Long Noncoding RNA MEG3 Interacts with p53 Protein and Regulates Partial p53 Target Genes in Hepatoma Cells

Maternally Expressed Gene 3 (MEG3) encodes a lncRNA which is suggested to function as a tumor suppressor. Previous studies suggested that MEG3 functioned through activation of p53, however, the functional properties of MEG3 remain obscure and their relevance to human diseases is under continuous investigation. Here, we try to illuminate the relationship of MEG3 and p53, and the consequence in hepatoma cells. We find that transfection of expression construct of MEG3 enhances stability and transcriptional activity of p53. Deletion analysis of MEG3 confirms that full length and intact structure of MEG3 are critical for it to activate p53-mediated transactivation. Interestingly, our results demonstrate for the first time that MEG3 can interact with p53 DNA binding domain and various p53 target genes are deregulated after overexpression of MEG3 in hepatoma cells. Furthermore, results of qRT-PCR have shown that MEG3 RNA is lost or reduced in the majority of HCC samples compared with adjacent non-tumorous samples. Ectopic expression of MEG3 in hepatoma cells significantly inhibits proliferation and induces apoptosis. In conclusion, our data demonstrates that MEG3 functions as a tumor suppressor in hepatoma cells through interacting with p53 protein to activate p53-mediated transcriptional activity and influence the expression of partial p53 target genes.

Introduction ncRNAs are functional RNA molecules that are not translated into proteins, including transfer RNAs (tRNAs), ribosomal RNAs (rRNAs), microRNAs (miRNAs), small nucleolar RNAs (snoRNAs), Piwi-RNAs and long non-coding RNAs (lncRNAs) [1]. A growing number of literatures point to the critical regulatory role of ncRNAs in normal cellular physiological processes as well as the contribution of aberrant ncRNA expression to cancer biology [2]. Among them, lncRNAs which are widely distributed in mammals, are a class of RNA molecules with

Materials and Methods
Patients and Tumor samples 23 pairs of HCC samples and adjacent non-tumor tissues were obtained from surgical specimens at General Hospital of the People's Liberation Army (Beijing, China) after informed consent. These samples used in this study have been described in previous publication [23]. Clear hepatocellular carcinoma was diagnosed histopathologically. All these specimens were snapfrozen in liquid nitrogen after excision. Clinical characteristics of the patients included in this study were in Table A in S1 File. This study was approved by the Ethics Committee of General Hospital of the People's Liberation Army. Informed consent was obtained from each participant and the methods were carried out in accordance with the approved guidelines.

Plasmid construction and Construction of hepatoma cell lines stably expressing MEG3
For the expression of MEG3, full-length Homo sapiens MEG3 sequences (NR_002766) were cloned into KpnⅠ and XhoⅠ site of the pcDNA3.0 vector. pG13L vector was a gift from Professor Lingqiang Zhang of Beijing Institute of Radiation Medicine.
To obtain cell line stably expressing MEG3, the expression vector pcDNA3.0-MEG3 was transfected into Hep-SK-1 cells and screened with G418 (1000 μg/ml). QRT-PCR was performed to examine the high expression of MEG3 at the RNA level. The cell line with stably expressing empty vector pcDNA3.0 served as control.

RNA extraction and Quantitative Reverse Transcription PCR
Total RNA from cells and 23 pairs of tissues were extracted using TRI regents (Sigma-Aldrich, St. Louis, MO) according to the manufacturer's protocol. Reverse transcription was performed using 1 μg reverse transcription primes Odigo(dT) and 2μg total RNA in ImProm-II TM Reverse Transcription System (Promega). Quantitative real-time PCR (qRT-PCR) was performed using SYBR Premix EX Taq TM (TaKaRa, Dalian, China). Primers of MEG3 and other genes as well as glyceraldehydes-3-phosphate dehydrogenase (GAPDH) as control were presented in Table B in S1 File.

Growth suppression assay
Growth suppression of hepatoma carcinoma cell line HepG2 by MEG3 was measured by cell proliferation assay and colony formation assay. The viability of cells in post-transfection was determined using Cell Counting Kit-8 (Dojindo, Japan). For colony formation assay, cells were harvested 24 hours post transfection and 2×10 3 cells were seeded in 6-well culture plates in DMEM containing 10% FBS and 1000 μg/ml neomycin, and incubated for 1 week for proper colony formation. Thereafter, culture media was discarded and the cells were washed two times with cold PBS. Then, the cells were fixed by treatment with 95% ethanol for 15 minutes and stained with 0.1% crystal violet solution for 15 minutes.

Flow-cytometric analysis of apoptosis
HepG2 cells transfected with pcDNA3.0 or pcDNA3.0-MEG3 after serum-starved for 24 hours were harvested by trypsinization. Following double staining with FITC-Annexin Ⅴ and Propidium isodide (PI), the cells were analyzed using flow cytometry (FACScan, BD, Biosciences). Cells were discriminated into four parts including viable cells, dead cells, early apoptotic cells, and apoptotic cells. This experiment was independently performed at least three times.

RNA pulldown assay
Biotin-labeled lncRNA-MEG3 were in vitro transcribed with a Biotin RNA Labeling Mix (Roche, Indianapolis, IN) and the T7 RNA polymerase (Promega), be treated with RNase-free DNase I (Promega) and purified with MEGAclear TM Kit (Ambion, Austin, USA). Five micrograms of biotinylated MEG3 RNA was heated to 70°C for 5 minutes, put on ice for 2 minutes, then supplied with RNA structure buffer (10 mM MgCl 2 , 10 mM Tris pH 7, 0.1 M KCl) and then shifted to room temperature (RT) for 20 minutes to allow proper secondary structure formation. 10 7 HEK293 cell pellets were re-suspended in 1ml lysis buffer (150 mMKCl, 25 mMTris pH 7.4, 0.5 mM DTT, 0.5% NP40, 1 mM PMSF, RNase inhibitor (TaKaRa) and protease inhibitor (Roche Complete Protease Inhibitor Cocktail Tablets). Debris was pelleted after centrifugation at 12,000 RPM for 15min. 10μl of the lyses was transferred to a microfuge tube as 1% input sample. Folded RNA was mixed with 1ml lyses of 10 7 cells extract in lysis buffer and incubated at RT for one hour. 20μl washed High Capacity Streptavidin Agarose Beads (Thermo, Rockford, IL) were added to the binding reaction and incubated for one hour at RT. Beads were washed two times in low salt wash buffer for 5 minutes with rotation, and two times in high salt wash buffer. Then, the beads were boiled in SDS buffer, and the associated protein was detected by western blot.
Recombinant GST-P53 was purified from BL21 bacterial using GST-agarose affinity chromatography. 2 μg of biotinylated RNA was incubated with 5μg of GST-P53 protein in 200μl binding buffer (50 mM TrisCl 7.9, 10% Glycerol, 100 mMKCl, 5 mM MgCl 2 , 10 mM β-ME, 0.1% NP- 40) for one hour at RT. 20 μl washed High Capacity Streptavidin Agarose beads (Thermo) were added to each binding reaction and incubated at RT for 30 minutes, and then beads were washed two times in low salt wash buffer for 5 minutes with rotation, and two times in high salt wash buffer. Retrieved protein was detected by standard western blot technique.

RNA immunoprecipitation
10 7 cells were washed twice in PBS with PMSF and pelleted. The pellet was lysed in 1 ml of RIPA buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 1% NP-40, and 0.5% sodium deoxycholate, 0.5 mM DTT and 1 mM PMSF/cocktail), incubated at 4°C for 30 minutes with rotation. The lysate was obtained after centrifugation at 13,000 RPM for 15 minutes at 4°C. Two parts of 10 μl of the supernatant of lysate were removed as 1% input. One part was used to test the expression of RNA-binding protein of interest by western blotting. The other part was utilized to retrieve RNA for comparison in RT-PCR. 5~10μg antibody and equal negative control IgG of the same species were added and binding samples were incubated for at 4°C for 4 hours with rotation. Samples were washed two times in low salt wash buffer (RIPA buffer), two times in high salt wash buffer (50 mMTris, pH 7.4, 1 M NaCl, 1 mM EDTA, 0.1% SDS, 1% NP-40, and 0.5% sodium deoxycholate). 100 μl each out of 1ml of the beads suspension during the last wash was removed to test the efficiency of immunoprecipitation by western blotting. Remained beads and 1% input sample were treated with proteinase K buffer containing 117 μl of RIP wash buffer, 15 μl of 10% SDS, 18 μl of 10 mg/mL proteinase K at 55°C for 30 minutes with shaking to digest the protein. RNA samples were extracted with phenol and chloroform. Co-precipitated RNAs were detected by qRT-PCR.

Western Blot
Western blot analysis was performed as described previously [24]. Briefly, cells were lysed in RIPA buffer supplemented with protease inhibitor cocktail (Roche) for 30 min on ice to extract total protein. Protein concentration was determined by BCA protein assay (Pierce, Rockford, IL). The protein samples were boiled in 1× sodium dodecyl sulfate buffer for 10 minutes. Protein with the same amount was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto poly-vinylidene fluoride membranes. The membranes were blocked with 5% skim milk, and probed with the antibody. It was visualized by enhanced chemiluminescence reagents super signal (Pierce). Mouse monoclonal anti-p53 antibody (sc-126; Santa Cruz, Santa Cruz Biotechnology, CA), mouse monoclonal anti-β-actin antibody (Proteintech, Proteintech group, USA), anti-GST antibody (MBL, Medical & Biological Laboratories Co., Japan), mouse monoclonal anti-FLAG antibody (MBL) were used for western blotting.

Microarray analysis and computational analysis
Sample preparation and microarray hybridization were performed by Emei Tongde Bio-tech, Beijing P.R. China. The Illumina TotalPrep RNA Amplification Kit was used to for generating biotinylated, amplified RNA for hybridization with Human HT-12 V4.0 Expression Beadchip. After having washed the slides, the arrays were scanned using Illumina BeadChip Reader and data analyzed using Illumina BeadStudio software. Average was carriey for sample normalization. Sample intensities are scaled by a factor equal to the ratio of average intensity of virtual sample to the average intensity of the given sample. Background is subtracted prior to the scaling. Differentially expressed genes were identified through fold change filtering (Fold Change 2.0 or 0.5) and paired t-test (p < 0.01). The microarray data discussed in this paper has been deposited in NCBI Gene Expression Omnibus and the GEO accession number is GSE69916. Compared with 287 p53 target genes whose data come from transcriptional regulatory element database (TRED), we found 34 of these p53 target genes showed consequent changes in gene expression in stable cell line overexpressing MEG3 (Tables 1 and 2).

Statistical analyses
All experiments were independently performed at least three times. The values are presented as mean ± SD. Differences were assessed by two-tailed Student's t-test. Analysis was used Microsoft Excel 2008 software and GraphPad prism (prism 5 for windows). P<0.05 was considered statistically significantly.

MEG3 enhances stability and transcriptional activity of p53
Previous studies suggested that MEG3 functioned as a tumor suppressor through activation of p53, leading to increase in p53 protein levels and stimulate p53-dependent transcription in colon and brain cancer cells [22]. To confirm the possibility that p53 may be involved in antiproliferation function of MEG3 in hepatoma cells, we co-transfected a p53-responsive reporter plasmid (pG13L) with MEG3 expression vector into HepG2 cells. Results of luciferase assay revealed that MEG3 could enhance the transcriptional activity of p53 ( Fig 1A). MEG3 contains three conserved motifs, designated as M1, M2 and M3 [16]. To validate the functional importance of these motifs, we generated a series of MEG3 deletion mutants. Each mutant was tested for its ability to activate p53-mediated transcriptional activity in luciferase reporter assay. As shown in Fig 1A, compared with full length of MEG3, each deletion mutants failed to stimulate the transcriptional activity of p53 transactivation in HepG2 cells, indicating that full length of MEG3 and intact structure are critical to activate the transcriptional activity of p53. Meanwhile, we also found that overexpression of MEG3 lead to an increased level of p53 protein in HepG2 (Fig 1B), but had no effect on the p53 transcript (data was not shown). Based on this observation, we speculated that enforced expression of MEG3 might affect the stability of the p53 protein. Then, we directly measured the stability of p53 protein under conditions where MEG3 was overexpressed by treating HepG2 cells and doxorubicin induced-HepG2 cells with cycloheximide in different time points and noted that overexpressed of MEG3 significantly stabilized p53 expression (Fig 1C and 1D).

MEG3 interacts with p53 protein directly
Our results indicated that MEG3 could enhance stability and transcriptional activity of p53 in hepatoma cells. One important way that lncRNAs play their functions is to interact with certain critical proteins, which imply that MEG3 might affect p53 transcriptional activity and target gene expression via interacting with p53 protein. Therefore, RNA pulldown was performed to identify whether p53 protein was associated with lncRNA MEG3. As was shown in Fig 2A  and 2B, purified biotinylated MEG3 RNA specifically retrieved p53 from HEK293 cell extract which was transfected with Flag-p53 expressing vector and from doxo-induced HepG2 cells extract. Further, biotinylated MEG3 bound to purified GST-p53 but not GST in vitro (Fig 2C).
Results of RIP assay revealed that antibody against p53 could result in higher enrichment of MEG3 RNA in doxo-induced HepG2 cells, compared with the IgG control ( Fig 2D). To further identify which domain of p53 is necessary for interacting with lncRNA MEG3, a series of vectors expressing p53 deletion mutants were constructed and RNA pulldown experiment was performed. The result showed that DBD of p53 was responsible for the direct association with MEG3 ( Fig 2E). Moreover, in order to identify which domain of MEG3 is critical for the interaction, we used MEG3 deletion mutants to bind p53 and found that M3 domain is important for the interaction (Fig 2B). In summary, results of in vivo RIP and in vitro RNA pulldown demonstrated the direct interaction between p53 and MEG3.

MEG3 regulates the expression of partial p53 target genes
The most significant function of p53 is to act as a transcription factor that directly or indirectly regulates many target genes. To quantify further analysis of the functional mechanism of MEG3, we constructed SK-Hep-1 hepatoma cell lines that stably over-expressed MEG3 ( Fig  3A). In order to investigate which p53 target genes were affected by MEG3, we compared the differential genes between stable SK-Hep-1 cell line overexpressed MEG3 and control SK-Hep-1 cell line transfected with pcDNA3.0 by mRNA expression microarray analysis. Through bioinformatics analysis of microarray data, compared with 287 p53 target genes whose data come from transcriptional regulatory element database (TRED) [25,26], we found 34 of these p53 target genes showed consequent changes in gene expression in stable cell line overexpressing MEG3 (Tables 1 and 2) including 21 up-regulated genes which principally involved in negative regulation of apoptosis, cell death, cell cycle and cell proliferation and 13 down-regulated genes which mainly participated in positive regulation of cell cycle and DNA replication. For example, GADD45A which is a p53-responsive stress protein leads to growth arrest and functions as tumor-and autoimmune suppressors [27]; CDC25C phosphatase which triggers cellular entry into mitosis is a target for transcriptional downregulation by p53 and this repression can be shown to contribute to p53-dependent growth arrest [28]. Using qRT-PCR, we validated the changes in the cellular levels of part of p53 target genes (Fig 3B and Figure A in S1 File.). The results showed that GADD45A, EGR1, SESN2 and TGFA were up- regulated in SK-Hep-1 after overexpression of MEG3, which is consistent with the microarray data. Moreover, to further identify whether the regulations of these genes by MEG3 depended on p53, we used siRNA-p53 to knock down p53 level and validate the changes of these genes by RT-qPCR (Fig 3C). The results showed that MEG3 overexpression could not regulate genes level after knockdown p53. All these data suggest that MEG3 not only affects p53 stability but regulates, at least in part, the transactivation of selected p53 target genes.

MEG3 is down-regulated in HCC tissues and inhibits the growth of hepatoma cells
To identify MEG3 that is altered in expression in HCC, we detected the expression levels of MEG3 in 23 pairs of HCC tissues and adjacent non-tumor tissues. The results showed that  Fig 1A). (C) In vitro transcribed biotinylated MEG3 retrieved purified GST-p53 but not GST. (D) RIP experiments were performed using an antibody against the p53 on extracts from doxo-induced HepG2 cells. The purified RNA was used for qRT-PCR, and the enrichment of the lncRNA MEG3 was normalized to GAPDH (upper, western blot of p53 protein after immunoprecipitation). Data was relative to mock-IP (IgG). (E) A series of p53 deletion mutants which were flag-tagged was treated as in (A), and association was detected by anti-Flag. Up represents successful expression of p53 deletion mutants. Down represents in vitro transcribed biotinylated MEG3 RNA was incubated with HEK293 cell extract which was transfected with p53 deletion mutants and associated proteins were detected by anti-Flag. MEG3 were remarkably reduced twofold or more in 14 HCC samples compared with the adjacent non-tumoral samples, and not detected in the other 3 HCC samples (Fig 4A). To assess the functional consequences of deregulated MEG3 expression, we transfected hepatoma cells with MEG3 expressing vector (pcDNA3.0-MEG3), and detected its ability to inhibit cell growth by cell-counting kit-8 assay. The results indicated that overexpression of the MEG3 in HepG2 cells resulted in a decrease in cell growth rate compared with the control vector ( Fig 4B). Moreover, enforced expression of MEG3 resulted in suppression of colony formation in SK-Hep-1 and HepG2 (Fig 4C). Taken together, these results demonstrated that MEG3 can inhibit the proliferation of human hepatoma cells. Next, to determine why MEG3 can inhibit hepatoma cell growth, cell cycle and apoptosis were analyzed by flow cytometry. The results showed that there was no significant change in cell cycle distribution of HepG2 cells transfected with MEG3 compared with control vector (data was not shown). The analysis of apoptosis indicated that ectopic expression of MEG3 could enhance serum starvation induced apoptosis in HepG2 cells (Fig 4D).

Discussion
In this study, we have identified that ectopic expression of MEG3 gene in human tumor cells significantly inhibits proliferation and induces apoptosis. Moreover, overexpression of MEG3 resulted in increasing of p53 protein and activated its transcription activity. Furthermore, we found that MEG3 could interact with DBD of p53, and play regulatory roles. Normally, half-life of p53 is short according to the degradation mediated by ubiquitin complex [29]. We observed that overexpression of MEG3 significantly stabilized p53 expression which leads to an increased level of p53 protein. Therefore, it is likely that MEG3 activates p53 through inhibition of p53 ubiquitination and blockage of p53 degradation. Y zhou et al. have found that MDM2 expression was repressed by MEG3, suggesting that down-regulation in MDM2 protein levels contributes to p53 activation by MEG3 [22]. Meanwhile, there is other possible mechanism by which MEG3 activates p53. A dizzying number of cofactors involved in the pathway of p53 network coordinate with p53. Indeed, as the most in-depth transcription factor, p53 tetramer associated with multiple regulators binds to its response elements where it could recruit myriad of transcriptional co-regulators such as chromatin remodeling complexes, histone modifying enzymes, and components of transcription machinery to modulate activity of RNA polymerase at target loci [30]. So it is possible that MEG3 which can be folded into a complicated structure can act as co-activator of p53 to activate its transcriptional activity. While p53 is active as a homotetramer whose tetramerization is pivotal for its function and plays a vital role in the regulation of p53 activity [31], it is possible that MEG3 contributes to the tetramization of p53 leading to activation of p53.
The results of RIP, RNA pulldown and deletion mapping demonstrated the direct interaction between MEG3 and DNA binding domain of p53 is responsible for the direct association of p53 with MEG3. The architecture of p53 protein contains three major domains: N-terminal transactivation domain (TAD), central sequence-specific DNA binding domain (DBD), and Cterminal regulator region containing oligomerization domain (OD) [32]. Disruption of p53   pathway occurring in a variety of cancers correlated with the tumorigenic mutants that are defective in DNA binding and consequently lose the function of transcriptional activation [33]. Therefore, it is clear that DBD is essential for the activity of p53 to bind specific response element. According to our result that DBD of p53 is responsible for the direct association with MEG3, it would lead to inhibition of p53, however, we observed that MEG3 enhance transcriptional activity of p53. Thus, we speculated that this is a dynamic process in cells that the compound which is formed by association of p53 and MEG3 was recruited to the specific target gene, then MEG3 dissociated and p53 started to activate target gene.
In this article, it is intriguing that although p53 protein level was increased after ectopic expression of MEG3, results showed that part of p53 target genes were varied, such as GADD45A, EGR1, SESN2 and TGFA. The p53 transcriptional program is regulated in a stimulus-specific fashion [34,35], whereby distinct subset of p53 target genes are induced in response to different stimulation [30]. Therefore, it is possible that interaction between MEG3 and p53 alters p53 binding propensity which may be due to the primary, secondary or senior structure of MEG3. GADD45A, a p53-regulated and DNA damage responsive gene, plays an important role in suppressing cell proliferation, mediating cell cycle arrest, promoting apoptosis, inducing DNA repair, and stabilizing genomics [36]. TGFA, a direct target gene of p53, encodes a growth factor that is a ligand for the epidermal growth factor receptor, which activates a signaling pathway for cell proliferation, differentiation and development [37]. These genes have been associated with many types of cancers. Tumor suppression of MEG3 may be implemented by the function of these target genes of p53.
P53 is a crucial tumor suppressor which is involved in preventing cancer through acting as a transcription factor to regulate its target genes [38][39][40]. In this article, we first found that lncRNA MEG3 could interact with p53 protein to activate p53 target genes. Association between a transcription factor and lncRNA is not yet the first discovery, as previous studies have found that lncRNA PANDA could interact with the transcription factor NF-YA to limit expression of proapoptotic genes under DNA damage [41]. Though there were many studies on the regulatory mechanism of p53 function before, lncRNA involved in the regulation of p53 function is a wholly novel field. How to build the relationship among lncRNA, protein and chromatin remains to be studied. Our data proposed a hypothesis that the association of p53 and MEG3 formed a compound to enhance stability of p53, and the compound was recruited to the specific target gene according to the selectivity of MEG3 sequence or structure, then MEG3 dissociated and p53 started to activate target gene to play a tumor suppressor role (Fig 3D).
Supporting Information S1 File. Figure