Introduction
It is estimated that approximately 90% of human genome DNA sequence is actively transcribed, but only 2% of the genes encode proteins. The majority of the transcripts that do not encode proteins are defined as non-coding RNAs (ncRNAs) [
1]. Most of ncRNAs belong to long non-coding RNAs (lncRNAs) or microRNAs (miRNAs). lncRNAs are another type of regulatory ncRNAs. LncRNAs are usually longer than 200 nucleotides with lack of protein-coding capability, but lncRNAs can interact with proteins [
2]. LncRNAs were previously considered as transcriptional noise, but recent studies have shown that lncRNAs play important roles in diverse epigenetic processes, including transcriptional regulation, histone modification, RNA splicing, and chromatin remodeling [
3,
4]. LncRNAs have been shown to regulate a wide variety of cellular processes, including cell growth, stem cell differentiation, cell cycle, apoptosis, metabolism, and cancer progression [
5‐
9].
Particularly, studies have demonstrated that lncRNAs are aberrantly expressed in a variety of cancers and play key vital roles in promoting tumor initiation and progression [
10,
11]. Therefore, functional lncRNAs are potential targets for cancer diagnosis and prognosis. Gastric cancer (GC) is a considerable health issue worldwide. It is the fifth most frequently diagnosed cancer and one of the leading causes of death due to cancer in the world [
12]. In China, GC is the second most frequently diagnosed cancer and the second leading causes of cancer death [
13]. In 2018, over 1,000,000 cases of newly diagnosed cancer are GC. It is estimated that 783,000 people died of GC in 2018, equating to 1 in every 12 deaths globally. The outlook for patients with metastatic GC is unfavorable, with median survival usually less than 1 year [
14,
15]. According to American Cancer Society, the 5-year relative survival rate is 69% for localized stage, 31% for regional state, and 5% for distant stage. Thereby, it is urgent to explore novel biomarkers and therapeutic targets for GC.
Several lncRNAs such as UCA1, MALAT1, HOXA11-AS, and ZEB1-AS1 have been proposed as potential diagnostic and prognostic biomarkers in GC [
16]. Gao et al. have shown by meta-analysis that numerous lncRNAs are abnormally over-expressed in GC [
17]. These lncRNAs includes AFAP1-AS1, ANRIL, CASC15, CCAT2, GAPLINC, H19, HOTAIR, HOTTIP, LINC00673, MALAT1, MEG3, PANDAR, PVT1, Sox2ot, UCA1, XIST, ZEB1-AS1 and ZFAS1, and the lncRNAs are associated with GC patient survival. Particularly, the expression levels of AFAP1-AS1, CCAT2, LINC00673, PANDAR, PVT1, Sox2ot, ZEB1-AS1 and ZFAS1 can be used to predict the GC prognosis. However, only a few lncRNAs that are associated with GC have so far been functionally characterized. Furthermore, the mechanism by which lncRNAs regulate GC remains to be fully elucidated.
LINC01564 is a lncRNA which is closely related to glutamate-cysteine ligase catalytic subunit [
18]. LINC01564 has been shown to be associated with a variety of cancers. For example, Ke et al. found that high expression level of LINC01564 is associated with poor overall survival in testicular cancer patients [
19]. Lee et al. reported that LINC01564 is elevated in prostate cancer patient samples and cell lines [
20]. LINC01564 is also upregulated urine samples of prostate cancer patients compared to healthy individuals. The integrity of LINC01564 has been confirmed by 3′RACE and 5′RACE experiments [
21].
In the present study, we aim to identify the function of lncRNA LINC01564 in promoting GC metastasis. In addition, we investigate whether LINC01564 affects the GC tumor progression through its interaction with transcription factor POU2F1. POU2F1, also known as OCT-1, has been reported to be associated with GC proliferation, migration, and epithelial-to-mesenchymal transition (EMT) via the enhancement of DLX6-AS1 expression by targeting the promoter region [
22].
Materials and methods
Go annotation of survival-related genes
Gene Ontology (GO) enrichment analysis was performed on the up-regulated genes in metastatic GC. DAVID (The Database for Annotation, Visualization, and Integrated Discovery) software was used to annotate and visualize GO terms. The data used for Go enrichment analysis was from Gene Expression Profiling Interactive Analysis (GEPIA) and TCGA.
Patients
Tumor samples and matched adjacent normal tissues were collected with informed consent from gastric cancer patients during surgeries performed between January 2019 to January 2020 at Hunan Cancer Hospital (Changsha, China). The current study was approved by the medical ethical review committees at Hunan Cancer Hospital, and written informed consent were obtained by all patients. There was no other therapy before surgery in recruited patients. The tumor tissues were immediately preserved in liquid nitrogen after surgery.
The human GC cell lines, including SGC-7901, MKN-45, BGC-823, and MKN-28, as well as the normal gastric epithelial cell line (GES-1) were purchased from American Type Culture Collection (ATCC). BGC823, MKN-45 and MKN28 cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Invitrogen, Waltham, MA, USA); GES-1 and SGC7901 cells were cultured in a Dulbecco-modified Eagle medium (Invitrogen). All the media were supplemented with 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO, USA) and 100 U penicillin–streptomycin. Cells were cultured under 5% CO2 at 37 °C.
Total RNA was extracted from cells using Trizol reagent (TaKaRa Biotechnology, Dalian, China) according to the manufacturer’s protocol. Total RNA was reverse transcribed with a PrimerScript RT-PCR kit (Takara Biotechnology, Dalian, China). Real time qPCR was conducted using a standard SYBR Green PCR kit (Roche, Upper Bavaria, Germany) protocol with a CFX real-time instrument (Bio-rad, Hercules, CA, USA). The relative expression was calculated using the 2−∆∆Ct method. The transcription level of GAPDH was used as an internal control. The primer sequences are listed in Additional file
1.
RNA FISH
The procedure was performed by using a Ribo FISH Hybridization Kit by RiboBio (Guangzhou, China). MKN-45 and SGC-7901 cells seeded on the glass coverslips (0.8 × 0.8 cm) were cultured to 60–70% confluence. The coverslips were washed with a solution of 0.5% Triton X-100 in 1× PBS. The coverslips were incubated with LINC01564, POU2F1 mRNA or U6 oligodeoxy-nucleotide probes by RiboBio (Guangzhou, China) with a hybridization solution containing 1% blocking solution in a humid chamber at 37 °C overnight. The following day, the coverslips were rinsed 3 times for 15 min (5 min each time) at 42 °C with a solution of 0.1% Tween-20 in 4× sodium citrate buffer (SSC), once for 5 min in 2× SSC and once for 5 min in 1× SSC in dark conditions. Finally, after rinsing with 1× PBS for 5 min 3 times at room temperature and re-staining by DAPI (Cell Signaling Technologies, Danvers, USA), the coverslips were observed and photographed using a fluorescent microscope (Leica, Wetzlar, Germany). Images were analyzed using Image-Pro Plus 6.0 software (Media Cybernetics, Rockville, USA). For double-fusion FISH, the protocol was similar. The only difference was that the coverslips were incubated with both LINC01564 and POU2F1 oligodeoxy-nucleotide probes.
Immunofluorescence (IF)
Cells were seeded into 6-well plates with the autoclaved cover glasses placed at the bottom of the wells. After cells grown to 50% confluence on the cover-slips, PBS rinsing was performed, followed by 4% paraformaldehyde (PFA) fixation at RT for 10 min. After washing, cells were permeated with PBS which containing 0.5% Triton X-100 for 20 min and blocked with 3% bovine serum albumin (BSA, Thermo Fisher Scientific) for 1 h at RT. After rinsing with PBS, primary antibody, including anti-E-cad (#20874-1-AP, 1:1000; Ptgcn), and anti-Vimentin (#60330-1-Ig, 1:1000; Ptgcn) antibodies were added to the cells attached on the cover-slips at 4 °C for overnight incubation. Alexa Fluor® 488-conjugated secondary antibody (ab150077, Abcam) incubation was conducted at RT for 1 h. Then, the cell nuclei were counterstained with 4-6-diamidino-2-phenylindole (DAPI) reagent (Sangon Biotech) for 10 min and the slices were mounted. Finally, images were observed and photographed using a confocal microscope (Zeiss, Jena, Germany).
Cell transfection
The short interfering RNA (siRNA) targeting LINC01564 were synthesized by RiboBio (Guangzhou, China). The siRNA sequences for LINC01564 were as follows: si-LINC01564-69: guide 5′-UUUCAGUUUGUUUUCCAUCCC-3′, passenger 5′-GAUGGAAAACAAACUGAAACU-3′; si-LINC01564-214: guide 5′-UGAUUUCUUGCCUUUUGAGCU-3′, passenger 5′-CUCAAAAGGCAAGAAAUCAGC-3′; si-LINC01564-506: guide 5′-UUUUGUCAAUAAAUAGGUGAU-3′, passenger 5′-CACCUAUUUAUUGACAAAAUU-3′. Oligonucleotide transfection was performed using Lipofectamine 2000 reagent (Invitrogen, United States). Short hairpin RNA (shRNA) directed against LINC01564 were synthesized by GeneChem (Shanghai, China) and were inserted into the vector. Lentiviral vector DNA and package vectors were transfected into HEK-293T cells. At 72 h after transfection, lentivirus supernatants were harvested and used to infect SGC-7901 and MKN-45 cells. Stable shRNA cell lines were generated following selection with 1 μg/mL puromycin.
Protein extraction and Western blot analysis
Cells were washed three times with PBS and collected in RIPA lysis buffer (Beyotime Biotechnology, China) supplemented with a protease inhibitor cocktail (Calbiochem, United States). Protein concentration was determined by staining with Coomassie Blue (Beyotime Biotechnology, China). After electrophoresis, the protein was transferred to a polyvinylidene fluoride membrane (Merck Millipore, Germany). After blocking with 0.1% Tween 20 (TBS-T) in Tris-buffered saline containing 5% skim milk for 1 h at room temperature, the primary monoclonal antibody, including anti-POU2F1 (#10387-1-AP, 1:500; Ptgcn), anti-E-cad (#20874-1-AP, 1:1000; Ptgcn), anti-Vimentin (#60330-1-Ig, 1:1000; Ptgcn), and anti-β-actin (#66009-1-Ig, 1:2000; Ptgcn), was added to the membrane and incubated overnight at 4 °C. The next day, the membrane was incubated with secondary antibody Goat Anti Rabbit IgG/HRP for 1 h at room temperature and the signal was detected in a Bio-Rad ChemiDoc XRS imaging system. The ratio of the gray value of the target protein to the gray value of β-actin indicates the relative amount of protein.
RNA stability assay
SGC-7901 and MKN-45 cells with stably expressed siRNAs against LINC01564, LINC01564 overexpression or NC were seeded into 6-well plates to get 50% confluency after 24 h. Cells were treated with 5 μg per mL actinomycin D and collected at indicated time points. The total RNA was extracted by miRNeasy Kit (Qiagen) and analyzed by RT-PCR. RNA stability profiling was generated from two biological replicates. The turnover rate and half-life of mRNA was estimated according to a previously published paper [
23].
RNA pull down
LINC01564 was in vitro transcribed and biotin‐labeled using the Biotin RNA Labeling Mix (Roche) and T7 RNA polymerase (Roche). After being treated with DNase I (Takara) to remove DNA and purified using RNeasy Mini Kit (Qiagen, Shenzhen, China), 3 µg of purified RNA was incubated with 1 mg of whole‐cell lysate from SGC-7901 and MKN-45 cells for 1 h at 25 °C. Next, the complexes were extracted by streptavidin agarose beads (Invitrogen) and the RNA present in the pull‐down material was detected by qPCR as described above.
RNA immunoprecipitation (RIP)
SGC-7901 and MKN-45 cells were used to carry out RIP assay with the Magna RIP RNA‐Binding Protein Immunoprecipitation Kit (Millipore, Bedford, MA) and an AGO2 specific antibody (Millipore) following the instructions. RIP‐derived RNA was detected by qPCR as described above.
Chromatin immunoprecipitation assay (ChIP)
ChIP assay was performed via a commercially purchased chromatin immunoprecipitation kit (Millipore, Temecula, CA), using anti-POU2F1 or anti-IgG antibodies. Cells were first cross-linked for 10 min by adding formaldehyde directly to culture medium to a final concentration of 1%. Cross-linked cells were then washed twice with cold PBS (with protease inhibitors), scraped, pelleted, resuspended in 200 μL SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris–HCl, pH 8.0), and incubated for 10 min on ice. The lysates were then sonicated for five cycles of 30 s each, resting on ice for 1 min between cycles. After sonication, the samples were centrifuged and the supernatants diluted tenfold in ChIP dilution buffer with protease inhibitors. Cross-linked chromatin was incubated overnight with 5 μg anti-POU2F1 or anti-IgG of 1 mL at 4 °C. Antibody-protein-DNA complexes were isolated by immunoprecipitation. After extensive washing, pellets were eluted by freshly prepared elution buffer (1% SDS, 0.1 M NaHCO
3). Formaldehyde cross-linking was reversed by 5–12-h incubation at 65 °C after adding 20 μL 5 M NaCl. Samples were purified through PCR purification kit columns (Qiagen, Chatsworth, CA) and used as a template in PCR. The primer sequences are listed in Additional file
1.
DNA affinity precipitation assay (DAPA)
The 31-wt-S oligonucleotide biotinylated at the 5′ end (Operon Biotech., Inc.) was annealed with the anti-sense oligonucleotide. DAPA was performed as described previously [
24] with some modifications. Briefly, the assays were done in a final volume of 400 μL of buffer D (20 mM HEPES, 10% glycerol, 50 mM KCl, 0.2 mM EDTA, 1.5 mM MgCl
2, 10 μM ZnCl
2, 1 mM DTT and 0.25% Triton X100, pH 7.9), by mixing 4 μg of biotinylated double stranded 31-wt oligonucleotides with 20–30 μg of nuclear extracts. The mix was incubated on ice for 45 min and then was added to the buffer D with equilibrated Streptavidin coated Magnetosphere particles (SMPs) (Promega). The mixture was incubated at room temperature for 2 h with continuous agitation. SMPs were then captured using the magnetic stand and the supernatant removed without disturbing the SMPs pellet. Particles were washed four times with the buffer D and the final pellet obtained was resuspended in 2× SDS-PAGE loading buffer and boiled for 5 min to uncouple the oligonucleoide bound proteins. After capturing the SMPs using the magnetic stand, the supernatant was loaded on SDS-PAGE gel and Western analysis was performed.
MTT assay
Media was discarded from cell cultures. 50 µL of serum-free media and 50 µL of MTT solution were added into each well. The plate was incubated at 37 °C for 3 h. After incubation, 150 µL of MTT solvent was added into each well. The plate was wrapped in foil and shaken on an orbital shaker for 15 min. Absorbance was read at OD = 590 nm. The absorbance was proportional to cell number.
Cell invasion assay
Cells were cultured to 70 to 80% confluency and sub-cultured into the invasion chamber. Cell invasion chambers were incubated overnight in a humidified tissue culture incubator at 37 °C, 5% CO2 atmosphere. Non-invading cells were removed. Cells in the invasion chambers were stained by adding 0.5 mL of each solution from the Diff-Quik kit. The permeable supports were subsequently transferred through each stain solution and the two plates of water. The permeable support membrane was allowed to air dry. The invaded cells were observed under the microscope.
Exponentially growing cells were re-plate in dishes. The dishes were left in the incubator until cells in control dishes had formed sufficiently large clones. After treatment, the number of cells in the resulting cell suspension was counted using a Coulter counter, and the cell suspension was diluted in sterile tubes so that 100 or up to 104 cells after the treatment could be pipetted into the test wells. Cells were re-plated immediately after treatment. The dishes were then put in an incubator and leave there until cells in control dishes had formed sufficiently large clones. The medium above the cells was removed and the cells were rinse with PBS. After the PBS was removed, 2–3 mL of a mixture of 6.0% glutaraldehyde and 0.5% crystal violet were added. After 30 min, the glutaraldehyde crystal violet mixture was carefully removed, and the cells were rinsed with tap water. The dishes with colonies were allowed to dry in normal air at room temperature.
Tumor xenograft in nude mice
SGC-7901 cells with POU2F1 (OE), POU2F1 (OE) plus LINC01564 (KD) or control vector, A total of 40 BALB/C nude mice of either gender (age: 4 weeks, weight: 18–22 g) (Hunan SJA Laboratory Animal Co., Ltd., Changsha, Hunan, China) were reared in the Specific Pathogen Free (SPF)-level condition and randomly classified into 3 groups with 4 mice in each group. The 3 groups were injected the cells as follows: (a) SGC-7901 cells with POU2F1 (OE); (b) SGC-7901 cells with POU2F1 (OE) and LINC01564 (KD); (c) SGC-7901 cells with control vector. All mice were euthanized by cervical dislocation at the end of experiment, and GC tumors were collected.
Hematoxylin and eosin (HE) staining
Xenografted tumor tissues were fixed with 4% paraformaldehyde (BOSTER, Wuhan, China), embedded in paraffin, sectioned, and visualized at 40× and 100× magnification after HE staining. We analyzed the number of metastatic foci and the relative area of metastatic foci in the field of HE staining of each section, using the method described in the reference [
25]. The area of metastatic lesions displayed by HE staining was measured by pathological graphic analysis software, and compared with the total HE staining area. The results were presented as a percentage (the total he area was set as 100%).
Immunohistochemistry (IHC)
The IHC analyses was conducted as described previously [
26]. Anti-MMP9 and anti-KI67 antibodies were used to detect the expression of MMP9 and KI67 in GC tissues.
Statistical analysis
The statistical analysis for all data was analyzed using SPSS 21.0 software (IBM, Armonk, NY, USA). Each experiment was repeated 3 times. The measurement data were expressed as mean ± standard deviation. The comparison between two groups was analyzed by t test, and among multiple groups was analyzed by one-way analysis of variance (ANOVA) followed by post hoc Dunnett’s test. A P value < 0.05 was considered to be significantly different.
Discussion
So far, the clinical treatment of GC remains limited. It is urgent to explore novel biomarkers and therapeutic targets for GC. Several lncRNAs, such as UCA1, MALAT1, HOXA11-AS, and ZEB1-AS1 have been reported to play important roles in GC [
16]. However, only several lncRNAs have been functionally characterized in GC metastasis. The molecular mechanisms exerted by lncRNAs are diverse. In this study, we systematically integrated GC clinical information and gene expression data from TCGA and GEPIA. By comparing cancer samples with normal samples, we identified 5 GC-related lncRNAs, including CATIP-AS2, TTC3-AS1, LINC01993, LINC01564 and LINC02015. Predicted using JASPAR database, we identified that the level of LINC01564 is associated with GC metastasis and LINC01564 thus may serve as a potential biomarker for diagnosis and prognosis of GC. By exploring the associations between expression of LINC01564 and clinicopathological features in GC patients, we found that the expression of LINC01564 was associated with GC metastasis. This study for the first time confirmed the interaction between LINC01564 and POU2F1.
Using INTARNA algorithm as described by Mann et al. [
27], which enables fast and accurate prediction of RNA–RNA hybrids by incorporating seed constraints and interaction site accessibility, we predicted the binding site of LINC01564 on POU2F1-mRNA. Using immunoprecipitation and RNA pull down assay, we confirmed that LINC01564 complementary bond to the 3′UTR of transcription factor POU2F1 to form an RNA duplex and thus stabilizing POU2F1 and increasing POU2F1 protein level. Due to positive feedback, the level LINC01564 also increased. We hypothesize that the binding of LINC01564 to the 3′UTR of POU2F1 prevents the mRNA of POU2F1 from binding miRNAs and thus stabilizing the mRNA. Collectively, our findings suggested that LINC01564 interacted with POU2F1 mRNA 3′UTR and increased POU2F1 level by stabilizing its mRNA. Similar mechanism has also been reported in previous studies. Jia et al. have demonstrated that long noncoding RNA PXN‐AS1‐L interacts with SAPCD2 3′UTR and stabilizes SAPCD2 mRNA [
28]. PXN‐AS1‐L thus promotes the malignancy of nasopharyngeal carcinoma cells via upregulation of SAPCD2. Yue et al. found that Down syndrome cell adhesion molecule antisense RNA 1 (DSCAM-AS1) binds to the 3′UTR mRNA of dCTP pyrophosphatase 1 (DCTPP1) and thus inhibits the binding of miRNAs [
29]. Moreover, we found that POU2F1, in turn, modulates the transcription of LINC01564. The results indicate the tumorigenic activity of POU2F1 in GC and the association with the increase of LINC01564. Similarly, Xie et al. have found that the expression of POU2F1 is regulated by long non-coding RNA TUG1 and their interaction is associated with tumorigenesis of human osteosarcoma [
30].
In vitro and in vivo experiments were performed to access the impacts of LINC01564 and POU2F1 on GC proliferation and metastasis. The results showed that LINC01564 promoted the proliferation, invasion and migration of GC cells. Previous studies highlighted the positive role of EMT in tumor metastasis, meanwhile lncRNAs could modulate cancer metastasis via affecting EMT [
22]. Particularly, we further measured the expression level of EMT markers. The results demonstrated that LINC01564 could regulate GC cells metastasis via affecting EMT. In vivo assays showed that tumor with higher expression of POU2F1 had stronger staining of ki67 and produced more MMP9. ki67 is an important mark to determine cell proliferation especially in tumor tissue. MMP9 is a member of the MMP family. Studies have shown that MMP9 is overexpressed in various cancers and is related to tumor metastasis and invasion [
31‐
35]. However, silencing of LINC01564 significantly reversed the oncogenic roles of POU2F1 in GC in vitro and in vivo, which supported that the cancer-promoting effects of POU2F1 was associated with LINC01564. Collectively, the results supported that POU2F1 was an important mediator of the roles of LINC01564 in GC. Functional experiments further revealed that overexpression of LINC01564 promoted GC cell proliferation, migration, and invasion in vitro
. While LINC01564 knocking down repressed GC cell proliferation, migration, and invasion.
Similarly, several recent studies also show that the interactions between long non-coding RNAs and protein-coding genes are associated with cancer cell proliferation and metastasis. For example, the investigation by Jia et al. has shown that long non-coding RNA PXN-AS1-L promotes the malignancy of nasopharyngeal carcinoma cells via upregulation of SAPCD2 [
28].
This study has some limitations. Although it has been demonstrated that LINC01564 promotes GC proliferation and metastasis via POU2F1, the involvement of other miRNAs or transcription factors cannot be excluded. Further research is needed to address the question.
In summary, our results indicate that the interaction between LINC01564 and POU2F1 promotes the proliferation, migration and invasion of GC cells through positive feedback. By systematically integrating bioinformatics and experimental methods, our findings firstly revealed that LINC01564 and POU2F1 have important oncogenic functions in GC. Our study provides novel insights on the functional characterization of LINC01564 in GC. In addition, LINC01564 is probably a potential biomarker for diagnosis and prognosis of GC in future studies.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.