Background
With an annual incidence of 0.5 per 100,000 populations in Western countries, mantle cell lymphoma (MCL) is an aggressive subtype of non-Hodgkin’s lymphoma (NHL), comprising about 6% of NHL cases [
1]. Despite an initial therapeutic response, patients consistently develop recurrence and chemoresistance [
2]. Moreover, elderly patients with MCL do not tolerate the toxicities of chemotherapy. Thus, there is a substantial need to identify novel markers for prognosis and explore alternative therapies for MCL patients.
Over half of the human genome is actively transcribed as noncoding RNAs [
3]. The noncoding transcripts that are more than 200 nucleotides in length are termed long noncoding RNAs (lncRNAs). lncRNAs have been recognized to play important roles in pathologic conditions, such as cancer [
4,
5]. It has been demonstrated that lncRNAs influence tumor progression through modulating cell cycle, survival, immune response or pluripotency, through their interactions with DNA, protein, and RNA [
6,
7].
Recent studies in lymphoma have shown that lncRNA FAS-AS1 regulates skipping of exon 6 through direct interaction with RBM5, in order to produce membrane associated Fas or decoy soluble Fas [
8]. In addition, expression of lncRNAs (PEG10, LUNAR1 and HULC) is correlated with clinical poor prognosis in diffuse large B cell lymphoma (DLBCL). Reduced expression of these same lncRNAs inhibited cell proliferation of DLBCL in vitro [
9‐
11]. Additional lncRNAs have been discovered to regulate oncogenic process in lymphomas. For example, Verma Asyu et al. examined RNA-seq data from primary DLBCL tumors and identified 2632 novel lncRNAs, which are implicated to interact with EZH2, H3K4me3, NFκB and STAT3 [
12]. Another study using microarray analysis reported 189 lncRNAs that were significantly dysregulated in follicular lymphoma (FL), which are related to the TNF signaling and virus related carcinogenesis [
13].
Metastasis-associated lung adenocarcinoma transcript 1(MALAT1) is an evolutionarily conserved, long non-coding RNA 8.7 kb transcript, located on chromosome 11q13, a site in vicinity of t (11;14) of MCL. Chromosomal translocations are known to influence expression of local genes. MALAT1 plays a critical role in maintaining the proliferation potential of early-stage hematopoietic cells [
14]. MALAT1 is differentially regulated during the activation of B-cells, and recent reports implicate it as a target of activation-induced deaminase (AID), which can induce oncogenic translocations in germinal center B-cells [
15]. The expression level of MALAT1 is correlated with tumorigenesis and metastasis in solid cancers and multiple myeloma, suggesting a universal cancer role [
16]. Despite many previous studies, the role of MALAT1 in lymphoma is not yet fully understood. In this study, we will analyze the role of MALAT1 in the pathophysiologic process of MCL.
Methods
Patient samples
Mantle cell lymphoma patients’ samples and clinical information were collected and published under The University of Texas MD Anderson Cancer Center IRB-approved clinical protocol LAB08-0190 for use of human tissues, with the written informed consent of all patients. Information about the patients is shown in Additional file
1: Table S1. Normal B lymphocytes were isolated from peripheral blood of healthy donors’ blood, obtained from Gulf Coast Blood Center (Houston, TX, USA), with CD19+ magnetic beads and released with DETACHaBEAD CD19 (Invitrogen, Grand Island, NY, USA).
Cell culture
Mantle cell lymphoma cell lines (Mino and Jeko-1) were obtained from ATCC (Manassas, VA, USA). Cell lines were regularly tested for mycoplasma (Lonza MycoAlert) and authenticated at the Cell Line Core Facility at MD Anderson Cancer Center, University of Texas, Houston, TX, USA. The cell lines were cultured in RPMI 1640 medium with 10% fetal bovine serum (FBS, Gibco). The human cell lines were validated based on short tandem repeats (STR). STR repeats are regions of microsatellite instability with defined tri- or tetrad-nucleotide repeats that are located throughout the chromosomes. PCR reactions using primers on non-repetitive flanking regions generate PCR products of different sizes based on the number of repeats in the region; the size of these PCR products are determined by capillary electrophoresis. This is performed by core facility at MD Anderson cancer center.
RNA isolation and quantitative real-time PCR
Total RNA was extracted from patients MCL tissue or cell lines using RNeasy kit (QIAGEN) according to the manufacturer’s instructions. Reverse transcription reactions were performed using a SuperScript III reverse transcriptase kit (Invitrogen-Life Technologies) according to the manufacturer’s protocol. Quantitative real-time RT-PCR was performed in triplicate using the StepOnePlus Real-Time PCR System (Applied Biosystems-Life Technologies) with TaqMan Universal PCR Master Mix according to the manufacture’s protocol (Applied Biosystems). The TaqMan Gene Expression Assays (probe and primers) were purchased from Invitrogen Life Technologies, including MALAT1 (ID: Hs00273907), EZH2 (ID: Hs00544833), CDKN1A/p21 (ID: Hs00355782) and CDKN1B/p27 (ID: Hs01597588). Human GAPDH was used as endogenous control. StepOne software version 2.0 (Applied Biosystems) was used to determine RNA expression levels.
RNA interference
Mantle cell lymphoma cell lines were transfected with Neo transfection (Invitrogen) using five human MALAT1 siRNAs (si-MALAT1, Life Technologies): si-MALAT1 No.1 (product ID: 272231), si-MALAT1 No.2 (product ID:272233), si-MALAT1 No.3 (product ID:272234), si-MALAT1 235 (product ID:272235) and si-MALAT1 236 (product ID:272236), or negative control siRNA (si-NC, product ID: AM4635; Life Technologies). Transfection was performed three times to confirm results. Briefly, MCL cells were resuspended in Buffer R (Invitrogen) and mixed with a siRNA. Each 100 μL aliquot contained 2 × 106 cells and 1 nmol of siRNA. After electroporation with program parameters (1550 V, 10 ms, 3 pulses) cells were cultured in RPMI 1640 and 10% FBS without antibodies. Quantitative real-time PCR was performed to evaluate expression levels of MALAT1 using total RNA extracted 48 h after transfection (si-NC, si-MALAT) in Mino and Jeko-1 cells.
Evaluation of proliferation, apoptosis and colony formation assay
Cell viability was measured 24, 48 and 72 h after transfection with MTT assay kit (Promega, Madison, WI). Flow cytometry analysis for apoptosis was performed using Annexin V-FITC Apoptosis Detection Kit 48 h after transfection according to the manufacturer’s protocol (Sigma–Aldrich). All experiments were performed in triplicate. The colony formation assay was performed as described previously [
17]. In brief, 5 × 10
4 cells were mixed with methylcellulose (H4100; Stemcell Technologies, Vancouver, BC, Canada) containing RPMI-1640 + 10% FBS and poured in 35 mm plate. The colonies were allowed to grow for 7–14 days, followed by staining with p-Iodonitrotetrazolium violet and counted using Fluorchem 8800 imaging system (Alpha-Innotech, San Leandro, CA).
Cell synchronization and cell cycle analysis
Mino and Jeko-1 cells were synchronized to G2/M phase by treatment with nocodazole (Selleckchem). To synchronize cells in G1, the synchronized G2/M cells were washed with PBS and grown in fresh medium for 12 h. Briefly, Mino and Jeko-1 cells were collected 24 h after transfection, and treated with 50 ng/ml nocodazole for 24 h, then released in fresh medium. Cells were collected at 12 and 18 h after release and processed to ethanol fixing (70% ethanol, ice-cold), RNase A-pretreating (0.5 mg/ml at 37 °C for 30 min) and propidium iodide staining (50 μg/ml). Cell-cycle progression was measured by flow cytometry.
RNA immunoprecipitation assay
RNA immunoprecipitation (RIP) was performed using a Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore) according to the manufacturer’s instructions. Antibodies EZH2 (#5246; Cell Signaling Technology) and SUZ12 (#3737; Cell Signaling Technology) used in RIP assays were diluted as 1:100. The coprecipitated RNAs were transcribed to cDNA, and detected by quantitative real-time RT-PCR using TaqMan Gene Expression Assays (Invitrogen Life Technologies).
Chromatin immunoprecipitation assay
Chromatin immunoprecipitation (ChIP) was performed using the EZ-ChIP Chromatin Immunoprecipitation Kit (Millipore) according to the manufacturer’s instruction. Briefly, cross-linked chromatin was sonicated on a Branson Sonifier 150 at setting 4 with 15 s pulses six times on ice. Then, the chromatin was immunoprecipitated using anti-EZH2 (#5246; Cell Signaling Technology) or anti-H3K27me3antibodies (#17-622, EMD Millipore). The quantitative real-time PCR was used to detect immunoprecipitated DNA using SYBR Green PCR Master Mix (Applied Biosystems) according to the method described above. Primers were used to identify promoter domain in immunoprecipitated DNA: CDKN1A/p21 (5′-TCTGGGGTCTCACTTCTTGG-3′; 5′-ATGTGAGGAAGGCTCAGTGG-3′) and CDKN1B/p27 (5′-GATGGGGTTCACCGTGTTAG-3′; 5′-CCCTTTCCAAACATCCATTG-3′).
Western blot analysis
Western blot analysis was performed as previously described [
8,
18]. The antibodies used were specific for p21 Waf1/Cip1 (#2947, Cell Signaling Technology), p27 Kip1 (#2552, Cell Signaling Technology), EZH2 (#5246; Cell Signaling Technology), H3K27me3 (#39155, Active Motif), Histone H3 (#9715, Cell Signaling Technology) and pEZH2 T345 (#61241, Active Motif, recognize human pEZH2 T350). Anti-β-actin horseradish peroxidase antibody (1:10,000; Sigma Aldrich, Buchs, Switzerland) was used as loading control. Visualization was achieved by Supersignal West Pico chemiluminescent (or Femto Maximum Sensitivity) substrate (Pierce).
Statistical analysis
Experimental data are presented as mean ± SD from three independent experiments, unless otherwise indicated. Differences between groups were calculated using the two-tailed Student’s t test (GraphPad Prism, La Jolla, CA, USA). Correlation between MALAT1 mRNA and EZH2 mRNA expression in human MCL tissues was examined with two-sided Pearson correlation. Overall survival was estimated with Kaplan–Meier method. P < 0.05 was considered statistically significant.
Discussion
Previous studies have reported high expression of lncRNA MALAT1 in solid tumors and hematologic malignancies and hinted to its role on transcription complexes [
16]. Here we showed that MALAT1 is overexpressed in MCL tissues, which correlates with high MIPI and poor overall survival of MCL patients. Silencing of MALAT1 expression inhibited cell proliferation and increased apoptosis rates of MCL cells. In addition, we observed that down-regulation of MALAT1 increased expression of p21 and p27 through an EZH2-associated mechanism. Moreover, decreased phosphorylation of EZH2 at T350 attenuated its binding to MALAT1. Collectively, our results implicate MALAT1 as a mechanistic and prognostic marker for MCL.
MALAT1 has been found to be up-regulated in several types of solid tumors [
16]. Our study investigated MCL, an aggressive type of non Hodgkins lymphoma, and found that the expression of MALAT1 was significantly higher in MCL compared to normal B-cells (P < 0.01). We additionally identified that the increased expression of MALAT1 was associated with high-risk group (by MIPI) and lower overall survival after current chemotherapy in patients with MCL. We probe for the role of MALAT1 using knockdown approach. After siRNA-mediated knockdown of MALAT1, cell proliferation was decreased and the percentage of apoptotic cells was significantly increased in MCL cells (P < 0.01). MALAT1 in another hematological cancer, multiple myeloma (MM) has been shown to also target latent transforming growth factor beta-binding protein 3 (LTB3) and suggest that MALAT1 may have widespread effects across different transcription complexes [
28‐
30]. Mechanistically, MALAT1 recruited the transcription factor Sp1 to the promoter of LTBP3, thereby resulting in elevated level of LTBP3 transcription and TGF-β1 secretion, which plays a role in the suppression of bone formation in MM bone lesions [
29]. MALAT1 was also found associated with molecular pathways involving cell cycle regulation, p53-mediated DNA damage response, and mRNA maturation processes in MM using microarray analysis [
30].
Our immunoprecipitation experiments revealed notably high affinity of binding between MALAT1 and EZH2, rather than SUZ12 in MCL. Recent studies reported that several lncRNAs (such as HOTAIR, Xist, and H19) act to repress genes by binding PRC2 [
31‐
34]. Further studies of over 3300 lncRNAs identified that about 20% of these lncRNAs are binding partners for PRC2 [
35]. Notably, binding of lncRNA to PRC2 is size dependent, with higher affinity belonging to longer RNAs [
36]. In line with this, lncRNA MALAT1, with a large size (∼8000 nt), has been found to interact with PRC2 through EZH2 or SUZ12. Specifically, EZH2 mediates the role of the MALAT1–PRC2 partnership in the epithelial–mesenchymal transition in renal cell carcinoma [
22], whereas SUZ12 promotes the role of the MALAT1-PRC2 binding in tumor metastasis in bladder cancer [
21].
Our results showed that the EZH2 mRNA expression was significantly higher in primary MCL tissues compared to normal B-cells (P < 0.01). Patients with a higher level of EZH2 expression had lower overall survival as compared to those with low EZH2 expression. Our data showed that there was a significant positive correlation between MALAT1 and EZH2 mRNA in MCL. Using Basso’s Lymphoma dataset from Oncomine database, we previously showed that EZH2 mRNA is highly expressed in lymphomas compared to healthy donor’s B-lymphocytes [
8]. Independent of tumor proliferation, the expression level of EZH2 was identified as a prognostic factor in MCL, using multivariate survival analysis [
37]. Thus, we sought to determine the underlying molecular mechanisms by which MALAT1 functions in concert with EZH2 to regulate downstream effector in MCL.
Through EZH2-mediated mechanisms, PRC2 catalyzes the trimethylation of histone H3 on lysine 27 (i.e. H3K27me3) to repress transcription of specific genes. Upregulation of EZH2 leads to silencing of the genes that are involved in the progression and metastasis of solid tumors [
38,
39], and malignant hematopoiesis and lymphoproliferative disorders [
40]. Our results revealed that after knockdown of MALAT1, the levels of EZH2 and H3K27me3 expression were both decreased, whereas that of p21 and p27 were increased at both mRNA and protein levels, resulting in cell cycle arrest at G1/S transition. p21 and p27 are both CDKs suppressors involved in regulating cell cycle progression, and have also been considered as candidates for tumor-suppressor genes. Reduced expression of p21 or p27 has shown to be correlated with increased malignancy, high Ki-67 index and poor prognosis in MCL patients [
37,
41,
42]. Upregulation of EZH2 was related with cell proliferation in the development of B lymphocyte, and tumor suppressor genes, CDKN1A/p21 and CDKN1B/p27, were histone-3 lysine27-trimethylated and repressed in those proliferating GC-B-cells [
24]. And siRNA-mediated knockdown of EZH2 increased the mRNA expression of CDKN1A/p21 and CDKN1B/p27, leading to cell cycle arrest at G1/S transition in DLBCL and MCL cell lines [
24,
43]. Consistently, in several solid tumors (such as prostate cancer, ovarian cancer and lung cancer), CDKN1A/p21 and CDKN1B/p27 were identified as EZH2 targets by ChIP-qPCR analysis, and inhibited due to elevated level of EZH2 [
44‐
47].
Intriguingly, some reports have identified an important role for MALAT1 in regulating p21 and p27, potentially mediated by p53 [
48,
49]. However, no p53 knockdown experiments were carried out to directly show that whether p53 is the only mediator of p21 regulation by MALAT1 [
48]. Indeed, another report found no correlation between the expression of p21 and p53 in MCL patients, suggesting the possibility of p53-independent mechanisms underlying MALAT1-mediated regulation of p21 [
50]. A recent study demonstrated that p21 regulation is determined by the methylation and acetylation status of histone H3 on the p21 (WAF-1) promoter in lymphoma [
51]. These previous observations lead to our hypothesis that p21 and p27 are regulated by MALAT1 through an EZH2-driven H3K27me3 mechanism in MCL.
In the present study, our ChIP analysis demonstrated that EZH2 occupancy and H3K27me3 level at the promoter domains of p21 and p27 were both significantly decreased in MALAT1-deficient cells compared with control cells (P < 0.01), suggesting that overexpressed MALAT1 down-regulates p21 and p27, a process possibly mediated by EZH2-driven H3K27me3 in MCL. Consistent with our results, EZH2 depletion by either DZNep or siRNA induced the expression of p21 and p27 in MCL cell lines [
43].
It has been established that CDK1 and CDK2 phosphorylate EZH2 at T350, which is important for its binding to lncRNAs, HOTAIR and XIST, and recruitment of the PRC2 complex to the EZH2 target genes [
25‐
27]. Our results suggested that the phosphorylation of EZH2 at T350 affected its binding to lncRNA MALAT1. This is consistent with a previous study showing that lncRNA MALAT1 preferentially binds to EZH2 in two regions (amino acids 1–173 and 336–554), including the phosphorylation site T350 [
52]. In the present study, we initially observed that the phosphorylation of EZH2 at T350 was consistently reduced after treatment with CDKs inhibitor roscovitine. To confirm consequence of reduced phosphorylation of EZH2 at T350, our RIP results show decreased EZH2–MALAT1 binding.
Our results showed that the phosphorylation of EZH2 at T350 was inhibited in MALAT1-deficient MCL cells. This could be a result of elevated expression of CDKs inhibitor p21 and p27. To collectively explain these observations, we propose a hypothesis of positive feedback loop. Specifically, up-regulated expression of MALAT1 leads to the recruitment of EZH2 to target gene loci, thus enhancing EZH2-mediated H3K27me3 and suppressing the expressions of p21 and p27 (Fig.
7b). This leads to the activation of CDK1 and CDK2 that promote the phosphorylation of EZH2 at T350, which further increases the binding of EZH2 with MALAT1. This increased binding in turn enhances EZH2-mediated H3K27me3 and gene repression (Fig.
7b). Given that CDK1 and CDK2 are highly activated at the S and G2 phase, we envision a model that CDK-induced phosphorylation of EZH2 would probably become decreased at the other phase of the cell cycle in normal cells, which might facilitate the expression of EZH2 target genes and thereby promote cell differentiation. In MCL cells, CDKs are constantly activated due to the repression of p21 and p27 genes by overexpressed lncRNA MALAT1, thereby activating EZH2 to induce uncontrolled cell proliferation. We showed that deletion of MALAT1 with siRNA interferes with this postulated positive feedback loop, resulting in cell cycle arrest.
This study shows that overexpression of the lncRNA MALAT1 provides some oncogenic properties, and may be a prognostic factor or therapeutic target in MCL. MALAT1 expression is significantly higher in MCL tissues than normal tissues (P < 0.01). This may be associated with the key translocation of MCL t (11;14) (q13;q32), the breakpoint of which is adjacent to the MALAT1 gene loci 11q13. Further experiments are required to delineate this hypothesis. Our study also needs to be interpreted with cautions due to the lack of in vivo experiments. Future experiments with appropriate animal models may be helpful to clearly understand the underlying molecular mechanism in MCL progression. Small interfering RNA is a good choice for deleting lncRNAs, which locate in the cytoplasm. But for suppressing nuclear lnRNAs, such as MALAT1 and NEAT1, it is more effective to use antisense oligonucleotides (ASOs) [
53]. Preclinical studies have shown the therapeutic efficacy of ASOs targeting MALAT1 in the mouse MMTV–PyMT breast cancer model. Systemic knockdown of MALAT1 through subcutaneous injection of ASOs inhibits tumor proliferation and metastasis, and induces differentiation [
54]. Moreover, RNA depletion is not the only way to inhibit lncRNA function. Using steric blocking oligonucleotides, lncRNAs may be blocked to interact with their binding partners, such as protein, DNA and miRNA [
55]. Future technical innovations will offer more effective lncRNA-targeting therapeutics.