LncRNA NR2F2‐AS1 promotes tumourigenesis through modulating BMI1 expression by targeting miR‐320b in non‐small cell lung cancer

Abstract Recently, long noncoding RNAs (lncRNAs) are attracting wide attention in the field of cancer research because of its important role in cancer diagnosis and prognosis. But studies on the biological effects and relevant mechanisms of lncRNAs in non‐small cell lung cancer (NSCLC) remain few and need to be enriched. Our study discussed the expression and biological effects of LncRNA NR2F2‐AS1, and further explored its possible molecular mechanisms. As a result, elevated expression of NR2F2‐AS1 was detected in NSCLC tissues and cells and was remarkably associated with the tumor, node, metastasis (TNM) stage and the status of lymphatic metastasis of patients. Down‐regulated NR2F2‐AS1 contributed to the promotion of cell apoptosis and the inhibition of cell proliferation and invasion in A549 and SPC‐A‐1 cells in vivo and vitro. Through bioinformatics analysis, NR2F2‐AS1 functions as a ceRNA directly binding to miR‐320b, BMI1 was a direct target of miR‐320b. Combined with the following cellular experiments, the data showed that NR2F2‐AS1 may influence the NSCLC cell proliferation, invasion and apoptosis through regulating miR‐320b targeting BMI1.


| INTRODUCTION
Lung cancer is the most commonly reported cancers worldwide, the new cases diagnosed are increasing year by year. 1 Despite the fact that comprehensive treatment, including surgical resection, radiotherapy and chemotherapy, continues to develop, the overall 5-year survival rate of lung cancer patients is less than 20% 2 which has not been obviously improved over the past few decades. Non-small cell lung cancer (NSCLC) is a main histological type of lung cancer with a constituent per cent of approximately 85%. 3 Notably, the proportion of patients who undergo surgical resection for NSCLC is in low level, for most of the patients are usually diagnosed at an advanced stage.
As a result, improving the diagnosis and treatment of NSCLC really needs more attention. In the process of cancer development, multiple factors are involved and may influence treatment and prognosis.
It is necessary to further understand relevant molecular mechanisms of NSCLC so that we can discover potential therapeutic targets which are helpful for cancer diagnosis or limiting cancer progress. Recently, long noncoding RNAs (lncRNAs) are attracting wide attention in the field of cancer research because of its important role in cancer diagnosis and prognosis. 4 LncRNAs itself are defined as non-protein coding transcripts longer than 200 nucleotides. [5][6][7] Research shows that abnormal expression of lncRNAs occurs in various cancer diseases, 8 including liver cancer, 9 gastric cancer, 10 oesophageal cancer, 11 breast cancer and prostate cancer. 12 Also, this kind of phenomenon can be found in coronary artery disease, 13 and nerve system disease. 14 Through the latest technology including bioinformatics analyses and next-generation sequencing technology, the regulatory effects and potential application of lncRNAs were slowly and partly come into our visual fields, such as lncRNA LINC01116, CASC9, ANRIL. [15][16][17][18] In the NSCLC which we have paid attention to, articles have demonstrated that long non-coding RNA MEG3 could inhibit NSCLC cell proliferation and induces apoptosis by affecting p53 expression 19 ; lncRNA AK126698 can affect the cisplatin resistance of NSCLC cells. 20 As a result, it is well worth exploring the LncRNAs and their relative mechanisms in NSCLC.
In this study, LncRNA NR2F2-AS1 was found to be abnormally expressed based on the previous analysis of lncRNAs microarray among NSCLC tissues. 21 NR2F2, also named COUP-TFII, belongs to the nuclear receptor family and has been reported to be important in the treatment and prevention of cancers. 22,23 So far, the function of LncRNA NR2F2-AS1 in NSCLC has not been explored. We aimed to confirm the expression of NR2F2-AS1 in NSCLC and dig into relevant mechanisms of its regulatory network.

| Patients and tissue specimens
For this study, we spent about a year to collect relative tissues.
Thirty-nine paired of tissue specimens and matched normal tissue samples obtained from the NSCLC patients underwent operation at the First Affiliated Hospital of Zhengzhou University from March 2016 to August 2017. The tissues were stored in liquid nitrogen until the isolation of total RNA. In this cohort of NSCLC patients, 22 were men and 17 were women, with a median age of 56 years.
According to the 7th TNM staging system, 24 23 cases of the NSCLC patients were classified as stage I/II and 16 cases were classified as stage III. Among all of the cases, 28 had lymph node metastasis.
Before we carried out this study, all the patients were informed and the informed consent forms were duly signed. Moreover, The Ethics Review Committee of Zhengzhou University approved the study.

| RNA isolation and quantitative real time PCR
Total RNA in NSCLC and normal tissues was extracted using TRIzol reagent (Invitrogen, USA). The uptake of total RNA in lung cancer

| Cell invasion assays
For the sake of cell invasion exploration, a transwell assay was per- After 48 hours, the membrane containing invaded cells was soaked in methanol and stained with 0.1% crystal violet. Cells were finally photographed and calculated under an inverted microscope (Olympus, Japan). The experiments were repeated in triplicate independently.

| Dual-luciferase assays
In order to investigate the relationship between LncRNA NR2F2-AS1 and miR-320b as well as miR-320b and BMI1, we first searched the potential binding sites through the informatics analysis of micro-RNA database and Target Scan. PCR and overlap PCR were performed to amplify LncRNA NR2F2-AS1 cDNA containing predicted miR-320b binding site or the 3ʹUTR of BMI1 and then the PCR products were cloned downstream of the luciferase gene in the pmirGLO vector. When it comes to the dual-luciferase assay, the combined plasmids pmirGLO-Wt-LncRNA NR2F2-AS1 or pmirGLO-Mt-LncRNA NR2F2-AS1 were transiently cotransfected to A549 and SPC-A-1 cells accompanied by miR-320b mimic or scrambled miRNA (miR-NC) through Lipofectamine ™ 3000 (Invitrogen, CA, USA), which is similar to the cotransfection of BMI1 combined plasmids and miR-320b mimic or miR-NC. Finally, the relative luciferase activity was determined with the luciferase assay kit (Promega, USA) at 48 hours after transfection.

| Western blot assays
We use Western blot assays to measure the expression of BMI1 protein level in A549 and SPC-A-1 cells. The RIPA lysis buffer (Solarbio, Beijing, China) was used to lyse cells to obtain total protein which was quantified by a BCA protein assay kit (Beyotime, Shanghai, China). About 50 µg of protein was separated by electrophoresis with 10% SDS-PAGE and transferred onto a PVDF membrane (Thermo Scientific, MA, USA). After being blocked in 5% skim milk powder for 1 hour, the membrane was probed with the primary antibody for BMI1 and GAPDH at a concentration of 1:500 and 1:1000, respectively at 4°C overnight. The next day the membrane was washed with TBST for four times (5, 10, 10, 15 minutes) at room temperature. The membrane was then incubated with the secondary antibody for 1 hour at a concentration of 1:5000. Wash it again with TBST for four times. Finally, the ECL Substrate (Thermo Scientific, MA, USA) was added to determine the protein immunoreactivity by FluorChem E (ProteinSimple, CA, USA).

| Statistical analysis
The whole data obtained above were analysed using the SPSS software version 22.0. Mean ± Standard deviation (SD) is the way to present the results of the statistical analysis. According to the data types, Student's t test and one-way ANOVA were carried out to compare statistical significance among groups. When it comes to evaluate the correlation between two variants, the spearman rank correlation test was applied. A P < 0.05 was considered statistically significance.

| The expression of LncRNA NR2F2-AS1 was in high level in NSCLC tissues and cells
In order to disclose the role of LncRNA NR2F2-AS1 in NSCLC carcinogenesis, the expression of LncRNA NR2F2-AS1 in NSCLC samples and cell lines was analysed through qRT-PCR. As shown in Figure 1A and Table 1, LncRNA NR2F2-AS1 expression in 39 surgically resected NSCLC tissues was significantly higher than that of the corresponding normal tissues (P < 0.05). Based on this finding, we carried out the statistical analysis in the field of age, gender, tumour stage (according to the 7th TNM staging system 24    found the putative binding sites between LncRNA NR2F2-AS1 and miR-320b ( Figure 3A). To further test this kind of relationship, a luciferase report assay was performed and as seen in Figure 3B, miR-320b mimic lessened the luciferase activity of pmirGLO-wt-LncRNA NR2F2-AS1 but not of pmirGLO-wt-LncRNA NR2F2-AS1 (P < 0.05).

Moreover, we examined the expression level of miR-320b in A549
and SPC-A-1 cells transfected with si-LncRNA NR2F2-AS1 and the level of LncRNA NR2F2-AS1 in cells transfected with miR-320b mimic respectively. The results showed that miR-320b was significantly up-regulated when decreasing the expression of LncRNA NR2F2-AS1 (P < 0.05, Figure 3C); in turn, when we artificially elevated the expression of miR-320b, LncRNA NR2F2-AS1 expression was in a relatively low level (P < 0.05, Figure 3D). Based on the inverse relationship between LncRNA NR2F2-AS1 and miR-320b (R 2 = 0.429, P < 0.05, Figure 3E), we concluded that in A549 and SPC-A-1 cells, LncRNA NR2F2-AS1 could act as a ceRNA through binding to miR-320b.

| Up-regulated miR-320b inhibited cell proliferation and invasion and induce cell apoptosis in A549 and SPC-A-1 cells
Based on the data obtained above, we have found that when LncRNA NR2F2-AS1 was down-regulated in A549 and SPC-A-1 cells, miR-320b was up-regulated accordingly; Down-regulated LncRNA NR2F2-AS1 contributed to the promotion of cell apoptosis and the inhibition of cell proliferation and invasion. Then, we wonder how to explore the effects of overexpressed miR-320b in lung cancer cells.
As shown in the Figure 4A, the expression of miR-320b was successfully increased after being transfected with miR-320b mimic.
Compared with the NC group, the A450 value of miR-320b mimic group was obviously decreased (P < 0.05, Figure 4B) along with the number of invasive cells (P < 0.05, Figure 4C). On the contrary, the number of apoptotic cells in miR-320b mimic group was significantly increased than that of the NC group (P < 0.05, Figure 4D). To further confirm the condition of cell apoptosis, the caspase3/7 activity was also detected, and the results were consistent with the FCM assay (P < 0.05, Figure 4E). Through the above experiments , up-regulated miR-320b turned out to inhibit cell proliferation and invasion and promote cell apoptosis in A549 and SPC-A-1 cells.

| BMI1 was a direct target of miR-320b
In order to elucidate the possible mechanisms of miR-320b in the apoptosis, proliferation and invasion of A549 and SPC-A-1 cells, the bioinformatics analysis of microRNA database and Target Scan was applied to search the putative binding sites of target genes. As indicated in Figure 5A, miR-320b may regulate BMI1 by binding to its 3ʹUTR. To identify this hypothesis, a mutant BMI1 3ʹUTR was designed ( Figure 5A) and amplified using PCR. After successfully constructing wild and mutant type of BMI1 3ʹUTR vector (pmirGLO-Wt-BMI1 3ʹUTR and pmirGLO-Mt-BMI1 3ʹUTR), the dual-luciferase report assay was performed in A549 and SPC-A-1 cells. As a consequence, a significant decrease in the relative luciferase activity was identified following cells co-transfected with pmirGLO-Wt-BMI1 3'UTR plasmid and miR-320b mimic. (P<0.05, Fig.5C); moreover, when cells were co-transfected with pmirGLO-Mt-BMI1 3'UTR and miR-320b mimic or co-transfected with pmirGLO-Mt-BMI1 3'UTR and miR-NC, the relative luciferase activity in the two groups showed no difference (P > 0.05, Figure 5C). The expression of BMI1 protein was also measured using Western Blot assay and was suppressed by overexpression of miR-320b (P < 0.05, Figure 5B). All these results indicated that miR-320b could down-regulate the expression of BMI1 for the fact that BMI1 is a direct target of miR-320b.

| Down-regulated LncRNA NR2F2-AS1, upregulated miR-320b and si-BMI1 shared similar effects in cell proliferation, invasion and apoptosis
Through the study above, we learned that either down-regulated LncRNA NR2F2-AS or up-regulated miR-320b inhibited cell proliferation, invasion and induce cell apoptosis in A549 and SPC-A-1 cells.
NR2F2-AS could serve as the ceRNA through binding to miR-320b.  Figure 6A). CCK-8 assay, transwell assay and FCM assay were applied to detect the cell activity. As shown in Figure 6B, C, compared to the negative control group, the number of live cells and invasive cells in the three experimental groups was significantly less (P < 0.05), on the contrary, the number of apoptotic cells in the three groups above was obviously increased (P < 0.05, Figure 6D).
Through the next statistical analysis among the three groups, there existed no difference in cells proliferation, invasion and apoptosis (P > 0.05) suggesting that down-regulated LncRNA NR2F2-AS1, upregulated miR-320b and si-BMI1 shared similar effects on the activity of A549 and SPC-A-1 cells.

| Down-regulated miR-320b or BMI1 lacking 3ʹUTR could reverse the effect of down-regulated LncRNA NR2F2-AS1 in A549 and SPC-A-1 cells
To determine whether NR2F2-AS1 produce biological effects through modulating the regulation effects of miR-320b on BMI1 in advance, rescue experiments were conducted. A549 and SPC-A-1 cells were divided into four groups respectively. The results of Western blot assay showed that in comparison with the NC group, the expression of BMI1 protein was obviously suppressed in si-LncRNA group (P < 0.05) but exerted no difference (P > 0.05, Figure 7A) from the two cotransfection groups (si-LncRNA+inhibitor miR-320b According to current research findings, the pool of mRNAs, lncRNAs and other non-coding RNAs shared common MREs with miRNAs; lncRNAs thus could serve as natural miRNA "sponges" and inhibit miRNA function through competitive binding to MREs. 25,29 For example, knockdown of lncRNA NEAT1 inhibited glioma cell migration and invasion via the modulation of miR-132. 30 lncRNA-H19 and miR-141 could compete with each other and affect their target genes in gastric cancer. 31 An lncRNA named CHRF was also found to act as an endogenous sponge of miR-489, which down-regulated miR-489 expression levels as a result to regulate cardiac hypertrophy. 32 In NSCLC patients, c-Myc-activated long non-coding RNA H19 could down-regulate miR-107 and promote cell cycle progression. 33 Then, in this study, we supposed there existed one or more miRNAs which could be regulated through the function of LncRNA NR2F2-AS1. The bioinformatics analysis of microRNA database and the results of dual-luciferase report assay showed that LncRNA NR2F2-AS1 could act as a ceRNA by binding to miR-320b. NSCLC patients was in a relatively low level, which is contrast to that of NR2F2-AS1. When we artificially up-regulate miR-320b expression, the proliferation and invasion of A549 and SPC-A-1 cells were significantly suppressed, while the apoptotic cells were highly increased, demonstrating that up-regulated miR-320b inhibit the progression of NSCLC cells in cell proliferation, invasion and apoptosis to some degree.
As we all know, miRNAs could act as oncogenes or anti-oncogenes within the modulation of various target genes, as a result, they could influence the transcription and translation of relative protein. 38 For the sake of regulatory mechanisms of miR-320b in NSCLC, the bioinformatics analysis of Target Scan and the results of dual-luciferase report assay revealed that BMI1 is a direct target of miR-320b. Besides, through Western blot assay, BMI1 protein expression was obviously inhibited when miR-320b was up-regulated, proving that miR-320b could down-regulate the expression of BMI1.
However, our results indicated that either down-regulated NR2F2-AS1 or up-regulated miR-320b inhibited cell proliferation, invasion and induced cell apoptosis in A549 and SPC-A-1 cells.
NR2F2-AS1 could serve as the ceRNA through binding to miR-320b.
Then, we supposed that miR-320b may be affected by NR2F2-AS1, and the signal pathway could be related to the effects of NR2F2-AS1 in NSCLC. In order to prove this hypothesis, the effects on cell proliferation, invasion and apoptosis of A549 and SPC-A-1 cells among si-LncRNA NR2F2-AS, miR-320b mimic and si-BMI1 were compared through cellular experiments. The data revealed that cell activities in the three groups were similarly suppressed. The results of further rescue experiments also supported the assumption for they showed that the down-regulated miR-320b or BMI1 lacking F I G U R E 7 Down-regulated miR-320b or BMI1 lacking 3ʹUTR could reverse the effect of down-regulated NR2F2-AS1 in A549 and SPC-A-1 cells. A, representative Western blots showed BMI1 levels in si-LncRNA group, si-LncRNA+inhibitor miR-320b group, si-LncRNA+pcDNA3.1-BMI1 group and NC group. (B-D) the effects on cell proliferation, cell invasion and cell apoptosis in si-LncRNA group, si-LncRNA+inhibitor miR-320b group, si-LncRNA+pcDNA3.1-BMI1 group and NC group through the application of CCK-8 assay, transwell detection and flow cytometry. *P < 0.05 3ʹUTR could reverse the effect of down-regulated NR2F2-AS1 on the growth, metastasis and apoptosis of A549 and SPC-A-1 cells.
In conclusion, we described down-regulated lncRNA NR2F2-AS1 as an anti-oncogenic lncRNA that contributed to the promotion of cell apoptosis and the inhibition of cell proliferation and invasion in A549 and SPC-A-1 cells through the regulation of miR-320b/BMI1 axis. Based on the abnormal expression of NR2F2-AS1 in NSCLC tissues, it may serve as a therapeutic target or useful marker in the treatment or diagnosis for NSCLC. For further understanding, more clinical and experimental studies need to be carried out.