LncRNA LIFR-AS1 promotes proliferation and invasion of gastric cancer cell via miR-29a-3p/COL1A2 axis

LncRNA was known to be closely associated with the progression of human tumors. The role of lncRNA LIFR-AS1 in the pathogenesis and progression of gastric tumor is still unclear. The aim of this study was to investigate the function of LIFR-AS1 and the underlying mechanism in the pathogenesis and progression of gastric cancer. QRT-PCR was used to evaluate the expression of LIFR-AS1, miR-29a-3p and COL1A2 in gastric tumor tissues and cells. Western blotting was used to evaluate the protein expression of COL1A2 in gastric tumor cells. CCK-8 assay, transwell assay and flow cytometry were used to evaluate the roles of LIFR-AS1, miR-29a-3p and COL1A2 in cell proliferation, invasion, migration and apoptosis. The relationship among LIFR-AS1, miR-29a-3p and COL1A2 was assessed by bioinformatics analyses and luciferase reporter assay. The expression levels of LIFR-AS1 were significantly increased in gastric tumor tissues and cells, while the expression levels of miR-29a-3p were decreased. The expression of miR-29a-3p was negatively correlated with the expression of LIFR-AS1 in gastric cancer tumor tissues. Knocking down of LIFR-AS1 inhibited proliferation, invasion and migration of gastric tumor cells, and induced apoptosis of gastric tumor cells. Bioinformatics analyses and integrated experiments revealed that LIFR-AS1 elevated the expression of COL1A2 through sponging miR-29a-3p, which further resulted in the progression of gastric tumor. LIFR-AS1 plays an important role as a competing endogenous RNA in gastric tumor pathogenesis and may be a potential target for the diagnosis and treatment of gastric tumor.


Background
The incidence of gastric cancer (GC) is not evenly distributed throughout the world [1]. Nearly 950,000 new cases of GC occur each year, and approximately 720,000 people die each year from GC [2]. Among all tumors, the incidence and mortality of GC is the fifth and the third, respectively [3]. It is seriously affecting the quality of life of the patients and is a heavy burden for the society [4]. GC is a multifactorial disease, and its onset is related to individual diet, living environment, and family genetics [5,6]. Despite the progress of medical technologies, most of the GC patients are diagnosed at late stage, which is accompanied by malignant proliferation and distant metastasis, leading to poor treatment effect and postoperative survival rate [7]. Accumulating studies have shown that lncRNAs involve in the pathogenesis and progression of many kinds of cancer [8]. But the involvement Pan et al. Cancer Cell Int (2021) 21:7 of lncRNAs in the pathogenesis and progression of GC is still unclear.
LncRNAs are a group of non-coding RNAs with the length of more than 200 nucleotides. It has been reported that lncRNAs play important roles in many life activities, such as dosage compensation effect, epigenetic regulation, and regulation of cell cycle and cell differentiation [9,10]. Studies have shown that lncRNAs are abnormally expressed in GC [11] and play a role in promoting or suppressing the occurrence and development of CC [12]. For example, lncRNAs including HOTAIR, HULC, LINC00152, MALAT2, H19, GHET1, and GACAT3 have been demonstrated to exert oncogene activities, while other lncRNAs, including LEIGC, GAS5, and FER1L4, function as tumor suppressors [12]. Another study reported that ALKBH5 promotes invasion and metastasis of gastric cancer by decreasing methylation of lncRNA NEAT1, suggesting that lncRNAs might serve as potential therapeutic targets for GC [11]. LncRNA LIFR-AS1, a newly discovered tumor-related intergenic lncRNA, is expressed in many malignant tumors such as colorectal cancer. It has been reported that knockout of LIFR-AS1 can inhibit the proliferation and apoptosis of HCT116 cells [13], suggesting that it may involve in regulating the biological processes of tumors. However, the function of LIFR -AS1 in GC has not been studied.
It has been found that lncRNAs can regulate the expression of genes through signaling, decoy, guidance, scaffold and competition of endogenous RNAs at transcriptional and post transcriptional levels [14]. MicroRNAs (miRNAs) are involved in the regulation of tumorigenesis, metastasis, and distant invasion [15]. Studies have confirmed that miRNAs are involved in the pathogenesis and treatment of cancers [16]. Abnormal expression of miRNAs may lead to the emergence of various tumors [17]. Studies have found that miR-NAs play critical roles in gastric cancer [18]. Studies have found that in ovarian cancer, many miRNAs have mutated, and some of these genes might be used for detection, diagnosis and or treatment of the cancer [19]. MiR-29a-3p has been reported to be involved in the pathogenesis of many kinds of tumor. For example, the expression of miR-29a-3p was markedly decreased in GC cell lines with stronger metastatic potential, and overexpression of miR-29a-3p inhibited GC cell proliferation, colony-forming, migration and invasion, suggesting that miR-29a-3p may be a potential therapeutic target for GC [20]. There are many members in the fibrous collagen family, among which type I collagen encoded by the α2 chain of type I collagen (COL1A2) gene is the most widely expressed collagen. It has been found that the COL1A2 affects cell proliferation, differentiation, adhesion and metastasis [21]. The expression of COL1A2 may be related to the occurrence of GC [22]. It was speculated that LIFR-AS1 regulated the progression of gastric cancer through miR-29a-3p/ COL1A2 axis. This study was carried out to investigate the function of LIFR-AS1 and its underlying mechanism in the pathogenesis and progression of GC.

Tissue samples
Between 2010 and 2011, 42 pairs of GC tissues and adjacent non-tumor tissues were obtained from patients undergoing surgery at the People's Hospital of Gaozhou. GC was confirmed by histopathology. Before the operation, the patients did not receive any local or systemic treatment. This study was approved by the Ethics Committee of the People's Hospital of Gaozhou. All patients signed the written informed consent.

Cell culture and transfection
Cell culture and transfection were carried out as previously described [23]. Human gastric adenocarcinoma cell lines (BSG823, HS746T, 9811, BGC803, MKN28, MGC803 and BSG823) and normal gastric cell lines (GSE1) were obtained from the Institute of Cell Research, Chinese Academy of Sciences (Shanghai, China). All cells were cultured in F12 or 1640 medium (Hyclone, South Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS) (Hyclone, South Logan, UT, USA), in which penicillin (100 U/mL) and streptomycin (100 mg/ mL) were added to prevent the contamination. Cell suspension was placed in a thermostat incubator (5% CO 2 , 37 °C), and digestion and subculture of cells were carried out. For transfection, cells were seeded in six-well plates (2 × 10 5 cells per well) and cultured at 37 °C with 5% CO 2 overnight and grown to 70% confluence prior to transfection. Cells were transiently transfected with the shLIFR-AS1 vector. A non-silencing shRNA (5′-GCG TGA AAG TCA ACT AAT TAA-3′) was used as the negative control (shNC). The shRNA sequence targeting the LIFR-AS1 gene was 5′-CTC CGA GCC TCC GGT GGT GA-3′. These vectors were obtained from GenePharma (Shanghai, China). LIFR-AS1 cDNA fragments containing either the predicted potential miRNA binding sites (wildtype, wt) or scrambled microRNA binding site sequences (mutation, mut) were amplified by PCR. MiR-29a-3p mimic, and MiR-29a-3p inhibitor or their parental negative controls were purchased from RiboBio (Guangzhou, China). COL1A2 pcDNA plasmid was chemically synthesized by GenePharma (Shanghai, China). They were transfected into A549 cells using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA).

RNA isolation and quantification
The TRIzol reagent (Invitrogen) was used to isolate total RNAs from tissue specimens and cells. All reagents for qRT-PCR were ordered from Takara (Dalian, China). For mRNA analyses, the first-strand cDNA was reverse transcribed from RNA samples (500 ng) using the PrimeScript ® RT reagent Kit following the manufacturer's instructions. SYBR Premix Ex Taq II (TaKaRa, Dalian, China) and 10 pg cDNA were used for quantitative PCR (qPCR). ABI 7500HT instrument (Applied Biosystems) was used to perform the polymerase chain reaction, which was set up to run the following program: 95 °C for 10 min, 95 °C for 30 s, and 60 °C for 30 s, followed by 38 cycles of 74 °C for 25 s. The expression levels of mRNAs were normalized to GAPDH, and the expression levels of miR-145-5p were normalized to U6. The 2 − ΔΔCT method was used to calculate the relative expression levels. The primer sequences were as follows:

Luciferase reporter assay
Luciferase reporter assay was carried out as previously described [13]. The wild-type or mutant LIFR-AS1 were PCR-amplified and sub-cloned into pmirGLO vector. The reporter vector pmirGLO-LIFR-AS1-wt (wild type) was inserted by the wild-type LIFR-AS1, which contained the predictive binding site of miR-29a-3p. The vector pmirGLO-LIFR-AS1-Mut (mutant) was inserted by the mutant LIFR-AS1, which contained point mutations of the miR-29a-3p seed region binding site. BSG823 cells were cultured and co-transfected with pmirGLO-LIFR-AS1 including wild-type or mutant fragments and miR-29a-3p or miR-NC. Likewise, miR-29a-3p or miR-NC was co-transfected with COL1A2-wt or COL1A2 Mut vector into BSG823 cells using Lipofectamine 3000 (Invitrogen). Among them, pRL-TK vector was co-transfected as a control. The Dual-Luciferase Reporter Assay System (Promega) was applied at 48 h after transfection following the manufacturer's instructions. After that, BSG823 cells were lysed with lysis buffer. After centrifuge, the luciferase activity was determined by a Modulus TD20/20 Luminometer (Turner Biosystems, CA). The relative luciferase activity was normalized to Renilla luciferase activity. The experiments were repeated for three times.

Cell proliferation detection
Cell proliferation detection was carried out as previously described [23,24]. Cells were seeded into 96-well plates at a density of 5 × 10 3 cells/well and incubated overnight. After transfection, 10 µL CCK8 solution (Liji, Shanghai, China) was added to each well to reach 10% final concentration at 4 h before the measurement of OD values at 450 nm, which was performed every 24 h for a total of 96 h. The absorbance at 450 nm was detected with a microplate reader (Bio-Tek EL 800, USA). The experiments were performed in triplicate.

In vivo tumor growth assay
In vivo tumor growth assay was carried out as previously described [24,25]. Female BALB/c nude mice (4-6 weeks, weighing 16-22 g) were obtained from the Experimental Animal Center of Sun Yat-sen University. The experiment was approved by the Experimental Animal Care Committee of the People's Hospital of Gaozhou and was performed in strict accordance with the NIH Laboratory Animal Care and Use Guidelines (Bethesda, Maryland). BSG823 cells transfected with sh-LIFR-AS1, LIFR-AS1 overexpression vector or empty vector. One side of each mouse was subcutaneously injected with 1.5 × 10 6 cell density for tumor formation assay. After injection, tumor formation and tumor volume were measured. After 4 weeks, the nude mice were euthanized, and tumor samples were taken out, then the weight and volume of the tumor were measured. Tumor volume was monitored by measuring the length and width with calipers and calculated with the following formula: tumor volume = length × width 2 /2. The experiments were repeated for three times.

Immunohistochemical staining
Immunohistochemical staining was carried out as previously described [25]. Paraffin-embedded tissue sections were deparaffinized and incubated with 0.1 M sodium citrate, PH 6.0 at 65 °C for antigen retrieval. Then the slides were sequentially exposed to 3% H 2 O 2 to block endogenous peroxidase activity, a buffer containing 5% BSA and 0.1% Triton X-100 to reduce non-specific binding, a specific primary antibody against Ki-67 (1:500; Abcam, UK) at 4 °C overnight, and HRP-conjugated secondary antibody at room temperature for 1 h. Signals of the antigen-antibody complexes were detected with a 3,3-N-diaminobenzidine tertrahydrochloride (DAB) kit (E670033, Sangon Biotech, Shanghai, China) following the manufacturer's instructions. Finally, the slides were counterstained in hematoxylin. For quantification, 10-20 fields were randomly selected from each tissue section and recorded by a microscope (Autostainer link48, Dako, Denmark). The percentage of positive staining tubules or numbers of positive staining cells per mm 2 were evaluated. The experiments were performed in triplicate and repeated at least three times.

Colony formation test
Colony formation test was carried out as previously described [25]. Transfected cells were seeded in 6-well plates at a density of 800 cells per well and then cultured at 37 °C in 5% CO 2 for 2 weeks. Then, the cells were washed with phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde for 20 min and stained with 0.5% crystal violet solution for another 10 min. The colonies were then counted and analyzed, and the average of three independent experiments was calculated.

Cell apoptosis assay
The flow cytometry assays were performed as previously described [14]. Cells were plated in 6-well plates at a density of 5 × 10 5 cells/well, and the cells were collected and counted when cells were grown to logarithmic growth phase. The cells were centrifuged and 195 μL of Annexin V-FITC binding solution was added. 5 μL Annexin V-FITC and 10 μL propidium iodide staining solution were add to mix. Cells were incubated in the dark for 20 min, and then placed in an ice bath (Apoptosis Detection kit; KeyGEN). Data were analyzed using FlowJo software (Tree Star, Ashland, OR, USA).

Bioinformatics analyses
The prediction of target genes of miR-29a-3p was performed by Starbase (http://starb ase.sysu.edu.cn), Tar-getScan (http://www.targe tscan .org/) and miRDB (mirdb. org/miRDB/) bioinformatics analyses, followed by further analysis with the KEGG system. Twenty predicted targets were identified during the first screening. Next, the secondary screening was performed by selection of genes with functional change that induced any changes in LIFR-AS1 and COL1A2 in previous studies, which were selected priority over other candidates.

Transwell assay
Transwell assay was carried out as previously described [25]. Gastric cancer cells were seeded in wells and then transfected with shLIFR-AS1, pcLIFR-AS1, miR-29a-3p mimic, and miR-29a-3p inhibitor or its negative control for 48 h. A cell migration test was carried out with 24-well Transwell system (Costar, Badhoevedorp, Netherlands). Cells (3 × 10 5 ) were plated onto a Matrigel-coated membrane in the upper chamber of a 24-well insert with the serum-free medium. Culture medium containing 10% fetal bovine serum was added into the basolateral chamber and left to incubate at 37 °C for 24 h. The transwell chamber was washed by PBS twice 5 min each time, fixed with 5% pentanediol at 4 °C, stained by 0.1% crystal violet for 30 min, washed with PBS twice, and observed under a microscope. The numbers of cells that passed through the Matrigel were used as the index to evaluate how well the cells were able to invade. Assays were performed in duplicate in three independent experiments.

Biotinylated RNA pull-down assay
The pull-down assay was performed as previously described [17]. Briefly, BSG823 cells were transfected with the 3′ end biotinylated miR-29a-3p or miR-29a-3pmut for 24 h at a final concentration of 20 nmol/L. Then, the cells were incubated in the cell lysate with streptavidin-coated magnetic beads (Ambion, Life Technologies). The biotin-coupled RNA complex was pulled down and analysis of the abundance of LIFR-AS1 in bound fractions was then conducted by RT-qPCR.

Western blotting
The western blotting was performed as previously described [25]. Proteins were extracted at 48 h after transfection or plating the cells onto a 6-well plate. The cells in each well were lysed in 500 μL of lysis buffer containing protease inhibitors (150 mM NaCl, 25 mM Tris-HCl, pH 8.0, 0.5 M EDTA, 20% Tri-ton X-100, 8 M urea, and 1X protease inhibitor cocktail). Protein samples (20 μg) were separated by SDS-PAGE analysis and transferred onto a PVDF membrane. The primary antibodies used were anti-GPC5 (1:1000; Abcam) and antiβ-actin (1:5000; Abcam). After washing, the membrane was incubated with goat anti-rabbit HRP-conjugated secondary antibody (1:10,000; Abcam) at room temperature for 2 h. The immunocomplexes were visualized by ECL detection (ThermoScientific).

Statistical methods
Data were statistically analyzed using SPSS version 19.0 (IBM, Armonk, NY, USA). Data were presented as mean ± standard deviation (SD) of at least three independent experiments. Unpaired Student t test was used to compare differences between two groups. One-way ANOVA with Students' Unpaired post hoc test was conducted to compare differences among multiple groups. Statistical significance was considered with p < 0.05.

The function of LIFR-AS1 in GC
To evaluate the expression of LIFR-AS1 in GC tissues and cell lines, qRT-PCR was performed. As shown in Fig. 1a, the expression of LIFR-AS1 was markedly upregulated in GC tissue samples in comparison to that in normal adjacent tissue samples (Fig. 1a, p < 0.01). Similarly, the expression of LIFR-AS1 was markedly upregulated in GC cell lines (BSG823, HS746T, 9811, BGC803, MKN28, MGC803 and BSG823) in comparison to that in normal GSE1 cells (Fig. 1b, p < 0.01 or p < 0.001). After down-regulating the expression of LIFR-AS1 in BSG823 cells and BGC803 cells with shLIFR-AS1 (Fig. 1c, p < 0.05), the cell viability, colony formation, invasion and migration of BSG823 cells and BGC803 cells were all significantly decreased ( Fig. 1d-f, p < 0.05), while the apoptosis of BSG823 cells and BGC803 cells was significantly increased (Fig. 1g, p < 0.01). These results suggested that LIFR-AS1 played an important role in the progression of GC.

The function of miR-29a-3p in GC
The expression of miR-29a-3p in GC tissues and cell lines was also detected by qRT-PCR. As shown in Fig. 2a, the expression levels of miR-29a-3p were markedly decreased in GC tissue samples in comparison to that in normal adjacent tissue samples (Fig. 2a, p < 0.01). The expression of LIFR-AS1 in GC tissues was negatively correlated with miR-29a-3p (Fig. 2b). In addition, the expression of LIFR-AS1 was markedly downregulated in GC cell lines (BSG823, HS746T, 9811, BGC803, MKN28, MGC803 and BSG823) in comparison to that in normal GSE1 cells (Fig. 2c, p < 0.01 or p < 0.001). After up-regulating the expression of miR-29a-3p in BSG823 cells and BGC803 cells with miR-29a-3p mimic (Fig. 2d, p < 0.05), the cell viability, colony formation, invasion and migration of BSG823 cells and BGC803 cells were all significantly decreased (Fig. 2e-g, p < 0.05), while the apoptosis of BSG823 cells and BGC803 cells was significantly increased (Fig. 2h, p < 0.01). These results suggested that miR-29a-3p played an important role in the progression of GC.

Effect of LIFR-AS1 on GC tumorigenesis
Finally, the effect of LIFR-AS1 on the growth of GC was studied through in vivo experiments. As shown in Fig. 6a-c, pcLIFR-AS1 increased the tumor volume and weight compared with that in the control group, while shLIFR-AS1 decreased tumor volume and weight (p < 0.05). It indicated that overexpression of LIFR-AS1 promoted GC tumor growth, while knockdown of LIFR-AS1 inhibited GC tumor growth. Furthermore, immunohistochemistry results showed that the number of Ki67-positive cells was increased in the pcLIFR-AS1 group, while it was decreased in the shLIFR-AS1 group (Fig. 6d). In addition, the expression levels of LIFR-AS1 in the pcLIFR-AS1 group were significantly raised (p < 0.01), while the expression levels of shIFR-AS1 in the shLIFR-AS1 group were significantly reduced (Fig. 6e, p < 0.01). Taken together, these results suggested that LIFR-AS1/miR-29a-3p/COL1A2 axis played an important role in tumorigenesis of GC.

Discussion
Although great improvement has been made in the treatment of GC, the mortality remains high in patients with advanced GC [26]. In order to improve the therapeutic effect of gastric cancer, new therapeutic strategies and targets should be further developed. In recent years, accumulating evidence has shown that lncRNAs are important regulators implicated in the tumor pathogenesis and progression of diverse human cancers, including GC [27]. In this study, we found that the expression of lncRNA LIFR-AS1 in GC was significantly upregulated. Down-regulation of LIFR-AS1 inhibited the proliferation, migration and invasion of GC cells and induced apoptosis of GC cells. Down-regulation of LIFR-AS1 promoted tumor growth, while upregulation of LIFR-AS1 inhibited tumor growth. Our results suggested that LIFR-AS1 plays an important role in the progression of GC. LIFR-AS1 is a recently discovered lncRNA, which is reported to be abnormally expressed in a variety of malignant tumors and involves in the progression of tumors. For example, one study reported that the expression of LIFR-AS1 was significantly upregulated in colorectal cancer tissues and could modulate the resistance of colorectal cancer to photodynamic therapy [28]. Another study reported that LIFR-AS1 was downregulated in breast cancer and associated with poor survival [25]. Moreover, in consistent with our results here that the expression levels of LIFR-AS1 in GC were significantly increased in GC, another study also reported that the expression levels of LIFR-AS1 were significantly higher in GC tumor samples, and it is correlated with the poor survival in GC. However, the underlying mechanism of the action of LIFR-AS1 in GC remains unknown.
LncRNA, as a competitive endogenous RNA (ceRNA), binds miRNA to lncRNA through sponge action, which affects the function of miRNA and regulates protein translation and cell activity. Recent studies have shown that miRNAs and the development of GC is closely associated [29]. By interacting with different signaling pathways and molecules, miRNA plays a part in the regulation of GC cell apoptosis, invasion and metastasis [30]. For instance, it was reported that the expression levels of miRNA-144-3p were reduced in GC, and microRNA-144-3p suppressed gastric cancer progression by inhibiting epithelial-to-mesenchymal transition through targeting PBX3 [31]. In this study, our bioinformatics analyses revealed that miR-29a-3p was a target gene of LIFR-AS1. Moreover, we found that the expression of miR-29a-3p in GC was also significantly reduced. LIFR-AS1 was negatively correlated with miR-29a-3p, which was found to be involved in tumor proliferation, migration, invasion, and apoptosis induced by LIFR-AS1. Relationship between LIFR-AS1 and miR-29a-3p expression in GC tissues. c MiR-29a-3p expression levels in GC cell lines. d The expression level of miR-29a-3p in BSG823 and BGC803 cells transfected with miR-NC, miR-29a-3p mimics. e Cell viability of BSG823 and BGC803 cells. f Colony formation of BSG823 and BGC803 cells. g BSG823 and BGC803 cell migration and invasion. h Flow cytometry for apoptosis. *p < 0.05; **p < 0.01, ***p < 0.001, n = 3 (See figure on next  Taken together, these results suggested that LIFR-AS1 plays important roles in the progression of GC by regulating miR-29a-3p. In addition, studies have shown that the COL1A2 gene is involved in the proliferation of various tumors, such as esophageal cancer, liver cancer and GC [32]. The COL1A2 gene is involved in encoding type I collagen, which affects tumor cell proliferation [33]. In this study, we found that the expression levels of COL1A2 in gastric cancer were raised. COL1A2 was a target gene of miR-29a-3p. Co-transfection of miR-29a-3p mimic and shLIFR-AS1 inhibited the expression of COL1A2 in BSG823 cells. Co-transfection of miR-29a-3p mimic and pcLIFR-AS1 reversed the effect of pcLIFR-AS1 on the expression of COL1A2. And co-transfection of miR-29a-3p mimic with pc-COL1A2 reversed the effects of miR-29a-3p mimic on cell proliferation. These results demonstrated that LIFR-AS1 regulated COL1A2 through miR-29a-3p in GC. Compared with small RNA, lncRNA has longer sequence, more complex spatial structure, and more complex and diverse mechanisms are involved in their regulation function. Clinically, even if the same cancer patients receive the same treatment, they often show different treatment results. Some studies have pointed out that the differential expression of lncRNA is one of the reasons for this phenomenon. Because of the complexity of the regulation of lncRNAs, the mechanism of lncRNAs in different types of cancer and even in the same type of cancer is not the same. For example, lncRNA XIST promotes cell growth and invasion through regulating miR-497/MACC1 axis in GC [34]. LncRNA H19 promotes GC via FADD/Caspase 8/Caspase 3 signaling pathway [35]. LncRNA CCAT1 contributes to the growth and invasion of GC via targeting miR-219-1 [36]. Moreover, LIFR-AS1 modulates Sufu to inhibit cell proliferation and migration by miR-197-3p in breast cancer [25]. LIFR-AS1 involved in colorectal cancer resistance to photodynamic therapy via LIFR-AS1/miR-29a/TNFAIP3 axis [13]. In this study, we showed that LIFR-AS1 modulates COL1A2 to promote cell proliferation and migration by miR-29a-3p in GC. Altogether, these studies suggest that the same lncRNA may play the same role through different mechanisms in different cancers, and there are probably more unknown mechanisms that need to be further studied. Meanwhile, our findings indicate a novel mechanism by which LIFR-AS1 promotes GC invasion and metastasis by miR-29a-3p/COL1A2 axis. They may be Cell viability of BSG823 cells. c Colony formation of BSG823 cells. d BSG823 cell apoptosis. *p < 0.05, **p < 0.01, ***p < 0.001, n = 3 potential therapeutic targets for GC. In addition, we also found that there were many other targets of LIFR-AS1 were detected by prediction tool analyses, such as miR-138-5p, miR-542-3p, miR-374c-5p, miR-144-3p, miR-655-3p and miR-4677-3p and so on (data not shown). Although they have a lower significant than mir-29a-3p, their role in the LIFR-AS1 signaling pathway should also be studied in the future.

Conclusion
Our results demonstrated that LIFR-AS1 modulates COL1A2 to promote cell proliferation and migration by miR-29a-3p in GC. LIFR-AS1 is a promising prognostic indicator or therapeutic targets in GC patients.