Long Noncoding RNA ANRIL Promotes Non–Small Cell Lung

Published OnlineFirst December 12, 2014; DOI: 10.1158/1535-7163.MCT-14-0492

Cancer Biology and Signal Transduction Molecular Cancer Therapeutics Long Noncoding RNA ANRIL Promotes Non–Small Cell Lung Cancer Cell Proliferation and Inhibits Apoptosis by Silencing KLF2 and P21 Expression Feng-qi Nie1, Ming Sun2, Jin-song Yang3, Min Xie2, Tong-peng Xu1, Rui Xia2, Yan-wen Liu2, Xiang-hua Liu2, Er-bao Zhang2, Kai-hua Lu1, and Yong-qian Shu1

Abstract

Recent evidence highlights long noncoding RNAs (lncRNA) was increased in NSCLC tissues, and its expression level was as crucial regulators of cancer biology that contribute to essen- significantly correlated with tumor–node–metastasis stages and tial cancer cell functions such as cell proliferation, apoptosis, tumor size. Moreover, patients with high levels of ANRIL and metastasis. In non–small cell lung cancer (NSCLC), several expression had a relatively poor prognosis. In addition, taking lncRNAs' expressions are misregulated and have been nomi- advantage of loss-of-function experiments in NSCLC cells, we nated as critical actors in NSCLC tumorigenesis. LncRNA ANRIL found that knockdown of ANRIL expression could impair cell was first found to be required for the PRC2 recruitment to and proliferation and induce cell apoptosis both in vitro and vivo. silencing of p15INK4B, the expression of which is induced by the Furthermore, we uncover that ANRIL could not repress p15 ATM–E2F1 signaling pathway. Our previous study showed that expression in PC9 cells, but through silencing of KLF2 and P21 ANRIL was significantly upregulated in gastric cancer, and it transcription. Thus, we conclusively demonstrate that lncRNA could promote cell proliferation and inhibit cell apoptosis by ANRIL plays a key role in NSCLC development by associating silencing of miR99a and miR449a transcription. However, its its expression with survival in patients with NSCLC, providing clinical significance and potential role in NSCLC is still not novel insights on the function of lncRNA-driven tumorigenesis. documented. In this study, we reported that ANRIL expression Mol Cancer Ther; 14(1); 268–77. 2014 AACR.

Introduction in the development, progression, and spread of the NSCLC is essential for the developing of specificdiagnosticmethodsand Lung cancer is the most common type of cancer and the designing of more individualized and effective therapeutic primary cause of cancer-related death worldwide (1). Non– strategies. small cell lung cancer (NSCLC) accounts for 80% of all lung Recently, studies using the great advances in genomic tech- cancer cases, represents the most prevalent class of this cancer nologies have revealed the majority of the human genome is type, and includes several histologic subtypes such as squa- transcribed, whereas only 2% of the transcribed genome codes mous cell carcinoma (SCC), adenocarcinoma and large cell for protein (5). Meanwhile, it is becoming increasingly appar- carcinoma (LCC; refs. 2, 3). Despite current advances in the ent that the large majority of genome is transcribed into treatments for NSCLC, including surgical therapy, chemother- noncoding RNAs (ncRNAs), including microRNAs and long apy, and molecular targeting therapy, the overall 5-year sur- ncRNAs (lncRNAs; ref. 6). The ENCODE Consortium has vival rate for patients with NSCLC has not been markedly elucidated the prevalence of thousands of human lncRNAs, improved over years and remains as low as 15% (4). Therefore, but only very few of them have been assigned with any biologic a greater understanding of the molecular mechanisms involved function (7). To date, studies showed that miRNAs play important roles in the posttranscriptional regulation of gene expression; however, the lncRNAs counterpart of ncRNA is not well 1Department of Oncology, First Affiliated Hospital, Nanjing Medical University, Nanjing, People's Republic of China. 2Department of Bio- characterized (8). Although very few are characterized in detail, chemistry and Molecular Biology, Nanjing Medical University, Nanjing, LncRNAs are involved in a large range of biologic processes, 3 People's Republic of China. Department of Oncology, Nanjing First including modulation of apoptosis and invasion, reprogram- Hospital, Nanjing Medical University, People's Republic of China. ming stem cell pluripotency, and parental imprinting through Note: Supplementary data for this article are available at Molecular Cancer the regulation of gene expression by chromatin remodeling, Therapeutics Online (http://mct.aacrjournals.org/). histone protein modification, regulation of mRNA splicing, and F.-q. Nie, M. Sun, and J.-s. Yang contributed equally and are joint first authors to acting as sponges for microRNAs (9–12). this article. In the past decade, lots of evidence have linked the dysregula- Corresponding Authors: Kai-hua Lu, Department of Oncology, First Affiliated tion of lncRNAs with diverse human diseases, in particular cancers Hospital, Nanjing Medical University, Han Zhong Road 140#, Nanjing 210029, (13–15). Therefore, identification of cancer-associated lncRNAs China. Phone/fax: 25-8686-2728; E-mail: [email protected]; and Yong-qian Shu, and investigation of their molecular and biologic functions are [email protected] important in understanding the molecular biology of NSCLC doi: 10.1158/1535-7163.MCT-14-0492 development and progression. Our previous study showed that 2014 American Association for Cancer Research. lncRNA ANRIL was significantly upregulated in gastric cancer,

268 Mol Cancer Ther; 14(1) January 2015

ANRIL Promotes Proliferation in NSCLC

and increased ANRIL promoted gastric cancer cells prolifera- cells proliferation and apoptosis partly via silencing of KLF2 tion and inhibited apoptosis by epigenetic silencing of miR99a and P21 transcription. This study advances our understanding and miR449a transcription (16). Moreover, ANRIL can bind of the role of lncRNAs, such as a regulator of pathogenesis of to and recruits PRC2 to repress the expression of p15INK4B locus, NSCLC and facilitates the development of lncRNA-directed which resulted in increased cell proliferation (17, 18). However, diagnostics and therapeutics. the ANRIL clinical signiﬁcance and potential role in NSCLC development and progression is still not documented. In this study, we found that lncRNA ANRIL expression was Materials and Methods increased in NSCLC tissues compared with adjacent normal Tissue collection tissues. Its expression level was signiﬁcantly correlated with We obtained 68 paired NSCLC and adjacent nontumor lung tumor–node–metastasis (TNM) stages and tumor size. More- tissues from patients who underwent surgery at Jiangsu Province over, patients with higher level of ANRIL expression had a Hospital between 2010 and 2011, and were diagnosed with relatively poor prognosis. Furthermore, we investigated the NSCLC based on histopathologic evaluation. Clinicopathologic effects of ANRIL expression on NSCLC cell phenotype in vitro characteristics, including TNM staging, were recorded. No local or and in vivo with the loss-of-function study. Moreover, we also systemic treatment was conducted in these patients before sur- showed that ANRIL could bind to PRC2 to repress KLF2 and gery. All collected tissue samples were immediately snap-frozen in P21 transcription, but not regulate P15INK4 expression in liquid nitrogen and stored at 80 C until required. Our study was NSCLC PC9 cells, which indicated that ANRIL affected NSCLC approved by the Research Ethics Committee of Nanjing Medical

A 50 B 50 Low ANRIL expression High ANRIL expression

40 40

30 30

20 20

10 10 Relative ANRIL expression (fold change) (fold Relative ANRIL expression Relative ANRIL expression (fold change) (fold Relative ANRIL expression

0 0

CD 1.0 1.0 Low ANRIL expression 0.8 Low ANRIL expression 0.8

0.6 0.6

0.4 0.4 Cum. survival Cum. High ANRIL expression High ANRIL expression survival Cum. 0.2 P < 0.001 0.2 P = 0.019

0.0 0.0

.00 10.00 20.00 30.00 40.00 .00 10.00 20.00 30.00 40.00 Months after surgery PFS Months after surgery overall survival

Figure 1. Relative ANRIL expression in NSCLC tissues and its clinical signiﬁcance. A, relative expression of ANRIL in NSCLC tissues (n ¼ 68) compared with corresponding nontumor tissues (n ¼ 68). ANRIL expression was examined by qPCR and normalized to GAPDH expression. Results are presented as the fold change in tumor tissues relative to normal tissues. B, ANRIL expression was classiﬁed into two groups. C and D, Kaplan–Meier DFS and OS curves according to ANRIL expression levels.

www.aacrjournals.org Mol Cancer Ther; 14(1) January 2015 269

Nie et al.

University, China. Written informed consent was obtained from early apoptotic cells, and apoptotic cells, and then the relative all patients. ratio of early apoptotic cells were compared with control trans- fectant from each experiment. Cells for cell–cycle analysis were Cell lines stained with PI using the CycleTEST PLUS DNA Reagent Kit (BD Five NSCLC adenocarcinoma cell lines (PC9, SPC-A1, NCI- Biosciences) following the protocol and analyzed by FACScan. H1975, H1299, and H358), and one NSCLC squamous carcino- The percentage of the cells in G0–G1, S, and G2–M phase were mas cell lines (H520) were purchased from the Institute of counted and compared. Biochemistry and Cell Biology of the Chinese Academy of Sciences (Shanghai, China). A549, H1975, H1299, and H520 Tumor formation assay in a nude mouse model cells were cultured in RPMI-1640; 16HBE, PC9, and SPC-A1 cells Female athymic BALB/c nude mice (4-weeks-old) were main- were cultured in DMEM (GIBCO-BRL) medium supplemented tained under pathogen-free conditions and manipulated with 10% FBS, 100 U/mL penicillin and 100 mg/mL streptomycin according to protocols approved by the Shanghai Medical o (Invitrogen, Carlsbad) at 37 C/5% CO2. All cell lines were Experimental Animal Care Commission. PC9 cells were stably authenticated by short tandem repeat DNA profiling. transfected with sh-ANRIL and empty vector and harvested from 6-well cell culture plates, washed with PBS, and resus- RNA extraction and qPCR assays pended at a concentration of 1 108 cells/mL. A total of 100 mL Total RNA was isolated with TRizol reagent (Invitrogen) of suspended cells was s.c. injected into a single side of the according to the manufacturer's instructions. Total RNA (500 posterior flank of each mouse. Tumor growth was examined ng) was reverse transcribed in a final volume of 10 mLusing every 3 days, and tumor volumes were calculated using the random primers under standard conditions for the PrimeScript equation V ¼ 0.5 D d2(V,volume;D,longitudinal RT reagent Kit (TaKaRa, Dalian, China). We used the SYBR diameter; d, latitudinal diameter). At 18 days after injection, Premix Ex Taq (TaKaRa, Dalian, China) to determine ANRIL mice were euthanized, and the subcutaneous growth of each expression levels, following the manufacturer's instructions. tumor was examined. This study was carried out in strict Results were normalized to the expression of GAPDH.The accordance with the recommendations in the Guide for specific primers used are shown in Supplementary Table S1. theCareandUseofLaboratoryAnimalsoftheNational The qPCR assays were conducted on an ABI 7500, and data Institutes of Health. The protocol was approved by the Com- collected with this instrument. Our qPCR results were analyzed mittee on the Ethics of Animal Experiments of the Nanjing and expressed relative to threshold cycle (Ct) values, and then medical University. converted to fold changes. RNA immunoprecipitation Cell transfection For immunoprecipitation (IP) of endogenous PRC2 complexes Human ANRIL cDNA clone L6ChoCKO-2-E10 with functional from whole-cell extracts, cells were lysed. The supernatants were region was provided by the Functional Genomics Research Cen- incubated with protein A Sepharose beads coated with antibodies ter, KRIBB, Korea. Plasmid vectors (pCNS-ANRIL, sh-ANRIL and that recognized EZH2, SNRNP70, or with control IgG (Millipore) empty vector) for transfection were prepared using DNA Midiprep or Midiprep kits (Qiagen), and transfected into16HBE, SPC-A1, H1299, or PC9cells. The si-ANRIL or si-NC was transfected into Table 1. Correlation between ANRIL expression and clinicopathologic SPC-A1, H1299, or PC9 cells. SPC-A1, H1299, or PC9 cells were characteristics of NSCLC patients fl grown on 6-well plates to con uency and transfected using ANRIL Lipofectamine 2000 (Invitrogen) according to the manufacturer's High, number of Low, number of c2 test instructions. At 48 hours after transfection, cells were harvested for Characteristics cases (34) cases (34) (P value) qPCR or Western blot analysis. Age, y 0.627 65 17 19 >65 17 15 Cell viability assays Gender 0.625 Cell viability was monitored using a Cell Proliferation Reagent Male 18 20 Kit I (MTT; Roche Applied Science). The SPC-A1, H1299, PC9, or Female 16 14 A549 cells transfected with si-ANRIL (3000 cells/well) were grown Histologic subtype 0.324 in 96-well plates. Cell viability was assessed every 24 hours SCC 22 18 following the manufacturer's protocol. All experiments were Adenocarcinoma 12 16 TNM stage 0.007a performed in quadruplicate. For each treatment group wells were Ia þ Ib 4 15 assessed in triplicate. IIa þ IIb 14 12 IIIa 16 7 Flow cytometry Tumor size 0.001a SPC-A1, H1299, or PC9 cells transfected with si-ANRIL were 5cm 13 26 harvested 48 hours after transfection by trypsinization. After the >5cm 21 8 double staining with FITC–Annexin V and Propidium iodide (PI) Lymph node metastasis 0.051 Negative 11 19 was done using the FITC–Annexin V Apoptosis Detection Kit (BD Positive 23 15 Biosciences) according to the manufacturer's recommendations, Smoking history 0.793 the cells were analyzed with a flow cytometry (FACScan; BD Smokers 23 24 Biosciences) equipped with a CellQuest software (BD Bio- Never smokers 11 10 sciences). Cells were discriminated into viable cells, dead cells, aOverall P < 0.05.

270 Mol Cancer Ther; 14(1) January 2015 Molecular Cancer Therapeutics

ANRIL Promotes Proliferation in NSCLC

Figure 2. Effects of knockdown of ANRIL on NSCLC cell viability in vitro. A, ANRIL expression levels of NSCLC cell lines (PC9, SPC-A1, NCI-H1975, H1299, and H358 and H520) compared with that in normal human bronchial epithelial cells (16HBE). B, SPC-A1, H1299, and PC9 cells were transfected with si-ANRIL. C, MTT assays were used to determine the cell viability for si-ANRIL–transfected SPC-A1, H1299, and PC9 cells. Values represent the mean SD from three independent experiments. D, colony-forming assays were conducted to determine the proliferation of si-ANRIL–transfected SPC-A1, H1299, and PC9 cells. Flow-cytometry assays were performed to analysis the cell-cycle progression and apoptosis when NSCLC cells transfected with si-ANRIL; , P < 0.05 and , P < 0.01. for 6 hours at 4C. After the beads were washed with wash buffer, with specific antibodies. ECL chromogenic substrate was used the complexes were incubated with 0.1% SDS/0.5 mg/mL Pro- and quantified by densitometry (Quantity One software; Bio- teinase K (30 minutes at 55C) to remove proteins, respectively. Rad). GAPDH antibody was used as control, anti-P21, CDK2, The PRC2 isolated from the IP materials was further assessed by CDK4, CDK6, P15, and PARP (1:1,000) were purchased from qPCR analysis (19). Cell Signaling Technology, Inc.; anti-KLF2 was purchased from Sigma. Chromatin immunoprecipitation PC9 cells were treated with formaldehyde and incubated for Statistical analysis 10 minutes to generate DNA–protein cross-links. Cell lysates All statistical analyses were performed using SPSS 17.0 software were then sonicated to generate chromatin fragments of 200 to (IBM). The significance of differences between groups was esti- 300 bp and immunoprecipitated with EZH2 and H3K27me3- mated by the Student t test, Wilcoxon test, or c2 test. Disease-free specific antibody (Cell Signaling Technology) or IgG as control. survival (DFS) and overall survival (OS) rates were calculated by Precipitated chromatin DNA was recovered and analyzed by the Kaplan–Meier method with the log-rank test applied for qPCR. comparison. The date of survival was evaluated by univariate and multivariate Cox proportional hazards models. Variables Western blot assay and antibodies with P < 0.05 in univariate analysis were used in subsequent Cells protein lysates were separated by 10% SDSPAGE, multivariate analysis on the basis of Cox regression analyses. transferredto0.22mm NC membranes (Sigma), and incubated Kendall Tau-b and Pearson correlation analyses were used to

www.aacrjournals.org Mol Cancer Ther; 14(1) January 2015 271

Nie et al.

investigate the correlation between ANRIL and KLF2 expressions. Association of ANRIL expression with patients' survival Two-sided P values were calculated, and a probability level of Kaplan–Meier survival analysis was conducted to investigate 0.05 was chosen for statistical significance. the correlation between ANRIL expression and NSCLC patient prognosis. According to relative ANRIL expression in tumor tissues, the 68 patients with NSCLC were classified into two Results groups: the high ANRIL group (n ¼ 34, fold-change mean ANRIL expression was upregulated and correlated with poor ratio); and the low ANRIL group (n ¼ 34, fold-change mean prognosis of NSCLC ratio; Fig. 1B). With respect to progression-free survival (PFS), this ANRIL expression levels were investigated in 68 paired was 35.3% for the low ANRIL group, and 13.6% for the high NSCLC samples and adjacent histologically normal tissues ANRIL group. Median survival time for the low ANRIL group was using qPCR assays. ANRIL expression was significantly upre- 31 months, and 14 months for the high ANRIL group (Fig. 1C). gulated (fold change >1.5, P < 0.01) in 76% (52/68) of The OS rate over 3 years for the low ANRIL group was 44.4%, and cancerous tissues compared with normal tissues (Fig. 1A); 20.8% for the high ANRIL group. Median survival time for the low the ANRIL expression level in each patient was shown in ANRIL group was 32 months, and 18 months for the high ANRIL Supplementary Table S2. Increased ANRIL expression levels group (Fig. 1D). in NSCLC were significantly correlated with tumor size (P ¼ Univariate analysis identified three prognostic factors: lymph 0.001), and advanced pathologic stage (P ¼ 0.007). However, node metastasis; TNM stage; and ANRIL expression level. Other ANRIL expression was not associated with other parameters clinicopathologic features such as gender and age were not sta- such as gender (P ¼ 0.625) and age (P ¼ 0.627) in NSCLC tistically significant prognosis factors (Supplementary Table S3). (Table 1). Multivariate analysis of the three prognosis factors confirmed that To investigate whether upregulation of ANRIL is caused by HR for ARAIL expression is 3.509 (95% confidence interval, DNA copy-number variation, we referred to the array comparative 1.619–7.607) of PFS, indicating that ANRIL expression may serve genomic hybridization (aCGH) database in GSE20393, where as a potential independent prognostic value in NSCLC (Supple- deposits of 52 lung cancer copy-number alteration data were mentary Table S4). generated by Agilent Human Genome CGH 244A Microarrays. We investigated 26 probes representing region of ANRIL and Modulation of ANRIL expression in NSCLC cells extracted GLAD-segmented copy number of these probes. The We next performed qPCR analysis to examine the expression results showed that there is no significant gain of DNA copy of ANRIL in 6 human NSCLC cell lines, including both number in this region, suggesting that upregulation of ANRIL in adenocarcinoma and squamous carcinoma subtypes (Fig. lung cancer is not due to copy-number variation (Supplementary 2A). To investigate the functional effects of ANRIL in NSCLC Fig. S1A). cells, we modulated its expression through RNAi. qPCR

Figure 3. Effects of knockdown of ANRIL on NSCLC cell cycle and apoptosis in vitro. A, the bar chart represents the percentage of cells in G0–G1,S,orG2–M phase, as indicated. B, apoptosis was determined by ﬂow cytometry; upper left, necrotic cells; upper right, terminal apoptotic cells; lower right, early apoptotic cells. C, apoptosis was determined by Tunel staining. All experiments were performed in biologic triplicates with three technical replicates; , P < 0.01.

272 Mol Cancer Ther; 14(1) January 2015 Molecular Cancer Therapeutics

ANRIL Promotes Proliferation in NSCLC

analysis of ANRIL levels was performed 48 hours after trans- that NSCLC cells transfected with ANRIL siRNA promoted apo- fection. ANRIL expression was knocked down by 74% in ptosis in comparison with that in control cells (Fig. 3B and 2C). SPC-A1 cells, 75% in H1299 cells, and 94% in PC9 cells These data indicate that ANRIL could promote the proliferation by si-ANRIL transfection when compared with control cells phenotype of NSCLC cells. (si-NC; Fig. 2B). Decreased ANRIL expression inhibits NSCLC cells migration Knockdown of ANRIL impaired NSCLC cells proliferation and To investigate the effect of ANRIL knockdown on NSCLC cells induced apoptosis migration, Transwells assays were performed. The results showed To assess the role of ANRIL in NSCLC, we investigated the effect that decreased ANRIL expression levels impeded the migration of of targeted knockdown of ANRIL on cell proliferation. MTT assays SPC-A1 and PC9 cells compared with controls (Supplementary revealed that cell growth was inhibited in SPC-A1, H1299, and Fig. S1D). PC9 cells transiently transfected with si-ANRIL compared with controls (Fig. 2C). Meanwhile, knockdown of ANRIL expression Downregulation of ANRIL inhibits NSCLC cells tumorigenesis could also inhibit A549 cells (with relative low endogenous in vivo ANRIL expression level) proliferation (Supplementary Fig. To explore whether the level of ANRIL expression could affect S1B). Colony formation assay results revealed that clonogenic tumorigenesis, PC9 cells stably transfected with sh-ANRIL or survival was inhibited following downregulation of ANRIL in empty vector were inoculated into female nude mice. Eighteen SPC-A1, H1299, and PC9cells (Fig. 2D). However, overexpression days after the injection, the tumors formed in the sh-ANRIL of ANRIL in 16HBE cells showed no significant effect on cell group were substantially smaller than those in the control group proliferation (Supplementary Fig. S1C). (Fig. 4A). Moreover, the mean tumor weight at the end of the To further examine whether the effect of knockdown ANRIL on experiment was markedly lower in the sh-ANRIL group (0.62 proliferation of NSCLC cells reflected cell-cycle arrest, cell-cycle 0.35 g) compared with the empty vector group (1.41 0.57 g; Fig. progression was analyzed by flow-cytometry analysis. The results 4B). qPCR analysis found that the levels of ANRIL expression in revealed that SPC-A1 and PC9 cells transfected with si-ANRIL had tumor tissues formed from sh-ANRIL cells were lower than in an obvious cell-cycle arrest at the G1–G0 phase and had a tumors formed in the control group (Fig. 4C). Tumors formed decreased G2–S phase (Fig. 3A). To determine whether NSCLC from sh-ANRIL–transfected PC9 cells exhibited decreased positive cell proliferation was influenced by cell apoptosis, we performed for Ki67 than those from control cells (Fig. 4D). These findings flow-cytometry and Tunel staining analysis. The results showed indicate that knockdown of ANRIL inhibits tumor growth in vivo.

Figure 4. Effects of downregulation of ANRIL on tumor growth in vivo. A, the tumor volume was calculated once every 3 days after injection of PC9 cells stably transfected with sh-ANRIL or empty vector. Points, mean (n ¼ 7); bars, SD. B, tumor weights are represented as means of tumor weights SD. C, qPCR analysis of ANRIL expression in tumor tissues formed from PC9/sh-ANRIL, PC9/empty vector. D, tumors developed from sh-ANRIL–transfected PC9 cells showed lower Ki67 protein levels than tumors developed by control cells. Top, H&E staining; bottom, immunostaining; , P < 0.05 and , P < 0.01.

www.aacrjournals.org Mol Cancer Ther; 14(1) January 2015 273

Nie et al.

ANRIL silences KLF2 and P21 transcription by binding with Silence of KLF2 is potentially involved in the oncogenic EZH2 function of ANRIL Previously studies have indicated that ANRIL could silence To investigate whether KLF2 is involved in the ANRIL-induced p15INK4 transcription and contribute to cancer cells proliferation increase in NSCLC cell proliferation, we performed gain-of-func- (18). The results of qPCR showed that p15 and p16 expression tion assays. The results of Western blot analysis showed that KLF2 was increased in SPCA1 and H1299 cells with transfection of si- expression was significantly upregulated in PC9 cells transfected ANRIL; however, there was no significant difference of p15 with pCDNA–KLF2 compared with control cells (Fig. 7A). Mean- expression in PC9 cells when knockdown of ANRIL expression while, MTT and colony formation assay results revealed that (Fig. 5A and Supplementary Fig. S1E). There are evidence that overexpression of KLF2 could impaired NSCLC cells proliferation showed that EZH2 could regulate KLF2 and P21 expression (20, (Fig. 7B). Moreover, flow-cytometry analysis indicated that 21), and our qPCR results also showed that inhibition of ANRIL increased KLF2 expression could induce NSCLC cells apoptosis expression led to increased KLF2 and P21 expression. Moreover, (Fig. 7C). Furthermore, to determine whether ANRIL regulate knockdown of EZH2 or SUZ12 could also upregulate KLF2 and NSCLC cell proliferation via repressing KLF2 expression, rescue P21 expression (Fig. 5B). Meanwhile, the Western blot assays assays were performed. PC9 cells were cotransfected with si- showed the same results (Fig. 5C), which indicated that KLF2 ANRIL and si-KLF2, and this was shown to rescue the decreased and P21 could be ANRIL novel targets in PC9 cells. In addition, expression of ANRIL induced by knockdown of KLF2 (Fig. 7D). we found that ANRIL RNAs were mostly located in the nucleus The results of MTT and colony formation assay results indicated (Fig. 5D). that cotransfection could partially rescue si-ANRIL–impaired To further investigate whether ANRIL repress KLF2 and P21 proliferation in PC9 cells (Fig. 7E). These data indicate that expression through binding PRC2, we performed RIP analysis and ANRIL promotes NSCLC cell proliferation through the down- the results showed that ANRIL could directly bind with EZH2 in regulation of KLF2 expression. PC9 cells (Fig. 6A). Furthermore, the results of ChIP assays showed that EZH2 could directly bind to KLF2 and P21 promoter Discussion region and mediate H3K27me3 modification (Fig. 6B). However, knockdown of ANRIL reduced EZH2 binding with KLF2 and P21 Recently, numerous pieces of evidence show that many promoter (Fig. 6C). Finally, we detected the KLF2 expression in lncRNAs are characterized and play important roles in cancer NSCLC tissues, and found that there is an inverse relationship pathogenesis, suggesting that they could provide new insights between ANRIL and KLF2 expression (Fig. 6D). These data sug- into the biology of this disease. For example, increased lncRNA gested that ANRIL promotes NSCLC PC9 cells proliferation is not HOTTIP is associated with progression and predicts outcome dependent on regulation p15 expression, but also through silenc- in patients with hepatocellular carcinoma by regulating ing of KLF2 and P21 transcription. HOXA13 expression (22). However, the roles of lncRNAs in

Figure 5. ANRIL could silence KLF2 and P21 expression. A, the levels of p15INK4B and p16 mRNA were determined by qPCR in SPC-A1 and PC9 cells transfected with si-ANRIL and results are expressed relative to the corresponding values for control cells. B and C, the levels of p21 and KLF2 mRNA and protein levels were determinedby qPCR and Western blot when PC9 cells were transfected with si-EZH2 or si-SUZ12. D, ANRIL expression levels in cell cytoplasm or nucleus of NSCLC cell lines SPC-A1, PC9, and H520 were detected by qPCR. , P < 0.05 and , P < 0.01; N.S., not signiﬁcant.

274 Mol Cancer Ther; 14(1) January 2015 Molecular Cancer Therapeutics

ANRIL Promotes Proliferation in NSCLC

Figure 6. ANRIL could directly bind PRC2 and silence KLF2 and P21 transcription. A, RIP with rabbit monoclonal anti-EZH2, preimmune IgG, or 10% input from PC9 cell extracts. RNA levels in immunoprecipitates were determined by qPCR. Expression levels of ANRIL RNA are presented as fold enrichment in EZH2 relative to IgG immunoprecipitates; relative RNA levels of U1 snRNA in SNRNP70 relative to IgG immunoprecipitates were used as positive control. B and C, ChIP–qPCR of EZH2 occupancy and H3K27-3me binding in the KLF2 promoter in PC9 cells, and IgG as a negative control; ChIP–qPCR of EZH2 occupancy and H3K27-3me binding in the KLF2 promoter in PC9 cells treated with ANRIL siRNA (48 hours) or scrambled siRNA. D, analysis of the relationship between ANRIL expression and KLF2 mRNA level (DCt value) in 40 NSCLC tissues. The mean values and SDs were calculated from triplicates of a representative experiment.

NSCLC are still not well documented, and one of these sion in NSCLC PC9 cells, which indicated that ANRIL contributed lncRNAs is metastasis-associated lung adenocarcinoma tran- to NSCLC cell proliferation is not dependent on regulating script 1 (MALAT1), which is a highly conserved nuclear lncRNA p15INK4, but also could through silencing KLF2 and P21 and a predictive marker for metastasis development in lung transcription. cancer (23). In our previous studies, we found that increased The Kruppel-like factor (KLF) family transcription factors, lncRNA HOTAIR–promoted NSCLC cells invasion and metas- with Cys2/His2 zinc-finger domains, could function as sup- tasis via regulating HOXA5 expression, and lncRNA BANCR pressors or activators in a cell type and promoter-dependent overexpression could impaired NSCLC cells proliferation manner and involve in cell differentiation and proliferation and metastasis by affecting epithelial–mesenchymal transition (28, 29). Some KLF members are emerging as tumor-suppressor (24, 25). genes due to their roles in the inhibition of proliferation, In this study, we demonstrated that the expression of another migration, and induction of apoptosis (30, 31). KLF2, as a lncRNA, ANRIL, is significantly upregulated in NSCLC tissues. member of KLF family, is diminished in many cancers and Specifically, increased ANRIL expression appears to be a signifi- possesses tumor-suppressor features such as inhibition of cell cant, independent predictive value for patients with NSCLC. proliferation mediated by KRAS (32–34). Moreover, there is Moreover, knockdown of ANRIL expression led to the significant evidence that showed that EZH2 could silence KLF2 expression inhibition of cell proliferation and the promotion of apoptosis and block the tumor-suppressor features of KLF2, which is both in vitro and in vivo. These findings suggest that ANRIL plays a partly mediated by p21 (21). Our results also showed that direct role in the modulation of cell proliferation and NSCLC lncRNA ANRIL takes part in NSCLC cells' proliferation by progression, and could be a useful novel prognostic or progres- epigenetic-silencing KLF2 and P21 transcription, and KLF2 sion marker for NSCLC. As more and more lncRNAs are studied, inactivation further led to the decreased P21 expression. As many have been shown to function by binding with PRC2 and more and more studies indicated that lncRNAs are often silencing downstream target genes that involved in multiple expressed in a spatial- or temporal-specific pattern, and more cancers, including NSCLC (26, 27). ANRIL has been reported to cell- and tissue-specific pattern. Our results also showed that involve in cancer cells proliferation by silencing p15INK4 expres- even in NSCLC cells, lncRNA ANRIL could regulate different sion. In this study, we found that ANRIL is mostly located in cell target genes in different cell lines, which suggested that lncRNA, nucleus and could directly bind with EZH2, a core subunit of especially ANRIL, can influence the same cell biologic function PRC2, resulted in repressing KLF2 and P21 transcription. How- via regulating different target genes dependent on different cell ever, knockdown of ANRIL could not influence p15INK4 expres- lines.

www.aacrjournals.org Mol Cancer Ther; 14(1) January 2015 275

Nie et al.

Figure 7. Overexpression of KLF2 expression inhibits PC9 cells proliferation. PC9 cells were transfected with pCDNA–KLF2 or cotransfected with si-ANRIL and si-KLF2. A, the protein level of KLF2 in PC9 cells transfected with pCDNA–KLF2 was detected by Western blot analysis. B, MTT assays and colony-forming assays were used to determine the cell viability for pCDNA–KLF2-transfected PC9 cells. Values represent the mean SD from three independent experiments. C, apoptosis was determined by ﬂow cytometry. Upper left, necrotic cells; upper right, terminal apoptotic cells; lower right, early apoptotic cells. D, the protein level of KLF2 in PC9 cells cotransfected with si-ANRIL and si-KLF2 was detected by Western blot analysis. E, MTT assays and colony-forming assays were used to determine the cell viability for si-ANRIL and si-KLF2 cotransfected PC9 cells. Values represent the mean SD from three independent experiments; , P < 0.05 and , P < 0.01.

To date, although only a small number of lncRNAs have been Analysis and interpretation of data (e.g., statistical analysis, biostatis- well characterized, they have been shown to regulate gene expres- tics, computational analysis): F.-q. Nie, J.-s. Yang, M. Xie, R. Xia, Y.-w. Liu, sion at various levels, including chromatin modification and post- X.-h. Liu Writing, review, and/or revision of the manuscript: F.-q.Nie,K.-h.Lu, transcriptional processing (35, 36). Here, the possible other targets Y.-q. Shu and mechanisms that underlie such regulatory behaviors still Administrative, technical, or material support (i.e., reporting or organizing remain to be fully understood despite our observation of data, constructing databases): E.-b. Zhang ANRIL-induced NSCLC cell proliferation. In summary, the expres- Study supervision: M. Sun, K.-h. Lu, Y.-q. Shu sion of ANRIL was significantly increased in NSCLC tissues, suggesting that its upregulation may be a negative prognostic factor for Acknowledgments The authors are very grateful to professor Xiongbin Lu for providing the patients with NSCLC, indicative of poor survival rates, and a higher ANRIL overexpression plasmid. risk for cancer metastasis. We showed that ANRIL possibly regulates the proliferation ability of NSCLC cells, partially through its reg- Grant Support ulation of the KLF2 and P21, which indicated that lncRNAs con- This work was supported by grants from the National Natural Scientific tribute to different cancer cells biologic function maybe through Foundation of China (81372397; to K.-h. Lu); (81301824; to X.-h. Liu); regulating different target genes. Our findings further the under- (81172140 and 81272532; to Y.-q. Shu). This work was also supported by Priority standing of NSCLC pathogenesis, and facilitate the development of Academic Program Development of Jiangsu Higher Education Institutions lncRNA-directed diagnostics and therapeutics against cancers. (JX10231801). M. Sun was supported by a Jiangsu province ordinary university graduate student research innovation project for 2013 (CXZZ13_0562, Disclosure of Potential Conflicts of Interest JX22013265). No potential conflicts of interest were disclosed. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in Authors' Contributions accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Conception and design: F.-q. Nie, M. Sun, T.-p. Xu, K.-h. Lu, Y.-q. Shu Acquisition of data (provided animals, acquired and managed patients, Received June 13, 2014; revised October 21, 2014; accepted November 7, provided facilities, etc.): J.-s. Yang, R. Xia, X.-h. Liu, E.-b. Zhang 2014; published OnlineFirst December 12, 2014.

276 Mol Cancer Ther; 14(1) January 2015 Molecular Cancer Therapeutics

ANRIL Promotes Proliferation in NSCLC

References 1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin facilitating re-expression of p21(waf1/cip1) and by inhibiting mutant p53. 2013;63:11–30. Cancer Lett 2013;336:53–60. 2. Spira A, Ettinger DS. Multidisciplinary management of lung cancer. N Engl J 21. Taniguchi H, Jacinto FV, Villanueva A, Fernandez AF, Yamamoto H, Med 2004;350:379–92. Carmona FJ, et al. Silencing of Kruppel-like factor 2 by the histone 3. Wistuba II. Genetics of preneoplasia: lessons from lung cancer. Curr Mol methyltransferase EZH2 in human cancer. Oncogene 2012;31:1988–94. Med 2007;7:3–14. 22. Quagliata L, Matter MS, Piscuoglio S, Arabi L, Ruiz C, Procino A, et al. Long 4. Verdecchia A, Francisci S, Brenner H, Gatta G, Micheli A, Mangone L, et al. noncoding RNA HOTTIP/HOXA13 expression is associated with disease Recent cancer survival in Europe: a 2000–02 period analysis of EURO- progression and predicts outcome in hepatocellular carcinoma patients. CARE-4 data. Lancet Oncol 2007;8:784–96. Hepatology 2014;59:911–23. 5. International Human Genome Sequencing Consortium. Finishing the 23. Gutschner T, Hammerle M, Eissmann M, Hsu J, Kim Y, Hung G, et al. The euchromatic sequence of the human genome. Nature 2004;431:931–45. noncoding RNA MALAT1 is a critical regulator of the metastasis phenotype 6. Djebali S, Davis CA, Merkel A, Dobin A, Lassmann T, Mortazavi A, et al. of lung cancer cells. Cancer Res 2013;73:1180–9. Landscape of transcription in human cells. Nature 2012;489:101–8. 24. Liu XH, Liu ZL, Sun M, Liu J, Wang ZX, De W. The long non-coding RNA 7. Derrien T, Johnson R, Bussotti G, Tanzer A, Djebali S, Tilgner H, et al. The HOTAIR indicates a poor prognosis and promotes metastasis in non–small GENCODE v7 catalog of human long noncoding RNAs: analysis of their cell lung cancer. BMC Cancer 2013;13:464. gene structure, evolution, and expression. Genome Res 2012;22:1775–89. 25. Sun M, Liu XH, Wang KM, Nie FQ, Kong R, Yang JS, et al. Downregulation 8. Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. of BRAF activated non-coding RNA is associated with poor prognosis for Cell 2004;116:281–97. non–small cell lung cancer and promotes metastasis by affecting epithe- 9. Fatica A, Bozzoni I. Long non-coding RNAs: new players in cell differen- lial–mesenchymal transition. Mol Cancer 2014;13:68. tiation and development. Nat Rev Genet 2014;15:7–21. 26. Beckedorff FC, Ayupe AC, Crocci-Souza R, Amaral MS, Nakaya HI, Soltys 10. Geisler S, Coller J. RNA in unexpected places: long non-coding RNA DT, et al. The intronic long noncoding RNA ANRASSF1 recruits PRC2 to the functions in diverse cellular contexts. Nat Rev Mol Cell Biol 2013; RASSF1A promoter, reducing the expression of RASSF1A and increasing 14:699–712. cell proliferation. PLoS Genet 2013;9:e1003705. 11. Loewer S, Cabili MN, Guttman M, Loh YH, Thomas K, Park IH, et al. Large 27. Brockdorff N. Noncoding RNA and Polycomb recruitment. RNA 2013; intergenic non-coding RNA-RoR modulates reprogramming of human 19:429–42. induced pluripotent stem cells. Nat Genet 2010;42:1113–7. 28. Black AR, Black JD, Azizkhan-Clifford J. Sp1 and kruppel-like factor family 12. Mercer TR, Dinger ME, Mattick JS. Long non-coding RNAs: insights into of transcription factors in cell growth regulation and cancer. J Cell Physiol functions. Nat Rev Genet 2009;10:155–9. 2001;188:143–60. 13. Wapinski O, Chang HY. Long noncoding RNAs and human disease. Trends 29. Kaczynski J, Cook T, Urrutia R. Sp1- and Kruppel-like transcription factors. Cell Biol 2011;21:354–61. Genome Biol 2003;4:206. 14. Tsai MC, Spitale RC, Chang HY. Long intergenic noncoding RNAs: new 30. Shen P, Sun J, Xu G, Zhang L, Yang Z, Xia S, et al. KLF9, a transcription links in cancer progression. Cancer Res 2011;71:3–7. factor induced in ﬂutamide-caused cell apoptosis, inhibits AKT activation 15. Spizzo R, Almeida MI, Colombatti A, Calin GA. Long non-coding RNAs and suppresses tumor growth of prostate cancer cells. Prostate 2014; and cancer: a new frontier of translational research? Oncogene 2012;31: 74:946–58. 4577–87. 31. Hsu LS, Chan CP, Chen CJ, Lin SH, Lai MT, Hsu JD, et al. Decreased 16. Zhang EB, Kong R, Yin DD, You LH, Sun M, Han L, et al. Long noncoding Kruppel-like factor 4 (KLF4) expression may correlate with poor survival in RNA ANRIL indicates a poor prognosis of gastric cancer and promotes gastric adenocarcinoma. Med Oncol 2013;30:632. tumor growth by epigenetically silencing of miR-99a/miR-449a. Oncotar- 32. Wu J, Lingrel JB. KLF2 inhibits Jurkat T leukemia cell growth via upregula- get 2014;5:2276–92. tion of cyclin-dependent kinase inhibitor p21WAF1/CIP1. Oncogene 17. Kotake Y, Nakagawa T, Kitagawa K, Suzuki S, Liu N, Kitagawa M, et al. Long 2004;23:8088–96. non-coding RNA ANRIL is required for the PRC2 recruitment to and 33. Wang F, Zhu Y, Huang Y, McAvoy S, Johnson WB, Cheung TH, et al. silencing of p15(INK4B) tumor suppressor gene. Oncogene 2011;30: Transcriptional repression of WEE1 by Kruppel-like factor 2 is involved in 1956–62. DNA damage-induced apoptosis. Oncogene 2005;24:3875–85. 18. Chen D, Zhang Z, Mao C, Zhou Y, Yu L, Yin Y, et al. ANRIL inhibits p15 34. Fernandez-Zapico ME, Lomberk GA, Tsuji S, DeMars CJ, Bardsley MR, Lin through the TGFbeta1 signaling pathway in human esophageal squamous YH, et al. A functional family-wide screening of SP/KLF proteins identiﬁes a cell carcinoma. Cell Immunol 2014;289:91–6. subset of suppressors of KRAS-mediated cell growth. Biochem J 2011; 19. Yoon JH, Abdelmohsen K, Srikantan S, Yang X, Martindale JL, De S, et al. 435:529–37. LincRNA-p21 suppresses target mRNA translation. Mol Cell 2012;47: 35. Ponting CP, Oliver PL, Reik W. Evolution and functions of long noncoding 648–55. RNAs. Cell 2009;136:629–41. 20. Seward S, Semaan A, Qazi AM, Gruzdyn OV, Chamala S, Bryant CC, et al. 36. Nagano T, Fraser P. No-nonsense functions for long noncoding RNAs. Cell EZH2 blockade by RNA interference inhibits growth of ovarian cancer by 2011;145:178–81.

www.aacrjournals.org Mol Cancer Ther; 14(1) January 2015 277

Long Noncoding RNA ANRIL Promotes Non−Small Cell Lung Cancer Cell Proliferation and Inhibits Apoptosis by Silencing KLF2 and P21 Expression

Feng-qi Nie, Ming Sun, Jin-song Yang, et al.

Mol Cancer Ther 2015;14:268-277. Published OnlineFirst December 12, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-14-0492

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2014/11/25/1535-7163.MCT-14-0492.DC1

Cited articles This article cites 36 articles, 5 of which you can access for free at: http://mct.aacrjournals.org/content/14/1/268.full#ref-list-1

Citing articles This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/14/1/268.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/14/1/268. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.