytic domain of polycomb repressive complex 2 (PRC2) and interfere with PRC2 binding to target gene sites. This binding suppresses histone H3K27 methylation in the promoter region of cardiac hypertrophy-related genes, thereby affecting their expression (7). However, the roles of lncRNAs in CRC tumorigenesis remain to be elucidated. miR-133b was previously identified as a tumor suppressor and was shown to inhibit cell proliferation and tumorigenesis in several cancers, such as prostate cancer (8), breast cancer (9), ovarian cancer (10), and CRC (11). Additionally, a variety of validated miR-133b targets including nuclear pore protein nucleoporin 214 (NUP214) (12), CXC chemokine receptor-4 (CXCR4) (13), and B cell lymphoma-2 (BCL-2) (14) are involved in mitosis, tumorigenesis, and apoptosis. Although these studies ascribe an antitumor role for miR-133b, no studies to date have investigated lncRNA-mediated epigenetic mechanisms that regulate miR-133b expression in CRC.
In the current study, we elucidated the mechanism underlying the modulation of miR-133b expression by LINC00114, which was determined by bioinformatics analysis. We determined whether LINC00114 suppressed miR-133b expression via DNA methylation in CRC and demonstrated that LINC00114 could directly inhibit CRC progression via miRNA sponging, providing a potential mechanism that might be utilized in CRC diagnosis and treatment.

Tissues and Cells
The human colon epithelial cell line NCM460 and the CRC cell lines HT-29, HCT116, SW620, and LoVo were purchased from the Cell Bank of Chinese Academy of Sciences (Shanghai). NCM460 cells were incubated in McCoy's 5a medium with 10% fetal bovine serum (FBS; Gibco, California, USA). The CRC cell lines were cultured in DMEM medium (Hyclone, Logan, UT, USA) containing 10% FBS. All cells were cultured with 5% CO 2 at 37 • C.
CRC specimens were obtained from Affiliated Hospital of Guilin Medical University, and adjacent normal tissues at least 3 cm away from the tumor border were isolated for analyses. Tissue samples were preserved in liquid nitrogen for transportation and stored at −70 • C. The use of the specimens was approved by the Institutional Review Board of Affiliated Hospital of Guilin Medical University.

In situ Hybridization
For in situ hybridization (ISH), CRC, and adjacent normal tissue specimens were fixed with 4% para-formaldehyde, dehydrated, and embedded in paraffin. The specimens were sliced into 4-µmthick sections and mounted onto charged slides. After dewaxing and hydration, the sections were air-dried and immersed in distilled water containing 3% hydrogen peroxide, followed by immersion in pepsin solution for 30 min at 37 • C. The sections were treated with hybridization buffer for 2 h at 37 • C. Next, hybridization was performed by incubating the sections with the target DIG-labeled probe (AAGAAGCTGCTGAAGAACCCA) at 42 • C overnight, followed by washing with 2×, 0.5×, and 0.2× SSC solution, respectively. The sections were then blocked with biotinylated digoxin (Boster Biological Technology, Wuhan) for 1 h at 37 • C and the streptavidin-biotin-peroxidase complex (Boster) for 20 min at 37 • C. Hybridization signals were detected using DAB (3,3 ′ -diaminobenzidine; P013IH, Auragene), and the sections were counterstained with hematoxylin. After dehydration, the sections were mounted with neutral gum and observed under the microscope.

MTT Assay
Cells were plated into 96-well plates at a density of 5,000 cells/well and were treat with or without miR-133b inhibitor 12 h later. After a another incubation for 24, 48, and 72 h, 10 µl MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) solution (Sangon Biotech, China) was added to each well, the cells were cultured for 4 h at 37 • C, and the medium was removed. Next, the cells were incubated with 150 µl dimethyl sulfoxide solution (MP Biomedicals, USA) for 10 min for cell lysis. The absorbance was measured at 570 nm using the Multiskan MK microplate reader (ThermoFisher Scientific, USA). All experiments were repeated three times.

Flow Cytometry
Forty-eight hours after transfection, the cells were collected, washed three times with cold phosphate-buffered saline (PBS) and fixed with 70% ethanol overnight. Next, the cells were stained with propidium iodide (PI; Keygentec, China) for 30 min at 4 • C in the dark for cell cycle analysis. The cell cycle status was evaluated using a flow cytometer (Beckman Coulter, USA). Apoptosis of cells was assessed 48 h after transfection using an Annexin V-fluorescein isothiocyanate (FITC) apoptosis detection kit. Briefly, the cells were collected, washed twice with cold PBS, resuspended in 100 µL binding buffer, and incubated at room temperature in the dark for 30 min.

Animals
The right flanks of BALB/c nude mice (aged 4-5 weeks) were subcutaneously injected with 1 × 10 6 stable expression of LINC00114 -sh HCT116 cells (n = 4 per group). After 4 weeks, the animals were sacrificed and the weight and volume of tumors were measured. Next, the tumor was cut into two pieces. One piece was fixed with 4% paraformaldehyde and sliced into 5µm-thick sections for immunohistochemistry assay. The other was stored in liquid nitrogen. All animal experiments were approved by the Animal Care Committee of Affiliated Hospital of Guilin Medical University. Immunohistochemistry Staining was performed using 5-µm-thick sections that were blocked with 10% FBS. After washing with PBS, the sections were incubated in TBST containing Ki67 (3174S, Cell Signaling Technologies, Danvers, MA, USA) overnight at 4 • C. Next, the sections were washed with PBS and incubated with the secondary antibody for 30 min, followed by incubation with DAB and hematoxylin for 30 s to visualize the signal.

RNA Immunoprecipitation
RNA immunoprecipitation (RIP) experiments were performed using EZMagna RIP kit (Millipore, Billerica, MA, USA), according to the manufacturer's instructions. Briefly, HCT116 and LoVo cells were lysed using the RIP lysis buffer, and the lysates were incubated with magnetic beads containing specific antibodies (Ago, EZH2) or control IgG (Millipore). The antibody against EZH2 (Abcam) was used at a 1:200 dilution. Next, the beads were washed and incubated with Proteinase K for protein removal. Finally, RT-qPCR assay was used to measure the RNAs.

RNA Pull-Down
Briefly, T7 RNA polymerase (Promega, Madison, WI, USA) and biotinylated RNA-tagged mixtures (Roche, Basel, Switzerland) were used to label the transcriptional LINC00114 in vitro. Biotinylated LINC00114 was then mixed and incubated with cell lysates. Streptavidin magnetic beads (Life Technologies) were added to each binding reaction and further incubated at room temperature for 1 h. Beads were then washed briefly three times and boiled in SDS buffer, and the retrieved protein was detected by standard western blotting.

Chromatin Immunoprecipitation
For chromatin immunoprecipitation (ChIP), the Magna ChIP Kit (Millipore, Bedford, MA) was used in accordance with the manufacturer's instructions. Briefly, CRC cells were treated with 4% formaldehyde, and chromatin fragments with sizes ranging from 200 to 300 bp were generated by sonicating the cell lysates. Next, chromatin was immunoprecipitated with anti-EZH2, anti-DNMT1, anti-DNMT3B, and anti-He3mek27 antibodies, all from Abcam. IgG was used as control. RT-PCR was performed to recover and measure the precipitated chromatin DNA. The forward primer sequences are TCAATGCTATCCCCTTGTTAAACC and the reverse are ACCAA TACCCATGAACAACG. The forward primer sequences of negative control (around −7,000) are GCCAAGTGACACACCACATC and the reverse GCATAGGGCAAGGGCATCTA. The amount of immunoprecipitated DNA was calculated and normalized to the amount of input DNA.

Luciferase Reporter Assay
The predicted miR-133b combining site of LINC00114 was mutated and sequences were generated by chemical synthesis. The Wt-LINC0014 and Mut-LINC0014 overexpression constructs were provided by genepharma (Suzhou, China), with a psi-CHECK2 vector. HCT116 and LoVo cells were maintained and co-transfected with miR-133 mimics, inhibitors, or negative control (NC), respectively, using the Cellfectin R II reagent. After incubation for 48 h, the firefly luciferase activity in cells was determined using the dual luciferase reporter assay system (E1910; Promega).

Statistical Analysis
Data were presented as means ± standard error. All experiments were repeated at least three times. All data was analyzed using GraphPad Prism 7.0 (CA, USA). Multiple comparisons were analyzed using one-way analysis of variance, and differences between groups were compared using Dunnett's multiple comparisons test. P < 0.05 was considered statistically significant.

Expression Levels of LINC00114 and mIR-133b in CRC Tissues and Cell Lines
We previously demonstrated that miR-133b expression was reduced in CRC, although the upstream regulatory mechanism remained unclear (11). Therefore, we first analyzed the TCGA database to see if there was any lncRNAs regulating miR-133b as a upstream regulator and found that there were 1,046 lncRNAs upregulated in CRC. We next investigated the potential targets of miR-133b using the miRanda software and LINC00114 was identified as a potential target of miR-133b. The relatively high expression level of LINC00114 in CRC was also verified in the GEPIA database (http://gepia.cancer-pku.cn/) ( Figure 1A). To verify whether LINC00114 and miR-133b were both dysregulation in CRC, we next performed RT-PCR and  ISH assays and found that LINC00114 was upregulated while miR-133b was downregulated in the CRC specimens compared with the adjacent tissues (Figures 1B-D), suggesting a negative correlation between LINC00114 and miR-133b ( Figure 1E). Additionally, the expression trend of LINC00114 and miR-133b in CRC cells was same with it in tissues (Figures 1F,G).

LINC00114 Regulates CRC Progression Through Inhibiting miR-133b Expression in vitro and in vivo
To determine whether LINC00114 promoted the progression of CRC, we transfected sh-LINC00114 plasmids (LINC00114-sh1 and LINC00114-sh2) into HCT116 and LoVo cells (Figure S1). The results showed that LINC00114 knockdown inhibited the proliferation of these cells (Figures 2A,B) and the result was further confirmed by flow cytometry and TUNEL assays (Figures 2C,D). What's more, silencing LINC00114 disrupted the transition of G1/S in HCT116 and LoVo cells (Figure 2E). We also performed a nude mouse xenograft model to explore the effects of LINC00114 in vivo and found that, consistent with the findings above, LINC00114 suppression significantly reduced the tumor growth, compared with the control group (Figures 3A-C). RT-PCR and immunohistochemistry analysis revealed that the expression of miR-133b was promoted and the signal of Ki67 in the tumors of LINC00114 suppression group was weaker than that in the control group (Figures 3D,E). All these data firmly demonstrated that LINC00114 played vital roles in the growth of CRC in vivo and in vitro.

LINC00114 Regulates the Methylation of miR-133b Promoter Region
To explore the mechanism underlying LINC00114-mediated CRC cell proliferation, we then investigated the localization of LINC00114 (http://www.rna-society.org/rnalocate/) and found that LINC00114 was localized in both the cytoplasm and the nucleus (60 and 40%, respectively) ( Figure 1D). These were further confirmed by RT-PCR assay ( Figure 4A). We next used the dual luciferase reporter and RIP assays to verify that miR-133b and LINC00114 could bind to each other in a direct or indirect manner (Figures 4B-D). Together, these data identified that LINC00114 could bind to miR-133b. Previous studies demonstrated that the methylation level of miR-133b promoter region was high in CRC (11). We further confirmed these using MSP assay, as shown in Figure 4E. The miR-133b methylation level was low in NCM460 while high in CRC cells (LoVo, HCT116, HT-29, and SW620). Considering that LINC00114 depletion increases the expression of miR-133b (Figure 3D), we were urged to explore if LINC00114 regulates the expression of miR-133b by regulating methylation levels in its promoter region. To this end, we transfected LoVo and HCT116 cells with LINC00114-sh1 or treated with RG108, a DNA methyltransferase inhibitor, and found that LINC00114 knockdown and RG108 treatment reduced the methylation level of the miR-133b promoter region both (Figure 4F), as we expected. The treatments also increased the expression of miR-133b in CRC cells (Figure 4G), indicating that LINC00114 might regulate the methylation level in miR-133b promoter region. It is known that DNMT3A and DNMT3B enzymes are important for the DNA methylation (16,17), and DNMT1 also is a designated maintenance DNA methyltransferase, in a replication-dependent manner DNMTs (18,19). To investigate the roles of these DNMTs in the LINC00114-regulated change of miR-133b promoter methylation, we performed western interference plasmids were transfected into HCT116 and LoVo cells and then H3K27me3, DNMT1, and EZH2 were detected using western blotting assay. (C) miR-133b levels were detected by RT-PCR in cells transfected with sh-EZH2. (D) Co-IP analysis showed that EZH2 could bind to DNMT1 and H3K27me3. (E) RNA pull-down assay was performed to detect the binding of LINC00114 with EZH2 and DNMT1. *P < 0.05; **P < 0.01; ***P < 0.001. blotting assays and found that the expression of DNMT1 and DNMT3B were decreased in the cells with LINC00114 silenced or RG108 treatment (Figures 4H,I), suggesting that LINC00114 might regulate the methylation of the miR-133b promoter through methyltransferases.

LINC00114 Regulates Methylation Level of miR-133b Promoter by Binding to EZH2
Several studies had been reported that the function of some lncRNAs was EZH2-depended and participated in tumor progression by recruiting PRC2 (20,21). Therefore, we first explored whether LINC00114 could bind to DNA methyltransferase using online RNA-protein interaction prediction software (http://pridb.gdcb.iastate.edu/RPISeq/). As expected, it showed that LINC00114 might bind to EZH2 and DNMT1 (with RF and SVM scores of >0.5), which did bind to miR-133b (Table S1). We also confirmed that LINC00114 expression was positively correlated with EZH2 using GEPIA (http://gepia.cancer-pku.cn/index.html) (Figure 5A). Further, we found that the expression levels of EZH2, DNMT1, and H3K27me3 were reduced when EZH2 knocked down ( Figure 5B). Also, EZH2 depression promoted the expression of miR-133b ( Figure 5C) and EZH2 indeed binds with DNMT1 and H3K27me3 as shown in the result of Co-IP assay ( Figure 5D). Moreover, the RNA pull-down assay revealed that biotinylated LINC00114 could enrich the EZH2 and DNMT1 proteins in cells ( Figure 5E) and LINC00114 was enriched in the RNA-protein complexes immunoprecipitated by EHZ2 and DNMT1. These results confirmed the interplay between LINC00114, EZH2, and DNMT1.
To investigate whether DNMTs and EZH2 were involved in the LINC00114 -mediated methylation of miR-133b promoter, we performed ChIP assay and it showed that the knockdown of LINC00114 reduced the binding of the DNMTs and EZH2 to the miR-133b promoter region (Figure 6A). This result indicated that LINC00114 altered the EZH2-mediated H3K27me3 trimethylation, thereby leading to the miR-133b repression in hypermethylation-depended manner. In addition, sh-EZH2 neutralized the miR-133b depression-mediated by LINC00114 overexpression (Figure 6B). The TUNEL and MTT assays revealed that LINC00114 overexpression failed to promote the proliferation of CRC cells, owing to EZH2 depression (Figures 6C,D). Rescue experiments demonstrated that sh-EZH2 attenuated the effect of LINC00114 overexpression on DNMT1 expression ( Figure 6E). These results indicated that LINC00114 promoted miR-133b promoter methylation by inducing the EZH2/DNMT1 binding to the miR-133b promoter region, thereby facilitating the development of CRC.

LINC00114 Regulates CRC Cell Proliferation by Sponging miR-133b
Previous study reported that miR-133b could target NUP214, a member of the massive nuclear pore complex, illustrating that the miR-133b-mediated NUP214 downregulation arrested cells in the early mitotic phase and led to aberrant chromosome segregation and cell death (12). Thus, we assessed whether LINC00114 regulated cell cycle progression through NUP214. We next found that both NUP214 mRNA and LINC00114 shared the binding site for miR-133b ( Figure 4D). Furthermore, LINC00114 knockdown facilitated miR-133b expression while depressed NUP214 expression in cells (Figures 7A-C). In addition, the assays revealed that miR-133b silencing promoted the CRC cell growth and that sh-LINC00114 neutralized the proliferation promotion mediated by miR-133b inhibitor (Figures 7D,E), indicating that LINC00114 regulated cell proliferation via sponging miR-133b.

DISCUSSION
Although radiotherapy chemotherapy and primary treatment approaches for CRC can prolong life expectancy and reduce patient burden, the CRC prognosis remains poor (22,23). LncRNAs were recently demonstrated to play vital roles as major regulators of gene expression and cancer cell proliferation (24) in multiple cancers. Several lncRNAs have been identified as regulators in CRC cell proliferation and metastasis (25). In current study, we focused on a cancer-related lncRNA, LINC00114, which we found to be upregulated in CRC. Moreover, we demonstrated that LINC00114 negatively regulated miR-133b expression via DNA methylation, suggesting that LINC00114 contributed to the CRC development as an oncogene. We further found that LINC00114 served as an miR-133b sponge to regulate NUP214 expression, thus participating in the progression of CRC.
On one hand, lncRNAs can competitively bind to miRNAs, regulating the proliferation and metastasis of cancer cells (26). Gastric cancer associated transcript 3, an lncRNA, is upregulated in colon cancer and it combines with miR-3127 via ELAV like RNA binding protein 1 to promote CRC (26). Moreover, nuclear enriched abundant transcript 1 is another lncRNA that can compete with sirtuin 1 to bind miR-34a and regulate the Wnt/β-catenin signaling pathway and promote the development of CRC (27). In the current study, we identified that knockdown of LINC00114 significantly increased the expression of miR-133b, and LINC00114 was inversely correlated with miR-133b in CRC. Numerous studies demonstrate that miR-133b plays an important role in CRC progression by inhibiting cell proliferation and metastasis (11,12,28). Our results showed that the knockdown of LINC00114 inhibited CRC cell proliferation by negatively regulating miR-133b expression, indicating that LINC00114 might promote the development of CRC by suppressing miR-133b.
Conversely, lncRNAs can regulate gene expression through various mechanisms, such as chromatin modification and preand post-transcriptional regulation. Growing evidences indicate that some lncRNAs regulate the expression of downstream target genes by DNA methylation (29,30). In this study, we showed that silencing LINC00114 decreased DNA methylation in the miR-133b promoter region. Previous study reported that lncRNAs were associated with several RNA-binding proteins including EZH2, SUZ12, LSD1, staufen double-stranded RNA binding protein 1, and DNMT1 (31,32). EZH2 is a component of the methyltransferase in PRC2 and regulated by lncRNAs at the transcriptional level. EZH2 induces H3K27 methylation; it also induces LSD1 to regulate H3K4 demethylation (33). EZH2 usually mediates the methylation of the second lysine of histone H3 and is associated with tumor progression, angiogenesis, inflammation (34). Moreover, EZH2 may also recruit DNMT1 to the promoter region of its target genes. DNMTs and methyl-CpG binding protein 2 were reported to play an important role in methylation modification (35)(36)(37). These studies raise the possibility that lncRNAs may have a primary role in recruiting polycomb proteins to their target genes (32,36). In our study, we found that LINC00114, together with EZH2 and DNMT1, played a direct role by online RIP-seq data. We hypothesized that LINC00114 recruited EZH2 and DNMT1 to the miR-133b promoter region for its hypermethylation and repression of miR-133b expression. To validate the relationship among EZH2, LINC00114, and DNMTs, we performed an RNA pull-down assay and confirmed that EZH2 and DNMT1 were binding partners of LINC00114. RT-PCR analysis confirmed that knockdown of LINC00114 affected the expression of DNMTs and miR-133b. In agreement with previous studies which reported that EZH2 could regulate DNA methylation through DNMTs (38), we determined that silencing EZH2 resulted in downregulation of DNMT1 and upregulation of miR-133b expression. Furthermore, knockdown of EZH2 reversed the LINC00114 -induced DNMT1 upregulation and cell cycle progression. Additionally, knockdown of LINC00114 reduced the containing of EZH2 and DNMT1 that bind to the promoter region of miR-133b. These results indicated that LINC00114 might regulate the expression of miR-133b by binding to EZH2 and DNMT1.
It was reported that miR-133b impacted the development of CRC by regulating NUP214. Nucleopore is essential for many species to enter the nucleus, and NUP214 regulates mitosis and promotes carcinogenesis (39). NUP214 is involved in impeding mitotic progression and is a reported target of miR-133b; it is also highly expressed in HCT116 cells (12). In the current study, we demonstrated that LINC00114 could regulate NUP214 protein expression by sponging miR-133b. In addition, miR-133b and NUP214 act as key factors in the pathogenesis of CRC. These findings also suggest that LINC00114 affects the development of CRC by competitively binding to miR-133b through NUP214.
In conclusion, the current study demonstrated that LINC00114 expression was increased in CRC, which might represent an unfavorable prognostic factor for CRC. In addition, LINC00114 depletion inhibited cancer cell proliferation both in vitro and in vivo. Moreover, the oncogenic effects of LINC00114 were facilitated by its binding to EZH2 and DNMT1 and consequent induction of miR-133b promoter methylation which led to the downregulation of miR-133b. Our results also revealed that LINC00114 served as a molecular sponge for miR-133b, reversing its ability to repress NUP214 protein.
Overall, these findings reveal novel therapeutic targets that should be considered for CRC.

DATA AVAILABILITY STATEMENT
Publicly available datasets were analyzed in this study. This data can be found here: http://gepia.cancer-pku.cn/.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by The Ethics Committee of the Guilin Medical University. The patients/participants provided their written informed consent to participate in this study. The animal study was reviewed and approved by The Ethics Committee of the Guilin Medical University. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.