Sage Journals: Discover world-class research

Abstract

We aimed to identify key genes associated with prostate cancer using RNA-sequencing (RNA-seq) data. RNA-seq data, including 1 cancer sample and 1 adjacent normal sample, were downloaded from the NCBI SRA database and the differentially expressed genes (DEGs) were identified with the software Cufflinks. Functional enrichment analysis was performed to uncover the biological functions of DEGs. Regulatory information was retrieved from the IPA database and a network was established. A total of 147 DEGs were obtained, including 96 downregulated and 51 upregulated DEGs. Gene ontology (GO) function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis suggested that metabolism and signal transduction were the 2 major functions that were significantly influenced. Moreover, an interaction network was built. In conclusion, a number of DEGs was identified and their roles in the pathogenesis of cancer were supported by previous studies. More studies are necessary to further validate their usefulness in the diagnosis and treatment of prostate cancer.

Keywords

Differentially expressed genes Functional enrichment analysis Prostate cancer Regulatory network RNA-seq

Introduction

Prostate cancer is one of the most common cancers among men in the United States and is the second leading cause of cancer-related death in men (1). Localized prostate cancer may be cured with surgery or radiation therapy, but the disease recurs in approximately 20%–30% of patients. For men with metastatic castration-resistant prostate cancer, the median survival ranges between 1 and 2 years (2 –5). A wide range of treatments, such as chemotherapy (6), immunotherapy (7), and combined therapy has been developed, but the outcomes are hardly ideal.

Uncovering the pathogenesis of prostate cancer is an important premise to develop a therapy, and many achievements have been made in this direction. True et al proposed overexpressed CD90/THY1 as a biomarker for prostate cancer (8). Taylor et al carried out concordant assessment of DNA copy number, mRNA expression, and focused exon resequencing in prostate cancer and found NCOA2 as a new oncogene as well as a potential tumor suppressor (9). Sreekumar et al identified sarcosine as a potentially important metabolic intermediary of cancer cell invasion and aggressivity with metabolomics profiling (10). The study by Khan et al revealed a role for miR-128 in prostate cancer with quantitative proteomic profiling (11).

High-throughput technologies, such as transcriptomics, proteomics, and metabolomics are powerful tools to identify biomarkers for detection, prognosis, and treatment (12). RNA-seq is a revolutionary tool for transcriptomics using deep-sequencing technologies (13). It provides a far more precise measurement of the levels of transcripts and their isoforms than other methods. Therefore, RNA-seq data were analyzed in the present study to identify differentially expressed genes (DEGs) associated with prostate cancer. Functional enrichment analysis and regulatory network analysis were both performed, and were beneficial for improving our understandings of the underlying biological mechanisms of prostate cancer.

Materials and Methods

RNA-seq data and treatment

RNA-seq data were downloaded from NCBI SRA, and included 1 prostate cancer sample (No. ERR031018) and 1 adjacent normal sample (No. ERR031024). Complete transcriptomic landscapes of the samples were from the Chinese population. Raw data were collected with HiSeq 2000 (Illumina) and the reading length was 90 bp. Detailed information is shown in Table I.

Table I

Detailed Information about the RNA-SEQ Data

Sample	Total	Paired-end1	Paired-end2	Length
ERR031018	68,015,576	34,007,788	34,007,788	90
ERR031024	75,152,222	37,576,111	37,576,111	90

Screening of differentially expressed genes (DEGs)

The sequencing quality was assessed with Fastx_toolkit V2.1.12 (http://hannonlab.cshl.edu/fastx_toolkit) according to the following 2 aspects: sequencing quality distribution of bases, and base composition distribution. Then, transcriptome data was mapped into the reference genome with TopHat V1.3.2 (14) and Bowtie V0.12.9 (15). The Cuffdiff module from Cufflinks software package (16) was chosen to calculate the RPKM value (reads per kilobase of exon model per million mapped reads) in order to identify DEGs. The parameters were set up according to the default settings of Cuffdiff. FDR <0.05 was set as the cutoff criterion.

Functional enrichment analysis

Functional annotations were performed for each DEGs and then clusters of orthologous groups (COG) classification was conducted (17, 18). Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were performed with the package GO stats (19) of R language to identify significantly overrepresented terms. Cytoscape 2.8 (20) was adopted to visualize pathways to directly observe the distribution of DEGs in the whole pathway.

Establishment of regulatory networks

A range of factors regulated gene and protein expression levels and activity. To deepen our understandings of the underling mechanisms, regulatory relationships were retrieved from IPA (Ingenuity Pathway Analysis) and then a network was established via Cytoscape 2.8.

Results

Identification of DEGs

After data quality assessment and screening, about 66% reads were mapped to the reference genome (details are shown in Tab. II). The expression level was then determined with Cufflinks.

Table II

RNA-SEQ Reads after Quality Control Were Mapped to the Reference Genome

Sample	Total	Pair-end1 mapping	Mapping rate	Pair-end2 mapping	Mapping rate
ERR031018	57,607,598	19,733,353	67.58%	19,554,509	66.97%
ERR031024	58,403,790	19,157,574	66.51%	19,047,112	66.14%

A total of 147 genes meeting the requirements were determined as DEGs, including 51 upregulated genes and 96 downregulated genes. Detailed information of the top 10 DEGs is shown in Table III.

Table III

TOP 10 Differentially Expressed Genes

Gene	E18	E24	Log2 (fold change)	P value	Q value	FDR
GDEP	44.0446	0.099039	-8.79676	9.54E-13	7.21E-09	<0.05
CRISP3	0.107661	127.654	10.2115	8.35E-13	7.21E-09	<0.05
ALOX15B	176.804	0.26407	-9.38701	3.77E-12	1.90E-08	<0.05
CHRNA2	32.7668	0.121363	-8.07676	6.50E-12	2.46E-08	<0.05
ANPEP	463.132	1.70384	-8.08649	3.02E-10	9.13E-07	<0.05
LRRN1	0.366223	36.461	6.63749	2.14E-09	5.40E-06	<0.05
OLFM4	0.110857	8.1249	6.19558	2.66E-08	5.53E-05	<0.05
IGJ	1.00817	39.2853	5.28417	2.93E-08	5.53E-05	<0.05
NTNG1	0.845157	39.7908	5.55707	3.65E-08	6.13E-05	<0.05
OR51E1	0.608947	21.9193	5.16974	4.74E-08	7.16E-05	<0.05

Functional enrichment analysis

Functional annotations and enrichment analysis of DEGs was performed. KEGG enrichment analysis results are shown in Table IV. The top 5 terms were: neuroactive ligand-receptor interaction, phenylalanine metabolism, arachidonic acid metabolism, histidine metabolism, and mucin type O-Glycan biosynthesis. Notably, 4 of these terms were related to metabolism, while the other 1 was associated with signal transduction.

Table IV

Kegg Enrichment Analysis of DEGs (TOP 5)

KEGG ID	Odds ratio	Exp count	Count	Size	Terms	KEGG_p.fdr
4080	4.959893	1.9458355	8	272	Neuroactive ligand-receptor interaction	0.0308469
360	19.38	0.1216147	2	17	Phenylalanine metabolism	0.1508758
590	7.9299451	0.4220746	3	59	Arachidonic acid metabolism	0.1508758
340	10.744444	0.2074604	2	29	Histidine metabolism	0.2068024
512	10.358929	0.2146142	2	30	Mucin type O-Glycan biosynthesis	0.2068024

GO enrichment analysis results are shown in Tables V, VI, and VII. The top 4 biological pathways’ terms were: small molecule biosynthetic process, positive regulation of potassium ion transport, G-protein signaling coupled to cAMP nucleotide second messenger, and regulation of cyclase activity (Tab. V). Accordingly, metabolism and signal transduction were the 2 major functions affected in prostate cancer tissue compared with normal tissue. Cellular component analysis (Tab. VI) and molecular function analysis (Tab. VII) further confirmed the alterations in signal transduction.

Table V

Go Biological Process (BP) Enrichment Analysis of DEGs (TOP 4)

GOBP ID	Odds Ratio	Exp Count	Count	Size	Terms	BP_p.fdr
GO:0044283	3.485338726	4.8569464	15	660	small molecule biosynthetic process	0.1558202
GO:0043268	68.70952381	0.0441541	2	6	positive regulation of potassium ion transport	0.1879568
GO:0007188	7.097445039	0.765337	5	104	G-protein signaling, coupled to cAMP nucleotide second messenger	0.1879568
GO:0031279	7.025980392	0.772696	5	105	regulation of cyclase activity	0.1879568

Table VI

Go Cellular Component (CC) Enrichment Analysis of DEGs (TOP 5)

GO CC ID	Odds ratio	Exp count	Count	Size	Term	CC_p.fdr
GO:0044459	2.1550485	13.752661	26	1933	plasma membrane part	0.1218397
GO:0031224	1.7992764	38.582865	55	5423	intrinsic to membrane	0.1218397
GO:0005887	2.1533948	8.6087528	17	1210	integral to plasma membrane	0.2095294
GO:0045211	4.6252249	1.1454622	5	161	postsynaptic membrane	0.2095294
GO:0043025	3.8834545	1.6363745	6	230	neuronal cell body	0.2095294

Table VII

Go Molecular Function (MF) Enrichment Analysis of DEGs (TOP 5)

GO MF ID	Odds ratio	Exp count	Count	Size	Term	MF_p.fdr
GO:0051380	138.3211	0.0292298	2	4	norepinephrine binding	0.1072432
GO:0051379	69.151376	0.0438446	2	6	epinephrine binding	0.122865
GO:0004935	39.507208	0.065767	2	9	adrenergic receptor activity	0.122865
GO:0060089	2.1242915	12.239961	23	1675	molecular transducer activity	0.122865
GO:0038023	2.2019402	9.5873601	19	1312	signaling receptor activity	0.122865

The distribution of DEGs in a complete KEGG pathway was visualized, and a typical illustration is shown in Figure 1.

Fig. 1

A typical illustration of the distribution of DEGs in the KEGG pathway. Red color indicates upregulated DEGs and green color indicates down-regulated DEGs.

Regulatory networks

The interactions between regulators and DEGs were retrieved from the IPA database and were then displayed via Cytoscape 2.8 (Fig. 2). In total, 6 types of relationships were assigned: complex, enzyme, transcription regulator, ligand-dependent nuclear receptor, group, and other.

Fig. 2

The regulatory network of the DEGs. Green triangles represent regulators and red squares represent DEGs.

Discussion

Deep sequencing-based transcriptome analysis improves the detection of differentially expressed genes, sequence variants, and novel fusion transcripts, thus holding a promising future in cancer research. In the present study, the RNA-seq data was analyzed to investigate key genes, pathways, and regulatory mechanisms associated with prostate cancer. A total of 147 DEGs were acquired, of which 51 were upregulated and 96 downregulated. Functional enrichment analysis revealed that metabolism and signal transduction were the top 2 functions significantly disturbed. The detected genes might be potential biomarkers for prostate cancer.

Prostaglandin-endoperoxide synthase 2 (PTGS2) is a key enzyme in prostaglandin biosynthesis, and is involved in inflammation (21) and mitogenesis (22). Previous studies indicated that the PTGS2 gene is silenced in prostate cancer by epigenetic hypermethylation and that it can be used as a prognostic marker (23 –25). In the present study, PTGS2 was downregulated in prostate cancer. As shown in Figure 2, PTGS2 is regulated by a range of factors such as histone deacetylase 4 (HDAC4), forkhead box L2 (FOXL2), and estrogen receptor 1 (ESR1).

ESR1 is a ligand-activated transcription factor implicated in several pathological processes like cancer. It localizes into the nucleus and then forms a homodimer or a heterodimer with the estrogen receptor 2. It has been reported that ESR1 is downregulated in prostate cancer and is correlated with tumor progression (26). Several studies also pointed out that promoter methylation of ESR1 contributed to its altered expression (27, 28).

The cysteine-rich secretory protein 3 (CRISP3) is regarded as a potential biomarker for prostate cancer (29). It's a secreted protein produced in the male reproductive tract and is involved in sperm maturation. Bjartell et al reported that strong immunostaining for CRISP-3 is common in HG-PIN and is preserved in most PCa specimens (30). CRISP-3 and beta-microseminoprotein are independent predictors of recurrence after radical prostatectomy for localized prostate cancer (31).

Alanyl (membrane) aminopeptidase (ANPEP, CD13) is involved in the metabolism of regulatory peptides in different cell types. ANPEP is considered as a prognostic factor for non-small cell lung cancer (32, 33) and a therapeutic target in liver cancer stem cells (34). Chang et al found that ANPEP can form a complex with the tumor-associated antigen L6, thus modulating cell migration and invasion (35). CD-marker expression profiles can be used to distinguish heterogeneous prostate cancer cells (36). The upregulation of ANPEP is reported in gastric cancer (37) and is also validated in prostate cancer in the present study.

Olfactomedin 4 (OLFM4) is an antiapoptotic factor that promotes tumor growth and is an extracellular matrix glycoprotein that facilitates cell adhesion (38). Kobayashi et al reported that OLFM4 promotes S-phase transition in the proliferation of pancreatic cancer cells (39). In accordance with its roles in cell adhesion, Besson et al confirmed OLFM4 as a new non-metastatic tumor marker by analyzing different stages of colorectal cancer with a quantitative proteomic approach (40). OLFM4 suppresses prostate cancer cell growth and metastasis via negative interaction with cathepsin D and SDF-1 (41), a fact that may explain the downregulation of OLFM4 in prostate cancer.

Functional enrichment analysis revealed that several metabolism pathways are disturbed in prostate cancer (42). The involvement of arachidonic acid metabolism in prostate cancer PC3 cell growth has been reported by Hassan and Carraway (43). Chaudry et al found that diacylglycerol formation in prostate cancer is about 4-fold greater than that in benign prostatic tissue, which may promote tumorigenesis through activation of protein kinase C (44). Hong et al reported that modulation of arachidonic acid metabolism by curcumin may contribute to the anti-inflammatory and anti-carcinogenic actions (45).

Overall, a range of DEGs were found in the present study, some of which have been suggested as potential biomarkers for prostate cancer. Besides, functional enrichment analysis provided information about the changes in the process of tumorigenesis. Meanwhile, regulatory network analysis revealed the underlying regulatory mechanisms. Clearly, more experimental works is necessary to confirm the roles of DEGs in prostate cancer.

Footnotes

Informed Consent: Written informed consent was obtained from the patient for the publication of this report and any accompanying images.

Financial Support: None.

Conflict of Interest Statements: The authors have declared that no conflict of interest exists.

Meeting Presentation: None.

References

Jemal

, Siegel

, Ward

, Hao

, Xu

, Thun

MJ.

Cancer statistics, 2009.

CA Cancer J Clin. 2009;59(4):225–249.

Higano

, Schellhammer

, Small

Integrated data from 2 randomized, double-blind, placebo-controlled, phase 3 trials of active cellular immunotherapy with sipuleucel-T in advanced prostate cancer.

Cancer. 2009;115(16):3670–3679.

Saad

, Gleason

, Murray

Zoledronic Acid Prostate Cancer Study Group.

A randomized, placebo-controlled trial of zoledronic acid in patients with hormone-refractory metastatic prostate carcinoma. J Natl Cancer Inst. 2002;94(19):1458–1468.

Carducci

, Saad

, Abrahamsson

Atrasentan Phase III Study Group Institutions.

A phase 3 randomized controlled trial of the efficacy and safety of atrasentan in men with metastatic hormone-refractory prostate cancer. Cancer. 2007;110(9):1959–1966.

Sternberg

, Petrylak

, Sartor

Multinational, double-blind, phase III study of prednisone and either satraplatin or placebo in patients with castrate-refractory prostate cancer progressing after prior chemotherapy: the SPARC trial.

J Clin Oncol. 2009;27(32):5431–5438.

, Wang

, Reed

, Huang

, Cuff

CF.

Evaluation of cisplatin in combination with β-elemene as a regimen for prostate cancer chemotherapy.

Basic Clin Pharmacol Toxicol. 2010;107(5):868–876.

Kawalec

, Paszulewicz

, Holko

, Pilc

Sipuleucel-T immunotherapy for castration-resistant prostate cancer.

A systematic review and meta-analysis. Arch Med Sci. 2012;8(5):767–775.

True

, Zhang

, Ye

CD90/THY1 is overexpressed in prostate cancer-associated fibroblasts and could serve as a cancer biomarker.

Mod Pathol. 2010;23(10):1346–1356.

Taylor

, Schultz

, Hieronymus

Integrative genomic profiling of human prostate cancer.

Cancer Cell. 2010;18(1):11–22.

10.

Sreekumar

, Poisson

, Rajendiran

Metabolomic profiles delineate potential role for sarcosine in prostate cancer progression.

Nature. 2009;457(7231):910–914.

11.

Khan

, Poisson

, Bhat

Quantitative proteomic profiling of prostate cancer reveals a role for miR-128 in prostate cancer.

Mol Cell Proteomics. 2010;9(2):298–312.

12.

Goo

, Goodlett

DR.

Advances in proteomic prostate cancer biomarker discovery.

J Proteomics. 2010;73(10):1839–1850.

13.

Wang

, Gerstein

, Snyder

RNA-Seq: a revolutionary tool for transcriptomics.

Nat Rev Genet. 2009;10(1):57–63.

14.

Trapnell

, Pachter

, Salzberg

SL.

TopHat: discovering splice junctions with RNA-Seq.

Bioinformatics. 2009;25(9):1105–1111.

15.

Langmead

, Trapnell

, Pop

, Salzberg

SL.

Ultrafast and memory-efficient alignment of short DNA sequences to the human genome.

Genome Biol. 2009;10(3):R25.

16.

Trapnell

, Roberts

, Goff

Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks.

Nat Protoc. 2012;7(3):562–578.

17.

Kanehisa

, Goto

, Sato

, Furumichi

, Tanabe

KEGG for integration and interpretation of large-scale molecular data sets.

Nucleic Acids Res. 2012;40(D1)(Database issue):D109–D114.

18.

Ashburner

, Ball

, Blake

The Gene Ontology Consortium.

Gene ontology: tool for the unification of biology. Nat Genet. 2000;25(1):25–29.

19.

Falcon

, Gentleman

Using GOstats to test gene lists for GO term association.

Bioinformatics. 2007;23(2):257–258.

20.

Smoot

, Ono

, Ruscheinski

, Wang

, Ideker

Cytoscape 2.8: new features for data integration and network visualization.

Bioinformatics. 2011;27(3):431–432.

21.

Williams

, Mann

, DuBois

RN.

The role of cyclooxygenases in inflammation, cancer, and development.

Oncogene. 1999;18(55):7908–7916.

22.

Zembowicz

, Jones

, Wu

KK.

Induction of cyclooxygenase-2 in human umbilical vein endothelial cells by lysophosphatidylcholine.

J Clin Invest. 1995;96(3):1688–1692.

23.

Ellinger

, Bastian

, Jurgan

CpG island hypermethylation at multiple gene sites in diagnosis and prognosis of prostate cancer.

Urology. 2008;71(1):161–167.

24.

Woodson

, O'Reilly

, Ward

CD44 and PTGS2 methylation are independent prognostic markers for biochemical recurrence among prostate cancer patients with clinically localized disease.

Epigenetics. 2006;1(4):183–186.

25.

Phé

, Cussenot

, Rouprêt

Methylated genes as potential biomarkers in prostate cancer.

BJU Int. 2010;105(10):1364–1370.

26.

Bonkhoff

, Fixemer

, Hunsicker

, Remberger

Estrogen receptor expression in prostate cancer and premalignant prostatic lesions.

Am J Pathol. 1999;155(2):641–647.

27.

, Chui

, Nakajima

, Oh

, Au

, Dahiya

Frequent methylation of estrogen receptor in prostate cancer: correlation with tumor progression.

Cancer Res. 2000;60(3):702–706.

28.

, Shiina

, Deguchi

Age-dependent methylation of ESR1 gene in prostate cancer.

Biochem Biophys Res Commun. 2004;321(2):455–461.

29.

Kosari

, Asmann

, Cheville

, Vasmatzis

Cysteinerich secretory protein-3: a potential biomarker for prostate cancer.

Cancer Epidemiol Biomarkers Prev. 2002;11(11):1419–1426.

30.

Bjartell

, Johansson

, Björk

Immunohistochemical detection of cysteine-rich secretory protein 3 in tissue and in serum from men with cancer or benign enlargement of the prostate gland.

Prostate. 2006;66(6):591–603.

31.

Bjartell

, Al-Ahmadie

, Serio

Association of cysteine-rich secretory protein 3 and beta-microseminoprotein with outcome after radical prostatectomy.

Clin Cancer Res. 2007;13(14):4130–4138.

32.

Murakami

, Yokoyama

, Kondo

, Nakanishi

, Kohno

, Miyake

Circulating aminopeptidase N/CD13 is an independent prognostic factor in patients with non-small cell lung cancer.

Clin Cancer Res. 2005;11(24 Pt 1):8674–8679.

33.

Tokuhara

, Adachi

, Hashida

Neutral endopeptidase/CD10 and aminopeptidase N/CD13 gene expression as a prognostic factor in non-small cell lung cancer.

Jpn J Thorac Cardiovasc Surg. 2001;49(8):489–496.

34.

Haraguchi

, Ishii

, Mimori

CD13 is a therapeutic target in human liver cancer stem cells.

J Clin Invest. 2010;120(9):3326–3339.

35.

Chang

, Chen

, Cheng

CD13 (aminopeptidase N) can associate with tumor-associated antigen L6 and enhance the motility of human lung cancer cells.

Int J Cancer. 2005;116(2):243–252.

36.

Liu

, Roudier

, True

LD.

Heterogeneity in primary and metastatic prostate cancer as defined by cell surface CD profile.

Am J Pathol. 2004;165(5):1543–1556.

37.

Carl-McGrath

, Lendeckel

, Ebert

, Wolter

, Roessner

, Röcken

The ectopeptidases CD10, CD13, CD26, and CD143 are upregulated in gastric cancer.

Int J Oncol. 2004;25(5):1223–1232.

38.

Grover

, Hardingham

, Cummins

AG.

Stem cell marker olfactomedin 4: critical appraisal of its characteristics and role in tumorigenesis.

Cancer Metastasis Rev. 2010;29(4):761–775.

39.

Kobayashi

, Koshida

, Moriai

, Tsuji

, Watanabe

Olfactomedin 4 promotes S-phase transition in proliferation of pancreatic cancer cells.

Cancer Sci. 2007;98(3):334–340.

40.

Besson

, Pavageau

, Valo

A quantitative proteomic approach of the different stages of colorectal cancer establishes OLFM4 as a new nonmetastatic tumor marker.

Mol Cell Proteomics 2011;10(12):M111.009712.

41.

Chen

, Li

, Liu

Olfactomedin 4 suppresses prostate cancer cell growth and metastasis via negative interaction with cathepsin D and SDF-1.

Carcinogenesis. 2011;32(7):986–994.

42.

Olsson

, Bera

, Essand

GDEP, a new gene differentially expressed in normal prostate and prostate cancer.

Prostate. 2001;48(4):231–241.

43.

Hassan

, Carraway

RE.

Involvement of arachidonic acid metabolism and EGF receptor in neurotensin-induced prostate cancer PC3 cell growth.

Regul Pept. 2006;133(1–3): 105–114.

44.

Chaudry

, Wahle

, McClinton

, Moffat

LE.

Arachidonic acid metabolism in benign and malignant prostatic tissue in vitro: effects of fatty acids and cyclooxygenase inhibitors.

Int J Cancer. 1994;57(2):176–180.

45.

Hong

, Bose

, Ju

Modulation of arachidonic acid metabolism by curcumin and related beta-diketone derivatives: effects on cytosolic phospholipase A(2), cyclooxygenases and 5-lipoxygenase.

Carcinogenesis. 2004;25(9):1671–1679.

Investigation of Key Genes associated with Prostate Cancer using RNA-Seq Data

Abstract

Keywords

Introduction

Materials and Methods

RNA-seq data and treatment

Screening of differentially expressed genes (DEGs)

Functional enrichment analysis

Establishment of regulatory networks

Results

Identification of DEGs

Functional enrichment analysis

Regulatory networks

Discussion

Footnotes

References