- Email: [email protected]
Epigenetic regulation on the gene expression signature in esophagus adenocarcinoma
Accepted Manuscript Title: Epigenetic regulation on the gene expression signature in esophagus adenocarcinoma Author: Ting Xi Guizhi Zhang PII: DOI: Reference:
S0344-0338(16)30381-8 http://dx.doi.org/doi:10.1016/j.prp.2016.12.007 PRP 51694
To appear in: Received date:
25-8-2016
Please cite this article as: Ting Xi, Guizhi Zhang, Epigenetic regulation on the gene expression signature in esophagus adenocarcinoma, Pathology - Research and Practice http://dx.doi.org/10.1016/j.prp.2016.12.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Epigenetic regulation on the gene expression signature in esophagus adenocarcinoma Ting Xia, Guizhi Zhangb,* a
Department of Gastroenterology, First People’s Hospital of Liaocheng of Shandong
Province b
Department of Gastroenterology, Second People’s Hospital of Liaocheng of
Shandong Province *
Correspondence to: Guizhi Zhang
Department of Gastroenterology, Second People’s Hospital of Liaocheng of Shandong Province, No.306 JianKang Street, Linqing 252600, China Email: [email protected] Tel: 13676357009 Ting Xi Email: [email protected]
ABSTRACT Background: Understanding the molecular mechanisms represents an important step for the development of diagnostic and therapeutic measures of esophagus adenocarcinoma (NOS). The objective of this study is to identify the epigenetic regulation on gene expression in NOS, shedding light on the molecular mechanisms of NOS. Methods: In this study, 78 patients with NOS were included and the data of mRNA, miRNA and DNA methylation of were downloaded from The Cancer Genome Atlas (TCGA). Differential analysis between NOS and controls was performed in terms of gene expression, miRNA expression, and DNA methylation. Bioinformatic analysis was followed to explore the regulation mechanisms of miRNA and DNA methylationon gene expression. Results: Totally, up to 1320 differentially expressed genes (DEGs) and 32 differentially expressed miRNAs were identified. 240 DEGs that were not only the target genes but also negatively correlated with the screened differentially expressed miRNAs. 101 DEGs were found to be highlymethylated in CpG islands. Then, 8 differentially methylated genes (DMGs) were selected, which showed down-regulated expression in NOS. Among of these genes, 6 genes including ADHFE1, DPP6， GRIA4，CNKSR2，RPS6KA6 and ZNF135 were target genes of differentially expressed miRNAs (hsa-mir-335, hsa-mir-18a, hsa-mir-93, hsa-mir-106b and hsa-mir-21). Conclusions: The identified altered miRNA, genes and DNA methylation site may be
applied as biomarkers for diagnosis and prognosis of NOS. Keywords: esophagus adenocarcinoma, differentially expressed genes, miRNAs, DNA methylation 1. Introduction There are two main kinds of esophageal carcinoma: esophagus adenocarcinoma and esophagus squamous cell carcinoma. Esophagus adenocarcinoma (NOS) is one of the fastest rising cancers. Up to now, the reason of increasing occurrence of NOS is not entirely known but the principal risk factors are male, visceral obesity and chronic reflux disease [1]. Chemotherapy and surgery are common therapeutic methods for patients with NOS. It is pointed out that NOS is resistant to chemotherapeutic regimens and is characterized by a poor survival outcome [2]. Recently, microRNAs (miRNAs) were found to be candidates as biomarkers for diagnosis and prognosis in different cancers [3, 4]. MiRNAs are small, well-conserved and non-coding RNA molecules that regulate the expression of mRNAs [5]. Generally, miRNAs inhibit gene expression through targeting the complementary mRNA and blocking its translation [6]. It is well known that miRNAs is tissue-specific in terms of expression and function [7, 8].They can function as oncomir as well as tumor suppressors, and are involved in various biological processes, including proliferation, differentiation, and apoptosis [9]. Some miRNAs have already been identified in NOS, for instance, miR-21, miR-233, miR-192 and miR-194 was up-regulated, and the expression patterns were validated by qRT-PCR [10]. This further suggested that miRNAs plays an important role in the diagnosis and
prognosis of NOS. Epigenetic modifications such as DNA methylation can modify gene expression patterns and control amounts of cellular functions [11]. Aberrant hypermethylation has been associated with inactivation of tumor-related genes in a large-scale of human neoplasms [12]. DNA methylation and miRNA regulation on gene expression both belong to the domain of epigenetic modification. So, it is important to analyze the correlation between transcriptional sequencing data and epigenetic modification data to illustrate the explicit epigenetic regulation mechanisms of NOS. In this study, the comprehensive analysis of mRNA, miRNA and DNA methylation profiling data of NOS from The Cancer Genome Atlas (TCGA) was performed. Differentially expressed genes (DEGs), differentially expressed miRNAs and differentially methylated genes (DMGs) were identified in NOS. Function enrichment analysis of DEGs and DMGs was then performed. DEGs associated with differential methylation or targeting by differentially expressed miRNAs were identified. At last, 6 candidate genes may be used in the potential targeted treatment of NOS, and further study was needed to investigate their roles in NOS carcinogenesis. 2. Materials and Methods 2.1. Datasets In
this
study,
we
downloaded
the
mRNA
expression
(UNC
IlluminaHiSeqRNASeq), miRNA expression (BCGSC_IlluminaHiSeq_miRNASeq) and DNA methylation data (JHU_USC_HumanMethylation450) of all 185 patients with NOS from TCGA data portal (http://tcga-data.nci.nih.gov/) (January 2016).
According to providing clinical information, we selected 78 NOS patients without histories of other malignancy or neoadjuvant treatment. Among which, 68 men and 10 women were included and 79% were the white race. Moreover, the age range of all individuals was 27 to 86 and the average age is 65. Additionally, the tumor degree was classified as G1 (2 cases), G2 (25 cases), G3 (27 cases) and GX (24 cases). Detailed information of these 78 patients was showed in supplementary 1. 2.2. Analysis of DEGs and differentially expressed miRNAs The DEGs and differentially expressed miRNAs were evaluated in the R-bioconductor package DESeq [13]. The Limma package in R was used to calculate p-values by two-tailed Student's t-test. MetaMA package in R was used to combine p-values, and the false discovery rate (FDR) was obtained from multiple comparisons using the Benjamini and Hochberg method [14]. We selected DEGs and differentially expressed miRNAs with criterion of FDR < 0.0001. 2.3. Correlation analysis of DEGs and differentially expressed miRNAs First, pairwise Pearson correlation coefficients between differentially expressed miRNAs and DEGs was calculated, and P < 0.05was defined as statistical significance. Second, six miRNA-target prediction tools (RNA22, miRanda, miRDB, miRWalk, PICTAR2 and Targetscan) were utilized to predict target genes of differentially expressed miRNAs. Only those miRNA-target pairs which were predicted by more than four algorithms can be selected out. The miRNA-targets pairs verified by experiment in miRWalk database were also screened out. Finally, we selected the miRNA-target pairs with negative correlations (p < 0.05, r < 0) to establish the
miRNA-target regulatory network, which was visualized using Cytoscape software [15]. 2.4. Functional annotation of target genes of differentially expressed miRNAs To acquire the biological function of the miRNA target genes, we conducted Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis based on the online software GENECODIS [16]. FDR < 0.05 was set as the cut-off for selecting significantly enriched functional KEGG pathways. 2.5. Influence of DNA methylation on DEGs The COHCAP package in R (https://sourceforge.net/projects/cohcap) [17] was used to identify the differential methylation sites between tumors and normal tissues that are more likely to regulate downstream gene expression. FDR < 0.05 was considered as differentially methylated CpG sites. Based on the gene expression of NOS, we identified the aberrant DNA methylated CpG site which affected corresponding gene expression. 3. Results 3.1. Gene expression pattern in NOS Totally, up to 78 tissue samples were included in the present study, all of which with fully characterized mRNA profiles, miRNA profiles and DNA methylation data. The TCGA barcode IDs for each sample were showed in Table 1. There were 1320 DEGs with FDR < 0.0001 between NOS and normal tissue. Among these genes, 594 genes were up-regulated and 726 genes were down-regulated. The heatmap of top fifty DEGs was showed in Fig. 1. The top ten up-regulated genes were ESM1,
LOC541471, CKS2, GABRD, MMP11, MMP3, LINC00152, TPX2, IL8 and NUF2. By contrast, the top ten down-regulated genes were LOC388387, GPR155, DQ599327, RGMB, SIK2, AK131020, FAM165B, C22orf23, NDE1 and AK055981. 3.2. Correlations of differentially expressed miRNAs and DEGs in NOS The screened differentially expressed miRNAs were presented in Table 2. There were 32 differentially expressed miRNAs including 14 up-regulated and 18 down-regulated miRNAs. To investigate the correlations between differentially expressed miRNAs and DEGs, we first performed the correlation analysis. Depending on the analysis, we obtained 3112 miRNA-mRNA pairs which were negatively correlated (P < 0.05, r < 0). Then, we got 1828 miRNA-mRNA pairs through miRNA-target prediction algorithms. At last, we screened out 240 DEGs including 160 up-regulated and 80 down-regulated genes, all of which were not only negatively correlated with corresponding miRNA expression, but also predicted by bioinformatics methods. The established miRNA regulatory network was showed in Fig. 2. To understand the potential function of these miRNA targets genes, we conducted KEGG pathway analysis. The results demonstrated that these DEGs significantly enriched in propanoate metabolism, valine, leucine and isoleucine degradation, valine, leucine and isoleucine degradation, DNA replication, cell cycle, homologous recombination, mismatch repair, oocyte meiosis, p53 signaling pathway, pyrimidine metabolism, pathways in cancer, tuberculosis and systemic lupus erythematous. 3.3. Influence of DNA methylationon DEGs in NOS
The methylation of CpG islands in the promoter regions generally results in the gene silencing. There were 116 differentially methylated CpG islands with FDR < 0.05 (The methylation levels of all sites in the CpG islands were averaged), which involved 101 genes with hypermethylation. To investigate the influence of DNA methylation on gene expression, we compared the list of differentially methylated genes and differentially expressed genes to identify the differentially methylated genes
displaying
differential
expression.
Based
on
the
fact
that
DNA
hypermethylation suppressed corresponding gene expression, we screened out 8 down-regulated genes from 101 DMGs, including GPRASP1, DPP6, TNXB, GRIA4, ADHFE1, CNKSR2, RPS6KA6, ZNF135, which were showed in Table 3. 3.4. Key genes upon regulation of miRNA and DNA methylation We combined the results of miRNA regulation and DNA methylation regulation on mRNA expression, and found that six DEGs (ADHFE1, DPP6, GRIA4, CNKSR2, RPS6KA6 and ZNF135) were under miRNA and DNA methylation regulation, which involved five differentially expressed miRNAs (hsa-mir-335, hsa-mir-18a, hsa-mir-93, hsa-mir-106b and hsa-mir-21). The correlation between these five differentially expressed miRNAs and target genes in NOS was showed in Fig. 3. Among six genes, ADHFE1 expression was negatively correlated with that of five miRNAs. 4. Discussion A critical step in fighting against NOS is to identify changes of genes in different levels occurring in tumor development. In this study, we integrated mRNA expression, miRNA expression data and DNA methylation profiles of NOS to find the valuable
genes which were highly associated with NOS. 1320 genes were aberrantly expressed in NOS compare with normal tissues. To investigate the epigenetic modifications of gene regulation, we assessed the miRNA expression data, and identified 32 differentially expressed miRNAs. Depending on the miRNA-mRNA expression correlation analysis and miRNA-target prediction, we obtained 240 DEGs in NOS, which were negatively correlated with miRNAs. Depending on the DNA methylation-mRNA expression correlation analysis, we obtained 8 differentially expressed genes which were hypermethylated. After combining the results of miRNA regulation and DNA methylation regulation on gene expression, we finally selected 6 genes, all of which were not only targeting by correlated differentially expressed miRNAs, but also down-regulated under hypermethylation, which involving 5 up-regulated miRNAs. The results suggested that these genes under epigenetic regulation may be associated with carcinogenesis of NOS. Alcohol dehydrogenase, iron containing 1 (ADHFE1) was cloned from a human fetal brain cDNA library [18]. Recently, aberrant methylation in ADHFE1 promoter was associated with colorectal carcinomas [19, 20]. It is pointed out that hypermethylation of ADHFE1 is associated with colorectal cancer differentiation [21]. Dipeptidyl peptidase like 6 (DPP6) is a protein which regulates various biological functions and also plays a role in carcinogenesis [22, 23]. DPP6 can maintain cell-specific phenotype and its deregulation will result in carcinogenesis [24]. In addition, hypermethylation and reduced expression of DPP6 is found in melanoma
[25] and acute myeloid leukemia (AML) patients [26]. These studies indicate that DPP6 is tightly regulated during neoplastic development. In this study, we found that ADHFE1 and DPP6 were all down-regulated and high methylated, and the expression of ADHFE1 and DPP6 was significantly inversely related to hsa-mir-93. Our results provided evidences for their important roles in NOS. Further study may be needed to verify their crucial functions in NOS. Glutamate ionotropic receptor AMPA type subunit 4 (GRIA4, also called GLUR4) is the subunit of amino-3-hydroxy-5-methyl-4isoxazolepropionic acid (AMPA) receptor in the brain. It is indicated that AMPA receptor activity has been related to neuronal cancer cell migration [27-29]. GRIA4 is important for viability and invasion of distinct renal cell carcinoma [30]. In this study we found that GRIA4 was down-regulated in NOS, and the regulatory miRNAs, hsa-mir-18a and has-mir-106b, were up-regulated. The results suggested that epigenetic modification of GRIA4 in terms of miRNA regulation and DNA methylation modification is linked to NOS progression. Connector enhancer of kinase suppressor of Ras 2 (CNKSR2, also called CNK2 or KSR2) regulates glycolysis and oxidative metabolism in tumor cells [31]. It is reported that CNKSR2 can enhance the proliferative rate of mouse embryo fibroblasts, and CNKSR2 depletion reduces tumor cell growth, which indicates that CNKSR2 is a regulator of cellular metabolism affecting the tumorigenic potential of cells [31]. Ribosomal protein S6 kinase A6 (RPS6KA6, also called RSK4) is a tumor suppressor, and RPS6KA6 mutation is an unusual event in cancer. It has been noted that the
expression of RPS6KA6 reduced in colon, breast, colonic, renal and endometrial carcinomas [32-36]. Additionally, the high frequency of the CpG island methylated phenotype can also be found in endometrial carcinomas [36]. RPS6KA6 contributions to cancer phenotypes could be through regulation of cell cycle arrest and stress responses [33], tumor invasiveness and metastasis [34, 37]. Accordingly, we found that both CNKSR2 and RPS6KA6 were down-regulated and hypermethylated in NOS. Interestingly, CNKSR2 and RPS6KA6 were target genes of hsa-mir-21, and hsa-mir-21 is up-regulated with the expression trend opposite with its target genes CNKSR2 and RPS6KA6. It is a known fact that hsa-mir-21 is an oncomir, and is up-regulated in NOS [38]. It is reported that has-mir-21 can serve as diagnostic biomarker of NOS. In a word, the expression changes of CNKSR2 and RPS6KA6 under the regulation of has-mir-21 illuminated the mechanism of carcinogenesis which may play a crucial role in the progress of NOS. Zinc finger protein 135 (ZNF135) is a transcription factor which involved in angiogenesis and wound repair [39]. It is reported that ZNF135 is high methylation in breast cancer, colon cancer, glioblastoma, head and neck squamous carcinoma, clear cellrenal carcinoma, lung adenocarcinoma, lung squamous carcinoma, serous ovarian cancer, rectal cancer, endometrial carcinoma sites [40]. In this study, we found that ZNF135 was down-regulated and hypermethylated in NOS. ZNF135 is the target gene of hsa-mir-335, hsa-mir-93 and hsa-mir-106b, and all of these miRNAs were up-regulated. This further indicated that ZNF135 may be related to the development of NOS.
In brief, integrated analysis of the TCGA data led to a pool of DEGs, differentially expressed miRNAs and DMGs. The epigenetic modifications via miRNAs and methylation for ADHFE1, DPP6, GRIA4, CNKSR2, RPS6KA6 and ZNF135 may be involved in growth and tumorigenesis of NOS. The study of epigenetic alterations of these genes is of value to investigate the pathogenesis of NOS and have an important role in guiding the development of new therapies for NOS. Conflict of interest None declared. Acknowledgement None. References [1] A. Cowre, F. Noble, T. Underwood, Strategies to improve outcomes in esophageal adenocarcinoma, Expert. Rev. Anticancer. Ther. 14 (2014) 677-87. [2] D.H. Ilson, Epirubicin, cisplatin, and fluorouracil in gastric and esophageal cancer: a step
ahead?, J. Clin. Oncol. 20 (2002) 1962-4.
[3] J. Lu, G. Getz, E.A. Miska, E. Alvarez-Saavedra, J. Lamb, D. Peck, et al., MicroRNA expression profiles classify human cancers, Nature. 435 (2005) 834-8. [4]
P.S.
Mitchell, R.K.
Parkin, E.M.
Kroh, B.R.
Fritz, S.K.
Wyman, E.L.
Pogosova-Agadjanyan, et al., Circulating microRNAs as stable blood-based markers for cancer detection, Proc. Natl. Acad. Sci. U S A. 105 (2008) 10513-8. [5] D.P. Bartel, MicroRNAs: genomics, biogenesis, mechanism, and function, Cell. 116 (2004) 281-97.
[6] Y.W. Kong, D. Ferland-McCollough, T.J. Jackson, M. Bushell, microRNAs in cancer management, Lancet. Oncol. 13 (2012) e249-58. [7] H. Liu, I.S. Kohane, Tissue and process specific microRNA-mRNA co-expression in mammalian development
and malignancy, PLoS. One. 4 (2009)
e5436. [8] T. Babak, W. Zhang, Q. Morris, B.J. Blencowe, T.R. Hughes, Probing microRNAs with microarrays: tissue specificity and functional inference, RNA. 10 (2004) 1813-9. [9] M. Seven, O.F. Karatas, M.B. Duz, M. Ozen, The role of miRNAs in cancer: from pathogenesis to therapeutic implications, Future. Oncol. 10 (2014) 1027-48. [10] E.A. Mathé, G.H. Nguyen, E.D. Bowman, Y. Zhao, A. Budhu, A.J. Schetter, et al., MicroRNA expression in squamous cell carcinoma and adenocarcinoma of the esophagus: associations with survival, Clin. Cancer. Res. 15 (2009) 6192-200. [11] H. Wu, Y.E. Sun, Epigenetic regulation of stem cell differentiation, Pediatr. Res. 59 (2006) 21R-5R. [12] J.P. Issa, S.B. Baylin, J.G. Herman, DNA methylation changes in hematologic malignancies: biologic and clinical implications, Leukemia. 11 (1997) S7-11. [13] S. Anders, W. Huber, Differential expression analysis for sequence count data, Genome. Biol. 11(2010) R106. [14] Y. Benjamini, Y. Hochberg, Controlling the false discovery rate: a practical and powerful approach to multiple testing, Journal. of. the. Royal. Statistical. Society. Series B. (1995) 289-300.
[15] M.E. Smoot, K. Ono, J. Ruscheinski, P.L. Wang, T. Ideker, Cytoscape 2.8: new features for data integration and network visualization, Bioinformatics. 27 (2011) 431-2. [16] D. Tabas-Madrid, R. Nogales-Cadenas, A. Pascual-Montano, GeneCodis3: a non-redundant and modular enrichment analysis tool for functional genomics, Nucleic Acids. Res. 40 (2012) W478-83. [17] C.D. Warden, H. Lee, J.D. Tompkins, X. Li, C. Wang, A.D. Riggs, et al., COHCAP: an integrative genomic pipeline for single-nucleotide resolution DNA methylation analysis, Nucleic. Acids. Res. 41 (2013) e117. [18] Y. Deng, Z. Wang, S. Gu, C. Ji, K. Ying, Y. Xie, et al., Cloning and characterization of a novel human alcohol dehydrogenase gene (ADHFe1), DNA Seq. 13 (2002) 301-6. [19] Y.H. Kim, H.C. Lee, S.Y. Kim, Y.I. Yeom, K.J. Ryu, B.H. Min, et al., Epigenomic analysis of aberrantly methylated genes in colorectal cancer identifies genes commonly affected by epigenetic alterations, Ann. Surg. Oncol. 18 (2011) 2338-47. [20]
B.
Oster, K.
Thorsen, P.
Lamy, T.K.
Wojdacz, L.L.
Hansen, K.
Birkenkamp-Demtröder, et al., Identification and validation of highly frequent CpG island hypermethylation in colorectal adenomas and carcinomas, Int. J. Cancer. 129 (2011) 2855-66. [21] C.H. Tae, K.J. Ryu, S.H. Kim, H.C. Kim, H.K. Chun, B.H. Min, et al., Alcohol dehydrogenase, iron containing, 1 promoter hypermethylation associated with
colorectal cancer differentiation, BMC. Cancer. 13 (2013) 142. [22] A. Sedo, J. Kraml, Dipeptidyl peptidase IV in cell proliferation and differentiation, Sb. Lek. 95 (1994) 285-88. [23] L. Kotackova, E. Balaziova, A. Sedo, Expression pattern of dipeptidyl peptidase IV activity and/or structure homologues in cancer, Folia. Biol. 55 (2009) 77-84. [24] M.A. Sheikh, Y.S. Malik, H. Yu, M. Lai, X. Wang, X.
Zhu, Epigenetic
regulation of Dpp6 expression by Dnmt3b and its novel role in the inhibition of RA induced neuronal differentiation of P19 cells, PLoS. One. 8 (2013) e55826. [25] J. Jaeger, D. Koczan, H.J. Thiesen, S.M. Ibrahim, G. Gross, R. Spang, et al., Gene expression signatures for tumor progression, tumor subtype, and tumor thickness in laser-microdissected melanoma tissues, Clin. Cancer. Res. 13 (2007) 806-15. [26] M.H. Saied, J. Marzec, S. Khalid, P. Smith, T.A. Down, V.K. Rakyan, et al., Genome wide analysis of acute myeloid leukemia reveal leukemia specific methylome and subtype specific hypomethylation of repeats, PLoS. One. 7 (2012) e33213. [27] S. Ishiuchi, K. Tsuzuki, Y. Yoshida, N. Yamada, N. Hagimura, H. Okado, et al., Blockage of Ca(2+)-permeable AMPA receptors suppresses migration and induces apoptosis in human glioblastoma cells, Nat. Med. 8 (2002) 971-8. [28] Y. Piao, L. Lu, J. de Groot, AMPA receptors promote perivascular glioma invasion via beta1 integrin-dependent adhesion to the extracellular matrix, Neuro. Oncol. 11 (2009) 260-73.
[29] S.A. Lyons, W.J. Chung, A.K. Weaver, T. Ogunrinu, H. Sontheimer, Autocrine glutamate signaling promotes glioma cell invasion, Cancer. Res. 67 (2007) 9463-71. [30] C.A. von Roemeling, D.C. Radisky, L.A. Marlow, S.J. Cooper, S.K. Grebe, P.Z. Anastasiadis, et al., Neuronal pentraxin 2 supports clear cell renal cell carcinoma by activating the AMPA-selective glutamate receptor-4, Cancer. Res. 74 (2014) 4796-810. [31] M.R. Fernandez, M.D. Henry, R.E. Lewis, Kinase suppressor of Ras 2 (KSR2) regulates tumor cell transformation via AMPK, Mol. Cell. Biol. 32 (2012) 3718-31. [32] M.E. LLeonart, F. Vidal, D. Gallardo, M. Diaz-Fuertes, F. Rojo, M. Cuatrecasas, et al., New p53 related genes in human tumors: significant downregulation in colon and lung carcinomas, Oncol. Rep. 16 (2006) 603-8. [33] L. López-Vicente, G. Armengol, B. Pons, L. Coch, E. Argelaguet, M. Lleonart, et al., Regulation of replicative and stress-induced senescence by RSK4, which is down-regulated in human tumors, Clin. Cancer. Res. 15 (2009) 4546-53. [34] A. Thakur, K.W. Rahman, J. Wu, A. Bollig, H. Biliran, X. Lin, et al., Aberrant expression of X-linked genes RbAp46, Rsk4, and Cldn2 in breast cancer, Mol. Cancer. Res. 5 (2007) 171-81. [35] A. Thakur, H. Xu, Y. Wang, A. Bollig, H. Biliran, J.D. Liao, The role of X-linked genes in breast cancer, Breast. Cancer. Res. Treat. 93 (2005) 135-43. [36] S.B. Dewdney, B.J. Rimel, P.H. Thaker, D.M. Thompson Jr, A. Schmidt, P. Huettner, et al., Aberrant methylation of the X-linked ribosomal S6 kinase RPS6KA6 (RSK4) in endometrial cancers, Clin. Cancer. Res. 17 (2011) 2120-9.
[37] A. Thakur, Y. Sun, A. Bollig, J. Wu, H. Biliran, S. Banerjee, et al., Anti-invasive and antimetastatic activities of ribosomal protein S6 kinase 4 in breast cancer cells, Clin. Cancer. Res. 14 (2008) 4427-36. [38] A. Feber, L. Xi, J.D. Luketich, A. Pennathur, R.J. Landreneau, M. Wu, et al., MicroRNA expression profiles of esophageal cancer, J. Thorac. Cardiovasc. Surg. 135 (2008) 255-60. [39] P. van Baarlen, F. Troost, C. van der Meer, G. Hooiveld, M. Boekschoten, R.J. Brummer, et al., Human mucosal in vivo transcriptome responses to three lactobacilli indicate how probiotics may modulate human cellular pathways, Proc. Natl. Acad. Sci. U S A. 108 (2011) 4562-9. [40] O. Gevaert, R. Tibshirani, S.K. Plevritis, Pancancer analysis of DNA methylation-driven genes using MethylMix, Genome. Biol. 16 (2015) 17.
Table 1 NOS patients sample IDs
Sample ID
Sample ID
Sample ID
Sample ID
TCGA-2H-A9GF
TCGA-JY-A939
TCGA-L5-A4OW
TCGA-L5-A8NW
TCGA-2H-A9GG
TCGA-JY-A93C
TCGA-L5-A4OX
TCGA-L7-A6VZ
TCGA-2H-A9GH
TCGA-JY-A93D
TCGA-L5-A88V
TCGA-M9-A5M8
TCGA-2H-A9GI
TCGA-JY-A93E
TCGA-L5-A891
TCGA-Q9-A6FW
TCGA-2H-A9GJ
TCGA-L5-A43C
TCGA-L5-A893
TCGA-R6-A6DN
TCGA-2H-A9GK
TCGA-L5-A43E
TCGA-L5-A8NE
TCGA-R6-A6DQ
TCGA-2H-A9GL
TCGA-L5-A4OF
TCGA-L5-A8NF
TCGA-R6-A6KZ
TCGA-2H-A9GM
TCGA-L5-A4OH
TCGA-L5-A8NH
TCGA-R6-A6L4
TCGA-2H-A9GN
TCGA-L5-A4OJ
TCGA-L5-A8NI
TCGA-R6-A6L6
TCGA-2H-A9GO
TCGA-L5-A4ON
TCGA-L5-A8NJ
TCGA-R6-A6XG
TCGA-2H-A9GQ
TCGA-L5-A4OO
TCGA-L5-A8NL
TCGA-R6-A6XQ
TCGA-2H-A9GR
TCGA-L5-A4OP
TCGA-L5-A8NM
TCGA-R6-A6Y0
TCGA-IC-A6RE
TCGA-L5-A4OQ
TCGA-L5-A8NN
TCGA-R6-A6Y2
TCGA-IG-A4QS
TCGA-L5-A4OR
TCGA-L5-A8NR
TCGA-R6-A8W5
TCGA-IG-A7DP
TCGA-L5-A4OS
TCGA-L5-A8NS
TCGA-R6-A8W8
TCGA-JY-A6F8
TCGA-L5-A4OT
TCGA-L5-A8NU
TCGA-R6-A8WC
TCGA-JY-A6FH
TCGA-L5-A4OU
TCGA-L5-A8NV
TCGA-R6-A8WG
TCGA-RE-A7BO
TCGA-V5-A7RE
TCGA-VR-A8EQ
TCGA-ZR-A9CJ
TCGA-S8-A6BV
TCGA-V5-AASW
TCGA-VR-AA4D
TCGA-V5-A7RB
TCGA-V5-AASX
TCGA-X8-AAAR
Table 2 The significantly differentially expressed miRNAs miRNA
log2FC
FDR
Down-regulated miRNAs
miRNA
log2FC
FDR
Up-regulated miRNAs
hsa-mir-30e
-1.294966 5.81E-12 hsa-mir-196a-1
3.05026 3.23E-09
hsa-mir-30c-2
-1.391467 3.60E-11 hsa-mir-196a-2
2.87572 7.44E-09
hsa-mir-148a
-2.496089 9.01E-10 hsa-mir-93
1.4898 9.53E-08
hsa-mir-145
-2.454574 3.24E-09 hsa-mir-18a
1.81826 2.62E-07
hsa-mir-133a-1
-3.222641 5.93E-09 hsa-mir-21
1.16873 1.14E-05
hsa-mir-365-2
-1.840407 7.44E-09 hsa-mir-181b-1
1.35557 1.19E-05
hsa-mir-139
-1.966064 2.11E-08 hsa-mir-135b
2.14399 1.33E-05
hsa-mir-365-1
-1.805253 2.11E-08 hsa-mir-106b
1.30143 1.36E-05
-1.24492 5.69E-07 hsa-mir-335
1.55011 1.65E-05
hsa-mir-664
-1.297577 1.86E-06 hsa-mir-7-3
2.54756 2.00E-05
hsa-mir-26b
-1.017161 1.87E-06 hsa-mir-181a-1
hsa-mir-628
-1.185607 5.24E-06 hsa-mir-3648
2.73239 4.06E-05
hsa-mir-1-2
-2.650192 1.33E-05 hsa-mir-877
1.92604 8.02E-05
hsa-mir-574
-1.010267 2.00E-05 hsa-mir-3677
1.72988 9.08E-05
hsa-mir-30a
-1.539267 8.02E-05
hsa-mir-125a
-1.200214 8.25E-05
hsa-mir-204
-3.241612 9.79E-05
hsa-mir-28
1.0317 3.36E-05
Table 3 Eight differentially expressed genes under the regulation of methylation Gene
Fold Change
FDR
Chr
CpG Island
Num Methyl
Island FDR
GPRASP1
-2.26859
0.000000 X 0719
chrX:1019060 01-101907017
5
1.03121E-06
DPP6
-4.6226
0.000000 7 00011
chr7:15358331 7-153585666
5
3.05411E-06
TNXB
-2.8195
0.000000 6
5
4.90164E-06
0000298
chr6:32063533 -32065044
GRIA4
-4.10134
0.000000 11 00971
chr11:1054811 26-105481422
4
2.29214E-05
ADHFE1
-3.86946
0.000000 8 0000003 69
chr8:67344497 -67344989
3
0.000125086
CNKSR2
-3.99557
0.000627427
-3.24036
chrX:2139183 5-21393408 chrX:8344239 6-83443335
3
RPS6KA6
0.000000 X 00765 0.000000 X 6
3
0.008119453
ZNF135
-1.96936 9484
0.000033 19 8
chr19:5857039 3-58571779
3
0.008451589 59988643
Figure legends Fig. 1 The heat map of the top fifty DEGs in NOS. Fig. 2 The regulatory network of miRNAs and target genes in NOS. Diamonds and ellipses represent the miRNAs and target genes, respectively. The red and green colors represent up-regulation and down-regulation, respectively. Fig. 3 The correlation between miRNA and target genes in NOS.
Dataset ESM1 LOC541471 CKS2 SYNM NDE1 TACR2 RYR3 VIP LEPR DPT PEBP4 C3orf18 FNDC5 RBPMS2 SLC2A4 SORBS1 ADCYAP1R1 PLIN4 DUSP19 RGMB DQ599327 SIK2 SLC25A4 C7orf41 CRY2 TMEM161B STX12 DBT KIAA1191 ALAD LOC100129931 TBC1D14 C22orf23 AK055981 ESRRB LOC388387 AQP4 CKM FAM165B AK131020 TMED6 CKMT2 HRH2 SIGLEC11 LRP1B HMP19 PM20D1 ZRANB2−AS1 GPR155 ADHFE1
4
Dataset case normal
2
0
−2
−4
86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
S0344-0338(16)30381-8 http://dx.doi.org/doi:10.1016/j.prp.2016.12.007 PRP 51694
To appear in: Received date:
25-8-2016
Please cite this article as: Ting Xi, Guizhi Zhang, Epigenetic regulation on the gene expression signature in esophagus adenocarcinoma, Pathology - Research and Practice http://dx.doi.org/10.1016/j.prp.2016.12.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Epigenetic regulation on the gene expression signature in esophagus adenocarcinoma Ting Xia, Guizhi Zhangb,* a
Department of Gastroenterology, First People’s Hospital of Liaocheng of Shandong
Province b
Department of Gastroenterology, Second People’s Hospital of Liaocheng of
Shandong Province *
Correspondence to: Guizhi Zhang
Department of Gastroenterology, Second People’s Hospital of Liaocheng of Shandong Province, No.306 JianKang Street, Linqing 252600, China Email: [email protected] Tel: 13676357009 Ting Xi Email: [email protected]
ABSTRACT Background: Understanding the molecular mechanisms represents an important step for the development of diagnostic and therapeutic measures of esophagus adenocarcinoma (NOS). The objective of this study is to identify the epigenetic regulation on gene expression in NOS, shedding light on the molecular mechanisms of NOS. Methods: In this study, 78 patients with NOS were included and the data of mRNA, miRNA and DNA methylation of were downloaded from The Cancer Genome Atlas (TCGA). Differential analysis between NOS and controls was performed in terms of gene expression, miRNA expression, and DNA methylation. Bioinformatic analysis was followed to explore the regulation mechanisms of miRNA and DNA methylationon gene expression. Results: Totally, up to 1320 differentially expressed genes (DEGs) and 32 differentially expressed miRNAs were identified. 240 DEGs that were not only the target genes but also negatively correlated with the screened differentially expressed miRNAs. 101 DEGs were found to be highlymethylated in CpG islands. Then, 8 differentially methylated genes (DMGs) were selected, which showed down-regulated expression in NOS. Among of these genes, 6 genes including ADHFE1, DPP6， GRIA4，CNKSR2，RPS6KA6 and ZNF135 were target genes of differentially expressed miRNAs (hsa-mir-335, hsa-mir-18a, hsa-mir-93, hsa-mir-106b and hsa-mir-21). Conclusions: The identified altered miRNA, genes and DNA methylation site may be
applied as biomarkers for diagnosis and prognosis of NOS. Keywords: esophagus adenocarcinoma, differentially expressed genes, miRNAs, DNA methylation 1. Introduction There are two main kinds of esophageal carcinoma: esophagus adenocarcinoma and esophagus squamous cell carcinoma. Esophagus adenocarcinoma (NOS) is one of the fastest rising cancers. Up to now, the reason of increasing occurrence of NOS is not entirely known but the principal risk factors are male, visceral obesity and chronic reflux disease [1]. Chemotherapy and surgery are common therapeutic methods for patients with NOS. It is pointed out that NOS is resistant to chemotherapeutic regimens and is characterized by a poor survival outcome [2]. Recently, microRNAs (miRNAs) were found to be candidates as biomarkers for diagnosis and prognosis in different cancers [3, 4]. MiRNAs are small, well-conserved and non-coding RNA molecules that regulate the expression of mRNAs [5]. Generally, miRNAs inhibit gene expression through targeting the complementary mRNA and blocking its translation [6]. It is well known that miRNAs is tissue-specific in terms of expression and function [7, 8].They can function as oncomir as well as tumor suppressors, and are involved in various biological processes, including proliferation, differentiation, and apoptosis [9]. Some miRNAs have already been identified in NOS, for instance, miR-21, miR-233, miR-192 and miR-194 was up-regulated, and the expression patterns were validated by qRT-PCR [10]. This further suggested that miRNAs plays an important role in the diagnosis and
prognosis of NOS. Epigenetic modifications such as DNA methylation can modify gene expression patterns and control amounts of cellular functions [11]. Aberrant hypermethylation has been associated with inactivation of tumor-related genes in a large-scale of human neoplasms [12]. DNA methylation and miRNA regulation on gene expression both belong to the domain of epigenetic modification. So, it is important to analyze the correlation between transcriptional sequencing data and epigenetic modification data to illustrate the explicit epigenetic regulation mechanisms of NOS. In this study, the comprehensive analysis of mRNA, miRNA and DNA methylation profiling data of NOS from The Cancer Genome Atlas (TCGA) was performed. Differentially expressed genes (DEGs), differentially expressed miRNAs and differentially methylated genes (DMGs) were identified in NOS. Function enrichment analysis of DEGs and DMGs was then performed. DEGs associated with differential methylation or targeting by differentially expressed miRNAs were identified. At last, 6 candidate genes may be used in the potential targeted treatment of NOS, and further study was needed to investigate their roles in NOS carcinogenesis. 2. Materials and Methods 2.1. Datasets In
this
study,
we
downloaded
the
mRNA
expression
(UNC
IlluminaHiSeqRNASeq), miRNA expression (BCGSC_IlluminaHiSeq_miRNASeq) and DNA methylation data (JHU_USC_HumanMethylation450) of all 185 patients with NOS from TCGA data portal (http://tcga-data.nci.nih.gov/) (January 2016).
According to providing clinical information, we selected 78 NOS patients without histories of other malignancy or neoadjuvant treatment. Among which, 68 men and 10 women were included and 79% were the white race. Moreover, the age range of all individuals was 27 to 86 and the average age is 65. Additionally, the tumor degree was classified as G1 (2 cases), G2 (25 cases), G3 (27 cases) and GX (24 cases). Detailed information of these 78 patients was showed in supplementary 1. 2.2. Analysis of DEGs and differentially expressed miRNAs The DEGs and differentially expressed miRNAs were evaluated in the R-bioconductor package DESeq [13]. The Limma package in R was used to calculate p-values by two-tailed Student's t-test. MetaMA package in R was used to combine p-values, and the false discovery rate (FDR) was obtained from multiple comparisons using the Benjamini and Hochberg method [14]. We selected DEGs and differentially expressed miRNAs with criterion of FDR < 0.0001. 2.3. Correlation analysis of DEGs and differentially expressed miRNAs First, pairwise Pearson correlation coefficients between differentially expressed miRNAs and DEGs was calculated, and P < 0.05was defined as statistical significance. Second, six miRNA-target prediction tools (RNA22, miRanda, miRDB, miRWalk, PICTAR2 and Targetscan) were utilized to predict target genes of differentially expressed miRNAs. Only those miRNA-target pairs which were predicted by more than four algorithms can be selected out. The miRNA-targets pairs verified by experiment in miRWalk database were also screened out. Finally, we selected the miRNA-target pairs with negative correlations (p < 0.05, r < 0) to establish the
miRNA-target regulatory network, which was visualized using Cytoscape software [15]. 2.4. Functional annotation of target genes of differentially expressed miRNAs To acquire the biological function of the miRNA target genes, we conducted Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis based on the online software GENECODIS [16]. FDR < 0.05 was set as the cut-off for selecting significantly enriched functional KEGG pathways. 2.5. Influence of DNA methylation on DEGs The COHCAP package in R (https://sourceforge.net/projects/cohcap) [17] was used to identify the differential methylation sites between tumors and normal tissues that are more likely to regulate downstream gene expression. FDR < 0.05 was considered as differentially methylated CpG sites. Based on the gene expression of NOS, we identified the aberrant DNA methylated CpG site which affected corresponding gene expression. 3. Results 3.1. Gene expression pattern in NOS Totally, up to 78 tissue samples were included in the present study, all of which with fully characterized mRNA profiles, miRNA profiles and DNA methylation data. The TCGA barcode IDs for each sample were showed in Table 1. There were 1320 DEGs with FDR < 0.0001 between NOS and normal tissue. Among these genes, 594 genes were up-regulated and 726 genes were down-regulated. The heatmap of top fifty DEGs was showed in Fig. 1. The top ten up-regulated genes were ESM1,
LOC541471, CKS2, GABRD, MMP11, MMP3, LINC00152, TPX2, IL8 and NUF2. By contrast, the top ten down-regulated genes were LOC388387, GPR155, DQ599327, RGMB, SIK2, AK131020, FAM165B, C22orf23, NDE1 and AK055981. 3.2. Correlations of differentially expressed miRNAs and DEGs in NOS The screened differentially expressed miRNAs were presented in Table 2. There were 32 differentially expressed miRNAs including 14 up-regulated and 18 down-regulated miRNAs. To investigate the correlations between differentially expressed miRNAs and DEGs, we first performed the correlation analysis. Depending on the analysis, we obtained 3112 miRNA-mRNA pairs which were negatively correlated (P < 0.05, r < 0). Then, we got 1828 miRNA-mRNA pairs through miRNA-target prediction algorithms. At last, we screened out 240 DEGs including 160 up-regulated and 80 down-regulated genes, all of which were not only negatively correlated with corresponding miRNA expression, but also predicted by bioinformatics methods. The established miRNA regulatory network was showed in Fig. 2. To understand the potential function of these miRNA targets genes, we conducted KEGG pathway analysis. The results demonstrated that these DEGs significantly enriched in propanoate metabolism, valine, leucine and isoleucine degradation, valine, leucine and isoleucine degradation, DNA replication, cell cycle, homologous recombination, mismatch repair, oocyte meiosis, p53 signaling pathway, pyrimidine metabolism, pathways in cancer, tuberculosis and systemic lupus erythematous. 3.3. Influence of DNA methylationon DEGs in NOS
The methylation of CpG islands in the promoter regions generally results in the gene silencing. There were 116 differentially methylated CpG islands with FDR < 0.05 (The methylation levels of all sites in the CpG islands were averaged), which involved 101 genes with hypermethylation. To investigate the influence of DNA methylation on gene expression, we compared the list of differentially methylated genes and differentially expressed genes to identify the differentially methylated genes
displaying
differential
expression.
Based
on
the
fact
that
DNA
hypermethylation suppressed corresponding gene expression, we screened out 8 down-regulated genes from 101 DMGs, including GPRASP1, DPP6, TNXB, GRIA4, ADHFE1, CNKSR2, RPS6KA6, ZNF135, which were showed in Table 3. 3.4. Key genes upon regulation of miRNA and DNA methylation We combined the results of miRNA regulation and DNA methylation regulation on mRNA expression, and found that six DEGs (ADHFE1, DPP6, GRIA4, CNKSR2, RPS6KA6 and ZNF135) were under miRNA and DNA methylation regulation, which involved five differentially expressed miRNAs (hsa-mir-335, hsa-mir-18a, hsa-mir-93, hsa-mir-106b and hsa-mir-21). The correlation between these five differentially expressed miRNAs and target genes in NOS was showed in Fig. 3. Among six genes, ADHFE1 expression was negatively correlated with that of five miRNAs. 4. Discussion A critical step in fighting against NOS is to identify changes of genes in different levels occurring in tumor development. In this study, we integrated mRNA expression, miRNA expression data and DNA methylation profiles of NOS to find the valuable
genes which were highly associated with NOS. 1320 genes were aberrantly expressed in NOS compare with normal tissues. To investigate the epigenetic modifications of gene regulation, we assessed the miRNA expression data, and identified 32 differentially expressed miRNAs. Depending on the miRNA-mRNA expression correlation analysis and miRNA-target prediction, we obtained 240 DEGs in NOS, which were negatively correlated with miRNAs. Depending on the DNA methylation-mRNA expression correlation analysis, we obtained 8 differentially expressed genes which were hypermethylated. After combining the results of miRNA regulation and DNA methylation regulation on gene expression, we finally selected 6 genes, all of which were not only targeting by correlated differentially expressed miRNAs, but also down-regulated under hypermethylation, which involving 5 up-regulated miRNAs. The results suggested that these genes under epigenetic regulation may be associated with carcinogenesis of NOS. Alcohol dehydrogenase, iron containing 1 (ADHFE1) was cloned from a human fetal brain cDNA library [18]. Recently, aberrant methylation in ADHFE1 promoter was associated with colorectal carcinomas [19, 20]. It is pointed out that hypermethylation of ADHFE1 is associated with colorectal cancer differentiation [21]. Dipeptidyl peptidase like 6 (DPP6) is a protein which regulates various biological functions and also plays a role in carcinogenesis [22, 23]. DPP6 can maintain cell-specific phenotype and its deregulation will result in carcinogenesis [24]. In addition, hypermethylation and reduced expression of DPP6 is found in melanoma
[25] and acute myeloid leukemia (AML) patients [26]. These studies indicate that DPP6 is tightly regulated during neoplastic development. In this study, we found that ADHFE1 and DPP6 were all down-regulated and high methylated, and the expression of ADHFE1 and DPP6 was significantly inversely related to hsa-mir-93. Our results provided evidences for their important roles in NOS. Further study may be needed to verify their crucial functions in NOS. Glutamate ionotropic receptor AMPA type subunit 4 (GRIA4, also called GLUR4) is the subunit of amino-3-hydroxy-5-methyl-4isoxazolepropionic acid (AMPA) receptor in the brain. It is indicated that AMPA receptor activity has been related to neuronal cancer cell migration [27-29]. GRIA4 is important for viability and invasion of distinct renal cell carcinoma [30]. In this study we found that GRIA4 was down-regulated in NOS, and the regulatory miRNAs, hsa-mir-18a and has-mir-106b, were up-regulated. The results suggested that epigenetic modification of GRIA4 in terms of miRNA regulation and DNA methylation modification is linked to NOS progression. Connector enhancer of kinase suppressor of Ras 2 (CNKSR2, also called CNK2 or KSR2) regulates glycolysis and oxidative metabolism in tumor cells [31]. It is reported that CNKSR2 can enhance the proliferative rate of mouse embryo fibroblasts, and CNKSR2 depletion reduces tumor cell growth, which indicates that CNKSR2 is a regulator of cellular metabolism affecting the tumorigenic potential of cells [31]. Ribosomal protein S6 kinase A6 (RPS6KA6, also called RSK4) is a tumor suppressor, and RPS6KA6 mutation is an unusual event in cancer. It has been noted that the
expression of RPS6KA6 reduced in colon, breast, colonic, renal and endometrial carcinomas [32-36]. Additionally, the high frequency of the CpG island methylated phenotype can also be found in endometrial carcinomas [36]. RPS6KA6 contributions to cancer phenotypes could be through regulation of cell cycle arrest and stress responses [33], tumor invasiveness and metastasis [34, 37]. Accordingly, we found that both CNKSR2 and RPS6KA6 were down-regulated and hypermethylated in NOS. Interestingly, CNKSR2 and RPS6KA6 were target genes of hsa-mir-21, and hsa-mir-21 is up-regulated with the expression trend opposite with its target genes CNKSR2 and RPS6KA6. It is a known fact that hsa-mir-21 is an oncomir, and is up-regulated in NOS [38]. It is reported that has-mir-21 can serve as diagnostic biomarker of NOS. In a word, the expression changes of CNKSR2 and RPS6KA6 under the regulation of has-mir-21 illuminated the mechanism of carcinogenesis which may play a crucial role in the progress of NOS. Zinc finger protein 135 (ZNF135) is a transcription factor which involved in angiogenesis and wound repair [39]. It is reported that ZNF135 is high methylation in breast cancer, colon cancer, glioblastoma, head and neck squamous carcinoma, clear cellrenal carcinoma, lung adenocarcinoma, lung squamous carcinoma, serous ovarian cancer, rectal cancer, endometrial carcinoma sites [40]. In this study, we found that ZNF135 was down-regulated and hypermethylated in NOS. ZNF135 is the target gene of hsa-mir-335, hsa-mir-93 and hsa-mir-106b, and all of these miRNAs were up-regulated. This further indicated that ZNF135 may be related to the development of NOS.
In brief, integrated analysis of the TCGA data led to a pool of DEGs, differentially expressed miRNAs and DMGs. The epigenetic modifications via miRNAs and methylation for ADHFE1, DPP6, GRIA4, CNKSR2, RPS6KA6 and ZNF135 may be involved in growth and tumorigenesis of NOS. The study of epigenetic alterations of these genes is of value to investigate the pathogenesis of NOS and have an important role in guiding the development of new therapies for NOS. Conflict of interest None declared. Acknowledgement None. References [1] A. Cowre, F. Noble, T. Underwood, Strategies to improve outcomes in esophageal adenocarcinoma, Expert. Rev. Anticancer. Ther. 14 (2014) 677-87. [2] D.H. Ilson, Epirubicin, cisplatin, and fluorouracil in gastric and esophageal cancer: a step
ahead?, J. Clin. Oncol. 20 (2002) 1962-4.
[3] J. Lu, G. Getz, E.A. Miska, E. Alvarez-Saavedra, J. Lamb, D. Peck, et al., MicroRNA expression profiles classify human cancers, Nature. 435 (2005) 834-8. [4]
P.S.
Mitchell, R.K.
Parkin, E.M.
Kroh, B.R.
Fritz, S.K.
Wyman, E.L.
Pogosova-Agadjanyan, et al., Circulating microRNAs as stable blood-based markers for cancer detection, Proc. Natl. Acad. Sci. U S A. 105 (2008) 10513-8. [5] D.P. Bartel, MicroRNAs: genomics, biogenesis, mechanism, and function, Cell. 116 (2004) 281-97.
[6] Y.W. Kong, D. Ferland-McCollough, T.J. Jackson, M. Bushell, microRNAs in cancer management, Lancet. Oncol. 13 (2012) e249-58. [7] H. Liu, I.S. Kohane, Tissue and process specific microRNA-mRNA co-expression in mammalian development
and malignancy, PLoS. One. 4 (2009)
e5436. [8] T. Babak, W. Zhang, Q. Morris, B.J. Blencowe, T.R. Hughes, Probing microRNAs with microarrays: tissue specificity and functional inference, RNA. 10 (2004) 1813-9. [9] M. Seven, O.F. Karatas, M.B. Duz, M. Ozen, The role of miRNAs in cancer: from pathogenesis to therapeutic implications, Future. Oncol. 10 (2014) 1027-48. [10] E.A. Mathé, G.H. Nguyen, E.D. Bowman, Y. Zhao, A. Budhu, A.J. Schetter, et al., MicroRNA expression in squamous cell carcinoma and adenocarcinoma of the esophagus: associations with survival, Clin. Cancer. Res. 15 (2009) 6192-200. [11] H. Wu, Y.E. Sun, Epigenetic regulation of stem cell differentiation, Pediatr. Res. 59 (2006) 21R-5R. [12] J.P. Issa, S.B. Baylin, J.G. Herman, DNA methylation changes in hematologic malignancies: biologic and clinical implications, Leukemia. 11 (1997) S7-11. [13] S. Anders, W. Huber, Differential expression analysis for sequence count data, Genome. Biol. 11(2010) R106. [14] Y. Benjamini, Y. Hochberg, Controlling the false discovery rate: a practical and powerful approach to multiple testing, Journal. of. the. Royal. Statistical. Society. Series B. (1995) 289-300.
[15] M.E. Smoot, K. Ono, J. Ruscheinski, P.L. Wang, T. Ideker, Cytoscape 2.8: new features for data integration and network visualization, Bioinformatics. 27 (2011) 431-2. [16] D. Tabas-Madrid, R. Nogales-Cadenas, A. Pascual-Montano, GeneCodis3: a non-redundant and modular enrichment analysis tool for functional genomics, Nucleic Acids. Res. 40 (2012) W478-83. [17] C.D. Warden, H. Lee, J.D. Tompkins, X. Li, C. Wang, A.D. Riggs, et al., COHCAP: an integrative genomic pipeline for single-nucleotide resolution DNA methylation analysis, Nucleic. Acids. Res. 41 (2013) e117. [18] Y. Deng, Z. Wang, S. Gu, C. Ji, K. Ying, Y. Xie, et al., Cloning and characterization of a novel human alcohol dehydrogenase gene (ADHFe1), DNA Seq. 13 (2002) 301-6. [19] Y.H. Kim, H.C. Lee, S.Y. Kim, Y.I. Yeom, K.J. Ryu, B.H. Min, et al., Epigenomic analysis of aberrantly methylated genes in colorectal cancer identifies genes commonly affected by epigenetic alterations, Ann. Surg. Oncol. 18 (2011) 2338-47. [20]
B.
Oster, K.
Thorsen, P.
Lamy, T.K.
Wojdacz, L.L.
Hansen, K.
Birkenkamp-Demtröder, et al., Identification and validation of highly frequent CpG island hypermethylation in colorectal adenomas and carcinomas, Int. J. Cancer. 129 (2011) 2855-66. [21] C.H. Tae, K.J. Ryu, S.H. Kim, H.C. Kim, H.K. Chun, B.H. Min, et al., Alcohol dehydrogenase, iron containing, 1 promoter hypermethylation associated with
colorectal cancer differentiation, BMC. Cancer. 13 (2013) 142. [22] A. Sedo, J. Kraml, Dipeptidyl peptidase IV in cell proliferation and differentiation, Sb. Lek. 95 (1994) 285-88. [23] L. Kotackova, E. Balaziova, A. Sedo, Expression pattern of dipeptidyl peptidase IV activity and/or structure homologues in cancer, Folia. Biol. 55 (2009) 77-84. [24] M.A. Sheikh, Y.S. Malik, H. Yu, M. Lai, X. Wang, X.
Zhu, Epigenetic
regulation of Dpp6 expression by Dnmt3b and its novel role in the inhibition of RA induced neuronal differentiation of P19 cells, PLoS. One. 8 (2013) e55826. [25] J. Jaeger, D. Koczan, H.J. Thiesen, S.M. Ibrahim, G. Gross, R. Spang, et al., Gene expression signatures for tumor progression, tumor subtype, and tumor thickness in laser-microdissected melanoma tissues, Clin. Cancer. Res. 13 (2007) 806-15. [26] M.H. Saied, J. Marzec, S. Khalid, P. Smith, T.A. Down, V.K. Rakyan, et al., Genome wide analysis of acute myeloid leukemia reveal leukemia specific methylome and subtype specific hypomethylation of repeats, PLoS. One. 7 (2012) e33213. [27] S. Ishiuchi, K. Tsuzuki, Y. Yoshida, N. Yamada, N. Hagimura, H. Okado, et al., Blockage of Ca(2+)-permeable AMPA receptors suppresses migration and induces apoptosis in human glioblastoma cells, Nat. Med. 8 (2002) 971-8. [28] Y. Piao, L. Lu, J. de Groot, AMPA receptors promote perivascular glioma invasion via beta1 integrin-dependent adhesion to the extracellular matrix, Neuro. Oncol. 11 (2009) 260-73.
[29] S.A. Lyons, W.J. Chung, A.K. Weaver, T. Ogunrinu, H. Sontheimer, Autocrine glutamate signaling promotes glioma cell invasion, Cancer. Res. 67 (2007) 9463-71. [30] C.A. von Roemeling, D.C. Radisky, L.A. Marlow, S.J. Cooper, S.K. Grebe, P.Z. Anastasiadis, et al., Neuronal pentraxin 2 supports clear cell renal cell carcinoma by activating the AMPA-selective glutamate receptor-4, Cancer. Res. 74 (2014) 4796-810. [31] M.R. Fernandez, M.D. Henry, R.E. Lewis, Kinase suppressor of Ras 2 (KSR2) regulates tumor cell transformation via AMPK, Mol. Cell. Biol. 32 (2012) 3718-31. [32] M.E. LLeonart, F. Vidal, D. Gallardo, M. Diaz-Fuertes, F. Rojo, M. Cuatrecasas, et al., New p53 related genes in human tumors: significant downregulation in colon and lung carcinomas, Oncol. Rep. 16 (2006) 603-8. [33] L. López-Vicente, G. Armengol, B. Pons, L. Coch, E. Argelaguet, M. Lleonart, et al., Regulation of replicative and stress-induced senescence by RSK4, which is down-regulated in human tumors, Clin. Cancer. Res. 15 (2009) 4546-53. [34] A. Thakur, K.W. Rahman, J. Wu, A. Bollig, H. Biliran, X. Lin, et al., Aberrant expression of X-linked genes RbAp46, Rsk4, and Cldn2 in breast cancer, Mol. Cancer. Res. 5 (2007) 171-81. [35] A. Thakur, H. Xu, Y. Wang, A. Bollig, H. Biliran, J.D. Liao, The role of X-linked genes in breast cancer, Breast. Cancer. Res. Treat. 93 (2005) 135-43. [36] S.B. Dewdney, B.J. Rimel, P.H. Thaker, D.M. Thompson Jr, A. Schmidt, P. Huettner, et al., Aberrant methylation of the X-linked ribosomal S6 kinase RPS6KA6 (RSK4) in endometrial cancers, Clin. Cancer. Res. 17 (2011) 2120-9.
[37] A. Thakur, Y. Sun, A. Bollig, J. Wu, H. Biliran, S. Banerjee, et al., Anti-invasive and antimetastatic activities of ribosomal protein S6 kinase 4 in breast cancer cells, Clin. Cancer. Res. 14 (2008) 4427-36. [38] A. Feber, L. Xi, J.D. Luketich, A. Pennathur, R.J. Landreneau, M. Wu, et al., MicroRNA expression profiles of esophageal cancer, J. Thorac. Cardiovasc. Surg. 135 (2008) 255-60. [39] P. van Baarlen, F. Troost, C. van der Meer, G. Hooiveld, M. Boekschoten, R.J. Brummer, et al., Human mucosal in vivo transcriptome responses to three lactobacilli indicate how probiotics may modulate human cellular pathways, Proc. Natl. Acad. Sci. U S A. 108 (2011) 4562-9. [40] O. Gevaert, R. Tibshirani, S.K. Plevritis, Pancancer analysis of DNA methylation-driven genes using MethylMix, Genome. Biol. 16 (2015) 17.
Table 1 NOS patients sample IDs
Sample ID
Sample ID
Sample ID
Sample ID
TCGA-2H-A9GF
TCGA-JY-A939
TCGA-L5-A4OW
TCGA-L5-A8NW
TCGA-2H-A9GG
TCGA-JY-A93C
TCGA-L5-A4OX
TCGA-L7-A6VZ
TCGA-2H-A9GH
TCGA-JY-A93D
TCGA-L5-A88V
TCGA-M9-A5M8
TCGA-2H-A9GI
TCGA-JY-A93E
TCGA-L5-A891
TCGA-Q9-A6FW
TCGA-2H-A9GJ
TCGA-L5-A43C
TCGA-L5-A893
TCGA-R6-A6DN
TCGA-2H-A9GK
TCGA-L5-A43E
TCGA-L5-A8NE
TCGA-R6-A6DQ
TCGA-2H-A9GL
TCGA-L5-A4OF
TCGA-L5-A8NF
TCGA-R6-A6KZ
TCGA-2H-A9GM
TCGA-L5-A4OH
TCGA-L5-A8NH
TCGA-R6-A6L4
TCGA-2H-A9GN
TCGA-L5-A4OJ
TCGA-L5-A8NI
TCGA-R6-A6L6
TCGA-2H-A9GO
TCGA-L5-A4ON
TCGA-L5-A8NJ
TCGA-R6-A6XG
TCGA-2H-A9GQ
TCGA-L5-A4OO
TCGA-L5-A8NL
TCGA-R6-A6XQ
TCGA-2H-A9GR
TCGA-L5-A4OP
TCGA-L5-A8NM
TCGA-R6-A6Y0
TCGA-IC-A6RE
TCGA-L5-A4OQ
TCGA-L5-A8NN
TCGA-R6-A6Y2
TCGA-IG-A4QS
TCGA-L5-A4OR
TCGA-L5-A8NR
TCGA-R6-A8W5
TCGA-IG-A7DP
TCGA-L5-A4OS
TCGA-L5-A8NS
TCGA-R6-A8W8
TCGA-JY-A6F8
TCGA-L5-A4OT
TCGA-L5-A8NU
TCGA-R6-A8WC
TCGA-JY-A6FH
TCGA-L5-A4OU
TCGA-L5-A8NV
TCGA-R6-A8WG
TCGA-RE-A7BO
TCGA-V5-A7RE
TCGA-VR-A8EQ
TCGA-ZR-A9CJ
TCGA-S8-A6BV
TCGA-V5-AASW
TCGA-VR-AA4D
TCGA-V5-A7RB
TCGA-V5-AASX
TCGA-X8-AAAR
Table 2 The significantly differentially expressed miRNAs miRNA
log2FC
FDR
Down-regulated miRNAs
miRNA
log2FC
FDR
Up-regulated miRNAs
hsa-mir-30e
-1.294966 5.81E-12 hsa-mir-196a-1
3.05026 3.23E-09
hsa-mir-30c-2
-1.391467 3.60E-11 hsa-mir-196a-2
2.87572 7.44E-09
hsa-mir-148a
-2.496089 9.01E-10 hsa-mir-93
1.4898 9.53E-08
hsa-mir-145
-2.454574 3.24E-09 hsa-mir-18a
1.81826 2.62E-07
hsa-mir-133a-1
-3.222641 5.93E-09 hsa-mir-21
1.16873 1.14E-05
hsa-mir-365-2
-1.840407 7.44E-09 hsa-mir-181b-1
1.35557 1.19E-05
hsa-mir-139
-1.966064 2.11E-08 hsa-mir-135b
2.14399 1.33E-05
hsa-mir-365-1
-1.805253 2.11E-08 hsa-mir-106b
1.30143 1.36E-05
-1.24492 5.69E-07 hsa-mir-335
1.55011 1.65E-05
hsa-mir-664
-1.297577 1.86E-06 hsa-mir-7-3
2.54756 2.00E-05
hsa-mir-26b
-1.017161 1.87E-06 hsa-mir-181a-1
hsa-mir-628
-1.185607 5.24E-06 hsa-mir-3648
2.73239 4.06E-05
hsa-mir-1-2
-2.650192 1.33E-05 hsa-mir-877
1.92604 8.02E-05
hsa-mir-574
-1.010267 2.00E-05 hsa-mir-3677
1.72988 9.08E-05
hsa-mir-30a
-1.539267 8.02E-05
hsa-mir-125a
-1.200214 8.25E-05
hsa-mir-204
-3.241612 9.79E-05
hsa-mir-28
1.0317 3.36E-05
Table 3 Eight differentially expressed genes under the regulation of methylation Gene
Fold Change
FDR
Chr
CpG Island
Num Methyl
Island FDR
GPRASP1
-2.26859
0.000000 X 0719
chrX:1019060 01-101907017
5
1.03121E-06
DPP6
-4.6226
0.000000 7 00011
chr7:15358331 7-153585666
5
3.05411E-06
TNXB
-2.8195
0.000000 6
5
4.90164E-06
0000298
chr6:32063533 -32065044
GRIA4
-4.10134
0.000000 11 00971
chr11:1054811 26-105481422
4
2.29214E-05
ADHFE1
-3.86946
0.000000 8 0000003 69
chr8:67344497 -67344989
3
0.000125086
CNKSR2
-3.99557
0.000627427
-3.24036
chrX:2139183 5-21393408 chrX:8344239 6-83443335
3
RPS6KA6
0.000000 X 00765 0.000000 X 6
3
0.008119453
ZNF135
-1.96936 9484
0.000033 19 8
chr19:5857039 3-58571779
3
0.008451589 59988643
Figure legends Fig. 1 The heat map of the top fifty DEGs in NOS. Fig. 2 The regulatory network of miRNAs and target genes in NOS. Diamonds and ellipses represent the miRNAs and target genes, respectively. The red and green colors represent up-regulation and down-regulation, respectively. Fig. 3 The correlation between miRNA and target genes in NOS.
Dataset ESM1 LOC541471 CKS2 SYNM NDE1 TACR2 RYR3 VIP LEPR DPT PEBP4 C3orf18 FNDC5 RBPMS2 SLC2A4 SORBS1 ADCYAP1R1 PLIN4 DUSP19 RGMB DQ599327 SIK2 SLC25A4 C7orf41 CRY2 TMEM161B STX12 DBT KIAA1191 ALAD LOC100129931 TBC1D14 C22orf23 AK055981 ESRRB LOC388387 AQP4 CKM FAM165B AK131020 TMED6 CKMT2 HRH2 SIGLEC11 LRP1B HMP19 PM20D1 ZRANB2−AS1 GPR155 ADHFE1
4
Dataset case normal
2
0
−2
−4
86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1