TGF-β–Induced Upregulation of malat1 Promotes Bladder Cancer Metastasis by Associating with suz12 Free

Human bladder cancer is one of the most common cancers worldwide (1, 2). The majority of bladder cancers are urothelial cell carcinomas evolved from the epithelial lining of the bladder wall (3, 4). The urothelial carcinomas are noninvasive tumors that commonly recur but rarely progress. Invasive bladder tumors are more aggressive. Patients with invasive cancer have a much worse prognosis, with a 50% 5-year survival (5). The advances in suitable therapy for the purpose of increasing survival rate have been limited because the molecular mechanisms causing metastasis are not entirely known. Revealing the underlying mechanism of bladder cancer metastasis is indispensable for developing effective therapy.

Long noncoding RNAs (lncRNA) are a class of noncoding RNA longer than 200 nucleotides with no protein-coding capacity (7). Through regulating gene expression by a variety of mechanisms, including transcription, posttranscriptional processing, genomic imprinting, chromatin modification, and the regulation of protein function, lncRNAs have been shown to play important regulatory roles in diverse biologic processes such as development, cell growth, and tumorigenesis (8–10). HOTAIR expression level is higher in tumor tissues and HOTAIR overexpression is correlated with the presence of cancer metastasis (6, 7). Forced expression of HOTAIR in cancer cells leads to altered histone H3 lysine 27 (H3K27) methylation and abnormal gene expression, and increases cancer invasiveness and metastasis in a manner dependent on polycomb repressive complex 2 (PRC2; ref. 6).

Suz12 or EZH2 functions as a H3K27 methyltransferase to selectively repress gene expression when comprising the PRC2 (8). Suz12 is also required for E-cadherin repression by the Snail1 transcription factor (9). In cancer cells, suz12 inhibition prevents the ability of Snail1 to downregulate E-cadherin and partially suppresses E-cadherin (9). The means by which PRC2 targets specific chromatin regions is currently unclear, but lncRNAs have been shown to interact with PRC2 and facilitate its recruitment to the promoters of some target genes (8, 10).

malat1 (metastasis associated lung adenocarcinoma transcript 1) is highly expressed in lung, pancreas, and other healthy organs as well as in non–small cell lung cancer (NSCLC; ref. 11). malat1 overexpression is associated with NSCLC (11). malat1 silencing impairs cell motility of lung cancer cells by regulating expression of migration-related genes (12, 13). More recently, the role of malat1 in regulating bladder cancer progression is revealed. malat1 contributes to bladder cancer cell migration by regulating EMT-associated ZEB1, ZEB2, and Slug levels (14). However, the underlying mechanisms of malat1 regulating cancer metastasis remain unclear.

On the basis of these findings, we further investigated the molecular mechanism that malat1 regulates bladder cancer progression. In the study, we found that TGF-β–induced malat1 promotes EMT and subsequent bladder cancer metastasis by associating with suz12.

Human bladder cancer cells were purchased from the American Type Culture Collection (ATCC) and maintained in RPMI-1640 with 10% FBS (Gibco), and cultured at 37°C with 5% CO₂. Human bladder specimens were obtained from First People's Hospital of Shanghai with informed consent. The protocols used in the study were approved by the Hospital's Protection of Human Subjects Committee. Of note, 95 specimens of pathologically and normally diagnosed biopsy specimens (≥3 cm away from bladder cancer tissues) were obtained from patients with bladder tumors (Supplementary Table S1).

Total RNA was extracted from bladder cancer tissues or cells using TRizol reagent (Invitrogen), and the reverse-transcription reactions were performed using random primers. Real-time PCR was carried out using a standard SYBR Green PCR kit (Toyobo) protocol on Applied Biosystems 7300 Real-Time PCR system (Applied Biosystems). β-Actin was used as reference for lncRNAs or mRNA. The results were expressed as log₁₀ (⁠|$2^{-{\rm \Delta \Delta} C_{\rm t}}$|⁠).

Western blot analysis to assess protein expression was performed as previously described (15). The anti-E-cadherin, fibronectin, N-cadherin, and matrix metalloproteinase 9 (MMP9) were purchased from Santa Cruz Biotechnology. The anti-β-actin primary antibodies were obtained from Sigma.

RNA immunoprecipitation (RIP) or RNA pulldown was performed as described previously (2, 16). RIP experiments were carried out using a Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore) according to the manufacturer's instructions. Antibody for RIP assays of suz12 (Cell Signaling Technology) was diluted as 1:1,000. The coprecipitated RNAs were detected by reverse transcription PCR. The primes for detecting malat1 are listed in Supplementary Table S2.

For RNA pulldown assay, biotin-labeled RNAs were in vitro transcribed with the Biotin RNA Labeling Mix (Roche Diagnostics) and T7/SP6 RNA polymerase (Roche Diagnostics). Cell nuclear extract (2 mg) was mixed with biotinylated RNA Biotin-labeled RNAs (100 pmol). Washed streptavidin agarose beads (100 μL) were added to each binding reaction and further incubated at room temperature for 1 hour. Beads were washed briefly three times and boiled in SDS buffer, and the retrieved protein was detected by standard Western blot technique.

Chromatin immunoprecipitation (ChIP) was carried out by using the EZ-ChIP Chromatin Immunoprecipitation Kit (Millipore) according to the manufacturer's protocol. Briefly, cross-linked chromatin was sonicated into 200- to 1,000-bp fragments. Then, the chromatin was immunoprecipitated using anti-suz12 or anti-H3K27me3 antibodies. Quantitative PCR was conducted according to the method described above. Primers are listed in Supplementary Table S2.

The siRNA sequences and RNA interference (RNAi) method are listed in Supplementary Table S3.

The male BALB/C nude mice were purchased from the Shanghai Experimental Animal Center of the Chinese Academy of Sciences, and were bred and maintained in a specific pathogen-free facility. For experimental metastasis assays, mice were implanted with 2 × 10⁶/50 μL tumor cells by lateral tail vein injection. Metastatic progression was monitored weekly and quantified using a noninvasive bioluminescence In-Vivo Imaging System (IVIS; Xenogen) as described previously (17, 18). For spontaneous metastasis experiments, C57BL/6 mice were injected subcutaneously in one flank with 2 × 10⁵ syngeneic MB49 cells, and lung metastasis were monitored after 6 weeks. Animals were divided into three groups, which include malat1-siRNA treatment group, suz12-siRNA treatment group, and MB49 control group.

RT4 or T24 (1to 2 × 10⁶ cells per well) cells were treated with malat1-siRNA or suz12-siRNA. Identical scratches were made in parallel wells 48 hours after transfection using a 1,000-μL plastic pipette tip. The size of wound was measured and the percentage of the cells that had migrated was calculated.

Cell invasion was administrated using the Transwell invasion assay with inserts of 8-μm pore size (Corning Costar), as described previously (19).

Data were presented as mean ± SD. The Student t test, χ² test, or Fisher exact test were used for comparisons between groups. The Kaplan–Meier method was used to estimate overall survival (OS), and multivariate Cox regression analysis with backward stepwise approach was used to test for independent prognostic factors. The difference was deemed statistically significant at P < 0.05.

TGF-β functions as a potent stimulator of bladder cancer cell migration and invasion (20). Recent studies showed that malat1 is an lncRNA associated with NSCLC and bladder cancer (11, 14). Therefore, we thought that TGF-β may upregulate malat1 to increase bladder cancer metastasis. We first confirmed that malat1 and EMT levels are regulated upon TGF-β treatment in bladder cancer cells. Figure 1A and B shows that malat1 is induced and peaked at 36 hours after TGF-β treatment, which is followed by E-cadherin mRNA reduction in RT4 cells. Meanwhile, TGF-β treatment increases the mesenchymal markers (N-cadherin and fibronectin) mRNA levels (Supplementary Fig. S1). Moreover, E-cadherin protein level is reduced, and N-cadherin and fibronectin protein level is increased after TGF-β treatment (Fig. 1C), indicating that EMT is induced by TGF-β. Similarly, malat1 is induced by TGF-β in another bladder cancer cell line T24 with concurrent induction of N-cadherin and fibronectin, as well as reduction of E-cadherin (Fig. 1D and E).

We next examined the malat1 expression levels in bladder cancer tissues and adjacent normal control. Figure 1F showed that malat1 expression levels are significantly upregulated in most bladder cancer tissues compared with normal control (81%, n = 95). When the tumor tissues were stratified on the basis of clinical progression, we found that the TGF-β1 and malat1 levels are higher in primary tumors that subsequently metastasized than those in nonmetastatic bladder cancer (Fig. 1G and Supplementary Table S1). In addition, multivariate Cox regression analysis showed that malat1 expression, clinical stage, and histologic grade were independent prognostic indicators for OS (Supplementary Table S4). Expectedly, E-cadherin mRNA levels are lower in primary tumors that subsequently metastasized than those in nonmetastatic bladder cancer (Fig. 1H). These data suggest that TGF-β induces malat1 and EMT in bladder cancer.

To define functional links between malat1 and EMT, we examined the effects of malat1 knockdown on E-cadherin expression. We first assayed the expression level of malat1 and E-cadherin in both noninvasive bladder cancer cell and invasive bladder cancer cell. Among the four invasive cell lines, the malat1 levels are relatively higher with concurrent low levels of E-cadherin compared with those in the three noninvasive ones (Fig. 2A). A significant negative correlation is also observed between the E-cadherin mRNA levels and the malat1 expression levels in vivo (r² = 0.1582; P = 0.0029; Fig. 2B). We then investigated whether malat1 regulates E-cadherin expression in bladder cancer cells. As shown in Fig. 2C and D, malat1 knockdown significantly increases E-cadherin expression at both transcript and protein levels in T24 cells. Similarly, malat1 knockdown also enhances E-cadherin expression in another bladder cancer cell line RT4 (Fig. 2E and F). More important, malat1 inhibition partially abrogates TGF-β–induced downregulation of E-cadherin in bladder cancer cells (Fig. 2G and H). These results demonstrate that TGF-β–induced malat1 promotes EMT in bladder cancer.

LncRNA typically functions by binding to specific protein partners, serving key regulatory roles to influence the activity of the proteins they bind (17). Recent studies reported that lncRNA recruits polycomb-group proteins to regulate gene expression, and about 20% of all human lncRNAs have been shown to physically associate with PRC2 (6, 21). For example, HOTAIR is proved to achieve this by targeting ezh2 and suz12 to govern the cells' epigenetic state and subsequent gene expression (6, 22). We, therefore, speculated that malat1 may regulate metastasis-related gene expression in such a manner. To test this, we performed RIP with antibodies against ezh2 or suz12 from nuclear extracts of T24 and RT4 cells. We observed a significant enrichment of malat1 with the suz12 antibody compared with the nonspecific immunoglobulin G (IgG) antibody control (Fig 3A). These results were confirmed by using a different antibody against suz12 and another primer pair for malat1 to exclude potential nonspecific association (Fig. 3B), but no enrichment of malat1 with the ezh2 antibody was confirmed (Fig. 3C). We then performed RNA pulldown to validate the association between malat1 and suz12. Figure 3D shows that malat1 can specifically combine and precipitate suz12, but shows no precipitation of suz12 by using lncRNA control. These data demonstrate that malat1 is specifically associated with suz12.

To check whether expression of the E-cadherin is controlled by suz12, we analyzed the E-cadherin expression after suz12 overexpression or knockdown. As shown in Fig. 4A, suz12 overexpression results in a significant reduction of E-cadherin mRNA level in RT4 cells. Similarly, suz12 overexpression decreases E-cadherin expression in T24 cells (Fig. 4B). Inversely, suz12 knockdown increases E-cadherin mRNA and protein level in bladder cancer cells (Fig. 4C and D). In addition, ezh2 overexpression also resulted in reduction of E-cadherin expression (Supplementary Fig. S2). To further address how malat1 is involved in EMT through enrichment of suz12, we performed ChIP analysis in T24 cells treated with malat1-siRNA. ChIP arrays showed that malat1 inhibition decreases the binding of suz12 with the E-cadherin promoter in T24 cells (Fig. 4E and F). Similar results were observed when we determined the levels of H3K27me3 in the E-cadherin promoter (Fig. 4G). We do not detect the binding of IgG with E-cadherin promoter after malat1 inhibition, suggesting that the association of suz12 and H3K27me3 in the E-cadherin promoter is specific (Fig. 4H). These results suggest that malat1 represses E-cadherin expression by associating with suz12.

To further evaluate the role of malat1 in TGF-β–induced EMT, we examined the effects of malat1 silencing on TGF-β–induced EMT markers. As shown in Fig. 5A, TGF-β–induced inhibition of E-cadherin transcript expression is potently suppressed by malat1 silencing in RT4 cells. Consistent with above results, suz12 silencing increases E-cadherin mRNA and protein level in RT4 cells treated with TGF-β (Fig. 5B and C). Moreover, malat1, as well as suz12, silencing suppresses the TGF-β induction of EMT markers such as N-cadherin and fibronectin (Fig. 5C). Similarly, using in T24 bladder cancer cells, malat1 or suz12 knockdown increases E-cadherin expression, confirming in another cell line that malat1 or suz12 silencing can suppress TGF-β induction of EMT (Fig. 5D and E).

The biologic consequences of malat1 and suz12 in TGF-β regulation of invasion and migration were then examined using cell biology assays. Scratch assays showed that TGF-β–induced cell migration is almost abolished by malat1 and suz12 knockdown in RT4 and T24 cells (Fig. 6A and B). Moreover, we found that malat1 or suz12 silencing significantly decreases basal MMP9 activity (Fig. 6C and D). So, we further investigated the role of malat1 in cell invasion. The bladder cancer cells were treated with malat1-siRNA or suz12-siRNA and cell invasion was analyzed. As shown in Fig. 6E and F, malat1 or suz12 knockdown in noninvasive bladder cancer cell RT4 significantly reduces TGF-β–induced cell invasion compared with N.C. Similarly, malat1 or suz12 silencing in invasive bladder cancer cell T24 inhibits TGF-β–induced increases in cell invasion (Fig. 6E and F). These data suggest that TGF-β promotes bladder cancer cell migration and invasion by the malat1/suz12 pathway in vitro.

On the basis of the above findings that malat1 or suz12 inhibition reduced TGF-β induction of markers of EMT as well as cell migration and invasion, we, therefore, investigated the effects of malat1 or suz12 inhibition on bladder cancer metastatic ability in vivo. To establish a metastatic cancer model in vivo, malat1-downregulated T24 cells, suz12-downregulated T24 cells, or T24 cells stably expressing luciferase were injected into tail vein of male nude mice. A bioluminescent signal first became detectable at 9 weeks after tail vein injection. At 9 weeks, bioluminescent signals are lower in malat1 or suz12 knockdown group compared with those in control group. Moreover, bioluminescent signals in control group increase over time, whereas those in malat1 or suz12 knockdown group remain suppressed (Fig. 6G). We then inoculated the syngeneic MB49 cells subcutaneously into C57BL/6 mice and observed lung metastases after 6 weeks. Consistent with the above observations, malat1 or suz12 knockdown reduces the number of visible lung metastases (Fig. 6H). Furthermore, Kaplan–Meier analysis revealed that higher malat1 expression level is significantly correlated with a markedly reduced OS in patients with bladder cancer (P = 0.0083, log-rank test; Fig. 6I). These data suggest that malat1 or suz12 inhibition can suppress bladder cancer metastasis in this preclinical study.

Large-scale complementary DNA cloning projects have identified that majority of the mammalian genome is transcribed, and only minority of these transcripts represent protein-coding genes (23). LncRNAs have been shown to play an important role in diverse biologic processes such as development, cell growth, and tumorigenesis (24, 25). More recently, lncRNAs have also been implicated in regulating specific steps in the metastatic cascade. Because metastasis is the major cause of death for patients with bladder cancer, it stands to reason that defining the molecular mechanisms whereby lncRNAs have an impact on metastasis may provide novel opportunities to treat metastatic bladder cancers.

malat1 is one of the first cancer-associated lncRNAs. malat1 level is upregulated in cancer tissues and its dysregulation is discovered as a marker for metastasis development in early stages of lung adenocarcinoma (11) and more recently in bladder cancer (14). However, its functional role in this process was only beginning to emerge by virtue of its link to cell migration. malat1 positively regulates cell motility through the concomitant regulation of motility-related genes (12). Ying and colleagues showed that upregulated malat1 promotes bladder cancer cell migration by inducing EMT (14). It is also well known that increased levels of TGF-β in patients with bladder cancer have significant prognostic value for highly aggressive metastatic disease and are considered a poor prognosis marker (26, 27). One mechanism by which TGF-β contributes to cancer progression is through induction of EMT. On the basis of these findings, we speculated that upregulated TGF-β may increase EMT by the regulation of malat1. Here, we found that TGF-β induces malat1 expression and EMT in bladder cancer cells. malat1 overexpression is significantly correlated with poor survival in patients with bladder cancer. We further demonstrated that malat1 and E-cadherin expression is negatively correlated in vitro and in vivo. malat1 inhibition increases E-cadherin expression with concurrent downregulation of N-cadherin and fibronectin in bladder cancer cells. More importantly, malat1 inhibition abrogates TGF-β–induced EMT.

Recent studies reported that about 20% of all human lncRNAs have been shown to physically associate with the PRC2 (6, 21). HOTAIR is proved to achieve this by targeting ezh2 and suz12 to govern the cells' epigenetic state and subsequent gene expression (6, 22). PRC2 is also required for E-cadherin repression (9). Herranz and colleagues showed that Snail1 recruits PRC2 to the E-cadherin promoter and requires the activity of this complex to repress E-cadherin expression. Therefore, we thought that malat1 may regulate EMT by recruiting PRC2. In fact, our results demonstrated that malat1 is associated with suz12, and that this association results in E-cadherin downregulation and N-cadherin and fibronectin upregulation. malat1 inhibition decreases the binding of suz12 with the E-cadherin promoter. Similar results were observed when we assayed the levels of H3K27me3 in the E-cadherin promoter. Finally, we demonstrated that TGF-β promotes bladder cancer cell migration and invasion in vitro and bladder cancer metastasis in vivo by the malat1/suz12 pathway. TGF-β–induced cell migration and invasion is almost abolished by malat1 and suz12 knockdown in vitro. Moreover, malat1 or suz12 knockdown markedly suppresses bladder cancer metastasis, suggesting that malat1 or suz12 inhibition can suppress bladder cancer metastasis in this preclinical study.

These data confirm that malat1 is an important mediator of TGF-β–induced EMT, and suggest that malat1 inhibition may represent a promising therapeutic option for suppressing bladder cancer metastatic progression.

No potential conflicts of interest were disclosed.

Conception and design: Y. Fan, B. Shen

Development of methodology: M. Tan, X. Mu, Y. Qin, F. Zhang, Y. Liu

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): B. Shen

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Fan, F. Zhang

Writing, review, and/or revision of the manuscript: Y. Fan, B. Shen, M. Tan, X. Mu, Y. Qin, Y. Liu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): X. Mu, F. Zhang, Y. Liu

Study supervision: B. Shen