Downregulated Reprimo by LINC00467 participates in the growth and metastasis of gastric cancer.

Gastric cancer (GC) as an aggressive malignancy still causes a global health problem. It has been documented that long noncoding RNAs are involved in GC development. Therefore, this research was designed to explore the role of LINC00467 in the growth and metastasis of GC. The expression of LINC00467 and Reprimo in GC tissues and cells was detected. The binding relationship among LINC00467, DNA methyltransferase 1 (DNMT1) and Reprimo was assessed following. Reprimo promoter methylation was detected by methylation sequencing. GC cell lines overexpressing or knock downing LINC00467 were constructed for pinpointing the effect of LINC00467 on cell functions as well as growth and metastasis of GC cells in vivo. LINC00467 was highly expressed, whereas Reprimo was poorly expressed in GC tissues and cells. Mechanically, LINC00467 promoted the methylation and decreased the expression of Reprimo promoter by recruiting DNMT1 in GC cells. Knockdown of LINC00467 diminished the malignant properties of GC cells. Knockdown of LINC00467 reduced tumorigenesis and metastasis of GC cells in vivo. LINC00467 might exert oncogenic effects in GC via Reprimo downregulation by recruiting DNMT1.

Highlights

LINC00467 is negatively correlated with Reprimo in GC.

LINC00467 recruits DNMT1 to promote Reprimo promoter region methylation.

LINC00467 aggravates GC cell malignant behaviors via the DNMT1/Reprimo axis.

LINC00467 contributes to growth and metastasis of GC via the DNMT1/Reprimo axis.

The study provides novel theoretical basis on occurrence and development of GC.

Introduction

As a malignancy with high aggressiveness, gastric cancer (GC) is heterogeneous and still poses a global health problem to date [1]. Because of its frequent diagnosis at advanced stage, GC exhibits high mortality, which makes it rank as the third cancer regarding to deaths, with 783,000 deaths reported worldwide in 2018 [2]. Moreover, chronic infection with Helicobacter pylori is the leading cause of GC, accounting for approximately 89% of distal GC cases globally [3]. Currently, there exist several treatment modes for GC, like surgery, chemotherapy, chemoradiotherapy, targeted therapy, and immune checkpoint inhibition [4]. Unfortunately, although there has been a steady decline in the incidence and mortality rates of GC in most countries, more cases of GC may be seen in the future owing to aging populations [5]. Therefore, there is ongoing need for more effective diagnosis and treatment for GC to deepen the knowledge about the molecular mechanism underlying GC.

As reported, there has been extensive discussion about the influence of long noncoding RNAs (lncRNAs) on GC progression and metastasis [6]. For instance, lncRNA gastric cancer metastasis-associated lncRNA has emerged as an oncogene in GC [7]. Additionally, lncRNA HOXA11-AS has been documented to accelerate GC cell proliferation and invasion [8]. Importantly, the research conducted by Deng et al. revealed that LINC00467 might assume a role in GC development [9]. Furthermore, lncRNAs may function as guides to lead modulatory proteins to the promoter region of targeted genes, thus orchestrating gene expression [10]. Interestingly, a prior research indicated that LINC00467 bound to DNA methyltransferase 1 (DNMT1) to decrease p53 expression in glioma cells [11]. Reprimo is a cytoplasmic protein in the family of molecules manipulated by p53 that depresses cell cycle progression [12]. The Reprimo gene family is a group of single exon genes existing only within the vertebrate lineage, two out of three members of which appear in humans and promote cell cycle arrest at G2/M in response to p53 expression [13]. Intriguingly, as a DNA damage-inducible gene, Reprimo exerts anti-oncogenic effects and is suppressed by promoter methylation in GC cells [14].

In this context, we hypothesized that LINC00467 might recruit DNMT1 into promoter region of Reprimo to mediate Reprimo expression, thus participating in GC development. Therefore, tissue, cell, and animal experiments were implemented here to verify this hypothesis, thus displaying a novel insight into candidate targets for GC treatment.

Materials and methods

Ethics statement

This study was ratified by the Ethics Committee of Cangzhou Central Hospital (No. 2017–0928-36) with conforming to the Declaration of Helsinki. All participants or their guardians provided signed informed consent prior to research. Animal experiments were implemented under ratification of Animal Ethics Committee of Cangzhou Central Hospital (No. 2017–1012-54) and in the light of the recommendations of the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health. We made adequate measures to limit animals’ pain.

Bioinformatics analysis

The Gene Expression Profiling Interactive Analysis (GEPIA) database was adopted to retrieve differentially expressed genes (DEGs) of GC samples and normal samples collected in The Cancer Genome Atlas (TCGA) and The Genotype-Tissue Expression (GTEx). Genes significantly negatively correlated with DNMT1 in GC included in TCGA were searched by LinkedOmics database. The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of DNMT1 downstream candidate genes was analyzed by KOBAS3.0 database. Subsequently, the co-expression of LINC00467 and DNMT1 with Reprimo in GC included by TCGA was searched by Starbase database.

Clinical samples

From January 2018 to January 2019, 52 cases of GC tissues and corresponding normal tissues were obtained by surgical resection from patients in Cangzhou Central Hospital. The clinical data are shown in Supplementary Table 1. All patients did not receive preoperative radiotherapy or chemotherapy before surgical resection. After tissue samples were obtained, total RNA was extracted by Trizol kit (Invitrogen, Carlsbad, CA, USA), frozen in liquid nitrogen and stored in refrigerator at −80°C.

Cell culture

The immortalized human gastric epithelial cell line GES-1 and four human GC cell lines (NCI-N87, OCUM-1, MKN-74 and HGC-27) were acquired from Nanjing Cobioer Biosciences Co., Ltd., (Nanjing, Jiangsu, China). The incubation of all cells was implemented with Roswell Park Memorial Institute-1640 medium encompassing 10% fetal bovine serum (FBS), 100 U/mL penicillin and 100 mg/mL streptomycin) in 5% CO2 incubator at 37°C [15].

Lentivirus transfection

Lentiviral vector (LV)-LINC00467 lentivirus for LINC00467 overexpression, short hairpin RNA (shRNA) (sh-LINC00467) lentivirus for LINC00467 knockdown, LV-Reprimo lentivirus, sh-Reprimo lentivirus, sh-DNMT1, negative control (NC) lentivirus (LV-NC) and sh-NC were bought from Obio Technology Corp., Ltd. (Shanghai, China). The OCUM-1 cells were seeded into a 6-well plate at 5 × 105 cells/well. Following 24-hour adherence, the virus solution and 6 μg Polybrene were supplemented to cells, and the solution was renewed after 24 hours. Subsequent to 48-hour culture, medium containing 5 μg/mL puromycin was adopted for screening and amplification, and its expression was detected 72 hours later. The stably infected cell line was constructed and cryopreserved in liquid nitrogen for a long time [16].

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

Total RNA was extracted with Trizol Kit (Invitrogen), from which cDNA was generated in the light of the protocols of PrimeScript RT reagent Kit (Takara Holdings Inc., Kyoto, Japan). The synthesized cDNA was detected by RT-qPCR with a Fast SYBR Green PCR kit (Applied Biosystems, Carlsbad, CA, USA) on an ABI PRISM 7500 RT-PCR system (Applied Biosystems). All RT-qPCR were set up with 3 duplicated wells. The 2−ΔΔCt was adopted to calculate the relative expression of genes with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a normalizer [15]. The primers are manifested in Supplementary Table 2.

Cell counting kit (CCK)-8 assay

The OCUM-1 cells were seeded into a 96-well plate with 100 μL cells (3 × 104 cells) per well, and cultured continuously. Then, 10 μL CCK-8 (Solarbio, Beijing, China) and 90 μL serum-free medium were supplemented to each well. After 1-hour incubation, the absorbance was measured at 450 nm [15].

Transwell experiment

The cell invasion was detected by Transwell assay as below. Transwell chambers were prepared (Merck Millipore, 12 μm aperture). After the chamber was put into the culture plate, 300 μL serum-free medium was added into the upper chamber, and the Matrigel was rehydrated by standing at ambient temperature for 15–30 minutes. Then the remaining medium was sucked out. OCUM-1 cell suspension was prepared and re-suspended in serum-free medium appended to 0.2% bovine serum albumin (BSA) to adjust the cell density to 5 × 105 cells/mL. Transwell upper chamber was supplemented with 200 μL cell suspension. The 500 μL medium encompassing 20% FBS was added into the 24-well plate of the lower chamber. Subsequent to 24 hours, the lower chamber was taken out, stained with 0.1% crystal violet, and counted under the microscope. The cell migration was detected by Transwell assay as above without Matrigel [15].

Flow cytometry

OCUM-1 cells were seeded into a 6-well plate (1 × 106 cells/well), attained by detachment with trypsin and washed with phosphate buffered solution (PBS) twice. The cell precipitate was re-suspended in 400 μL Annexin V binding solution with reference to the manual of Annexin V fluorescein isothiocyanate/propidium iodide (FITC/PI) double staining apoptosis detection kit (BestBio Co., Ltd., Shanghai, China), followed by addition of 10 μL PI and 5 μL Annexin V-FITC. Cell apoptosis was detected by FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA) following incubation for 10 minutes at 4°C void of light [17].

Chromatin immunoprecipitation (ChIP) assay

When OCUM-1 cells reached 70–80% confluence, cells were attained by a ChIP kit (Millipore, Billerica, MA, USA). The cells were immobilized with 1% formaldehyde at ambient temperature for 10 minutes to cross-link DNA and protein. After that, crosslinking was terminated by glycine. After added with lysis buffer, cells were treated with ultrasound and cut into 500–1000 bp fragments. Then the supernatant was obtained by cell centrifugation at 4°C and 30,000 × g. Cells were probed overnight with specific antibody to DNMT1 (ab92314, Abcam, Cambridge, UK) at 4°C with Immunoglobulin G (IgG) as NC. Protein G agarose beads were added to the antibody-bound DNA-protein compound and dissociated overnight at 65°C. Finally, the DNA fragment was extracted and purified by hydroxybenzene or chloroform. The enrichment of DNMT1b in the Reprimo promoter region was detected by RT-qPCR [18].

RNA immunoprecipitation (RIP) assay

When OCUM-1 cells were cultured to 90% confluence, the cells were harvested by trypsin detachment and re-suspended in PBS, freshly prepared nuclear separation buffer (2 mL) and water (6 mL). The cells were placed on ice for 20 minutes. The nuclei were centrifuged and precipitated (2500 × g, 15 minutes), after which nuclear precipitate was resuspended in the freshly prepared RIP buffer (1 mL). The re-suspended nucleus was divided into two parts. The chromatin was mechanically sheared 15–20 times with a Dunns homogenizer and centrifuged at 20,000 × g for 10 minutes to precipitate the nuclear membrane and debris. DNMT1 (ab92314, Abcam) and IgG (5 μg) were supplemented to the supernatant respectively, stirred gently at 4°C and incubated for 2 hours. Next, the supernatant was added with protein G magnetic beads (40 μL), stirred gently at 4°C, and incubated for 1 hour. The magnetic beads were precipitated by centrifugation at 3000 × g for 30 seconds, followed by removal of the supernatant. The magnetic beads were re-suspended in 500 mL RIP buffer. The RNA was isolated by Trizol, and then reversely transcribed for RT-qPCR analysis [19].

Methylation-specific PCR (MSP)

Genomic DNA was extracted from OCUM-1 cells after overexpression of LINC00467 and/or knockdown of DNMT1. After that, the DNA was treated with sodium sulfite using EZ DNA Methylation Kit (Zymo Research, Orange, CA, USA), followed by desulfurization and purification with the reaction column. Then PCR was carried out with the following conditions: 5-minute pre-denaturation at 95°C, 30 cycles of 30-second denaturation at 95°C, 60-second denaturation at 62°C, and 30-second annealing at 72°C, and 10-minute extension at 72°C. The products were analyzed by agarose gel electrophoresis. After that, image analysis was implemented by gel electrophoresis imaging and image analysis system. The design of CpG Island at the Reprimo promoter and MSP primer was from MethPrimer website. The primers of Reprimo MSP were as follows: forwards, 5'- GTGGTGCAGATCGCAGTCAT −3’, and reverse, 5'- CGGTCCTTCACTAGGAAGTTGA −3'. Specific experimental conditions refer to the instructions of the kit [20].

Tumor formation and metastasis in nude mice

The 4-week-old BALB/c female nude mice (Charles River Laboratories, Beijing, China) were separately raised in specific pathogen free animal laboratories with humidity of 60–65% at 22–25°C under 12-hour light/12-hour darkness with free access to drinking water and food. The experiment was implemented after 1-week acclimation with the health status of nude mice under observation.

After detachment, logarithmically growing OCUM-1 cells were re-suspended in serum-free medium and counted. The concentration of cell suspension was adjusted to 2 × 107 cells/mL. The nude mice were subcutaneously inoculated on the back with 100 μL cell suspension (2 × 106 cells) harboring shCon and sh-LINC00467 by a micro syringe, with 10 nude mice in each group. The growth of transplanted tumor were observed every 4 days. Four weeks later, the mice were euthanized and the tumor was isolated, weighed and fixed in 10% formaldehyde for subsequent experiments [21].

Furthermore, 2 × 106 OCUM-1 cells stably infected with shCon and sh-LINC00467 were injected into nude mice via tail vein (n = 10). The liver tissues were then isolated to observe tumor nodules, followed by hematoxylin and eosin (H&E) staining [21].

Ki67 immunohistochemical staining

The tissue samples were sliced into appropriate sizes, fixed in 4% paraformaldehyde for 24 hours, dehydrated with ethanol and sliced into sections (4 μm thickness), which were embedded with paraffin. The paraffin-embedded tissue sections were stored at 60°C for 2 hours, dewaxed by xylene, and hydrated by ethanol (100%, 95%, 85% and 70%) and deionized water. After heated in citric acid buffer (0.01 mol/L, pH 6.0) at 95–100°C for 30 minutes, sections were incubated with 0.5% Triton × 100 for 30 minutes and stained with biotin streptatin horseradish peroxidase detection system (ZSGB-Bio, Beijing, China). Antibody to Ki67 (#9449, 1: 200, CST, Danvers, MA, USA) was applied to incubation with section at 4°C overnight, followed by 1-hour incubation with secondary antibody. According to microscopic observation, brown staining was indicative of immunodominant. The images were visualized utilizing Nikon ECLIPSE Ti microscope (Fukasawa, Japan) and processed using Nikon software [22].

Terminal deoxyribonucleotidyl transferase (TdT)-mediated 2’-deoxyuridine 5’-triphosphate (dUTP)-biotin nick end-labeling (TUNEL) staining

The tissue sections were dewaxed and hydrated. After antigen recovery with protease K working solution and osmotic treatment of osmotic working solution, the sections were reacted with TdT and dUTP mixed solution at the ratio of 1: 9 at 37°C for 2 hours. After that, the activity of endogenous peroxidase (POD) was blocked and the tissues were covered with Transform-POD. The sections were reacted with newly prepared diaminobenzidine (DAB) chromogenic agent, stained with hematoxylin for 3 minutes, dehydrated in the gradient of 70%, 80%, 95% and 100% ethanol and xylene, and then fixed with resin. The nuclei stained with hematoxylin were blue, while the positive cells cultured with DAB reagent were brown. All sections were observed by a microscope and analyzed by Image-Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, USA) [23].

H&E staining

The paraffin-embedded slices were dewaxed by xylene I and II (5 minute each), dewaxed by ethanol (100%, 95%, 80% and 75%, 1 minute each) and hydrated for 5 minutes. Then slices were stained with Harris hematoxylin for 3 minutes before 30-second differentiation with 1% hydrochloric acid ethanol (75%), blued with 0.25% ammonia for 1 minute and treated with 75% ethanol for 1 minute. The slices were treated with 0.5% water-soluble Eosin Y ethanol for 1 minute, treated with 85% ethanol, 95% ethanol, anhydrous ethanol, xylene carbonate, xylene I, and xylene II for 1 minute respectively, and sealed with gum [24].

Statistical analysis

The data of this study were analyzed by SPSS 21.0 software (IBM Corp., Armonk, NY, USA). The measurement data were summarized as mean ± standard deviation. Unpaired t-test was adopted for comparison between the two groups. One-way analysis of variance (ANOVA) was implemented for comparison among multiple groups, followed by Tukey’s post-hoc test. Two-way ANOVA was utilized to compare the optical density (OD) values at different time points, and repeated measures ANOVA to compare the tumor volume at different time points, followed by Bonferroni post-hoc test. Pearson correlation analysis was adopted to evaluate the correlation between LINC00467 and Reprimo. Values of p < 0.05 and p < 0.01 were concluded as significant difference.

Results

LINC00467 is a critical mediator in tumor progression. We here focused on the regulatory role of LINC00467 in GC. Through in vitro and in vivo experimentations, we unfolded that LINC00467 might promote the growth and metastasis of GC by recruiting DNMT1 to elevate the methylation of Reprimo promoter and downregulate Reprimo.

LINC00467 might recruit DNMT1 to manipulate Reprimo expression in GC

The DEGs in GC were retrieved using GEPIA database (Figure 1a). Among these DEGs, we noticed that LINC00467 was highly expressed in GC (Figure 1b). Moreover, it has been documented that LINC00467 orchestrates p53 gene expression through the recruitment of DNMT1, thus affecting the development of tumor [11].

Figure 1.

Prediction of molecular mechanism of LINC00467 in GC. A, The chromosome map of significantly differentially expressed genes in GC included by TCGA. Each number represented a chromosome, the red line represented a highly expressed gene, the green line represented a poorly expressed gene, and the position of the line represented the position of the gene on the chromosome. B, Differential expression of LINC00467 in GC and normal samples collected by TCGA and GTEx. The red box graph showed tumor samples, and the gray box graph showed normal samples. C, The intersection of genes significantly negatively correlated with DNMT1 in GC samples of TCGA and significantly downregulated genes in GC samples of TCGA. The middle part represented the intersection of the two groups of data. D, KEGG pathway enrichment analysis of intersection gene. The abscissa indicated the GeneRatio, the ordinate indicated the KEGG pathway entry, the color indicated the enrichment p value, the circle size indicated the number of genes enriched in the entry, and the histogram on the right was the color scale. E, Differential expression of Reprimo in GC and normal samples collected by TCGA and GTEx. F, The co-expression of DNMT1 and Reprimo in GC samples of TCGA. G, The co-expression of LINC00467 and Reprimo in GC samples of TCGA. The correlation p value and correlation coefficient were in the upper left corner of the figure. * p < 0.01.

To further understand the mechanism of LINC00467 mediating downstream genes by recruiting DNMT1 in GC, we searched genes significantly negatively correlated with DNMT1 in GC included in TCGA, results of which were intersected with the significantly downregulated genes in GC of TCGA, revealing 148 candidate genes (Figure 1c). Subsequently, the KEGG pathway enrichment analysis showed that these candidate genes were mainly enriched in tumor related pathway and p53 signaling pathway (Figure 1d). Among them, Reprimo (RPRM) was enriched in the p53 signaling pathway. Moreover, in GC and normal samples included by TCGA and GTEx, Reprimo showed significantly low expression in GC (Figure 1e). Correlation analysis indicated that Reprimo was negatively correlated with LINC00467 and DNMT1 in GC samples from TCGA (figure 1f-g).

These results and previous reports suggested that LINC00467 might affect the methylation of Reprimo through the recruitment of DNMT1, thereby regulating Reprimo expression and affecting GC development.

LINC00467 expression was highly expressed in GC tissues and cells in correlation with Reprimo expression inversely

It has been uncovered that Reprimo may be a potential marker of GC and LINC00467 is aberrantly expressed in GC tissues and cells [14,25]. Based on bioinformatics analysis, we further detected the expression of LINC00467 and Reprimo in 52 cases of clinical sample tissues by RT-qPCR. The results also indicated that LINC00467 was highly expressed in GC tissues (Figure 2a) and Reprimo was poorly expressed (Figure 2b). Pearson correlation analysis illustrated that LINC00467 shared negative correlation with Reprimo expression in GC tissues (Figure 2c).

Figure 2.

LINC00467 upregulation and Reprimo downregulation are observed in GC tissues and cells. A, RT-qPCR to detect the expression of LINC00467 in GC and adjacent normal tissues (n = 52). B, The expression of Reprimo in GC and adjacent normal tissues determined by RT-qPCR (n = 52). C, Correlation between LINC00467 and Reprimo expression analyzed by Pearson correlation analysis (n = 52). D, RT-qPCR to measure the expression of LINC00467 in human normal gastric epithelial cells GES-1 and four GC cell lines (NCI-N87, OCUM-1, MKN-74 and HGC-27). E, The expression of Reprimo mRNA in human normal gastric epithelial cells GES-1 and four GC cell lines (NCI-N87, OCUM-1, MKN-74 and HGC-27) detected by RT-qPCR. * p < 0.05. All experiments were repeated three times independently.

In different GC cell lines, we also discovered that LINC00467 was highly expressed, while Reprimo was poorly expressed in four GC cell lines (NCI-N87, OCUM-1, MKN-74 and HGC-27) (Figure 2d-e). Among the four GC cell lines, OCUM-1 cell line had the most obvious difference regarding their expression, and was therefore selected for the following experiments.

Collectively, LINC00467 was elevated while Reprimo was limited in GC tissues and cells.

LINC00467 recruited DNA methyltransferase DNMT1 to promote methylation of the Reprimo promoter region for Reprimo downregulation

Most lncRNAs in human have been reported with ability of binding to DNMT1 for mediation on methylation of downstream gene promoter region [19]. Given the aforementioned bioinformatics analysis and clinical experiments, it was inferred that LINC00467 might recruit DNMT1 to mediate Reprimo expression.

For verification purpose, RIP results documented that LINC00467 enrichment was significantly increased in presence of DNMT1 relative to IgG, indicating that LINC00467 could bind to DNMT1 (Figure 3a). ChIP results displayed that Reprimo promoter enrichment was increased in presence of DNMT1 relative to IgG, suggesting that DNMT1 can bind to Reprimo promoter region (Figure 3b).

Figure 3.

DNMT1 is recruited by LINC00467 for Reprimo downregulation. A, RIP experiment verifying the interaction between LINC00467 and DNMT1. B, ChIP experiment verifying the interaction between DNMT1 and Reprimo promoter region. C, MSP experiment determining the methylation of Reprimo in OCUM-1 cells. D, RT-qPCR to detect the expression of LINC00467, DNMT1 and Reprimo in OCUM-1 cells. * p < 0.05. All experiments were repeated three times independently.

Furthermore, MSP assay was performed to detect whether LINC00467 mediated Reprimo methylation by recruiting DNMT1 and OCUM-1 cells were treated with oe-LINC00467 and sh-DNMT1 at the same time (Figure 3c). It was found that the methylation level of the Reprimo promoter region was significantly diminished in OCUM-1 cells treated with sh-DNMT1 yet elevated under oe-LINC00467 treatment. Moreover, the inhibited methylation level of the Reprimo promoter region induced by sh-DNMT1 was significantly reversed by additional treatment of oe-LINC00467.

Subsequently, LINC00467 was overexpressed and DNMT1 was knocked down in OCUM-1 cells to investigate the effect of LINC00467 on Reprimo expression by promoting Reprimo methylation through DNMT1 recruitment. According to RT-qPCR, DNMT1 expression was diminished and Reprimo expression was upregulated in OCUM-1 cells treated with sh-DNMT1 while upregulated LINC00467 and downregulated Reprimo was observed in OCUM-1 cells treated with oe-LINC00467 where DNMT1 expression was not significantly different. Moreover, co-treatment of oe-LINC00467 + sh-DNMT1 led to low DNMT1 expression and high Reprimo expression when compared with oe-LINC00467 treatment alone (Figure 3d).

In conclusion, LINC00467 inhibited Reprimo expression by promoting the methylation of Reprimo promoter region through DNMT1 recruitment in GC cells.

Knockdown of LINC00467 depressed GC cell malignant properties via the DNMT1/Reprimo axis

To explore the effects of LINC00467 on GC cell functions, we constructed three kinds of shRNAs to knock down LINC00467 in OCUM-1 cells. Based on RT-qPCR results (Figure 4a), all the three shRNAs markedly reduced LINC00467 expression in GC cells, among which sh-LINC00467-3 had the highest knockdown efficiency. Therefore, sh-LINC00467-3 was adopted for the follow-up experiments.

Figure 4.

LINC00467 knockdown causes repression of GC cell proliferation, invasion, and migration and promotion of their apoptosis via the DNMT1/Reprimo axis. A, The transfection efficiency of three kinds of shRNAs against LINC00467 detected by RT-qPCR. B, RT-qPCR to detect the expression of DNMT1 and Reprimo in OCUM-1 cells. C, CCK-8 assay to detect the effect of LINC00467 knockdown on the proliferation of OCUM-1 cells. D, Effect of LINC00467 knockdown on invasion of OCUM-1 cells detected by Transwell assay. E, Transwell assay to measure the effect of LINC00467 knockdown on the migration of OCUM-1 cells. F, Effect of LINC00467 knockdown on apoptosis of OCUM-1 cells determined by flow cytometry. * p < 0.05. All experiments were repeated three times independently.

Then, LINC00467 was knocked down in OCUM-1 cells, followed by quantification of DNMT1 and Reprimo expression using RT-qPCR (Figure 4b). Results revealed no significant difference regarding DNMT1 expression in response to sh-LINC00467 treatment while Reprimo expression was increased in OCUM-1 cells. As depicted in (Figure 4c-f), knockdown of LINC00467 reduced the proliferative, invasive, and migrative capabilities of OCUM-1 cells but substantially augmented cell apoptosis.

Collectively, knockdown of LINC00467 inhibited the malignant features of GC cells through the DNMT1/Reprimo axis.

LINC00467 knockdown suppressed tumorigenesis and metastasis of GC cells in vivo

The tumorigenesis and liver metastasis models were established in nude mice to explore the effect of LINC00467-mediated DNMT1/Reprimo axis on GC cells in vivo. The results of in vivo experiments demonstrated that after knockdown of LINC00467, the growth rate of tumor in nude mice was evidently inhibited, and the weight and volume of tumor were distinctly reduced (Figure 5(a-d)). The results of Ki67 staining and TUNEL staining depicted that knockdown of LINC00467 obviously decreased Ki-67 expression and increased the apoptosis in tumor tissues (Figure 5e).

Figure 5.

Tumorigenesis and metastasis of GC cells are inhibited by LINC00467 knockdown in vivo. A, The morphology of tumor in nude mice after knockdown of LINC00467 observed by naked eye. B, Tumor growth curve in nude mice after knockdown of LINC00467. C, Changes of tumor weight in nude mice after knockdown of LINC00467. D, Changes of tumor volume in nude mice after knockdown of LINC00467. E, Ki67 staining and TUNEL assay to test the Ki67 expression and apoptosis of GC cells after knockdown of LINC00467. F, The tumor nodules in liver tissues of nude mice after knockdown of LINC00467 observed by naked eye. G, H&E staining to observe the tumor nodules in liver tissues of nude mice after knockdown of LINC00467. n = 10 mice/group. * p < 0.05.

In addition, the experimental results of GC liver metastasis model in nude mice induced by injection of OCUM-1 cells via tail vein (figure (5f-g)) exhibited that knockdown of LINC00467 clearly reduced the tumor nodules in liver tissues.

Taken together, knockdown of LINC00467 exerted inhibitory effect on tumorigenesis and metastasis of GC cells in vivo.

Discussion

GC is the final result of a series of events that take decades to occur and result from the accumulation of various genetic and epigenetic changes [26]. Due to the insignificant early symptoms of GC, most GC patients are diagnosed at a late stage and have a poor prognosis [27]. Therefore, it is necessary to acquire the deep understanding of molecular mechanism underlying GC for treatment. Moreover, it has been manifested that lncRNAs exhibit promising values in GC diagnosis or prognosis evaluation [10]. Therefore, we conducted this research to explore the role of an lncRNA, LINC00467, in GC development and the potential mechanism. Consequently, our data illustrated that LINC00467 might promote the growth and metastasis of GC by recruiting DNMT1 to elevate the methylation of Reprimo promoter and downregulate Reprimo.

Initially, one of our major findings was that LINC00467 was high in GC tissues and cells, and that LINC00467 silencing caused decline of the malignant properties of GC cells. Consistently, overexpression of several lncRNAs, like lncRNA NALT1, lncRNA BLACAT1, and lnc01614, has been detected in GC tissues [28-30]. Intriguingly, the pivotal role of LINC00467 in GC development has been predicted by a meta-analysis of a prior work [31]. More importantly, overexpressed LINC00467 has been indicated to strengthen the viability and proliferation yet suppress apoptosis of GC cells by elevating the level of integrin subunit beta 3 [32]. Additionally, overexpression of LINC00467 has been observed in osteosarcoma tissues and cells [33]. Also, LINC00467 was upregulated in non-small cell lung cancer tissues [34]. These results partially supported the upregulation of LINC00467 in GC cells and tissues. Moreover, concordant with our results, silence of LINC00467 has been demonstrated to curb malignant properties of GC cells by targeting microRNA-7-5p [9]. Of note, LINC00467 knockdown diminished glioma U87 and U251 cell malignant properties [35,36]. In addition, ectopically expressed LINC00467 contributed to elevation of cell malignant properties in osteosarcoma [37]. Also, another work clarified that ectopic expression of LINC00467 accelerated the malignant properties of lung adenocarcinoma cells [38]. Collectively, LINC00467 might possess tumor-promoting potential in GC.

As widely recognized, lncRNAs have the ability to bind DNMT1 to modulate the methylation of downstream gene promoter [19]. Specifically, it has been unraveled in the research of Zhang et al. that LINC00467 reduced p53 expression to induce glioma cell proliferation and invasion by binding to DNMT1 [11]. Similarly, our study also discovered that LINC00467 was capable of recruiting DNMT1 into Reprimo promoter to decrease its expression in GC cells. Further analysis noted the low expression of Reprimo in glioma tissues. As reported, Reprimo assumes a crucial role in tumor suppression in accordance to its increased expression after X-ray irradiation and its identification as a downstream target of p53 [39]. Reprimo has been identified as a target for hypermethylation in various cancers, including prostate cancer [40], hepatocellular carcinoma [41], and GC [25] because of the negative relationship between methylation and transcription. More importantly, Saavedra et al. observed that elevated Reprimo methylation and decreased Reprimo expression were detected in GC cells, and that Reprimo downregulation was associated with invasive stage of tumor progression in GC [42]. In addition, the research of Luo et al. unveiled the downregulation of Reprimo in gastric adenocarcinoma tissues, and that the loss of Reprimo was correlated to promoted tumor invasion, lymphatic vessel invasion, and lymph node metastasis in gastric adenocarcinoma [43]. Therefore, Reprimo might be involved in the oncogenic effect of LINC00467 on GC.

Conclusion

Conclusively, this study found that LINC00467 elevated Reprimo methylation level by recruiting DNMT1 into its promoter to downregulate Reprimo, thereby augmenting GC cell malignant properties (Figure 6). This research provided a new theoretical basis for revealing the mechanism of the occurrence, development, early diagnosis, and treatment of GC.

Figure 6.

The mechanism of LINC00467 promoting the growth and metastasis of GC by recruiting DNMT1 to promote the methylation of Reprimo promoter.

Limitation

However, the physiological and patho-physiological differences concerning the cell and animal results to the human clinical condition need further exploration. Further experimentations regarding humans are needed to confirm the value of clinical application involving other signaling pathways.

Click here for additional data file.