Long non‐coding RNA MYOSLID functions as a competing endogenous RNA to regulate MCL‐1 expression by sponging miR‐29c‐3p in gastric cancer

Abstract Objective Long non‐coding RNA (lncRNA) has become an important regulator of many human malignancies. However, the biological role and clinical significance of most lncRNA in gastric cancer (GC) remain unclear. Methods We investigate the biological function, mechanism of action and clinical expression of lncRNA MYOSLID in GC. First, we analysed the differential expression of lncRNA MYOSLID in GC tissues and non‐cancerous tissues by analysing the sequencing data obtained from The Cancer Genome Atlas. Subsequently, we verified that lncRNA MYOSLID regulates the proliferation and apoptosis of GC cells by acting as a ceRNA against miR‐29c‐3p. The nude mouse xenograft was used to further confirm the functional significance of lncRNA MYOSLID in vivo. Results We found for the first time that the expression of lncRNA MYOSLID was significantly up‐regulated in GC tissues, and the up‐regulation of lncRNA MYOSLID in GC was correlated with tumour size, AJCC stage, depth of invasion and survival time. In addition, apoptosis and growth arrest can be induced in vitro after knockdown of lncRNA MYOSLID, which inhibits tumorigenesis in mouse xenografts in vivo. Further in‐depth studies revealed that lncRNA MYOSLID acts as a ceRNA of miR‐29c‐3p, resulting in de‐repression of its downstream target gene MCL‐1. Conclusion The lncRNA MYOSLID‐miR‐29c‐3p‐MCL‐1 axis plays a key role in the development of GC. Our findings may provide potential new targets for the diagnosis and treatment of human GC.

Long non-coding RNAs (lncRNAs) are a class of transcripts greater than 200 nucleotides in length with limited protein-coding ability or no protein-coding ability. 5 It has been found that lncRNA is abundantly transcribed in mammalian cells and in plant cells. 6,7 These lncRNAs can be involved in a number of important cellular biological processes, including regulation of cell growth, 8 apoptosis, 9 cell differentiation, 10 and cell invasion and metastasis. 11 Increasing evidence showed that lncRNA can be used as a biomarker for diagnosis and prognosis in a variety of cancers, such as colorectal cancer, 12 breast cancer, 13 liver cancer, 14 prostate cancer 15 and GC. 16 Typically, lncRNAs exert their biological functions by regulating epigenetic, 17 transcriptional 18 and post-transcriptional levels 19 that regulate potential target gene expression. In recent years, more and more studies have shown that lncRNA plays an important role in human cancer. 6 For example, LINC00941 is significantly up-regulated in liver cancer and is significantly associated with poor clinical outcomes, and regulates the metastasis and proliferation of liver cancer by binding ANXA2 to affect the activity of the Wnt/β-catenin signalling pathway. 20 Furthermore, in colorectal cancer, lncRNA UICLM inhibits the expression of miR-150-5p by competitive endogenous RNA action, thereby promoting liver metastasis of colorectal cancer. 21 It is well documented that a number of important lncRNAs have been proved to be significant survival prognosis of GC. For example, lncRNA HOTAIR promotes gastric cancer metastasis by binding to the epigenetic transcriptional regulator polycomb inhibitor complex 2 (PRC2). 21 HOTAIR also regulates cisplatin resistance in GC by acting as a competitive endogenous RNA (ceRNA) of miR-126. 22 In addition, LINC01234 functions as a competing endogenous RNA to regulate CBFB expression by sponging miR-204-5p regulates the malignant proliferation of GC. 4 LncRNA MYOSLID was first reported in human VSMC-selective and serum-responsive factor/CArG-dependent lncRNA, which regulates VSMC differentiation through the MKL1 and transforming growth factor-beta/SMAD pathways. 23 We have previously studied the differentially expressed lncRNA in GC from the Cancer RNA-Seq Nexus database and found that the expression of lncRNA MYOSLID in GC is significantly different and is associated with the survival prognosis of GC. 24 However, the mechanism of lncRNA MYOSLID in GC remains elusive.
In this study, we studied lncRNA MYOSLID in GC. We first discovered that lncRNA MYOSLID is significantly up-regulated in GC tissues and is associated with poor prognosis. Loss and functional gain assays showed that lncRNA MYOSLID promotes GC cell proliferation and inhibits apoptosis by acting as a miR-29c-3p ceRNA, thereby preventing miR-29c-3p from binding to the target protein MCL-1. Collectively, the results suggested that lncRNA MYOSLID is an oncogenic regulator of tumorigenesis in GC and may be a potential target for the diagnosis and treatment of patients with GC.

| Tissue samples
Seventy-five patients with gastric adenocarcinoma and paired normal tissues were obtained from patients undergoing GC surgery at Xijing Digestive Disease Hospital. All samples were clinically and pathologically validated. This study was approved by the Xijing Hospital Human Body Protection Committee. Informed consent was obtained from each patient.

| Cell culture
Human GC cell lines MKN45, AGS, SGC-7901 and BGC-823 were purchased from Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences (Shanghai, China). The immortal normal gastric epithelial cell line GES-1 was purchased from the Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences (Shanghai, China). All cells were cultured in DMEM basic containing 10% foetal bovine serum (Gibco) and 1% penicillin-streptomycin (Gibco). All cells were incubated with 5% (v/v) CO 2 at 37°C. All cells were tested for mycoplasma contamination before the experiments.

| RNA extraction, reverse transcription and realtime RT-PCR
Total RNA from cells was extracted using an RNA isolation kit (TaKaRa, Tokyo, Japan) according to the manufacturer's instructions.
Subsequently, the RevertAid First Strand cDNA Synthesis Kit (TaKaRa, Tokyo, Japan) was used to reverse-transcribe the messenger RNA (mRNA) from the total mRNA; primers for miR-29c-3p and U6 were purchased from RiboBio (Guangzhou, China); the specific primer (Table   S2) and the SYBR premix Ex Taq (TaKaRa, Tokyo, Japan) were used to expand by real-time qPCR (Bio-Rad, CA, USA). It was carried out with the following parameters: pre-denaturation at 95°C for 5 minutes, denaturation at 95°C for 10 seconds, annealing at 62°C for 20 seconds and extension at 72°C for 30 seconds for 40 cycles. Glyceraldehyde-3phosphate dehydrogenase (GAPDH) was used as an internal control.

| Western blot analysis
The cells were washed three times with PBS and collected in RIPA lysis buffer (Beyotime Biotechnology, Shanghai, China) supplemented with a protease inhibitor cocktail (Calbiochem, San Diego, USA). Protein concentration was determined by staining with Coomassie Blue (Beyotime Biotechnology, Shanghai, China). Cellular protein lysates were separated by 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred to a 0.22 mm polyvinylidene fluoride membrane (Millipore) and probed with specific antibodies. Specific bands were detected by ECL chromogenic substrate and quantified by densitometry (Quantity One software, Bio-Rad). The GAPDH antibody was used as a control. Anti-caspase-3, cleaved caspase-3, poly (ADP ribose) polymerase protein (PARP), cleaved PARP, cyclin D1, CDK2 and MCL-1 (1:1000) were purchased from Cell Signalling Technology. GAPDH antibody was purchased from Proteintech. All antibodies are listed in Table S3.

| RNA fluorescent in situ hybridization
The subcellular localization of lncRNA MYOSLID was detected by FISH kit (RiboBio, Guangzhou, China) according to the manufacturer's instructions. The Cy3 labelled lncRNA MYOSLID probe was obtained from RiboBio (Guangzhou, China). Briefly, gastric cancer cells (2 × 10 4 ) were seeded on cell slides in 24-well culture plates. After waiting for the cells to adhere, the cells were fixed in 4% paraformaldehyde for 30 minutes at room temperature. After permeabilization, the cells are pre-hybridized with the pre-hybridization solution and the hybridization solution and then incubated with the cy3-labelled lncRNA MYOSLID oligonucleotide probe. The nuclei were stained with DAPI for 10 minutes at room temperature.

| RNA immunoprecipitation
RNA immunoprecipitation was performed using the EZ-Magna RIP kit (Millipore, Billerica, MA, USA) according to the manufacturer's instructions. First, we lysed SGC-7901 and BGC-823 cells and incubated with Protein A magnetic beads; next, we conjugated the magnetic beads to the antibody at 4°C. After 3-6 hours, the beads were washed with washing buffer and then incubated with 0.1% SDS/0.5 mg/mL proteinase K for 30 minutes at 55°C to remove proteins. Finally, we performed qRT-PCR analysis of immunoprecipitated RNA using primers specific for lncRNA MYOSLID.

| Statistical analysis
We performed statistical analysis using Prism 5 (San Diego, CA, USA) and SPSS 18.0 (Chicago, IL, USA) software. Differences between the two groups were assessed using Student's t test. A P value < .05 was considered to represent a significant difference. The overall survival curve was estimated by the Kaplan-Meier method and the Cox proportional hazard model. All values are expressed as mean ± SD unless otherwise stated.

| lncRNA MYOSLID is up-regulated in GC and associated with poor prognosis
To investigate the expression of lncRNA MYOSLID in human GC, we searched the Cancer Genome Atlas (TCGA) database and found that the lncRNA MYOSLID gene copy number was significantly elevated in GC tissues compared with normal gastric tissue ( Figure 1A). We next analysed publicly available data and found that lncRNA MYOSLID expression is closely related to the overall survival of patients with GC ( Figure 1B). Then, the expression of lncRNA MYOSLID in GC tissues was detected by real-time PCR and found that the expression of lncRNA MYOSLID was higher in GC tissues than in matched nontumour tissues (n = 75, P < .0001, Figure 1C). To assess the clinical significance of lncRNA MYOSLID overexpression in GC, we evaluated the association between lncRNA MYOSLID levels and clinicopathological features. As shown in Table S1, high lncRNA MYOSLID expression was associated with patients age (P = .018), larger tumour size (P = .001), invasion-related depth (P = .010) and AJCC staging (P = .001), while lncRNA MYOSLID was no significant correlation between expression and other factors including gender (P = 1.000).
We also examined the association between lncRNA MYOSLID expression levels and prognosis in patients with GC. Kaplan-Meier survival analysis showed that patients with higher lncRNA MYOSLID levels had shorter overall survival than patients with lower lncRNA MYOSLID levels ( Figure 1D). In addition, lncRNA MYOSLID was significantly overexpressed in GC cell lines compared with human normal gastric epithelial cells (GES-1) ( Figure 1E). These data suggested that IncRNA MYOSLID is involved in the pathogenesis of GC.

| Knockdown of lncRNA MYOSLID inhibits gastric cancer cell tumorigenesis in vivo
To determine whether lncRNA MYOSLID affects tumour growth in   Figure 4B). From these results, we initially hypothesized that the molecular mechanism of lncRNA MYOSLID in GC cells may act as a ceRNA of miRNA. To further confirm this hypothesis, we used an online bioinformatics database (Cancer RNA-Seq Nexus database) to analyse predicted miRNAs that bind to the lncRNA MYOSLID sequence. The data indicate the presence of a putative binding site between lncRNA MYOSLID and miR-29c-3p. Then, we analysed RNA sequencing data from GC tissues from TCGA and found that miR-29c-3p was significantly down-regulated in gastric tissue ( Figure 4C). Additionally, we analysed the expression levels of lncRNA MYOSLID and miR-29c-3p in GC tissue RNA sequencing data from TCGA and found a negative correlation ( Figure 4D). At the same time, we analysed the expression of miR-29c-3p in 75 pairs of GC tissues by qRT-PCR and found that the expression level of miR-29c-3p in cancer tissues was significantly lower than that in adjacent tissues ( Figure 4E). In addi-  Figure 4I). Then, we determined the change in the activity of lncRNA MYOSLID by the luciferase reporter gene after sitedirected mutagenesis by the putative miR-29c-3p binding site in the lncRNA MYOSLID sequence. As expected, the luciferase reporter assay showed that miR-29c-3p directly targets the 3'UTR of lncRNA MYOSLID-WT to negatively regulate the luciferase activity of lncRNA MYOSLID-wt-3'UTR, rather than lncRNA MYOSLID-MUT's 3'UTR ( Figure 4J). As shown in Figure 4K, the expression level of miR-29c-3p in GC cell lines was significantly lower than that in normal gastric epithelial cells.

| The biological function of lncRNA MYOSLID is partly mediated by the negative regulation of miR-29c-3p
To determine the role of miR-29c-3p in GC cells, we transfected miR-29c-3p mimics or miR-29c-3p inhibitors in GC cell increasing the proportion of apoptotic cells ( Figure S1D). These results indicated that the function of lncRNA MYOSLID in GC cells is at least partly mediated by the negative regulation of miR-29c-3p.

| MCL-1 is a miR-29c-3p target gene and is indirectly regulated by lncRNA MYOSLID
The role of ceRNA regulatory networks in GC has been widely reported. To determine the ceRNA regulatory network between lncRNA MYOSLID, miR-29c-3p and downstream targets, we used a network database (miRWalk, miRtarbase and Diana) to predict potential miR-29c-3p target genes. In addition, we predicted the lncRNA MYOSLID-miR-29c-3p targeting ceRNA network was effectively reversed by miR-29c-3p inhibitors ( Figure 6E). In addition, we analysed the association between miR-29c-3p and MCL-1 expression in 75 pairs of GC tissues and found a negative correlation between miR-29c-3p and MCL-1 ( Figure 6F). Taken together, these data suggest that MCL-1 expression regulation is primarily mediated by post-transcriptional regulation of miR-29c-3p via lncRNA MYOSLID.

| MCL-1 expression is up-regulated in GC tissues and promotes GC cell growth
To investigate the potential role of MCL-1 in GC, we analysed its MCL-1 protein in GC tissues ( Figure S2A). Interestingly, Kaplan-Meier analysis found that higher MCL-1 expression was significantly associated with shorter overall survival in patients with gastric cancer ( Figure S2B). Then, gastric cancer cell lines SGC-7901 and BGC-823 were transfected with MCL-1 siRNA to knockdown their expression, which was confirmed by qRT-PCR and Western blot ( Figure 7A,B). Simultaneously, CCK8 and colony formation assay incorporation assays showed knockdown of MCL-1 expression significantly reduced cell growth viability and colony formation ( Figure 7C,G). In addition, cycle arrest and apoptotic cell rates of GC cells SGC-7901 and BGC-823 transfected with si-MCL-1 or scrambled siRNA were analysed by flow cytometry (Figure 7D,7).
Western blot analysis showed that the expression of cycle-regulated proteins such as cyclin D1 and CDK2 was significantly de-  32 In this study, we found that lncRNA MYOSLID is mainly localized in the cytoplasm by in situ hybridization and interacts with Ago2 in GC cells, suggesting that ln-cRNA MYOSLID may act as an endogenous miRNA sponge. Then, bioinformatics analysis and luciferase reporter analysis revealed that miR-29c-3p is a new target for lncRNA MYOSLID. By analysing TCGA data, it was found that miR-29c-3p is down-regulated in human GC and acts as a tumour suppressor. There is evidence that the expression of miR-29c-3p in GC is lower than that in adjacent tissues, and it can significantly inhibit the proliferation of GC cells by down-regulating the expression of ITGB1. 33 In addition, it has been reported that miR-29c is significantly down-regulated in colon cancer. 34 In this study, we also found that miR-29c-3p was significantly down-regulated in GC, and we found that increased expression of miR-29c-3p inhibited GC cell proliferation and induced apoptosis. At the same time, our results reveal that lncRNA MYOSLID plays an important role in GC cells by sponging miR-29c-3p during tumorigenesis and progression.
In general, lncRNA acts primarily by inhibiting miRNAs to affect downstream miRNA targets in the mechanism of the competing endogenous RNAs of lncRNA. Therefore, miRNA targets are an important part of the ceRNA network. 35,36 Next, we used three online prediction databases and found that MCL-1 is one of the potential miR-29c-3p targets not reported in GC. Meanwhile, to elucidate that miR-29c-3p directly targets MCL-1, we performed a luciferase reporter assay and confirmed that the 3'UTR region of MCL-1 mRNA is the target site for miR-29c-3p. In addition, overexpression of miR-29c-3p in GC cells significantly reduced the protein level of MCL-1. MCL-1 is a unique anti-apoptotic BCL-2 family member that is overexpressed in many tumour types. 37 For example, Chen G et al reported that targeting MCL-1 enhanced the sensitivity of DNA replication stress to cancer treatment. 38 In addition, Zhan Z et al also reported that MCL-1 can increase the anti-apoptotic ability of GC cells, thereby promoting the proliferation of GC cells. 39 In our study, we found that MCL-1 was significantly up-regulated in GC tissues compared with normal samples. At the same time, Kaplan-Meier analysis found that higher MCL-1 expression was significantly associated with poor overall survival in patients with GC. Then, we found that knockdown of MCL-1 expression significantly inhibited GC cell proliferation and induced apoptosis. Furthermore, by co-transfection of MCL-1 siRNA with the miR-29c-3p inhibitor, we found that the function of the miR-29c-3p inhibitor can be reversed by MCL-1 knockdown. These results indicated that miR-29c-3p inhibits GC cell proliferation depending on inhibition of MCL-1 expression.
In conclusion, we report a novel gastric cancer-associated ln-cRNA MYOSLID and first discovered that lncRNA MYOSLID is a carcinogenic lncRNA that promotes cell proliferation and inhibits apoptosis in human GC via the miR-29c-3p-MCL-1 axis. This study provided better understanding of lncRNA-miRNA-mRNA ceRNA network in the development of GC. lncRNA MYOSLID may be a potential important target for the diagnosis and treatment of GC.