Clinical Cancer Research
Biology of Human Tumors
DNMT1–MicroRNA126 Epigenetic Circuit Contributes to Esophageal Squamous Cell Carcinoma Growth via ADAM9–EGFR–AKT Signaling Ronghua Liu1,2, Jie Gu3, Pei Jiang1,2, Yijie Zheng1,2, Xiaoming Liu1,2, Xuechao Jiang1,2, Enyu Huang1,2, Shudao Xiong1,4, Fengkai Xu3, Guangwei Liu1,2, Di Ge3, and Yiwei Chu1,2
Abstract Purpose: MicroRNAs (miRNA) are involved in and are controlled by epigenetic regulation, and thereby form a reciprocal regulatory circuit. Using next-generation sequencing (NGS)– based miRNA profiling, this study aimed to discover esophageal squamous cell carcinoma (ESCC)–specific miRNAs and miRNArelated epigenetic modulations. Experimental Design: NGS-based miRNA profiles were generated for four pairs of ESCC tissues and adjacent normal tissues. In situ hybridization was used to assess miRNA expression and its correlation with prognosis. miRNA-related DNA methylations were identified using bisulfite genomic sequencing, and the role of DNA methyltransferase 1 (DNMT1) was investigated using RNA interference. miRNA targets were screened by mRNA sequencing, and functional validation was performed in vitro and in vivo. Results: NGS-based miRNA profiling identified 78 differentially expressed miRNAs in ESCC. Among them, microRNA126-
3p (miR-126) was significantly downregulated, and its downregulation correlated with poor ESCC prognosis. Downregulation of miR-126 was due to promoter hypermethylation of its host gene, Egfl7. DNMT1 was aberrantly upregulated in ESCC and responsible for the hypermethylation of Egfl7. Intriguingly, DNMT1 was suppressed by overexpression of miR-126, indicating the existence of a regulatory feedback circuit. ADAM9 was identified as a key target of miR-126. Ectopic expression of miR-126 or silencing of ADAM9 reduced ESCC cell proliferation and migration by inhibiting epidermal growth factor receptor–AKT signaling. Conclusions: Our results indicate that miR-126 is a potential prognostic indicator for ESCC and suggest that a novel "DNMT1– miR-126 epigenetic circuit" is involved in ESCC progression. Consequently, miR-126–based epigenetic modulations may provide a basic rationale for new approaches to antitumor therapeutics. Clin Cancer Res; 21(4); 854–63. 2014 AACR.
Introduction
esophageal cancer remains less than 30% (2). Previous studies have described complicated genetic and epigenetic alterations involved in ESCC progression (3, 4), but the underlying mechanisms have not been fully elucidated. Therefore, to develop better approaches for ESCC diagnosis and therapy, a comprehensive understanding of the molecular variables that affect ESCC pathways is needed. MicroRNAs (miRNA) have emerged as key players in human tumors; their dysregulation affects cell proliferation (5), differentiation (6), invasion (7), and the epithelial–mesenchymal transition (EMT; ref. 8). miRNA expression profiles have been developed for use as prognostic/predictive biomarkers and been used to identify the molecular signatures and underlying functional mechanisms of various cancers. Previous miRNA profiling studies were conducted using miRNA expression microarray platforms (9, 10). However, in recent decades, the emergence of nextgeneration sequencing (NGS) platforms has revolutionized genomic studies. In comparison with microarray, NGS-based sequencing has advantages for the analysis of widely expressed miRNAs, the identification of miRNA sequence variations, and the discovery of novel miRNAs (11, 12). Although, microarray-based miRNA profiles of ESCC have been described (13), to our knowledge, no studies have used an NGS-based platform to identify ESCC-specific miRNAs. Furthermore, the pathologic mechanisms underlying their dysregulation remain incompletely understood.
Esophageal squamous cell carcinoma (ESCC) is one of the most aggressive malignancies of the gastrointestinal tract (1). Despite improvements in diagnostic approaches and the introduction of adjuvant chemotherapy, the 5-year survival rate in patients with 1 Department of Immunology and Key Laboratory of Medical Molecular Virology of MOE/MOH, School of Basic Medical Sciences, Fudan University, Shanghai, China. 2Biotherapy Research Center, Fudan University, Shanghai, China. 3Department of Thoracic Surgery, Zhongshan Hospital, Fudan University, Shanghai, China. 4Department of Hematology/Oncology, The Second Hospital of Anhui Medical University, Hefei, Anhui Province, China.
Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/). R. Liu and J. Gu contributed equally to this article. Corresponding Authors: Yiwei Chu, Department of Immunology and Key Laboratory of Medical Molecular Virology of MOE/MOH, School of Basic Medical Sciences, Fudan University, No. 138, Yi Xue Yuan Road, Mail Box 226, Shanghai 200032, China. Phone/Fax: 86-21-54237324; E-mail:
[email protected]; and Di Ge, Department of Thoracic Surgery, Zhongshan Hospital, Fudan University, 180 Fenglin Road, Shanghai 200032, China. Phone/Fax: 86-21-64041990-2559; E-mail:
[email protected] doi: 10.1158/1078-0432.CCR-14-1740 2014 American Association for Cancer Research.
854 Clin Cancer Res; 21(4) February 15, 2015
DNMT1–MicroRNA126 Epigenetic Circuit Regulates ESCC Growth
Translational Relevance Esophageal squamous cell carcinoma (ESCC) is one of the most aggressive malignancies of the gastrointestinal tract. Previous studies have described complicated genetic and epigenetic alterations involved in ESCC progression, but the underlying mechanisms have not been fully elucidated. Recent evidence shows that microRNAs (miRNA) are involved in and are controlled by epigenetic regulation, and thereby form a reciprocal regulatory circuit in many systems. In this study, using NGS-based miRNA profiling, we found the downregulation of miR-126 in ESCC and proposed a "DNMT1–miR-126 epigenetic circuit" involved in ESCC progression. Moreover, miR-126 acted as a novel tumor suppressor that effectively suppresses ESCC cell proliferation and migration by repressing ADAM9–EGFR–AKT signaling transduction and as an independent prognostic indicator for patients with ESCC. Taken together, our funding of "DNMT1–miR-126 epigenetic circuit" suggests another insight into pathologic mechanisms underlying ESCC initiation and progression, implying its clinical significance in developing targets for ESCC prediction and therapy.
Epigenetic modulations, including DNA methylation and histone acetylation, play a critical role in various diseases by regulating gene expression without changing the DNA sequence (14). Recent evidence indicates that miRNAs integrate into the epigenetic machinery via the transcriptional regulation of epigenetic factors, creating an "epigenetic–miRNA feedback loop" (15, 16). In this study, we propose that a "DNMT1–miR-126 epigenetic circuit" is involved in ESCC progression. MiR-126 was silenced by promoter hypermethylation of its host gene in ESCC tissues, whereas DNMT1 was aberrantly overexpressed. Silencing of DNMT1 increased miR-126 expression by reducing the hypermethylation, whereas miR-126 suppressed DNMT1 expression via a negative feedback process. MiR-126 was also identified as a novel tumor suppressor and a potential prognostic indicator in ESCC. Overexpression of miR-126 repressed ESCC cell proliferation and migration by inhibiting the activation of the ADAM9– EGFR–AKT pathway. Understanding the epigenetic–miRNA network in ESCC progression and modulating miRNA-related epigenetic dynamics hold promise for ESCC-targeted therapy.
Materials and Methods Samples and cells ESCC specimens embedded in paraffin were obtained from 185 patients who underwent esophagectomy at Zhongshan Hospital (Shanghai, China) in 2007 with informed consent, and clinicopathologic characteristics of patients were listed in Supplementary Table S1. Charlson comorbidity index (CCI) was used for the analysis of comorbidities (17). Each of the paraffin blocks was well formalin-fixed, paraffin-embedded, and without necrosis and hemorrhage. Follow-up was conducted every 6 months, and ended in December 2012, with a median follow-up of 32 months. The clinical endpoint was the survival state of patient at the last follow-up. Disease-specific survival was defined as the interval between surgery and death of ESCC. The data were censored for patients who died of causes other than ESCC or survived at the last
www.aacrjournals.org
follow-up. Another 30 pairs of frozen human primary human ESCC tissues (EC) and matched adjacent noncancerous tissues (NM) were obtained from 30 patients at Zhongshan Hospital in 2012. Tumor stage was determined according to seventh edition of the Union for International Cancer Control-American Joint Committee on Cancer tumor, node, metastasis staging system (18). Approval for the study was obtained from the Research Ethics Committee of Zhongshan Hospital. ESCC cell line ECa-109 was obtained from the Institute of Biochemistry and Cell Biology of the Chinese Academy of Science (TCHu 69, Shanghai, China), and ESCC cell line KYSE-510 (JCRB1436, National Institute for Biomedical Innovation, Japan) and normal human esophageal epithelial cell line HEEpiC (HEEC; #2720; ScienCell Research Laboratories) were supplied by Dr. Lu (Department of Thoracic Surgery, Zhongshan Hospital, Shanghai, China) (19), which were authenticated using short tandem repeat (STR) profiling and used in the sixth month after receipt. All the cells were cultured in RPMI-1640 supplemented with 10% FBS, 2 mmol/L glutamine, 100 IU/mL penicillin, and 100 mg/mL streptomycin sulfates. NGS-based miRNA profiling Total RNA from four pairs of ECs and matched NMs were isolated using a mirVana miRNA Isolation Kit (Ambion) according to the manufacturer's instructions. After PAGE purification, small RNA molecules were amplified and sequenced on an Illumina platform according to the manufacturer's instructions (20). Tissue microarray and in situ hybridization Tissue microarrays were constructed as described in a previous study (21). Digoxin (DIG)-labeled oligonucleotide probes complementary to hsa-miR-126 were designed as previously described and purchased from ExonBIO (22). The probe binding sequence was 50 -UCGUACCGUGAGUAAUAAUGCG-30 . The slides were hybridized with DIG-conjugated hsa-miR-126 probes overnight at 42 C and stained using an Enhanced Sensitive in situ hybridization (ISH) Detection Kit II (AP) with phosphatase-conjugated anti-DIG antibody (Boster) according to the manufacturer's instructions. Statistical analysis Data are presented as the mean SEM of replicate experiments (n 3). Results were analyzed using the SPSS 16.0 software (SPSS) and PRISM 5.0 (GraphPad Software Inc.). The Student unpaired t test or unpaired t test with Welch's correction was used to analyze intergroup differences for two groups, ANOVA was used to analyze more than two groups, and Pearson's correlation coefficient was used to analyze the correlation between groups. The cumulative survival time was calculated using the Kaplan–Meier method and analyzed using the log-rank test. Univariate and multivariate analyses were based on the Cox proportional hazards regression model. P values < 0.05 were considered statistically significant.
Results NGS-based discovery of miR-126 and validation of its downregulation in ESCC To identify ESCC-specific miRNA profiles, NGS-based miRNA profiling was performed on four pairs of frozen ECs and adjacent NMs. When ECs and NMs were compared, 78 miRNAs were
Clin Cancer Res; 21(4) February 15, 2015
855
Liu et al.
differentially expressed, of which 44 were downregulated and 34 were upregulated in ECs (P < 0.05 with fold-change > 3; Fig. 1A; Supplementary Table S2). Among the differentially regulated miRNAs, the downregulation of miR-126-3p in ECs exhibited the strongest statistical significance. For validation, qPCR was used to assess the expression of miR-126 in 30 pairs of frozen ECs and matched NMs, as well as in ESCC cell lines (ECa-109 and KYSE-510) and HEECs, normal human esophageal epithelial cells. The results showed that miR-126 expression was also downregulated in ESCC tissues (P < 0.001; Fig. 1B) and cell lines (P < 0.001; Fig. 1C), when compared with expression in NMs and HEECs, respectively. Downregulation of miR-126 correlates with poor prognosis in ESCC patients To address the clinical significance of miR-126 in ESCC, ISH was performed in 185 ESCC specimens. Patients with negative
miR-126 expression exhibited lower 5-year disease-specific survival rate than that with positive miR-126 expression (P ¼ 0.014, Fig. 1D). In multivariate analysis, CCI, tumor stage, and miR-126 expression were identified as independent prognostic factors in patients' disease-specific survival (Supplementary Table S3). These results suggest that downregulation of miR-126 is an independent indicator of prognosis in ESCC. Hypermethylation of the Egfl7 promoter induces miR-126 downregulation in EC Given that the miR-126 gene is located in an intron of Egfl7 and its expression is reportedly upregulated in cancer cells treated with inhibitors of DNA methylation (23), we analyzed the sequence of Egfl7 and found that a large CpG island was present in the promoter region (Fig. 2A). To understand the correlation between methylation of the Egfl7 promoter and the downregulation of miR-126, we analyzed CpG island methylation in ECs and
Figure 1. NGS-based discovery of miR-126 and validation of its downregulation in ESCC. A, heatmap and sample clustering analysis of the miRNAs differentially expressed in ECs and NMs, and the Pearson correlation coefficient are shown. The expression of miR-126 detected by RNA-Seq was confirmed by qPCR. B, levels of miRNA-126 in 30 matched ECs and NMs were detected using qPCR and normalized to U6 levels. The percentage of ESCC samples with upregulated, downregulated, and unchanged miR-126 expression was analyzed. The qPCR results are shown as the mean SD ( , P < 0.001). C, relative levels of miR-126 were detected in ECa-109, KYSE-510, and HEECs by qPCR and normalized to U6 levels. The mean SD was determined from three replicates ( , P < 0.001). D, ISH was used to detect the presence of miR-126 in ECs. Kaplan–Meier analysis showed that negative miR126 expression was associated with poor disease-specific survival in ECs (P ¼ 0.014, log-rank test).
856 Clin Cancer Res; 21(4) February 15, 2015
Clinical Cancer Research
DNMT1–MicroRNA126 Epigenetic Circuit Regulates ESCC Growth
Figure 2. Hypermethylation of the Egfl7 promoter induces miR-126 downregulation in ESCC. A, the location of the miR-126 gene (MIR 126) in Egfl7 was analyzed. The CpG island consisting of 12 CpG sites was used for bisulfite genomic sequencing (BSP) and MSP analyses. B, BSP was used to determine the methylation ratio at the Egfl7 promoter in ECs (n ¼ 4) and matched NMs (n ¼ 4). Black circles, methylated (M) CpG dinucleotides; open circles, unmethylated (U) CpG dinucleotides. Data shown are the mean SD ( , P < 0.01). C, the correlation between miR-126 levels (bars) and Egfl7 promoter methylation levels (lines) was analyzed in the four pairs of esophageal samples. D, Egfl7 promoter methylation levels and miR126 expression were determined by MSP and qPCR, respectively, both in ECa-109 and KYSE-510 cells treated with 5-aza-dC (5 mmol/L) or vehicle. MSP was carried out using the specific primers for the unmethylated (U) or methylated (M) Egfl7 promoter sequence. The mean SD was determined from three replicates ( , P < 0.001).
matched NMs. The results showed that Egfl7 promoter methylation was higher in ECs than in NMs (Fig. 2B). Samples with higher methylation exhibited lower miR-126 expression (Fig. 2C). To determine the effect of methylation of the Egfl7 promoter on the expression of miR-126, genomic DNA was extracted from ECa109 and KYSE-510 cells treated with 5-aza-dC or vehicle. After sodium-bisulfite modification, methylation-specific PCR (MSP) was carried out using the specific primers for the unmethylated (U) or methylated (M) Egfl7 promoter sequence. As a result, compared with vehicle-treated cells in which the Egfl7 promoter was hypermethylated, cells treated with the demethylating agent 5-aza-dC to induce demethylation of the Egfl7 promoter showed more unmethylated sequences and an increase in miR126 (Fig. 2D). These results indicate that hypermethylation of the Egfl7 promoter directly downregulates miR-126 in ESCC. DNMT1–miR-126 regulatory circuit contributes to Egfl7 promoter hypermethylation DNA methyltransferases (DNMT) are critical for gene-specific methylation and the consequent transcriptional silencing of cancer-related genes (24). In this study, we found that DNMT1 expression was dramatically increased in ESCC tumor tissues and cell lines (Fig. 3A; Supplementary Fig. S1A). Furthermore, correlation analysis showed that the expression of DNMT1 inversely correlated with that of endogenous miR-126 (Fig. 3A). To investigate the role of DNMT1 in the regulation of miR-126 expression, we silenced DNMT1 expression in cells using siRNA. DNMT1 knockdown reduced the hypermethylation of the miR-126 host
www.aacrjournals.org
gene (Fig. 3B), leading to miR-126 upregulation (Fig. 3C). Interestingly, overexpression of miR-126 suppressed DNMT1 expression in both KYSE-510 and ECa-109 cells (Fig. 3D). Collectively, overexpression of DNMT1 resulted in the hypermethylation and downregulation of miR-126, whereas overexpression of miR-126 suppressed DNMT1 expression. Our findings indicate the involvement of a novel DNMT1–miR-126 regulatory loop in ESCC. Overexpression of miR-126 suppresses ESCC cell proliferation and migration Given the low expression of miR-126 in ESCC, we further evaluated the biologic role played by miR-126 in ESCC cell lines. miR-126 mimics were transfected into ECa-109 and KYSE-510 cells to upregulate miR-126 expression (Supplementary Fig. S2A). Overexpression of miR-126 markedly suppressed cell proliferation (Fig. 4A) and the rate of colony formation (Fig. 4B) in ECa109 and KYSE-510 cells when compared with the negative control. Further analysis showed that miR-126 mainly promoted cell apoptosis, but without a marked effect on the cell cycle (Supplementary Fig. S2B and C). In addition, overexpression of miR-126 inhibited ESCC cell migration in vitro (Fig. 4C; Supplementary Fig. S3A). To confirm the suppressive role of miR-126 in ESCC, we next investigated the effect of miR-126 on ESCC tumor growth in vivo. ECa-109 and KYSE-510 cells transfected with either miR-126 or miR-NC (control group) were implanted subcutaneously into the bilateral posterior flank of nude mice. Compared with tumor growth in the control group, tumor growth in the
Clin Cancer Res; 21(4) February 15, 2015
857
Liu et al.
Figure 3. DNMT1–miR-126 regulatory circuit contributes to Egfl7 promoter hypermethylation. A, DNMT1 RNA levels in 30 matched ECs and their NMs were detected using qPCR and normalized to GAPDH expression. Data shown are the mean SD values ( , P < 0.01). DNMT1 RNA levels in cancer tissues negatively correlated with miR-126 levels (P ¼ 0.0074; R ¼ 0.4787). DNMT1 protein levels were measured in six pair-matched ESCC samples and in cell lines. B, ECa-109 and KYSE-510 cells were transfected with control RNA (si-Ctrl, 100 nmol/L) or DNMT1 siRNA (si-DNMT1, 100 nmol/L) as indicated. DNA from each sample was extracted, and Egfl7 promoter methylation levels were analyzed using MSP. M, methylated; U, unmethylated. C, miR-126 expression was determined using qPCR and normalized to U6 expression. DNMT1 expression was detected and normalized to GAPDH. The mean SD was determined from three replicates ( , P < 0.001). D, DNMT1 expression in ECa-109 and KYSE-510 cells transfected with miR-126 and miR-NC mimics was determined by Western blotting.
miR-126–transfected group was markedly suppressed, as reflected by the smaller tumor sizes (Fig. 4D, Supplementary Fig. S3B). Collectively, these results indicate that miR-126 functions as a tumor suppressor by inhibiting ESCC cell growth and migration. The ADAM9–EGFR–AKT signaling directly mediates the suppressive role of miR-126 To investigate the targets and downstream signaling pathway of miR-126, we compared mRNA profiles of ECa-109 cells transfected with miR-126 (126-EC cells) or miR-NC (NC-EC cells) mimics using NGS-based mRNA sequencing. In 126-EC cells, 31 genes were significantly downregulated relative to the levels in NC-EC cells (Supplementary Table S4), including ADAM9. Consistently, in bioinformatics analysis, ADAM9, a putative target of miR-126, was identified with four prediction algorithms (PicTar, TargetScan, miRanda, and miRDB; Supplementary Table S4). Therefore, ADAM9 was selected for further validation and functional analyses. Results from qPCR, Western blot analysis, and immunohistochemistry (IHC) showed that ADAM9 was overexpressed in ECs and ESCC cell lines and that its mRNA levels inversely correlated with miR-126 levels in ECs (Fig. 5A-B; Supplementary Fig. S4A). This inverse correlation strongly suggested the involvement of ADAM9 in the miR-126– related network. Bioinformatics analysis showed that the binding sites in the 30 -UTR of ADAM9 and miR-126 are highly conserved (Fig. 5C). We constructed the ADAM9-30 -UTR reporter plasmid and performed luciferase reporter assays in cells transfected with the 30 -UTR plasmid. As expected, miR-126 repressed luciferase activity in cells expressing ADAM9-30 -UTR, indicating a direct interaction between miR-126 and ADAM9 in ESCC cells (Fig. 5C). Furthermore, a decrease in ADAM9 was detected in miR-126–overexpressing ESCC cells (P < 0.05; Fig. 5D). To address the direct involvement of ADAM9
858 Clin Cancer Res; 21(4) February 15, 2015
in miR-126–regulated tumor suppression, ECa-109, KYSE-510, and KYSE-510 cells were transfected with siRNAs against ADAM9 mRNA; si-1 (si-ADAM9) was selected for further study because of its efficient interference (Supplementary Fig. S4B). ESCC cell proliferation was inhibited in ADAM9–siRNA–transfected ECa-109 and KYSE-510 cells and the cell migration rate was lower than in si-Ctrl–transfected cells (Fig. 6A). Moreover, after silencing DNMT1, the expression of ADAM9 in protein level was also reduced, with decreasing cell viability and migration (Fig. 5D; Supplementary Fig. S1B and S1C). These data indicate that ADAM9 is the target of miR-126 and is regulated by the "DNMT1–miR-126 circuit" in ESCC. ADAM9 activates EGFR by cleaving the transmembrane proheparin–binding epidermal growth factor-like growth factor (proHB-EGF) precursor to yield HB-EGF, a soluble EGFR ligand. HBEGF binds and activates intracellular signaling cascades downstream of EGFR, such as the AKT pathway, and thereby promotes the growth of normal and neoplastic cells (25, 26). In our study, we found that the EGFR–AKT pathway was involved in the suppressive action of miR-126/ADAM9 in ESCC. Both overexpression of miR-126 and blockage of ADAM9 reduced levels of p-EGFR and p-AKT (Fig. 6B), cell proliferation, and cell migration decreased (Fig. 6C). These results indicate that the ADAM9– EGFR–AKT pathway directly mediates the suppressive action of miR-126 in ESCC (Fig. 6D).
Discussion The "epigenetic–miRNA loop," wherein miRNAs regulate or are regulated by epigenetic factors, was proposed by Pagano and colleagues (27). They suggested that epigenetic–miRNA loops contributed to hematopoiesis by orchestrating gene expression in hematopoietic cells. In this study, we identified
Clinical Cancer Research
DNMT1–MicroRNA126 Epigenetic Circuit Regulates ESCC Growth
Figure 4. Overexpression of miR-126 suppresses ESCC cell proliferation and migration. A, ECa-109 and KYSE-510 cells were transfected with miR-126 (100 nmol/L) or miR-NC (100 nmol/L) mimics for 24 hours, and cell viability and proliferation were assessed with the CCK-8 assay at 24, 48, 72, and 96 hours. B, cell colonies were counted after transfection for 12 days. The colony formation ratio was calculated as follows: (visible colony number/ original number of cells) 100. C, the total number of migrated cells was calculated 48 hours after transfection. D, ECa-109 or KYSE-510 cells transfected with miR-126 or miR-NC mimics were injected subcutaneously into either side of the posterior flank of the same nude mouse, respectively (n ¼ 4). Photographs show mice and tumors 4 weeks after tumor cell implantation. Tumor sizes are shown as the mean SD. Values represent the mean SD ( , P < 0.01; , P < 0.001).
the DNMT1–miR-126 epigenetic loop in ESCC and investigated its contribution to ESCC development via modulation of the ADAM9–EGFR–AKT pathway. Dysregulation of miRNAs in diseases is frequently due to epigenetic changes in the miRNA genome (28, 29). Previous studies of miRNA genomic sequences have indicated that promoter hypermethylation of miRNAs occurs frequently as a result of the aberrant expression of DNMTs, histone deacetylases, and methyl-CpG–binding proteins (29–31). In the present study, we found that miR126 was downregulated in ECs and that its downregulation was induced by promoter hypermethylation of its host gene. Dysregulation of miR-126 has also been observed in other tumors, such as colon cancer (32) and lung cancer (33), but little is known about its function in ESCC. This is the first report to show that miR-126 is downregulated and acts as a new tumor suppressor in ESCC. Moreover, we found that downregulation of miR-126 is a potential prognostic indicator for ESCC. However, the results was different from the prior report that miRNA126 failed to show a relation to outcome of patients with esophageal adenocarcinoma (34). Such kind of conflict might be due to the distinct biology and clinical features of esophageal squamous carcinoma in our study and esophageal adenocarcinoma in the previous study. The differences in races (Asian vs. American) might also contribute to it. Although evidence from Zhang and colleagues indicated that the downregulation of miR-126 in breast cancer positively correlated with hypermethylation of the Egfl7 T2 promoter (35),
www.aacrjournals.org
the underlying molecular mechanism was not elucidated. Notably, our results demonstrate that DNMT1 is overexpressed in primary ECs and cell lines and that DNMT1 responsible for the hypermethylation of the Egfl7 promoter, leading to miR-126 downregulation. DNMT1, a member of the DNMT family (DNMT1, DNMT3A, and DNMT3B), is overexpressed in tumors and is essential for maintaining promoter hypermethylation of tumor-suppressive genes, leading to their silencing (36). In breast cancer, overexpressed DNMT1 induces miR-148a/152 silencing by maintaining the hypermethylation of their promoters and contributes to tumor transformation and angiogenesis (37). In ESCC, we demonstrate that overexpressed DNMT1 was responsible for the promoter hypermethylation of the miR-126 host gene, thereby downregulating miR-126 and inhibiting ESCC growth and cell migration. Interestingly, in a feedback mechanism, DNMT1 was suppressed by miR-126 overexpression in ESCC cells, indicating the presence of a novel "DNMT1–miR-126 epigenetic circuit." We also showed that DNMT1 was the direct target of miR-126 (Supplementary Fig. S5). Regulation of DNMT1 by miR-126 has also been described in CD4þ T cells from patients with systemic lupus erythematosus (37). Similar to the "DNMT1–miR-126 epigenetic circuit" in ESCC, a negative feedback regulatory loop between miR-148/152 and DNMT1 has been reported in breast cancer (38). Thus, epigenetic–miRNA loops are widely present in various tumors. Given that both miRNA and epigenetic modulations strongly regulate
Clin Cancer Res; 21(4) February 15, 2015
859
Liu et al.
Figure 5. ADAM9 is the target of miR-126 in ESCC. A, ADAM9 RNA levels in 30 matched esophageal cancer tissues and their adjacent normal tissues were determined using qPCR and normalized to GAPDH expression. Data shown are the mean SD values ( , P < 0.001). ADAM9 RNA levels in cancer tissues negatively correlated with endogenous miR-126 levels (P ¼ 0.0019, R ¼ 0.5446, n ¼ 30). B, ADAM9 protein levels were measured in six matched ESCC samples and three cell lines by Western blotting. IHC straining for ADAM9 was performed in ECs and matched NMs. Magnification, 200; 400. C, alignment of miR-126 with the ADAM9 30 -UTR. MiR-126 (100 nmol/L) or miR-NC (100 nmol/L) mimics were cotransfected with the ADAM9-30 -UTR reporter plasmid, along with a control Renilla luciferase pRL-TK vector, into ECa-109 or KYSE-510 cells. The relative luciferase activity in each group was analyzed 48 hours later and expressed relative to the activity in the miR-NC–transfected group. Data are representative of three independent experiments. Values represent the mean SD ( , P < 0.001). D, ADAM9 expression in ECa-109 and KYSE-510 cells transfected with miR-126 mimics or siRNA against DNMT1 (si-D) was determined by Western blotting.
the expression of various disease-associated genes (14), controlling the dynamic balance of epigenetic–miRNA loops might be of great clinical significance. The expression of ADAM9 has been reported to correlate with cell adherence and migration (25, 33, 39, 40). This study found that ADAM9 functioned as a direct target of miR-126 and contributed to miR-126 repressing cell migration in ESCC. Consistently, the targeting of miR-126 to ADAM9 and the role of ADAM9 in cell migration was also reported in a previous study by Hamada and colleagues (39). A similar approach with our study using restoration of miR-126 and silencing of ADAM9 showed inhibition of the invasive growth. To systematically investigate the target of miR-126 in ESCC, RNA sequencing was applied in the present study to examine the differential gene expression in miR-126–overexpressed ESCC EC-109 cells versus control cells. ADAM9 was showed to be a functional target of miR-126, whereas other potential targets are under exploration. Therefore, apart from pancreatic cancer in Hamada's study, this study also identified the interaction between miR-126 and ADAM9 in ESCC. In addition, we investigated the downstream pathway and found that ADAM9 was critical for promoting cell proliferation in ESCC by targeting EGFR–AKT signaling. ADAM9 cleaved the pro-HB-EGF precursor to yield soluble HB-EGF, which binds to EGFR, thereby activating downstream signal transduction (26, 41). The EGFR– AKT pathway is important for cell growth and migration (42, 43). Our results showed that blocking ADAM9 inhibits
860 Clin Cancer Res; 21(4) February 15, 2015
EGFR–AKT activation by suppressing phosphorylation, but has no effect on ERK phosphorylation. We also found that the expression of phosphoinositide-3-kinase, regulatory subunit 2 (beta) was downregulated upon miR-126 overexpression (Supplementary Table S4 and Supplementary Fig. S6A). However, its expression in ECs and in NMs did not differ (P > 0.05), and its inverse correlation with miR-126 expression was poor (R ¼ 0.1791; P ¼ 0.5255; Supplementary Fig. S6B and S6C). Other growth-related targets of miR-126 identified in other tumors, such as Crk (44) and IRS (45), were not downregulated in miR-126–overexpressing ECa-109 cells (Supplementary Fig. S6A), implying they are not key targets of miR-126 in ESCC growth. With accumulating evidence demonstrating the importance of miRNAs in cancer biology, miRNA-based cancer therapy emerges as a hot issue and also a big challenge in the current miRNA research. To be excited, MRX34, as the first microRNA-based therapy for cancer, is entering the phase I trial (46). In the present study, our primary data showed that the exogenous miR-126 mimics suppressed ESCC cell growth, which implies its potential therapeutic significance in ESCC using a miRNAs' "replacement" strategy. However, miR-126–based therapy for ESCC in practice still needs to be further explored, considering some challenges, especially technical limitations, such as how to improve its accumulation in the target tissues and balance the safety and efficiency. In general, there is a long way ahead to realize the translation of fundamental research of miR-126 to clinical applications in ESCC.
Clinical Cancer Research
DNMT1–MicroRNA126 Epigenetic Circuit Regulates ESCC Growth
Figure 6. The ADAM9–EGFR–AKT signaling directly mediates the suppressive role of miR-126 in ESCC. A, ECa-109 and KYSE-510 cells were transfected with control siRNA (si-Ctrl, 100 nmol/L) or ADAM9 siRNA (si-ADAM9, 100 nmol/L) as indicated, and cell viability was assessed at different time points. Cell migration ability was determined with the Transwell assay, and the migrated cell number was calculated. B, ADAM9, pEGFR, EGFR, pAKT, AKT, pERK, and ERK protein levels were determined by Western blotting and compared with GAPDH levels. C, ECa109 cells were treated with an EGFR inhibitor (AG1478, 10 mmol/L), and pEGFR, EGFR, pAKT, and AKT protein levels in ECa-109 cells were analyzed by Western blotting. Cell viability and cell migration were determined at different time points after treatment with AG1478 (10 mmol/L). The data are representative of three independent experiments. D, diagram summarizing the findings. Values represent the mean SD ( , P < 0.05; , P < 0.01; , P < 0.001).
In summary, this study proposes a "DNMT1–miR-126 epigenetic circuit" in ESCC and sheds light on its importance for ESCC progression via regulation of ADAM9–EGFR–AKT signaling. Altering critical gene expression and signal transduction by modulating miR-126–related epigenetic dynamics might be of clinical significance in ESCC prediction and therapy.
Disclosure of Potential Conflicts of Interest No potential conflicts of interest were disclosed.
Authors' Contributions Conception and design: R. Liu, P. Jiang, Y. Zheng, D. Ge, Y. Chu Development of methodology: R. Liu, J. Gu, P. Jiang, X. Liu, X. Jiang, E. Huang, F. Xu Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R. Liu, J. Gu, X. Liu, E. Huang, Y. Chu Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R. Liu, J. Gu, P. Jiang, X. Jiang, F. Xu, G. Liu, Y. Chu
www.aacrjournals.org
Writing, review, and/or revision of the manuscript: R. Liu, J. Gu, Y. Zheng, X. Liu, S. Xiong, G. Liu, Y. Chu Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): R. Liu, J. Gu, P. Jiang, Y. Zheng, Y. Chu Study supervision: Y. Zheng, D. Ge, Y. Chu
Grant Support This work was supported by the National Science Foundation of China (81273215, 81372313, 81401876, and 81272259), the Major Research Plan of the National Natural Science Foundation of China (91229110), the National Basic Research Program of China (973 Programs, 2011CB910404), and the Specialized Research Fund for the Doctoral Program of Higher Education (20120071110046). 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 accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received July 10, 2014; revised October 9, 2014; accepted November 21, 2014; published OnlineFirst December 15, 2014.
Clin Cancer Res; 21(4) February 15, 2015
861
Liu et al.
References 1. Holmes RS, Vaughan TL. Epidemiology and pathogenesis of esophageal cancer. Semin Radiat Oncol 2007;17:2–9. 2. Yamasaki M, Miyata H, Miyazaki Y, Takahashi T, Kurokawa Y, Nakajima K, et al. Evaluation of the nodal status in the 7th edition of the UICC-TNM classification for esophageal squamous cell carcinoma: proposed modifications for improved survival stratification: impact of lymph node metastases on overall survival after esophagectomy. Ann Surg Oncol 2014; 21:2850–6. 3. Zhang W, Zhu H, Liu X, Wang Q, Zhang X, He J, et al. Epidermal growth factor receptor is a prognosis predictor in patients with esophageal squamous cell carcinoma. Ann Thorac Surg 2014;98:513–9. 4. Wang G, Zhang W, Jiang W, Luan L. Overexpression of her-2 associated with the progression of esophageal cancer patients. Hepatogastroenterology 2013;60:1972–8. 5. Xiong S, Zheng Y, Jiang P, Liu R, Liu X, Chu Y. MicroRNA-7 inhibits the growth of human non-small cell lung cancer A549 cells through targeting BCL-2. Int J Biol Sci 2011;7:805–14. 6. Yu X, Li H, Jiang X, Guo L, Jiang W, Lu SH. MiR-203 inhibits the proliferation and self-renewal of esophageal cancer stem-like cells by suppressing stem renewal factor Bmi-1. Stem Cells Dev 2013; 23:576–85. 7. Matsushima K, Isomoto H, Yamaguchi N, Inoue N, Machida H, Nakayama T, et al. MiRNA-205 modulates cellular invasion and migration via regulating zinc finger E-box binding homeobox 2 expression in esophageal squamous cell carcinoma cells. J Transl Med 2011;9:30. 8. Ohashi S, Natsuizaka M, Naganuma S, Kagawa S, Kimura S, Itoh H, et al. A NOTCH3-mediated squamous cell differentiation program limits expansion of EMT-competent cells that express the ZEB transcription factors. Cancer Res 2011;71:6836–47. 9. Di Leva G, Croce CM. miRNA profiling of cancer. Curr Opin Genet Dev 2013;23:3–11. 10. Zheng Y, Xiong S, Jiang P, Liu R, Liu X, Qian J, et al. Glucocorticoids inhibit lipopolysaccharide-mediated inflammatory response by downregulating microRNA-155: a novel anti-inflammation mechanism. Free Radic Biol Med 2012;52:1307–17. 11. Fehniger TA, Wylie T, Germino E, Leong JW, Magrini VJ, Koul S, et al. Nextgeneration sequencing identifies the natural killer cell microRNA transcriptome. Genome Res 2010;20:1590–604. 12. Tam S, de Borja R, Tsao MS, McPherson JD. Robust global microRNA expression profiling using next-generation sequencing technologies. Lab Invest 2014;94:350–8. 13. Kong KL, Kwong DL, Chan TH, Law SY, Chen L, Li Y, et al. MicroRNA-375 inhibits tumour growth and metastasis in oesophageal squamous cell carcinoma through repressing insulin-like growth factor 1 receptor. Gut 2012;61:33–42. 14. Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nature reviews Genetics 2002;3:415–28. 15. Sun X, He Y, Huang C, Ma TT, Li J. The epigenetic feedback loop between DNA methylation and microRNAs in fibrotic disease with an emphasis on DNA methyltransferases. Cell Signal 2013;25:1870–6. 16. Iorio MV, Piovan C, Croce CM. Interplay between microRNAs and the epigenetic machinery: an intricate network. Biochim Biophys Acta 2010;1799:694–701. 17. Tougeron D, Di Fiore F, Thureau S, Berbera N, Iwanicki-Caron I, Hamidou H, et al. Safety and outcome of definitive chemoradiotherapy in elderly patients with oesophageal cancer. Br J Cancer 2008;99: 1586–92. 18. Talsma K, van Hagen P, Grotenhuis BA, Steyerberg EW, Tilanus HW, van Lanschot JJ, et al. Comparison of the 6th and 7th Editions of the UICCAJCC TNM Classification for Esophageal Cancer. Ann Surg Oncol 2012; 19:2142–8. 19. Lu CL, Guo J, Gu J, Ge D, Hou YY, Lin ZW, et al. CXCR4 heterogeneous expression in esophageal squamous cell cancer and stronger metastatic potential with CXCR4-positive cancer cells. Dis Esophagus 2014;27: 294–302. 20. Hou J, Lin L, Zhou W, Wang Z, Ding G, Dong Q, et al. Identification of miRNomes in human liver and hepatocellular carcinoma reveals miR199a/b-3p as therapeutic target for hepatocellular carcinoma. Cancer Cell 2011;19:232–43.
862 Clin Cancer Res; 21(4) February 15, 2015
21. Gu J, Ding JY, Lu CL, Lin ZW, Chu YW, Zhao GY, et al. Overexpression of CD88 predicts poor prognosis in non-small-cell lung cancer. Lung Cancer 2013;81:259–65. 22. Li J, Li X, Li Y, Yang H, Wang L, Qin Y, et al. Cell-specific detection of miR-375 downregulation for predicting the prognosis of esophageal squamous cell carcinoma by miRNA in situ hybridization. PLoS ONE 2013;8:e53582. 23. Felli N, Felicetti F, Lustri AM, Errico MC, Bottero L, Cannistraci A, et al. miR126&126 restored expressions play a tumor suppressor role by directly regulating ADAM9 and MMP7 in melanoma. PLoS ONE 2013;8:e56824. 24. Di Ruscio A, Ebralidze AK, Benoukraf T, Amabile G, Goff LA, Terragni J, et al. DNMT1-interacting RNAs block gene-specific DNA methylation. Nature 2013;503:371–6. 25. Fry JL, Toker A. Secreted and membrane-bound isoforms of protease ADAM9 have opposing effects on breast cancer cell migration. Cancer Res 2010;70:8187–98. 26. Itabashi H, Maesawa C, Oikawa H, Kotani K, Sakurai E, Kato K, et al. Angiotensin II and epidermal growth factor receptor cross-talk mediated by a disintegrin and metalloprotease accelerates tumor cell proliferation of hepatocellular carcinoma cell lines. Hepatol Res 2008; 38:601–13. 27. Pagano F, De Marinis E, Grignani F, Nervi C. Epigenetic role of miRNAs in normal and leukemic hematopoiesis. Epigenomics 2013;5:539–52. 28. Bernstein BE, Meissner A, Lander ES. The mammalian epigenome. Cell 2007;128:669–81. 29. Weber B, Stresemann C, Brueckner B, Lyko F. Methylation of human microRNA genes in normal and neoplastic cells. Cell Cycle 2007;6: 1001–5. 30. Meng X, Wu J, Pan C, Wang H, Ying X, Zhou Y, et al. Genetic and epigenetic down-regulation of microRNA-212 promotes colorectal tumor metastasis via dysregulation of MnSOD. Gastroenterology 2013;145:426–36 e1–6. 31. Roy SS, Gonugunta VK, Bandyopadhyay A, Rao MK, Goodall GJ, Sun LZ, et al. Significance of PELP1/HDAC2/miR-200 regulatory network in EMT and metastasis of breast cancer. Oncogene 2013;33:3707–16. 32. Liu Y, Zhou Y, Feng X, An P, Quan X, Wang H, et al. MicroRNA-126 functions as a tumor suppressor in colorectal cancer cells by targeting CXCR4 via the AKT and ERK1/2 signaling pathways. Int J Oncol 2014;44:203–10. 33. Crawford M, Brawner E, Batte K, Yu L, Hunter MG, Otterson GA, et al. MicroRNA-126 inhibits invasion in non-small cell lung carcinoma cell lines. Biochem Biophys Res Commun 2008;373:607–12. 34. Hu Y, Correa AM, Hoque A, Guan B, Ye F, Huang J, et al. Prognostic significance of differentially expressed miRNAs in esophageal cancer. Int J Cancer 2011;128:132–43. 35. Zhang Y, Yang P, Sun T, Li D, Xu X, Rui Y, et al. miR-126 and miR-126 repress recruitment of mesenchymal stem cells and inflammatory monocytes to inhibit breast cancer metastasis. Nat Cell Biol 2013; 15:284–94. 36. Jin B, Robertson KD. DNA methyltransferases, DNA damage repair, and cancer. Adv Exp Med Biol 2013;754:3–29. 37. Xu Q, Jiang Y, Yin Y, Li Q, He J, Jing Y, et al. A regulatory circuit of miR-148a/ 152 and DNMT1 in modulating cell transformation and tumor angiogenesis through IGF-IR and IRS1. J Mol Cell Biol 2013;5:3–13. 38. Zhao S, Wang Y, Liang Y, Zhao M, Long H, Ding S, et al. MicroRNA-126 regulates DNA methylation in CD4þ T cells and contributes to systemic lupus erythematosus by targeting DNA methyltransferase 1. Arthritis Rheum 2011;63:1376–86. 39. Hamada S, Satoh K, Fujibuchi W, Hirota M, Kanno A, Unno J, et al. MiR126 acts as a tumor suppressor in pancreatic cancer cells via the regulation of ADAM9. Mol Cancer Res 2012;10:3–10. 40. Shintani Y, Higashiyama S, Ohta M, Hirabayashi H, Yamamoto S, Yoshimasu T, et al. Overexpression of ADAM9 in non-small cell lung cancer correlates with brain metastasis. Cancer Res 2004;64:4190–6. 41. Bakken AM, Protack CD, Roztocil E, Nicholl SM, Davies MG. Cell migration in response to the amino-terminal fragment of urokinase requires epidermal growth factor receptor activation through an ADAM-mediated mechanism. J Vasc Surg 2009;49:1296–303. 42. deGraauw M, Cao L, Winkel L, van Miltenburg MH, le Devedec SE, Klop M, et al. Annexin A2 depletion delays EGFR endocytic trafficking via cofilin
Clinical Cancer Research
DNMT1–MicroRNA126 Epigenetic Circuit Regulates ESCC Growth
activation and enhances EGFR signaling and metastasis formation. Oncogene 2013;33:2610–9. 43. Li Y, Huang X, Zhang J, Ma K. Synergistic inhibition of cell migration by tetraspanin CD82 and gangliosides occurs via the EGFR or cMet-activated Pl3K/Akt signalling pathway. Int J Biochem Cell Biol 2013;45: 2349–58. 44. Liu LY, Wang W, Zhao LY, Guo B, Yang J, Zhao XG, et al. Mir-126 inhibits growth of SGC-7901 cells by synergistically targeting the oncogenes
www.aacrjournals.org
PI3KR2 and Crk, and the tumor suppressor PLK2. Int J Oncol. 2014;45:1257–65. 45. Zhou Y, Feng X, Liu YL, Ye SC, Wang H, Tan WK, et al. Down-regulation of miR-126 is associated with colorectal cancer cells proliferation, Migration and invasion by targeting IRS-1 via the AKT and ERK1/2 signaling pathways. PLoS ONE 2013;8:e81203. 46. Bouchie A. First microRNA mimic enters clinic. Nat Biotechnol 2013; 31:577.
Clin Cancer Res; 21(4) February 15, 2015
863
