- Email: [email protected]
Screening of Differently Expressed miRNA and mRNA in Prostate Cancer by Integrated Analysis of Transcription Data
Accepted Manuscript Title: Screening of Differently Expressed miRNA and mRNA in Prostate Cancer by Integrated Analysis of Transcription Data Author: Yanan Sun, Xiaopeng Jia, Lianguo Hou, Xing Liu PII: DOI: Reference:
S0090-4295(16)30149-2 http://dx.doi.org/doi: 10.1016/j.urology.2016.04.041 URL 19771
To appear in:
Urology
Received date: Accepted date:
1-3-2016 28-4-2016
Please cite this article as: Yanan Sun, Xiaopeng Jia, Lianguo Hou, Xing Liu, Screening of Differently Expressed miRNA and mRNA in Prostate Cancer by Integrated Analysis of Transcription Data, Urology (2016), http://dx.doi.org/doi: 10.1016/j.urology.2016.04.041. 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.
Screening of differently expressed miRNA and mRNA in prostate cancer by integrated analysis of transcription data Yanan Sun1, Xiaopeng Jia2*, Lianguo Hou3, Xing Liu4 1
Department of Obstetrics and Gynecology, Bethune International Peace Hospital of PLA, Shijiazhuang, Hebei 050071, China
2
Department of Urology, The Third Hospital of Hebei Medical University, Shijiazhuang,
Hebei 050051, China 3
Department of Biochemistry and Molecular Biology, Hebei Medical University,
Shijiazhuang, Hebei 050011, China 4
Department of Orthopaedic Trauma, The Third Hospital of Shijiazhuang City, Shijiazhuang,
Hebei 050011, China *
Corresponding author: Xiaopeng Jia
Department of Urology, The Third Hospital of Hebei Medical University, Shijiazhuang, Hebei 050051, China Email: [email protected] Tel: +86-0311-88602033 Fax: +86-0311-88602033
Abstract Objective: The purpose of this study was to screen aberrantly expressed miRNAs and genes in prostate cancer (PCA), and further uncover the underlying mechanisms for the development of PCA. Methods: We searched GEO database for miRNA and gene expression datasets of PCA, and then separately integrated miRNA and gene expression datasets to identify miRNA and gene expression profiles in PCA. Target genes of differentially expressed miRNAs were predicted through miRWalk database. We matched these target genes with the list of differentially 1
Page 1 of 16
expressed genes to identify miRNA-target gene pairs whose expression was inversely correlated. The function of these target genes was annotated. Results: 29 differentially expressed miRNAs and 946 differentially expressed genes were identified between PCA and normal control. 751 miRNA-target gene pairs that showed inverse expression in PCA, were obtained to establish a regulatory network. In this regulatory network, ten genes (BCL2, BNC2, CCND2, EPM2A, MRAS, NAV2, RASL12, STK33, TCEAL1, WWC2) were co-regulated by five miRNAs (hsa-miR-106b, hsa-miR-130b, hsamiR-93, hsa-miR-153, hsa-miR-182). The expression of hsa-miR-182 was significantly associated with PCA survival through the online validation tool of SurvMicro, suggesting the potential use as a diagnostic or prognostic biomarker in PCA. Conclusions: This integrated analysis was performed to infer new miRNA regulation activities, which provides insights into the understanding of underlying molecular mechanisms of PCA, and guides for exploration of novel therapeutic targets. Keywords: miRNA; differentially expressed; target gene; microarray; prostate cancer
Introduction Prostate cancer (PCA) is the second most frequently diagnosed tumor and the sixth leading cause of cancer-related death occurring in men worldwide. Previous studies have reported that a number of factors, such as genetic back-ground, age, ethnicity and a family history have an impact on the risk of PCA. Currently, the clinical treatments of prostate cancer mainly contain surgery with adjuvant endocrine therapy, chemotherapy, and gene therapy. Despite the great progress in the treatment of PCA, clinical outcomes are unsatisfactory occasionally because of resistance to androgen deprivation therapy, heterogeneity of PCA and some other reasons. Therefore, a better understanding of the molecular events underlying the 2
Page 2 of 16
prostate cancer tumorigenesis may aid not only in the development of novel targets for PCA, but also in the improvement of the current strategies to prolong the survival of patients. Accumulation of genetic and epigenetic alterations is one of the most important factors in the tumorigenesis and progression of PCA. Many genes are identified to be involved in PCA, such as FNDC3B, PCAF, SOCS1, EZH2, PSGR et al. It has been reported that some of these genes are under the regulation of miRNAs. It is estimated that one-third of genes in the human genome are regulated by miRNAs. MiRNAs are small (~22 nucleotides) non-coding RNAs with gene regulatory functions by binding to the 3’-untranslated region (3’-UTR) of their target mRNA resulting in either translational repression or mRNA degradation1. It has been reported that miRNAs have distinct expression profiles in various human diseases, particularly in cancers 2, 3. Amounts of researches have indicated that miRNAs are involved in prostate cancer development as either tumor suppressors or oncogenes4-7. In this study, we identified differentially expressed miRNA and genes between PCA and normal control by combining miRNA and gene expression data from three microarray studies, and detected miRNA-target gene pairs with inverse expression based on miRWalk database to construct a miRNA-target gene regulatory network. Our findings may provide a contribution to elucidating the mechanisms of PCA. Materials and Methods Gene expression profiles We searched the Gene Expression Omnibus database (GEO, http://www.ncbi.nlm.nih.gov/geo) for miRNA and gene expression datasets. GEO served as a public repository for gene expression DataSets, initiated by the growing demand for a public repository for high-throughput gene expression data8. We only retained datasets that analyzed both miRNA and gene expression profiling of PCA in one study to minimize the heterogeneity. 3
Page 3 of 16
Differential analysis of miRNA and genes Due to the heterogeneity of multiple microarray datasets caused by different platforms, gene nomenclature and clinical simples, the raw data of each study need to be preprocessed with Quantile normalization and log2 transformation. The MATrix LABoratory (MATLAB) software was used to identify the differentially expressed probe sets in the tumor tissues compared to the normal tissues by two-tailed Student’s t-test, and then p-value and effect size of individual microarray study were calculated. Fisher’s method was used to combine p-value from individual microarray studies, and the random effects model was used to combine effect size of individual studies. We selected differently expressed miRNA with criterion of p-value < 0.01 and effect size > 0.8, while for differentially expressed genes, the criterion of p-value < 0.01 and effect size > 1 was applied. Identification of differently expressed miRNA target genes As the biological functions of miRNAs are preformed through their regulation of target protein expression, precise miRNA target prediction is important for the research of miRNA function. Putative targets of differentially expressed miRNAs were predicted by six bioinformatic algorithms (DIANAmT，miRanda，miRDB，miRWalk，PICTAR andTargetscan) with the online tools of miRWalk (http://www.umm.uniheidelberg.de/apps/zmf/mirwalk/)9, and the targets recorded by ≥ 4 algorithms or verified by experiment were selected to compare with the identified differentially expressed genes. As miRNAs tend to down-regulate the expression of their target genes, we selected target genes expressed inversely with corresponding miRNA to subject to further investigation10-12. Functional annotation Gene Ontology (GO) classification was performed to gain insights into the biological functions of predicted miRNA target genes. Kyoto Encyclopedia of Genes and Geno differently expressed mes (KEGG) pathway enrichment analysis13 was also employed to 4
Page 4 of 16
detect the potential pathway that the predicted miRNA target genes was involved in. Functional annotation was achieved based the online software GENECODIS14. Constructing regulatory network between miRNAs and their target genes The posttranscriptional regulatory network is defined as a directed and bipartite graph in which miRNAs and interacting target genes were expressed inversely. We conducted a PCAspecific regulatory network with miRNA-target gene interacting pairs we identified, and visualized by Cytoscape. Survival analysis through SurvExpress database SurvMicro (http://bioinformatica.mty.itesm.mx:8080/Biomatec/Survmicro.jsp), an online validation tool, could be used to identify survival miRNA biomarkers in human cancer datasets mainly from GEO (http://www.ncbi.nlm.nih.gov/geo/) and TCGA (https://tcgadata.nci.nih.gov) 15. In the present study, survival analysis was performed to evaluate the correlation between expression level of differentially expressed miRNA and the overall survival time of PCA patients through the online validation tool of SurvMicro. Result Differentially expressed miRNAs and genes in the PCA In this work, we collected a total of 3 microarray studies, and it contains197 samples of PCA and 43 samples of normal control, respectively (Supplementary Table 1). After normalization of the original miRNA and gene expression data, we performed differential expressed analysis between PCA and normal control samples using MATLAB. Finally, 29 miRNAs were regarded as significantly differentially expressed miRNA under the threshold of p-value < 0.01 and effect size > 0.8, with 10 up-regulated and 19 down-regulated miRNAs (Table 1). The up-regulated miRNA with the lowest p-value was hsa-miR-183, which was confirmed to be up-regulated in PCA as an oncogene targeting Dkk-3 and SMAD416. The down-regulated miRNA with the lowest p-value was hsa-miR-222, which was in accordance 5
Page 5 of 16
with previous study17. A total of 946 genes were identified to be aberrantly expressed in PCA, including 177 down-regulated genes and 769 up-regulated genes. The full list of these genes was provided in Supplementary material (Supplementary Table 2). Identification of differently expressed miRNA target genes The target genes of differentially expressed miRNA in the PCA were predicted by six bioinformatic algorithms, furthermore we searched miRWalk databases for miRNA-target gene pairs with experiment validation. To get more genuine target genes, we compared the target genes with the list of differentially expressed genes. The miRNA-target gene pairs that displayed inverse expression pattern in PCA were selected. As a result, we identified 600 miRNA-target gene pairs for the up-regulated miRNA with 24 validated miRNA-target gene pairs, and 151 miRNA-target gene pairs for the down-regulated miRNA with 35 validated miRNA-target gene pairs (Supplementary Table 3). GO classification and KEGG pathway enrichment of miRNA target genes We performed GO classification analysis for differently expressed miRNA target genes, and found that negative regulation of transcription from RNA polymerase II promoter (GO: 0000122, P=7.19E-07) and signal transduction (GO: 0007165, P=5.14E-06) were significantly enriched for biological processes. While for molecular functions, protein binding (GO: 0005515, P=2.84E-14) and metal ion binding (GO: 0046872, P=5.90E-11) were significantly enriched, and for cellular component, cytoplasm (GO: 0005737, P=5.15E-22) and plasma membrane (GO: 0005886, P=5.34E-18) were significantly enriched (Supplementary Table 4). We also performed the KEGG pathway enrichment analysis for differently expressed miRNA target genes. Hypergeometric test with p-value < 0.05 was used as the selected criteria for significantly enriched pathways. The most significant pathway in our analysis was
6
Page 6 of 16
Pathways in cancer (P=4.92E-06). Furthermore, Focal adhesion (P=1.82E-04) and Melanogenesis (P=5.19E-04) are also highly enriched (Table 2). The regulatory network of miRNAs and target genes The PCA-specific posttranscriptional regulatory network was constructed with the miRNA-target gene pairs by Cytoscape software. As a result, 751 miRNA-target gene pairs with inverse expression were consisted of 26 miRNAs and 385 target genes (Figure 1). In this network, ten genes (BCL2, BNC2, CCND2, EPM2A, MRAS, NAV2, RASL12, STK33, TCEAL1, WWC2) were co-regulated by five miRNAs (hsa-miR-106b, hsa-miR-130b, hsamiR-93, hsa-miR-153, hsa-miR-182) (Supplementary Table 5). Survival analysis through SurvExpress database With the dataset from GEO database (GSE26245), Kaplan–Meier curves indicated that the expression of hsa-miR-182 was significantly correlated with the overall survival time of PCA (p=0.001504). In addition, the expression level in high risk group was significantly higher than that in low risk group, which was in accordance with our findings (p=6.87e-21) (Figure 2). Discussion It has been widely accepted that oncogenesis and tumor progression is initiated through a deregulated expression of oncogenes and tumor suppressor genes which further triggers the malignant transformation of the affected cells. miRNAs, as post-transcriptional regulators of around 30% of the human genome, are becoming more and more necessary to understand the mechanisms leading to cancer. Widespread deregulation of miRNA expression occurs in human prostate cancer. In this study, after combining with miRNA and gene expression data in public database, we more accurately speculated new miRNA regulation activities in the process of PCA.
7
Page 7 of 16
29 differentially expressed miRNAs were identified by combining three miRNA expression datasets. Interestingly, most of them have previously been found to be involved in the development of PCA, such as miR-153, miR-183, miR-96, miR-221, miR-222, and miR133a, etc. miR-153 was identified to be up-regulated in PCA, and targeted PTEN to promote cell cycle transition and cell proliferation in PCA18. miR-183 was also significantly upregulated in PCA tissues, and in vitro and in vivo experiments demonstrated that the knockdown of miR-183 inhibited cell growth in PCA cells by targeting Dkk-3 and SMAD416. miR-96 was detected to be markedly up-regulated in prostate cancer cell and tissues by microarray method and RT-PCR analysis19. miR-221 and miR-222 was showed to affect the proliferation potential of the human PCA cell by targeting the tumor suppressor p27(Kip1)20. miR-133a was frequently down-regulated in PCA and functioned as tumor suppressors by targeting PNP21. However, in this study, we also discovered five novel miRNAs, such as hsa-miR-181c, hsa-miR-139-5p, hsa-miR-505, hsa-miR-324-5p, hsa-miR502-5p. Further studies are needed to uncover the role of those miRNAs in the development of PCA. miRNAs perform their regulatory function by degrading or inhibiting the translation of target genes. Consequently it is of vital importance to identify miRNA target genes to understand the biological functions of miRNAs. We predicted target genes of miRNA exhibited differential expression by the miRWalk database. Based on the fact that miRNA could negatively control the target gene expression, we matched target genes with the list of differentially expressed genes to identify miRNA-target gene pairs with an inverse expression in PCA. As a result, 751 miRNA-target gene pairs were identified, consisted of 26 miRNAs and 385 genes. A miRNA-target gene regulatory network was constructed with miRNA-target gene pairs. In this network, ten genes (BCL2, BNC2, CCND2, EPM2A, MRAS, NAV2, RASL12, STK33, TCEAL1, WWC2) were co-regulated by five miRNAs (hsa-miR-182, hsamiR-106b, hsa-miR-130b, hsa-miR-93, hsa-miR-153), suggesting that the five miRNAs and ten genes may play important roles in the tumorigenesis of PCA. The overexpression of miR8
Page 8 of 16
182 in PCA and other types of cancer suggested its linkage to malignancy20. Recently, more than one study revealed that miR-182 could be used as promising biomarkers for early diagnosis of PCA. Additionally, miR-182 was associated with biochemical and clinical progression-free survival22, 23, which was also consistent with our results of survival analysis, suggesting its potential use of prognostic biomarker in PCA. Apoptotic protein BCL2 was a potential target gene of miR-182 proved by the present and previous studies in PCA24. The overexpression of BCL2 was associated with adverse outcome in this malignancy25. BNC2 targeted by hsa-miR-106b, hsa-miR-130b, hsa-miR-93, hsa-miR-153, hsa-miR-182 in our analysis was found to be associated with PCA progression26 by evaluating the sequence variants in oestrogen response elements (EREs) of BNC2 in a cohort of 601 men with advanced prostate cancer treated with ADT. CCND2, a crucial cell cycle-regulatory gene, also targeted by the five miRNAs in our analysis is aberrantly expressed in PCA27 and many other cancers to regulate cancer cell growth. To estimate the biological roles of differentially expressed miRNAs in PCA, we performed GO and KEGG pathway enrichment analysis of the 385 miRNA target genes. The most enriched GO term of the target genes for biological processes was negative regulation of transcription from RNA polymerase II promoter. Some of the biological function may be correlated with the development of PCA including signal transduction, cell adhesion, response to hypoxia and canonical Wnt receptor signaling pathway28-30. KEGG pathway enrichment analysis showed that Pathways in cancer was statistically enriched. We found that many of the target genes were involved in Pathways in cancer, including TCF7L1, MET, LAMA3, RARB, COL4A6, WNT2B, etc. These genes functioned as oncogenes and tumor suppressor genes or involved in the regulation of oncogenes and tumor suppressor genes to trigger the tumorigenesis of PCA. In summary, we identified 29 differentially expressed miRNAs, 946 differentially expressed mRNAs, and constructed a regulatory network including 751 miRNA-target gene 9
Page 9 of 16
pairs. We also showed the potential use of hsa-miR-182 as a diagnostic or prognostic biomarker in PCA. Our data are helpful for elucidating the molecular mechanism of tumorigenesis of PCA. Acknowledgments None. Declaration of conflicting interests The Authors declare that there is no conflict of interest. References 1.
Winter J, Diederichs S. MicroRNA biogenesis and cancer. Methods in molecular biology. 2011;676:3-22.
2.
Volinia S, Calin GA, Liu CG, et al. A microRNA expression signature of human solid tumors defines cancer gene targets. Proceedings of the National Academy of Sciences of the United States of America. 2006;103:2257-2261.
3.
Lu J, Getz G, Miska EA, et al. MicroRNA expression profiles classify human cancers. Nature. 2005;435:834-838.
4.
Porkka KP, Pfeiffer MJ, Waltering KK, Vessella RL, Tammela TL, Visakorpi T. MicroRNA expression profiling in prostate cancer. Cancer research. 2007;67:61306135.
5.
Ambs S, Prueitt RL, Yi M, et al. Genomic profiling of microRNA and messenger RNA reveals deregulated microRNA expression in prostate cancer. Cancer research. 2008;68:6162-6170.
6.
Ozen M, Creighton CJ, Ozdemir M, Ittmann M. Widespread deregulation of microRNA expression in human prostate cancer. Oncogene. 2008;27:1788-1793.
7.
Chen X, Gong J, Zeng H, et al. MicroRNA145 targets BNIP3 and suppresses prostate cancer progression. Cancer research. 2010;70:2728-2738. 10
Page 10 of 16
8.
Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic acids research. 2002;30:207-210.
9.
Dweep H, Sticht C, Pandey P, Gretz N. miRWalk--database: prediction of possible miRNA binding sites by "walking" the genes of three genomes. Journal of biomedical informatics. 2011;44:839-847.
10.
Luo Z, Zhang L, Li Z, et al. An in silico analysis of dynamic changes in microRNA expression profiles in stepwise development of nasopharyngeal carcinoma. BMC medical genomics. 2012;5:3.
11.
Lionetti M, Biasiolo M, Agnelli L, et al. Identification of microRNA expression patterns and definition of a microRNA/mRNA regulatory network in distinct molecular groups of multiple myeloma. Blood. 2009;114:e20-26.
12.
Enerly E, Steinfeld I, Kleivi K, et al. miRNA-mRNA integrated analysis reveals roles for miRNAs in primary breast tumors. PloS one. 2011;6:e16915.
13.
Altermann E, Klaenhammer TR. PathwayVoyager: pathway mapping using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. BMC genomics. 2005;6:60.
14.
Tabas-Madrid D, Nogales-Cadenas R, Pascual-Montano A. GeneCodis3: a nonredundant and modular enrichment analysis tool for functional genomics. Nucleic acids research. 2012;40:W478-483.
15.
Aguirre-Gamboa R, Gomez-Rueda H, Martinez-Ledesma E, et al. SurvExpress: an online biomarker validation tool and database for cancer gene expression data using survival analysis. PloS one. 2013;8:e74250.
16.
Ueno K, Hirata H, Shahryari V, et al. microRNA-183 is an oncogene targeting Dkk-3 and SMAD4 in prostate cancer. British journal of cancer. 2013;108:1659-1667.
17.
Schaefer A, Jung M, Mollenkopf HJ, et al. Diagnostic and prognostic implications of microRNA profiling in prostate carcinoma. International journal of cancer. Journal international du cancer. 2010;126:1166-1176. 11
Page 11 of 16
18.
Wu Z, He B, He J, Mao X. Upregulation of miR-153 promotes cell proliferation via downregulation of the PTEN tumor suppressor gene in human prostate cancer. The Prostate. 2013;73:596-604.
19.
Yu JJ, Wu YX, Zhao FJ, Xia SJ. miR-96 promotes cell proliferation and clonogenicity by down-regulating of FOXO1 in prostate cancer cells. Medical oncology. 2014;31:910.
20.
Galardi S, Mercatelli N, Giorda E, et al. miR-221 and miR-222 expression affects the proliferation potential of human prostate carcinoma cell lines by targeting p27Kip1. The Journal of biological chemistry. 2007;282:23716-23724.
21.
Kojima S, Chiyomaru T, Kawakami K, et al. Tumour suppressors miR-1 and miR133a target the oncogenic function of purine nucleoside phosphorylase (PNP) in prostate cancer. British journal of cancer. 2012;106:405-413.
22.
Casanova-Salas I, Rubio-Briones J, Calatrava A, et al. Identification of miR-187 and miR-182 as biomarkers of early diagnosis and prognosis in patients with prostate cancer treated with radical prostatectomy. J Urol. 2014;192:252-259.
23.
Xiaoli Z, Yawei W, Lianna L, Haifeng L, Hui Z. Screening of Target Genes and Regulatory Function of miRNAs as Prognostic Indicators for Prostate Cancer. Medical science monitor : international medical journal of experimental and clinical research. 2015;21:3748-3759.
24.
Peng X, Li W, Yuan L, Mehta RG, Kopelovich L, McCormick DL. Inhibition of proliferation and induction of autophagy by atorvastatin in PC3 prostate cancer cells correlate with downregulation of Bcl2 and upregulation of miR-182 and p21. PloS one. 2013;8:e70442.
25.
Bachmann HS, Heukamp LC, Schmitz KJ, et al. Regulatory BCL2 promoter polymorphism (-938C>A) is associated with adverse outcome in patients with prostate carcinoma. International journal of cancer. Journal international du cancer. 2011;129:2390-2399. 12
Page 12 of 16
26.
Huang CN, Huang SP, Pao JB, et al. Genetic polymorphisms in oestrogen receptorbinding sites affect clinical outcomes in patients with prostate cancer receiving androgen-deprivation therapy. Journal of internal medicine. 2012;271:499-509.
27.
Dong Q, Meng P, Wang T, et al. MicroRNA let-7a inhibits proliferation of human prostate cancer cells in vitro and in vivo by targeting E2F2 and CCND2. PloS one. 2010;5:e10147.
28.
Bisson I, Prowse DM. WNT signaling regulates self-renewal and differentiation of prostate cancer cells with stem cell characteristics. Cell research. 2009;19:683-697.
29.
Hyder CL, Lazaro G, Pylvanainen JW, Roberts MW, Qvarnstrom SM, Eriksson JE. Nestin regulates prostate cancer cell invasion by influencing the localisation and functions of FAK and integrins. Journal of cell science. 2014;127:2161-2173.
30.
Bristow RG, Berlin A, Dal Pra A. An arranged marriage for precision medicine: hypoxia and genomic assays in localized prostate cancer radiotherapy. The British journal of radiology. 2014;87:20130753.
Figure 1 legend The regulatory network between miRNAs and target genes in PCA. The diamonds and ellipses represent the miRNAs and genes, respectively. The red and green colors represent the relatively high and low expression, respectively. The larger geometric drawing indicates the more miRNAs or genes interacted with it. Figure 2 legend The association between hsa-miR-182 gene expression and PCA survival in SurvExpress database. A: Survival analysis; B Risk assessment.
13
Page 13 of 16
Table 1 List of differentially expressed miRNAs MiRNAs
P-value
Eeffect Size
hsa-miR-183
2.71E-14
-1.3658
hsa-miR-153
2.74E-13
-1.3519
hsa-miR-96
2.33E-12
-1.2324
hsa-miR-25
2.26E-11
-1.1717
hsa-miR-93
1.69E-09
-1.0973
hsa-miR-182
1.80E-09
-0.98643
hsa-miR-663
2.62E-09
-0.87466
hsa-miR-106b
4.06E-09
-1.0368
hsa-miR-130b
3.52E-07
-0.98146
hsa-miR-18a
1.55E-06
-0.90428
hsa-miR-222
5.14E-13
1.3321
hsa-miR-224
1.25E-11
1.2899
hsa-miR-99b
3.18E-11
1.2091
hsa-miR-221
1.47E-10
1.0797
hsa-miR-204
4.83E-10
1.2168
hsa-miR-181c
6.80E-09
1.0934
hsa-miR-378
7.00E-09
0.82274
hsa-miR-452
7.18E-09
1.0563
hsa-miR-378*
1.64E-08
1.0709
hsa-miR-31
1.87E-08
1.0903
hsa-miR-139-5p
4.98E-08
0.80539
hsa-miR-505
1.73E-07
0.90227
Up-regulated miRNAs
Downregulated miRNAs
14
Page 14 of 16
hsa-miR-133a
3.72E-07
0.82767
hsa-miR-328
7.85E-07
0.96744
hsa-miR-27b
8.21E-07
0.86406
hsa-miR-154
1.81E-05
0.85747
hsa-miR-324-5p
2.06E-05
0.82328
hsa-miR-487b
3.69E-05
0.82271
hsa-miR-502-5p
3.86E-05
0.80137
Table 2. KEGG pathway enrichment analysis of differentailly expression miRNA target genes (Top 15) KEGG KEGG ID
Count
FDR
Genes
term TCF7L1,MET,LAMA3,RARB,COL4A6,W Pathways in hsa05200
NT2B,GLI3,COL4A5,FZD1,LAMB3,MYC, 18
4.92E-06
cancer
MITF,BCL2,SMAD3,DAPK1,STAT5B,STA T5A,FGFR2 MET,LAMA3,COL4A6,PDGFC,CCND2,C
hsa04510
Focal adhesion
12
1.82E-04
OL4A5,PARVA,CAV1,LAMB3,BCL2,ACT N4,PIP5K1C TCF7L1,ADCY2,GNAO1,WNT2B,PRKAC
hsa04916
Melanogenesis
8
5.19E-04 A,FZD1,GNAI2,MITF
Glutamatergic hsa04724
DLG4,ADCY2,GNAO1,CACNA1C,PRKAC 9
5.32E-04
synapse
A,GNB5,GNG11,GNAI2,TRPC1
Wnt signaling hsa04310
PRICKLE2,TCF7L1,WNT2B,CCND2,PRK 9
1.14E-03
pathway hsa05222
Small cell lung
ACA,FZD1,MYC,SMAD3,TBL1X 7
1.24E-03
LAMA3,RARB,COL4A6,COL4A5,LAMB3,
15
Page 15 of 16
cancer
MYC,BCL2
Purine hsa00230
ADCY2,PDE7B,PPAT,NPR2,PDE3B,PFAS, 9
1.54E-03
metabolism
PGM1,NT5E,PDE1C ROBO1,MET,SEMA5A,SEMA6D,CFL2,EF
hsa04360
Axon guidance
8
1.76E-03 NA5,SRGAP3,GNAI2
Pentose hsa00030
phosphate
4
3.34E-03
9
8.61E-03
ALDOC,PFKP,ALDOA,PGM1
pathway Regulation of hsa04810
MRAS,PDGFC,CFL2,MSN,PIP5K1B,TIAM
actin cytoskeleton
2,ACTN4,PIP5K1C,FGFR2 LAMA3,COL4A6,COL4A5,PRKACA,LAM
hsa05146
Amoebiasis
6
1.28E-02 B3,ACTN4
Chemokine hsa04062
ADCY2,PRKACA,CX3CL1,GNB5,GNG11, 8
1.32E-02
signaling pathway hsa04520
Adherens junction
GNAI2,TIAM2,STAT5B 5
1.38E-02
6
1.71E-02
TCF7L1,MET,SNAI2,SMAD3,ACTN4
Vascular smooth ADCY2,KCNMA1,CACNA1C,PRKACA,N hsa04270
muscle
PR2,KCNMB1 contraction ECM-receptor hsa04512
5
2.33E-02
SDC4,LAMA3,COL4A6,COL4A5,LAMB3
interaction
16
Page 16 of 16
S0090-4295(16)30149-2 http://dx.doi.org/doi: 10.1016/j.urology.2016.04.041 URL 19771
To appear in:
Urology
Received date: Accepted date:
1-3-2016 28-4-2016
Please cite this article as: Yanan Sun, Xiaopeng Jia, Lianguo Hou, Xing Liu, Screening of Differently Expressed miRNA and mRNA in Prostate Cancer by Integrated Analysis of Transcription Data, Urology (2016), http://dx.doi.org/doi: 10.1016/j.urology.2016.04.041. 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.
Screening of differently expressed miRNA and mRNA in prostate cancer by integrated analysis of transcription data Yanan Sun1, Xiaopeng Jia2*, Lianguo Hou3, Xing Liu4 1
Department of Obstetrics and Gynecology, Bethune International Peace Hospital of PLA, Shijiazhuang, Hebei 050071, China
2
Department of Urology, The Third Hospital of Hebei Medical University, Shijiazhuang,
Hebei 050051, China 3
Department of Biochemistry and Molecular Biology, Hebei Medical University,
Shijiazhuang, Hebei 050011, China 4
Department of Orthopaedic Trauma, The Third Hospital of Shijiazhuang City, Shijiazhuang,
Hebei 050011, China *
Corresponding author: Xiaopeng Jia
Department of Urology, The Third Hospital of Hebei Medical University, Shijiazhuang, Hebei 050051, China Email: [email protected] Tel: +86-0311-88602033 Fax: +86-0311-88602033
Abstract Objective: The purpose of this study was to screen aberrantly expressed miRNAs and genes in prostate cancer (PCA), and further uncover the underlying mechanisms for the development of PCA. Methods: We searched GEO database for miRNA and gene expression datasets of PCA, and then separately integrated miRNA and gene expression datasets to identify miRNA and gene expression profiles in PCA. Target genes of differentially expressed miRNAs were predicted through miRWalk database. We matched these target genes with the list of differentially 1
Page 1 of 16
expressed genes to identify miRNA-target gene pairs whose expression was inversely correlated. The function of these target genes was annotated. Results: 29 differentially expressed miRNAs and 946 differentially expressed genes were identified between PCA and normal control. 751 miRNA-target gene pairs that showed inverse expression in PCA, were obtained to establish a regulatory network. In this regulatory network, ten genes (BCL2, BNC2, CCND2, EPM2A, MRAS, NAV2, RASL12, STK33, TCEAL1, WWC2) were co-regulated by five miRNAs (hsa-miR-106b, hsa-miR-130b, hsamiR-93, hsa-miR-153, hsa-miR-182). The expression of hsa-miR-182 was significantly associated with PCA survival through the online validation tool of SurvMicro, suggesting the potential use as a diagnostic or prognostic biomarker in PCA. Conclusions: This integrated analysis was performed to infer new miRNA regulation activities, which provides insights into the understanding of underlying molecular mechanisms of PCA, and guides for exploration of novel therapeutic targets. Keywords: miRNA; differentially expressed; target gene; microarray; prostate cancer
Introduction Prostate cancer (PCA) is the second most frequently diagnosed tumor and the sixth leading cause of cancer-related death occurring in men worldwide. Previous studies have reported that a number of factors, such as genetic back-ground, age, ethnicity and a family history have an impact on the risk of PCA. Currently, the clinical treatments of prostate cancer mainly contain surgery with adjuvant endocrine therapy, chemotherapy, and gene therapy. Despite the great progress in the treatment of PCA, clinical outcomes are unsatisfactory occasionally because of resistance to androgen deprivation therapy, heterogeneity of PCA and some other reasons. Therefore, a better understanding of the molecular events underlying the 2
Page 2 of 16
prostate cancer tumorigenesis may aid not only in the development of novel targets for PCA, but also in the improvement of the current strategies to prolong the survival of patients. Accumulation of genetic and epigenetic alterations is one of the most important factors in the tumorigenesis and progression of PCA. Many genes are identified to be involved in PCA, such as FNDC3B, PCAF, SOCS1, EZH2, PSGR et al. It has been reported that some of these genes are under the regulation of miRNAs. It is estimated that one-third of genes in the human genome are regulated by miRNAs. MiRNAs are small (~22 nucleotides) non-coding RNAs with gene regulatory functions by binding to the 3’-untranslated region (3’-UTR) of their target mRNA resulting in either translational repression or mRNA degradation1. It has been reported that miRNAs have distinct expression profiles in various human diseases, particularly in cancers 2, 3. Amounts of researches have indicated that miRNAs are involved in prostate cancer development as either tumor suppressors or oncogenes4-7. In this study, we identified differentially expressed miRNA and genes between PCA and normal control by combining miRNA and gene expression data from three microarray studies, and detected miRNA-target gene pairs with inverse expression based on miRWalk database to construct a miRNA-target gene regulatory network. Our findings may provide a contribution to elucidating the mechanisms of PCA. Materials and Methods Gene expression profiles We searched the Gene Expression Omnibus database (GEO, http://www.ncbi.nlm.nih.gov/geo) for miRNA and gene expression datasets. GEO served as a public repository for gene expression DataSets, initiated by the growing demand for a public repository for high-throughput gene expression data8. We only retained datasets that analyzed both miRNA and gene expression profiling of PCA in one study to minimize the heterogeneity. 3
Page 3 of 16
Differential analysis of miRNA and genes Due to the heterogeneity of multiple microarray datasets caused by different platforms, gene nomenclature and clinical simples, the raw data of each study need to be preprocessed with Quantile normalization and log2 transformation. The MATrix LABoratory (MATLAB) software was used to identify the differentially expressed probe sets in the tumor tissues compared to the normal tissues by two-tailed Student’s t-test, and then p-value and effect size of individual microarray study were calculated. Fisher’s method was used to combine p-value from individual microarray studies, and the random effects model was used to combine effect size of individual studies. We selected differently expressed miRNA with criterion of p-value < 0.01 and effect size > 0.8, while for differentially expressed genes, the criterion of p-value < 0.01 and effect size > 1 was applied. Identification of differently expressed miRNA target genes As the biological functions of miRNAs are preformed through their regulation of target protein expression, precise miRNA target prediction is important for the research of miRNA function. Putative targets of differentially expressed miRNAs were predicted by six bioinformatic algorithms (DIANAmT，miRanda，miRDB，miRWalk，PICTAR andTargetscan) with the online tools of miRWalk (http://www.umm.uniheidelberg.de/apps/zmf/mirwalk/)9, and the targets recorded by ≥ 4 algorithms or verified by experiment were selected to compare with the identified differentially expressed genes. As miRNAs tend to down-regulate the expression of their target genes, we selected target genes expressed inversely with corresponding miRNA to subject to further investigation10-12. Functional annotation Gene Ontology (GO) classification was performed to gain insights into the biological functions of predicted miRNA target genes. Kyoto Encyclopedia of Genes and Geno differently expressed mes (KEGG) pathway enrichment analysis13 was also employed to 4
Page 4 of 16
detect the potential pathway that the predicted miRNA target genes was involved in. Functional annotation was achieved based the online software GENECODIS14. Constructing regulatory network between miRNAs and their target genes The posttranscriptional regulatory network is defined as a directed and bipartite graph in which miRNAs and interacting target genes were expressed inversely. We conducted a PCAspecific regulatory network with miRNA-target gene interacting pairs we identified, and visualized by Cytoscape. Survival analysis through SurvExpress database SurvMicro (http://bioinformatica.mty.itesm.mx:8080/Biomatec/Survmicro.jsp), an online validation tool, could be used to identify survival miRNA biomarkers in human cancer datasets mainly from GEO (http://www.ncbi.nlm.nih.gov/geo/) and TCGA (https://tcgadata.nci.nih.gov) 15. In the present study, survival analysis was performed to evaluate the correlation between expression level of differentially expressed miRNA and the overall survival time of PCA patients through the online validation tool of SurvMicro. Result Differentially expressed miRNAs and genes in the PCA In this work, we collected a total of 3 microarray studies, and it contains197 samples of PCA and 43 samples of normal control, respectively (Supplementary Table 1). After normalization of the original miRNA and gene expression data, we performed differential expressed analysis between PCA and normal control samples using MATLAB. Finally, 29 miRNAs were regarded as significantly differentially expressed miRNA under the threshold of p-value < 0.01 and effect size > 0.8, with 10 up-regulated and 19 down-regulated miRNAs (Table 1). The up-regulated miRNA with the lowest p-value was hsa-miR-183, which was confirmed to be up-regulated in PCA as an oncogene targeting Dkk-3 and SMAD416. The down-regulated miRNA with the lowest p-value was hsa-miR-222, which was in accordance 5
Page 5 of 16
with previous study17. A total of 946 genes were identified to be aberrantly expressed in PCA, including 177 down-regulated genes and 769 up-regulated genes. The full list of these genes was provided in Supplementary material (Supplementary Table 2). Identification of differently expressed miRNA target genes The target genes of differentially expressed miRNA in the PCA were predicted by six bioinformatic algorithms, furthermore we searched miRWalk databases for miRNA-target gene pairs with experiment validation. To get more genuine target genes, we compared the target genes with the list of differentially expressed genes. The miRNA-target gene pairs that displayed inverse expression pattern in PCA were selected. As a result, we identified 600 miRNA-target gene pairs for the up-regulated miRNA with 24 validated miRNA-target gene pairs, and 151 miRNA-target gene pairs for the down-regulated miRNA with 35 validated miRNA-target gene pairs (Supplementary Table 3). GO classification and KEGG pathway enrichment of miRNA target genes We performed GO classification analysis for differently expressed miRNA target genes, and found that negative regulation of transcription from RNA polymerase II promoter (GO: 0000122, P=7.19E-07) and signal transduction (GO: 0007165, P=5.14E-06) were significantly enriched for biological processes. While for molecular functions, protein binding (GO: 0005515, P=2.84E-14) and metal ion binding (GO: 0046872, P=5.90E-11) were significantly enriched, and for cellular component, cytoplasm (GO: 0005737, P=5.15E-22) and plasma membrane (GO: 0005886, P=5.34E-18) were significantly enriched (Supplementary Table 4). We also performed the KEGG pathway enrichment analysis for differently expressed miRNA target genes. Hypergeometric test with p-value < 0.05 was used as the selected criteria for significantly enriched pathways. The most significant pathway in our analysis was
6
Page 6 of 16
Pathways in cancer (P=4.92E-06). Furthermore, Focal adhesion (P=1.82E-04) and Melanogenesis (P=5.19E-04) are also highly enriched (Table 2). The regulatory network of miRNAs and target genes The PCA-specific posttranscriptional regulatory network was constructed with the miRNA-target gene pairs by Cytoscape software. As a result, 751 miRNA-target gene pairs with inverse expression were consisted of 26 miRNAs and 385 target genes (Figure 1). In this network, ten genes (BCL2, BNC2, CCND2, EPM2A, MRAS, NAV2, RASL12, STK33, TCEAL1, WWC2) were co-regulated by five miRNAs (hsa-miR-106b, hsa-miR-130b, hsamiR-93, hsa-miR-153, hsa-miR-182) (Supplementary Table 5). Survival analysis through SurvExpress database With the dataset from GEO database (GSE26245), Kaplan–Meier curves indicated that the expression of hsa-miR-182 was significantly correlated with the overall survival time of PCA (p=0.001504). In addition, the expression level in high risk group was significantly higher than that in low risk group, which was in accordance with our findings (p=6.87e-21) (Figure 2). Discussion It has been widely accepted that oncogenesis and tumor progression is initiated through a deregulated expression of oncogenes and tumor suppressor genes which further triggers the malignant transformation of the affected cells. miRNAs, as post-transcriptional regulators of around 30% of the human genome, are becoming more and more necessary to understand the mechanisms leading to cancer. Widespread deregulation of miRNA expression occurs in human prostate cancer. In this study, after combining with miRNA and gene expression data in public database, we more accurately speculated new miRNA regulation activities in the process of PCA.
7
Page 7 of 16
29 differentially expressed miRNAs were identified by combining three miRNA expression datasets. Interestingly, most of them have previously been found to be involved in the development of PCA, such as miR-153, miR-183, miR-96, miR-221, miR-222, and miR133a, etc. miR-153 was identified to be up-regulated in PCA, and targeted PTEN to promote cell cycle transition and cell proliferation in PCA18. miR-183 was also significantly upregulated in PCA tissues, and in vitro and in vivo experiments demonstrated that the knockdown of miR-183 inhibited cell growth in PCA cells by targeting Dkk-3 and SMAD416. miR-96 was detected to be markedly up-regulated in prostate cancer cell and tissues by microarray method and RT-PCR analysis19. miR-221 and miR-222 was showed to affect the proliferation potential of the human PCA cell by targeting the tumor suppressor p27(Kip1)20. miR-133a was frequently down-regulated in PCA and functioned as tumor suppressors by targeting PNP21. However, in this study, we also discovered five novel miRNAs, such as hsa-miR-181c, hsa-miR-139-5p, hsa-miR-505, hsa-miR-324-5p, hsa-miR502-5p. Further studies are needed to uncover the role of those miRNAs in the development of PCA. miRNAs perform their regulatory function by degrading or inhibiting the translation of target genes. Consequently it is of vital importance to identify miRNA target genes to understand the biological functions of miRNAs. We predicted target genes of miRNA exhibited differential expression by the miRWalk database. Based on the fact that miRNA could negatively control the target gene expression, we matched target genes with the list of differentially expressed genes to identify miRNA-target gene pairs with an inverse expression in PCA. As a result, 751 miRNA-target gene pairs were identified, consisted of 26 miRNAs and 385 genes. A miRNA-target gene regulatory network was constructed with miRNA-target gene pairs. In this network, ten genes (BCL2, BNC2, CCND2, EPM2A, MRAS, NAV2, RASL12, STK33, TCEAL1, WWC2) were co-regulated by five miRNAs (hsa-miR-182, hsamiR-106b, hsa-miR-130b, hsa-miR-93, hsa-miR-153), suggesting that the five miRNAs and ten genes may play important roles in the tumorigenesis of PCA. The overexpression of miR8
Page 8 of 16
182 in PCA and other types of cancer suggested its linkage to malignancy20. Recently, more than one study revealed that miR-182 could be used as promising biomarkers for early diagnosis of PCA. Additionally, miR-182 was associated with biochemical and clinical progression-free survival22, 23, which was also consistent with our results of survival analysis, suggesting its potential use of prognostic biomarker in PCA. Apoptotic protein BCL2 was a potential target gene of miR-182 proved by the present and previous studies in PCA24. The overexpression of BCL2 was associated with adverse outcome in this malignancy25. BNC2 targeted by hsa-miR-106b, hsa-miR-130b, hsa-miR-93, hsa-miR-153, hsa-miR-182 in our analysis was found to be associated with PCA progression26 by evaluating the sequence variants in oestrogen response elements (EREs) of BNC2 in a cohort of 601 men with advanced prostate cancer treated with ADT. CCND2, a crucial cell cycle-regulatory gene, also targeted by the five miRNAs in our analysis is aberrantly expressed in PCA27 and many other cancers to regulate cancer cell growth. To estimate the biological roles of differentially expressed miRNAs in PCA, we performed GO and KEGG pathway enrichment analysis of the 385 miRNA target genes. The most enriched GO term of the target genes for biological processes was negative regulation of transcription from RNA polymerase II promoter. Some of the biological function may be correlated with the development of PCA including signal transduction, cell adhesion, response to hypoxia and canonical Wnt receptor signaling pathway28-30. KEGG pathway enrichment analysis showed that Pathways in cancer was statistically enriched. We found that many of the target genes were involved in Pathways in cancer, including TCF7L1, MET, LAMA3, RARB, COL4A6, WNT2B, etc. These genes functioned as oncogenes and tumor suppressor genes or involved in the regulation of oncogenes and tumor suppressor genes to trigger the tumorigenesis of PCA. In summary, we identified 29 differentially expressed miRNAs, 946 differentially expressed mRNAs, and constructed a regulatory network including 751 miRNA-target gene 9
Page 9 of 16
pairs. We also showed the potential use of hsa-miR-182 as a diagnostic or prognostic biomarker in PCA. Our data are helpful for elucidating the molecular mechanism of tumorigenesis of PCA. Acknowledgments None. Declaration of conflicting interests The Authors declare that there is no conflict of interest. References 1.
Winter J, Diederichs S. MicroRNA biogenesis and cancer. Methods in molecular biology. 2011;676:3-22.
2.
Volinia S, Calin GA, Liu CG, et al. A microRNA expression signature of human solid tumors defines cancer gene targets. Proceedings of the National Academy of Sciences of the United States of America. 2006;103:2257-2261.
3.
Lu J, Getz G, Miska EA, et al. MicroRNA expression profiles classify human cancers. Nature. 2005;435:834-838.
4.
Porkka KP, Pfeiffer MJ, Waltering KK, Vessella RL, Tammela TL, Visakorpi T. MicroRNA expression profiling in prostate cancer. Cancer research. 2007;67:61306135.
5.
Ambs S, Prueitt RL, Yi M, et al. Genomic profiling of microRNA and messenger RNA reveals deregulated microRNA expression in prostate cancer. Cancer research. 2008;68:6162-6170.
6.
Ozen M, Creighton CJ, Ozdemir M, Ittmann M. Widespread deregulation of microRNA expression in human prostate cancer. Oncogene. 2008;27:1788-1793.
7.
Chen X, Gong J, Zeng H, et al. MicroRNA145 targets BNIP3 and suppresses prostate cancer progression. Cancer research. 2010;70:2728-2738. 10
Page 10 of 16
8.
Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic acids research. 2002;30:207-210.
9.
Dweep H, Sticht C, Pandey P, Gretz N. miRWalk--database: prediction of possible miRNA binding sites by "walking" the genes of three genomes. Journal of biomedical informatics. 2011;44:839-847.
10.
Luo Z, Zhang L, Li Z, et al. An in silico analysis of dynamic changes in microRNA expression profiles in stepwise development of nasopharyngeal carcinoma. BMC medical genomics. 2012;5:3.
11.
Lionetti M, Biasiolo M, Agnelli L, et al. Identification of microRNA expression patterns and definition of a microRNA/mRNA regulatory network in distinct molecular groups of multiple myeloma. Blood. 2009;114:e20-26.
12.
Enerly E, Steinfeld I, Kleivi K, et al. miRNA-mRNA integrated analysis reveals roles for miRNAs in primary breast tumors. PloS one. 2011;6:e16915.
13.
Altermann E, Klaenhammer TR. PathwayVoyager: pathway mapping using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. BMC genomics. 2005;6:60.
14.
Tabas-Madrid D, Nogales-Cadenas R, Pascual-Montano A. GeneCodis3: a nonredundant and modular enrichment analysis tool for functional genomics. Nucleic acids research. 2012;40:W478-483.
15.
Aguirre-Gamboa R, Gomez-Rueda H, Martinez-Ledesma E, et al. SurvExpress: an online biomarker validation tool and database for cancer gene expression data using survival analysis. PloS one. 2013;8:e74250.
16.
Ueno K, Hirata H, Shahryari V, et al. microRNA-183 is an oncogene targeting Dkk-3 and SMAD4 in prostate cancer. British journal of cancer. 2013;108:1659-1667.
17.
Schaefer A, Jung M, Mollenkopf HJ, et al. Diagnostic and prognostic implications of microRNA profiling in prostate carcinoma. International journal of cancer. Journal international du cancer. 2010;126:1166-1176. 11
Page 11 of 16
18.
Wu Z, He B, He J, Mao X. Upregulation of miR-153 promotes cell proliferation via downregulation of the PTEN tumor suppressor gene in human prostate cancer. The Prostate. 2013;73:596-604.
19.
Yu JJ, Wu YX, Zhao FJ, Xia SJ. miR-96 promotes cell proliferation and clonogenicity by down-regulating of FOXO1 in prostate cancer cells. Medical oncology. 2014;31:910.
20.
Galardi S, Mercatelli N, Giorda E, et al. miR-221 and miR-222 expression affects the proliferation potential of human prostate carcinoma cell lines by targeting p27Kip1. The Journal of biological chemistry. 2007;282:23716-23724.
21.
Kojima S, Chiyomaru T, Kawakami K, et al. Tumour suppressors miR-1 and miR133a target the oncogenic function of purine nucleoside phosphorylase (PNP) in prostate cancer. British journal of cancer. 2012;106:405-413.
22.
Casanova-Salas I, Rubio-Briones J, Calatrava A, et al. Identification of miR-187 and miR-182 as biomarkers of early diagnosis and prognosis in patients with prostate cancer treated with radical prostatectomy. J Urol. 2014;192:252-259.
23.
Xiaoli Z, Yawei W, Lianna L, Haifeng L, Hui Z. Screening of Target Genes and Regulatory Function of miRNAs as Prognostic Indicators for Prostate Cancer. Medical science monitor : international medical journal of experimental and clinical research. 2015;21:3748-3759.
24.
Peng X, Li W, Yuan L, Mehta RG, Kopelovich L, McCormick DL. Inhibition of proliferation and induction of autophagy by atorvastatin in PC3 prostate cancer cells correlate with downregulation of Bcl2 and upregulation of miR-182 and p21. PloS one. 2013;8:e70442.
25.
Bachmann HS, Heukamp LC, Schmitz KJ, et al. Regulatory BCL2 promoter polymorphism (-938C>A) is associated with adverse outcome in patients with prostate carcinoma. International journal of cancer. Journal international du cancer. 2011;129:2390-2399. 12
Page 12 of 16
26.
Huang CN, Huang SP, Pao JB, et al. Genetic polymorphisms in oestrogen receptorbinding sites affect clinical outcomes in patients with prostate cancer receiving androgen-deprivation therapy. Journal of internal medicine. 2012;271:499-509.
27.
Dong Q, Meng P, Wang T, et al. MicroRNA let-7a inhibits proliferation of human prostate cancer cells in vitro and in vivo by targeting E2F2 and CCND2. PloS one. 2010;5:e10147.
28.
Bisson I, Prowse DM. WNT signaling regulates self-renewal and differentiation of prostate cancer cells with stem cell characteristics. Cell research. 2009;19:683-697.
29.
Hyder CL, Lazaro G, Pylvanainen JW, Roberts MW, Qvarnstrom SM, Eriksson JE. Nestin regulates prostate cancer cell invasion by influencing the localisation and functions of FAK and integrins. Journal of cell science. 2014;127:2161-2173.
30.
Bristow RG, Berlin A, Dal Pra A. An arranged marriage for precision medicine: hypoxia and genomic assays in localized prostate cancer radiotherapy. The British journal of radiology. 2014;87:20130753.
Figure 1 legend The regulatory network between miRNAs and target genes in PCA. The diamonds and ellipses represent the miRNAs and genes, respectively. The red and green colors represent the relatively high and low expression, respectively. The larger geometric drawing indicates the more miRNAs or genes interacted with it. Figure 2 legend The association between hsa-miR-182 gene expression and PCA survival in SurvExpress database. A: Survival analysis; B Risk assessment.
13
Page 13 of 16
Table 1 List of differentially expressed miRNAs MiRNAs
P-value
Eeffect Size
hsa-miR-183
2.71E-14
-1.3658
hsa-miR-153
2.74E-13
-1.3519
hsa-miR-96
2.33E-12
-1.2324
hsa-miR-25
2.26E-11
-1.1717
hsa-miR-93
1.69E-09
-1.0973
hsa-miR-182
1.80E-09
-0.98643
hsa-miR-663
2.62E-09
-0.87466
hsa-miR-106b
4.06E-09
-1.0368
hsa-miR-130b
3.52E-07
-0.98146
hsa-miR-18a
1.55E-06
-0.90428
hsa-miR-222
5.14E-13
1.3321
hsa-miR-224
1.25E-11
1.2899
hsa-miR-99b
3.18E-11
1.2091
hsa-miR-221
1.47E-10
1.0797
hsa-miR-204
4.83E-10
1.2168
hsa-miR-181c
6.80E-09
1.0934
hsa-miR-378
7.00E-09
0.82274
hsa-miR-452
7.18E-09
1.0563
hsa-miR-378*
1.64E-08
1.0709
hsa-miR-31
1.87E-08
1.0903
hsa-miR-139-5p
4.98E-08
0.80539
hsa-miR-505
1.73E-07
0.90227
Up-regulated miRNAs
Downregulated miRNAs
14
Page 14 of 16
hsa-miR-133a
3.72E-07
0.82767
hsa-miR-328
7.85E-07
0.96744
hsa-miR-27b
8.21E-07
0.86406
hsa-miR-154
1.81E-05
0.85747
hsa-miR-324-5p
2.06E-05
0.82328
hsa-miR-487b
3.69E-05
0.82271
hsa-miR-502-5p
3.86E-05
0.80137
Table 2. KEGG pathway enrichment analysis of differentailly expression miRNA target genes (Top 15) KEGG KEGG ID
Count
FDR
Genes
term TCF7L1,MET,LAMA3,RARB,COL4A6,W Pathways in hsa05200
NT2B,GLI3,COL4A5,FZD1,LAMB3,MYC, 18
4.92E-06
cancer
MITF,BCL2,SMAD3,DAPK1,STAT5B,STA T5A,FGFR2 MET,LAMA3,COL4A6,PDGFC,CCND2,C
hsa04510
Focal adhesion
12
1.82E-04
OL4A5,PARVA,CAV1,LAMB3,BCL2,ACT N4,PIP5K1C TCF7L1,ADCY2,GNAO1,WNT2B,PRKAC
hsa04916
Melanogenesis
8
5.19E-04 A,FZD1,GNAI2,MITF
Glutamatergic hsa04724
DLG4,ADCY2,GNAO1,CACNA1C,PRKAC 9
5.32E-04
synapse
A,GNB5,GNG11,GNAI2,TRPC1
Wnt signaling hsa04310
PRICKLE2,TCF7L1,WNT2B,CCND2,PRK 9
1.14E-03
pathway hsa05222
Small cell lung
ACA,FZD1,MYC,SMAD3,TBL1X 7
1.24E-03
LAMA3,RARB,COL4A6,COL4A5,LAMB3,
15
Page 15 of 16
cancer
MYC,BCL2
Purine hsa00230
ADCY2,PDE7B,PPAT,NPR2,PDE3B,PFAS, 9
1.54E-03
metabolism
PGM1,NT5E,PDE1C ROBO1,MET,SEMA5A,SEMA6D,CFL2,EF
hsa04360
Axon guidance
8
1.76E-03 NA5,SRGAP3,GNAI2
Pentose hsa00030
phosphate
4
3.34E-03
9
8.61E-03
ALDOC,PFKP,ALDOA,PGM1
pathway Regulation of hsa04810
MRAS,PDGFC,CFL2,MSN,PIP5K1B,TIAM
actin cytoskeleton
2,ACTN4,PIP5K1C,FGFR2 LAMA3,COL4A6,COL4A5,PRKACA,LAM
hsa05146
Amoebiasis
6
1.28E-02 B3,ACTN4
Chemokine hsa04062
ADCY2,PRKACA,CX3CL1,GNB5,GNG11, 8
1.32E-02
signaling pathway hsa04520
Adherens junction
GNAI2,TIAM2,STAT5B 5
1.38E-02
6
1.71E-02
TCF7L1,MET,SNAI2,SMAD3,ACTN4
Vascular smooth ADCY2,KCNMA1,CACNA1C,PRKACA,N hsa04270
muscle
PR2,KCNMB1 contraction ECM-receptor hsa04512
5
2.33E-02
SDC4,LAMA3,COL4A6,COL4A5,LAMB3
interaction
16
Page 16 of 16