The silencing of long non-coding RNA ANRIL suppresses invasion, and promotes apoptosis of retinoblastoma cells through the ATM-E2F1 signaling pathway

As one of the most common primary intraocular carcinomas, retinoblastoma generally stems from the inactivation of the retinoblastoma RB1 gene in retinal cells. Antisense non-coding RNA in the INK4 locus (ANRIL), a long non-coding RNA (lncRNA), has been reported to affect tumorigenesis and progression of various cancers, including gastric cancer and non-small cell lung cancer. However, limited investigations emphasized the role of ANRIL in human retinoblastoma. Hence, the current study was intended to investigate the effects of ANRIL on the proliferation, apoptosis, and invasion of retinoblastoma HXO-RB44 and Y79 cells. The lentivirus-based packaging system was designed to aid the up-regulation of ANRIL and ATM expressions or employed for the down-regulation of ANRIL in human retinoblastoma cells. Afterward, ANRIL expression, mRNA and protein expression of ATM and E2F1, and protein expression of INK4b, INK4a, alternate reading frame (ARF), p53 and retinoblastoma protein (pRB) were determined in order to elucidate the regulation effect associated with ANRIL on the ATM-E2F1 signaling pathway. In addition, cell viability, apoptosis, and invasion were detected accordingly. The results indicated that the down-regulation of ANRIL or up-regulation of ATM led to an increase in the expressions of ATM, E2F1, INK4b, INK4a, ARF, p53, and pRB. The silencing of ANRIL or up-regulation of ATM exerted an inhibitory effect on the proliferation and invasion while improving the apoptosis of HXO-RB44 and Y79 cells. In conclusion, the key observations of our study demonstrated that ANRIL depletion could act to suppress retinoblastoma progression by activating the ATM-E2F1 signaling pathway. These results provide a potentially promising basis for the targetted intervention treatment of human retinoblastoma.


Introduction
Retinoblastoma is an aggressive form of an intraocular cancer usually occurring during childhood that is initiated by the mutation of the RB1 gene. Timely diagnoses along with early treatment may boast excellent outcome, however, retinoblastoma may also be a life-threatening condition if left without a swift and adequate treatment [1,2]. Although the etiology of retinoblastoma is relatively well-understood, the mortality rate of the condition sits at an alarming 70% in lower and middle-income countries (MICs); while the incidence rate of retinoblastoma has been found to be higher amongst Asian and African regions, and children were reported to have a greater susceptibility to it with a mortality rate of approximately 40-70% [3]. An investigation into retinoblastoma survival in less-developed countries, suggested there to be an estimated survival rate of 40% in lower income countries with survival rates approximately 77% and 79% in lower MICs and upper MICs, respectively [4]. The treatment for retinoblastoma generally

Cell culture
Human retinoblastoma HXO-RB 44 and Y79 cells were purchased from the Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences (Shanghai, China). Cells preserved in liquid nitrogen in a sealed tube were taken out and rapidly thawed in a water bath at 37 • C. Following sterilization using 75% ethanol, the cell sap in the sealed tube was transferred to a centrifuge tube containing Eagle's minimum essential medium (EMEM) using a clean bench (Thermo Fisher, CA, U.S.A.), after which it was centrifuged at 290×g at 4 • C for 3 min. The supernatant was then carefully removed with the cells then re-suspended and transferred to a culture flask (Corning Inc., NY, U.S.A.) containing EMEM supplemented with 10% FBS and 1% double antibody (Gibco, NY, U.S.A.). The cells were subsequently incubated in an incubator containing 5% CO 2 at 37 • C. When the cells had reached approximately 70-80% confluence, they were treated with trypsin and passed into six-well plates with the density adjusted to 1 × 10 5 cells/well.

Construction of vectors
Based on the full length of 1258 bp cDNA sequence of ANRIL obtained from the NCBI (GI: 641451086), the specific primers of target genes were designed and restriction enzyme Xba I and BamH I recognition sites were introduced. Reverse transcription quantitative PCR (RT-qPCR) methods were performed for amplification purposes with the recovered purified PCR products subsequently inserted into the pMD18-T vector (TaKaRa Biotechnology Co. Ltd., Liaoning, China) and transformed into Escherichia coli (E. coli) DH5α competent cells. After the positive clones were obtained by means of blue-white screening, the plasmids were extracted and confirmed by means of restricted digestion. Lentiviral vector pCD513B (System Biosciences Inc., CA, U.S.A.) carrying GFP as well as the puromycin resistance gene were used for ANRIL overexpression. The pCD513B vector was then linearized after the digestion of the restriction enzymes Xba I and BamH I, and the target gene fragments and the linearized vector ligated with T4 DNA ligase, after which the ligation mixture was transformed into E. coli DH5α competent cells. After the positive cloned cells were selected by virtue of the ampicillin resistance, plasmid extraction and identification was performed by means of Xba I and BamH I enzymes digestion. The method for the vector construction of overexpressed ANRIL was also applied to the vector construction of ATM. ANRIL-shRNA was transfected into cells in order to silence ANRIL using a Lipofectamine TM 2000, with ANRIL NC regarded as the control.

Cell transfection and grouping
Cells at the exponential phase were seeded in six-well plates (approximately 1-3 × 10 5 cells/well) and cultured with CO 2 overnight until the cells were confirmed to have reached 50-80% confluence. Next, the cells were assigned into six groups, namely control group (without any transfection), negative control (NC) group (transfected with empty vectors), overexpressed ANRIL group, sh-ANRIL group (transfected with ANRIL-shRNA), overexpressed ATM group, and overexpressed ATM + overexpressed ANRIL group. The following solutions were prepared in a simultaneous manner: (A) 100 μl serum-free medium added with 1-2 μg high quality vector DNA; (B) 100 μl serum-free medium added with 1-2 μl Lipofectamine 2000 (Invitrogen Inc., CA, U.S.A.). Solutions A and B were mixed and placed at room temperature for 15-45 min in order to facilitate the formation of DNA-liposome complex. A total of 2 ml serum-free medium was used to wash the cells and 0.8 ml serum-free medium was added into the DNA-liposome complex. The mixture was then gently shaken in a thorough manner and then slowly added dropwise into the six-well plates, followed by incubation in an incubator containing 5% CO 2 for over 6 h. Each six-well plate was then added with 1 ml medium with double concentration of serum and the culture containing serum then refreshed following a 24-h period of transfection. The total RNA and protein were extracted over a 72-h period post transfection.

RT-qPCR
The total RNA was extracted using TRIzol (Takara, Kyoto, Japan) from the cells at the 72 h point after transfection. The optical density (OD) value and OD260/280 ratio were measured using a micro-spectrophotometer (NanoDrop Technologies, DE, U.S.A.) after which cDNA was synthesized using a reverse transcription kit (Takara, Kyoto, Japan). The primers for ANRIL, ATM, and E2F1 were synthesized by Shanghai Sangon Biological Engineering Technology and Services Co., Ltd., (Shanghai, China), as illustrated in Table 1. The SYBR Green Real-time Fluorescence Quantitative PCR Kit was applied in order to determine the total volume of reaction system which was confirmed to be 20 μl. Glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) was considered to be the internal reference. The reaction system and conditions are illustrated in Tables 2 and 3. The quantitative results were calculated based on the method 2 − C t which demonstrated the relative gene expression ratio of the experimental group to the control group. The formula of 2 − C t method was as follows:

Western blot analysis
After transfection for 72 h, the protein levels of ATM, E2F1, INK4b, INK4a, alternate reading frame (ARF), p53 and retinoblastoma protein (pRB) were detected by Western blot analysis, with GAPDH employed as the internal reference. The total proteins were extracted with the protein concentration then determined accordingly. Spacer gel and separating gel were prepared and loading quantity of sample was set. The spacer gel ran at 60 V, which was then turned up to 100 V, followed by a wet-transfer process at 100 V for 70 min with the proteins transferred to the membrane. Membrane blockade was then conducted using a mixture containing 10% skim milk powder and TBS for 2 h after which the membrane was incubated overnight at 4 • C with the addition of the following diluted primary antibodies: Anti-ATM (p-S1981),

Cell counting kit-8 assay
The transfected cells were seeded in a 96-well plate at a density of 4 × 10 3 cells/100 μl, with each well supplemented with 20 μl Cell counting kit-8 (CCK-8) solutions. After incubation for 4 h, the supernatant was removed, followed by gentle shaking with the addition of 150 μl DMSO in each well for 10 min. An OD value at 499 nm was subsequently determined using an enzyme-labeling measuring instrument (Bio-Tek, VT, U.S.A.), and then the cell proliferation curves were constructed with the x-axis representing time and y-axis representing OD value.

Flow cytometry
Annexin V/propidium iodide (PI) double staining was performed in order to determine cell apoptosis, while HXO-RB 44 cells at the logarithmic phase of growth were adopted. The adopted HXO-RB 44 cells were then suspended following detachment with 0.25% trypsin and were seeded in culture plate with a density of 1 × 10 5 cells/ml, after which they were washed three times with pre-cooled PBS at 4 • C and treated with trypsin. Following 5 min of centrifugation at 290×g, the supernatant was removed and the cells were re-suspended with PBS. The concentration of the transfected cells was adjusted to 1 × 10 6 cells/ml. A total of 100 μl cell suspension was centrifuged at 290×g for 5 min, after which the supernatant was discarded and 500 μl 1× binding buffer, 5 μl FITC-labeled annexin V (annexinV-FITC), and 10 μl PI were added in a successive manner, followed by gentle shaking. The mixture was incubated at room temperature for 5

Transwell assay
A total of 20-30 μl of Matrigel was added into each well of the filter membrane of the chamber at 37 • C overnight with fibronectin added in an even manner to the other side of the membrane. Next, 1 × 10 5 cells were added in the Transwell chamber. After a 24-h period of incubation, the cells on the upper layer of membrane were removed and the membrane was obtained and subsequently fixed with formaldehyde and placed at room temperature for 30 min. Next, the cells were stained with Hematoxylin for 3-5 min, dehydrated with ethanol (5 min each time), and cleared with formaldehyde. A portion of the lower layer of membrane was subsequently selected and placed on a slide for further observation, with the average of four representative views selected followed by cell number counting under a microscope.

Statistical analysis
Statistical analysis was performed using SPSS 18.0 software (IBM Corp. Armonk, NY, U.S.A.). Categorical data were measured by chi-square test and the measurement data were expressed as mean + − S.D. One-way ANOVA was employed for the comparisons of multiple groups. A P-value <0.05 was considered to be indicative of statistical significance.

Successful construction and transfection of T-ANRIL and pCD513B-ANRIL vectors
Initially, T-ANRIL, an ANRIL cDNA clone, was constructed. After digestion by restriction enzymes Xba I and BamH I, the bands of 1700 and 1400 bp were observed accordingly. The length of pMD18-T was determined to be approximately 2700 bp with the full length of the ANRIL cDNA approximately 1400 bp. T-ANRIL was successfully constructed, as illustrated Figure 1A. The two bands of approximately 8000 and 1400 bp of pCD513B-ANRIL, a vector for ANRIL overexpression, were observed following the digestion of restriction enzymes Xba I and BamH I. The length of linearized vector pCD513B was approximately 8100 bp, while the full length of ANRIL cDNA was approximately 1400 bp. As depicted in Figure 1B, pCD513B-ANRIL was successfully constructed. As GFP was carried by the pCD513B-ANRIL, green fluorescence in HXO-RB 44 could be observed under a fluorescence microscope in the event that the cells were transfected. The transfection of HXO-RB 44 is illustrated in Figure 1C. No fluorescence appeared in the cells without transfection under a fluorescent microscope ( Figure 1D). Thus, T-ANRIL and pCD513B-ANRIL vectors were deemed to be successfully constructed.

ANRIL negatively regulates the ATM-E2F1 signaling pathway
RT-qPCR and Western blot analysis were employed for observation of the regulatory roles of ANRIL in the ATM-E2F1 signaling pathway. ANRIL expression and mRNA and protein levels of ATM and E2F1 were detected amongst the five groups with GAPDH considered as the internal reference, the results of which are depicted in Figure 2.  levels were similar to those in the control group (P>0.05). These findings indicated that ANRIL could down-regulate the ATM-E2F1 signaling pathway.

ANRIL silencing inhibits proliferation of retinoblastoma cells
CCK-8 assay was conducted in order to determine the effect of ANRIL on the proliferation of HXO-RB 44 and Y79 cells, while subsequently comparing the viability of transfected cells amongst 12 groups. As shown in Figure 3, on the sixth day, the cell viability increased in the overexpressed ANRIL group, while it decreased in the sh-ANRIL group and the overexpressed ATM group (all P<0.05) when compared with the control and NC groups; no significant difference was detected in the overexpressed ATM + overexpressed ANRIL group (P>0.05). These results demonstrated that ANRIL silencing could act to suppress the cell proliferation of HXO-RB 44 and Y79 by activating the ATM-E2F1 signaling pathway.

ANRIL silencing improves apoptosis of retinoblastoma cells
Annexin V/PI double staining method and flow cytometry were performed in order to evaluate the effects of ANRIL on the regulation of HXO-RB 44 and Y79 cell apoptosis (Figure 4). In comparison with the apoptotic index in the control and NC groups, overexpressed ANRIL group exhibited a remarkable reduction (P<0.01), and the sh-ANRIL group as well as the overexpressed ATM group displayed a significant increase (all P<0.05). The apoptotic index of the overexpressed ATM + overexpressed ANRIL group was observed to be strikingly similar to that of the control and NC groups (both P>0.05). The aforementioned findings suggested that ANRIL silencing promoted the apoptosis of HXO-RB 44 and Y79 cells by activating the ATM-E2F1 signaling pathway.

Down-regulated ANRIL arrests retinoblastoma cell cycle progression
Flow cytometry was applied in order to examine cell cycle and to determine the DNA content of the single cell in samples, followed by the detection of the cell ratio in G 0 /G 1 , S, and G 2 /M phases. The results ( Figure 5) revealed that compared with the control and NC groups, the overexpressed ANRIL group exhibited a lower cell ratio at the G 0 /G 1 phase but a higher cell ratio at the S phase (all P<0.05). The sh-ANRIL group and the overexpressed ATM group displayed a higher cell ratio at the G 0 /G 1 phase but a lower cell ratio at the S phase in contrast with the control and NC groups (all P<0.05). The cell ratio in each phase of the overexpressed ATM + overexpressed ANRIL group was observed to be similar to that of the control and NC group (all P>0.05). Thus, based on our results we concluded that ANRIL depletion led to the inhibition of the HXO-RB 44 and Y79 cell cycle progression by activating the ATM-E2F1 signaling pathway.

The down-regulation of ANRIL inhibits the invasion of retinoblastoma cells
During the following experiments, the roles of ANRIL in the invasion of retinoblastoma HXO-RB44 and Y79 cells was detected using a Transwell assay. As depicted in Figure 6, compared with the control group and the NC group, invasion of HXO-RB44 and Y79 cells in the overexpressed ANRIL group was significantly higher (P<0.01) while it was notably weakened in the sh-ANRIL group (P<0.05) and the overexpressed ATM group (P<0.01). In the overexpressed ATM + overexpressed ANRIL group, there was no significant difference in regard to the cell invasion ability between the control and NC groups (P>0.05). The results indicated that invasion ability of HXO-RB44 and Y79 cells was inhibited by ANRIL silencing.

Protein levels of INK4b, INK4a, ARF, p53, and pRB are increased by ANRIL silencing
At last, Western blot analysis was performed in order to determine the levels of INK4b, INK4a, ARF, p53, and pRB amongst the six groups using GAPDH as an internal reference. As illustrated in Figure 7, the results obtained demonstrated that compared with the control and NC groups, the overexpression of ANRIL as well as the protein levels of INK4b, INK4a, ARF, p53, and pRB decreased in the overexpressed ANRIL group (all P<0.05), while that of INK4b,   INK4a, ARF, p53, and pRB increased in the sh-ANRIL group and the overexpressed ATM group (all P<0.05). However, regarding the protein levels in the overexpressed ATM + overexpressed ANRIL group, there was no significant difference detected in comparison with the control and NC groups (P>0.05). The aforementioned results provided evidence suggesting that down-regulation of ANRIL or activation of ATM-E2F1 signaling pathway resulted in increased levels of INK4b, INK4a, ARF, p53, and pRB.

Discussion
The vectors for the overexpression of ANRIL and ATM were successfully constructed followed by the negative feedback regulation of the ATM-E2F1 signaling pathway by ANRIL. Through the application of RT-qPCR and Western blot analysis methods, the effects of ANRIL on the regulation of the proliferation, apoptosis, and invasion of retinoblastoma HXO-RB 44 cells were examined through the ATM-E2F1 signaling pathway. The obtained results demonstrated that ANRIL could influence the biological processes of proliferation, apoptosis, and invasion of the HXO-RB 44 cells owing to its ability to negatively regulate the ATM-E2F1 signaling pathway. Regardless of the wide spanning knowledge of the etiology of retinoblastoma, the mortality rate remains high at approximately 40% [3]. Previous studies have implicated lncRNAs such as lncRNA BANCR and lncRNA HOTAIR in cases of retinoblastoma with poor prognoses due to their ability to facilitate cell proliferation and invasion [15,16]. Our findings demonstrated that the silencing of ANRIL resulted in the inhibition of retinoblastoma HXO-RB 44 and Y79 cell proliferation, cell invasion ability, and promotion of apoptosis. A previously conducted study regarding the gastric cancer suggested that the up-regulation of ANRIL had the capability to facilitate the proliferation of cells and suppress cell apoptosis [14]. Similarly, ANRIL was found to strengthen cancer cell invasion and inhibit cell apoptosis, while the up-regulation of ANRIL could promote the hypoxic osteosarcoma cell invasion and inhibit cell apoptosis [17]. A key observation of our study revealed that, ATM overexpression could stimulate an increase in E2F1 expression. In cases where DNA damage is observed, activated ATM and ATR phosphorylation activated the transcription factor E2F1 and thus maintained the stability of the genome [18]. Reports have indicated that the up-regulation of E2F1 induces the expression of ANRIL, leading to the formation of a positive feedback loop, resulting in the continuation of gastric cancer cell proliferation [10]. ANRIL has been shown to be activated through the mutation of kinase-E2F1 transcription factor signaling pathway by the ataxia telangiectasia [19]. The silencing of ANRIL could potentially inhibit cell proliferation, migration, and invasion in hepatocellular carcinoma [20]. E2F1, an effector of the retinoblastoma tumor suppressor pathway was shown to induce suppression of apoptosis with a activation of DNA repair [21]. Moreover, overexpressed ANRIL down-regulated ATM and E2F1, suggesting that ANRIL negatively regulates the ATM-E2F1 signaling pathway. Based on the aforementioned findings, we subsequently concluded that silencing ANRIL suppresses cell proliferation, cell invasion ability, while acting to promote the apoptosis of retinoblastoma cells with a possible involvement of the ATM-E2F1 signaling pathway.
We also found that the underexpressed ANRIL or activated ATM-E2F1 signaling pathway could result in increased expressions of INK4b, INK4a, p53, and pRB. Two major tumor suppressors, the transcription factor p53 and pRB take control of cellular reaction as potential carcinogenic stimuli like the repeated division of cells, damage of DNA as well as improper mitogenic signals and their reactivation significantly takes part in tumor therapy [22,23]. With regard to the expressions of protein-coding genes, ANRIL, the antisense non-coding RNA of the INK4b-ARF-INK4a locus plays a central role [9]. ANRIL was transcribed from the INK4b-ARF-INK4a locus encoding for INK4b, INK4a, and ARF (also known as p15, p16, and p14, respectively), three tumor suppressors whose function was usually lost or attenuated in human cancers [24]. ANRIL has also been reported to inhibit the expression of a widely known tumor suppressor gene, INK4b after its interaction with suppressor of zeste 12 homolog (SUZ12) [8]. In the cells with overexpressed ANRIL, the mRNA and protein levels of INK4b, INK4a, and ARF have been reported to exhibit very low levels, while the anti-proliferative activities of pRB and p53 are then triggered by INK4a, INK4b, and ARF, protein levels [7,9]. p53 is regulated by INK4a and in the absence of p19ARF expression in C2C12 cells results in the loss of the INK4a locus based on the results of a genomic PCR analysis in the study [25]. Our study illustrated that overexpressed ATM acts to reverse the inhibition of INK4b, INK4a, ARF, p53, and pRB induced by the overexpression of ANRIL. Studies have shown that E2F1 induces ANRIL in a transcriptional manner, which leads to a depression in the expression of p16INK4 family, providing an alleviating effect for the p53 and pRB signaling pathways [22]. In this regard, we speculated that ANRIL silencing may possibly prevent retinoblastoma progression via the ATM-E2F1 signaling pathway activation.
Taken together, the key findings of our study suggesting that the silencing of ANRIL can result in suppressed invasion and proliferation, while acting to enhance the apoptosis of retinoblastoma cells by up-regulating the ATM-E2F1 signaling pathway, thus providing a fresh basis for which the treatment of retinoblastoma may be premised upon. Since the localization of lncRNA is beneficial in providing further understanding of its function, the localization of ANRIL on ATM needs further analyzation in future studies. In addition, the underlying molecular mechanism of ANRIL and ATM should be further investigated based on the findings of the present study. Moreover, further experiments should be conducted in order to identify novel therapeutic drugs and provide a new and improved therapeutic strategy for tumor intervention.