Research ArticleIdentification of Transcriptional Signatures of Colon TumorStroma by a Meta-Analysis

Md. Nazim Uddin ,1,2,3 Mengyuan Li,1,2,3 and XiaoshengWang 1,2,3

1Biomedical Informatics Research Lab, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University,Nanjing 211198, China2Cancer Genomics Research Center, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University,Nanjing 211198, China3Big Data Research Institute, China Pharmaceutical University, Nanjing 211198, China

Correspondence should be addressed to Xiaosheng Wang; xiaosheng.wang@cpu.edu.cn

Received 28 February 2019; Accepted 31 March 2019; Published 2 May 2019

Guest Editor: Nathaniel Weygant

Copyright © 2019 Md. Nazim Uddin et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Background. The tumor stroma plays pivotal roles in influencing tumor growth, invasion, and metastasis. Transcriptionalsignatures of colon tumor stroma (CTS) are significantly associated with prognosis of colon cancer. Thus, identification ofthe CTS transcriptional features could be useful for colon cancer diagnosis and therapy. Methods. By a meta-analysis of threeCTS gene expression profiles datasets, we identified differentially expressed genes (DEGs) between CTS and colon normalstroma. Furthermore, we identified the pathways, upstream regulators, and protein-protein interaction (PPI) network that weresignificantly associated with the DEGs. Moreover, we analyzed the enrichment levels of immune signatures in CTS. Finally, weidentified CTS-associated gene signatures whose expression was significantly associated with prognosis in colon cancer. Results. Weidentified numerous significantly upregulated genes (such asCTHRC1,NFE2L3, SULF1, SOX9, ENC1, andCCND1) and significantlydownregulated genes (such as MYOT, ASPA, KIAA2022, ARHGEF37, BCL-2, and PPARGC1A) in CTS versus colon normalstroma. Furthermore, we identified significantly upregulated pathways in CTS that were mainly involved in cellular development,immune regulation, andmetabolism, as well as significantly downregulated pathways in CTS that were mostly metabolism-related.Moreover, we identified upstream TFs (such as SUZ12, NFE2L2, RUNX1, STAT3, and SOX2), kinases (such as MAPK14, CSNK2A1,CDK1, CDK2, and CDK4), and master metabolic transcriptional regulators (MMTRs) (such as HNF1A, NFKB1, ZBTB7A, GATA2,and GATA5) regulating the DEGs. We found that CD8+ T cells were more enriched in CTS than in colon normal stroma.Interestingly, we found that many of the DEGs and their regulators were prognostic markers for colon cancer, including CEBPB,PPARGC1, STAT3,MTOR, BCL2, JAK2, and CDK1. Conclusions. The identification of CTS-specific transcriptional signatures mayprovide insights into the tumor microenvironment that mediates the development of colon cancer and has potential clinicalimplications for colon cancer diagnosis and treatment.

1. Background

The tumor stroma is an important component of the tumormicroenvironment (TME) and plays key roles in the tumordevelopment [1]. Stromal cells are composed of many dif-ferent types of cells, including vascular endothelial cells,pericytes, adipocytes, fibroblasts, osteoblasts, chondrocytes,extracellular matrix (ECM), and bone-marrowmesenchymalstromal cells [2]. The tumor stroma can promote ECMremodeling, cellular migration, neoangiogenesis, invasion,

immunosurveillance evasion, and drug resistance of tumors[3]. Colorectal cancer (CRC) is the fourth most commoncancer and a leading cause of cancer mortality world-wide [4]. Transcriptional signatures of CRC stromal cellshave been associated with poor prognosis in CRC [5].Isella et al. demonstrated that the gene signatures of CRCstromal cells (cancer-associated fibroblasts, leukocytes, andendothelial cells) were significantly upregulated in thestem/serrated/mesenchymal transcriptional subtype of CRCwhich had a poor prognosis [6]. Calon et al. showed that

HindawiJournal of OncologyVolume 2019, Article ID 8752862, 12 pageshttps://doi.org/10.1155/2019/8752862

2 Journal of Oncology

the CRC stromal transcriptional signatures correlated withdisease relapse [5]. These prior studies exhibited the signif-icant roles of tumor stroma in CRC growth, invasion, andmetastasis.

In this study, we performed a meta-analysis of threecolon tumor stromal transcriptome datasets using the bioin-formatics approach. We identified differentially expressedgenes (DEGs) between colon tumor stroma (CTS) andnormal stroma. On the basis of these DEGs, we identifiedtheir associated pathways, upstream regulators, and protein-protein interaction (PPI) network and certain prognosticmarkers that were associated with survival of colon cancerpatients. We also analyzed the enrichment levels of immunesignatures in CTS. This study provides insights into CTSmolecular features that could have clinical implications forcolon cancer diagnosis and treatment.

2. Methods

2.1. Datasets. We searched the NCBI Gene ExpressionOmnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) using the keywords “colon cancer,” “stroma,” and“tumor stroma” and identified three CTS gene expressionprofiles datasets (GSE31279, GSE35602, and GSE46824)[7–9]. In survival analyses, we used the TCGA colon cancerdataset (https://portal.gdc.cancer.gov/) and a SurvExpress(http://bioinformatica.mty.itesm.mx/SurvExpress) built-indataset (colon metabase) [10]. A summary of these datasetsis shown in Supplementary Table S1.

2.2. Identification of DEGs between CTS and Normal Stroma.We used the web tool Network Analyst [11] to identify theDEGs betweenCTS andnormal stroma.TheComBatmethod[12] in the tool was utilized to remove batch effects fromthe three CTS datasets (Supplementary Figure S1). Each indi-vidual dataset was normalized by base-2 log transformationand quantile normalization, and the R package “limma”was utilized to identify the DEGs between CTS and normalstroma. A meta-analysis of the three datasets was performedusing Cochran's combination test [13]. The false discoveryrate (FDR), calculated by the Benjamini–Hochberg method[14], was used to adjust for multiple tests. We determined theDEGs with a threshold of absolute combined effect size (ES)>0.82 and FDR<0.05.

2.3. Gene-Set Enrichment Analysis. We performed gene-setenrichment analysis of the DEGs by GSEA [15]. The KEGGpathways significantly associated with the upregulated andthe downregulated DEGs were identified (FDR < 0.05),respectively.

2.4. Identification of Transcription Factors (TFs), Kinases, andMaster Metabolic Transcriptional Regulators (MMTRs) �atAre Significantly Associated with the DEGs. To link geneexpression signatures to upstream cell signaling networks, weused eXpression2Kinases [16] to identify the upstream TFsand kinases that regulate the DEGs and utilized iRegulon [17]to identify the MMTRs of the DEGs.

2.5. Identification of PPI Network of the DEGs. We employedNetwork Analyst [11] to construct a PPI network of theDEGs [11]. Two types of modules (function-first modules andconnection-first modules) of the PPI network were extracted.The function-firstmodules (FFMs)were constructed by path-way enrichment analysis and the connection-first modules(CFMs) were identified by the randomwalk-based algorithm[18].

2.6. Comparison of the Enrichment Levels of CD8+ T Cellsbetween Two Classes of Samples. The enrichment level ofCD8+T cells in a samplewas evaluated by the expression levelofCD8A.We compared the enrichment levels of CD8+T cellsbetween two groups of samples using Student's t-test.

2.7. Identification of DEGs between High-Stroma-Content andLow-Stroma-Content TCGA Colon Cancer Samples. We usedESTIMATE [19] to quantify the intratumoral stromal content(stroma score) of TCGA colon cancer samples. We identifiedthe DEGs between high-stroma-content (stroma score >median) and low-stroma-content (stroma score < median)tumors using Student's t-test.

2.8. Survival Analyses. We compared the overall survival(OS) and the disease-free survival (DFS) of colon cancerpatients classified based on gene expression levels (expressionlevels > median versus expression levels < median). Kaplan-Meier survival curves were used to show the survival dif-ferences, and the log-rank test was utilized to evaluate thesignificance of survival differences.The individual prognosticgenes were identified and were fitted in a multivariate Coxregression model. SurvExpress [10] was used for the multi-variate survival analysis.

3. Results

3.1. Identification of DEGs between CTS and Normal Stroma.We identified 694 DEGs between CTS and normal stroma bythe meta-analysis. These DEGs included 295 downregulatedand 399 upregulated genes in CTS (Supplementary TablesS2 and S3). Figure 1 shows the top 25 upregulated andtop 25 downregulated genes in CTS ranked on the basis ofthe combined ES (the detailed results of statistical analysisfor the top 10 upregulated and top 10 downregulated genesin CTS are shown in Supplementary Tables S4). CTHRC1,a gene involved in vascular remodeling, bone formation,and developmental morphogenesis, was upregulated in CTSwith the highest ES. It has been shown that CTHRC1 couldpromote human CRC cell proliferation and invasion byactivating Wnt/PCP signaling [20]. This gene also plays animportant role in promoting ovarian cancer cell adhesion,migration, and metastasis through the activation of integrin𝛽3/FAK signaling [21]. NFE2L3, a gene regulating the cellcycle progression in colon cancer [22], was upregulated inCTS with the second highest ES. Interestingly, both CTHRC1and NFE2L3 have been indicated as useful biomarker candi-dates for CRC diagnosis because of their overexpression inadenomas and CRC relative to normal tissue [23]. SULF1,

Journal of Oncology 3

classdataset

Low High

Tumor stroma

Normal stroma

GSE46824

GSE35602

GSE31279

GSM

1138867G

SM1138870

GSM

1138871G

SM1138869

GSM

1138868G

SM1138866

GSM

1138865G

SM1138864

GSM

1138872G

SM871440

GSM

871441G

SM871437

GSM

871436G

SM775273

GSM

775268G

SM775262

GSM

775222G

SM775188

GSM

775182G

SM775247

GSM

775256G

SM775284

GSM

775277G

SM1138846

GSM

1138847G

SM1138848

GSM

1138845G

SM1138849

GSM

1138844G

SM1138841

GSM

1138863G

SM1138843

GSM

1138862G

SM1138861

GSM

1138860G

SM1138859

GSM

1138858G

SM1138857

GSM

1138856G

SM1138855

GSM

1138854G

SM1138853

GSM

1138852G

SM1138851

GSM

1138850G

SM1138842

GSM

1138840G

SM1138839

GSM

871433G

SM871432

GSM

871429G

SM871428

GSM

871426G

SM871423

GSM

871422G

SM871420

GSM

871418G

SM871416

GSM

871414G

SM871412

GSM

871410G

SM775287

GSM

775279G

SM775275

GSM

775271G

SM775265

GSM

775259G

SM775191

GSM

775250G

SM775225

GSM

775185

CTHRC1SULF1NFE2L3CCND1ENC1CLDN1SOX9SEMA5APCDH17BCL6BVCANNUAK1KDELR3COL5A1KIAA1217CKAP4DCBLD1PROX1PLAURPTBP3ESM1MLXIPLSLC16A3SEMA3GGZMBMYOTC7GNG7ARHGEF37ASPABCL2RFX2SLC6A16KIAA2022SRSF5CBX7METTL7AEFCAB1PELI2ANKRD13AC1QTNF9SEPP1FXYD1GATSPDK4TMOD1C16orf89PCF11FGF13CCDC7

Figure 1: Gene expression pattern of the top 25 upregulated and top 25 downregulated genes in colon tumor stroma (CTS) relative to colonnormal stroma ranked on the basis of the combined effect size (ES) identified by Network Analyst [11].

whose expression in tumor stroma is a prognostic markerin advanced pancreatic cancer [24], was upregulated in CTSwith the third highest ES. The overexpression of this genehas been associated with a poor prognosis in urothelialcarcinoma [25]. SOX9, the gene upregulated in CTS withthe fourth highest ES, has been shown to be overexpressedin CRC and its overexpression was an independent adverseprognosticator in CRC [26]. Some other genes upregulatedin CTS have been demonstrated to be overexpressed in CRCand their expression was negatively associated with CRCprognosis, such as ENC1, CCND1, VCAN, SEMA5A, andNOS3 [27–31]. Interestingly, both PCDH17 and BCL6B wereupregulated in CTS, while they had reduced expression inCRC [32, 33]. It indicates that PCDH17 and BCL6B could bespecifically expressed in CTS cells but not in colon cancercells.

Many of the significantly downregulated genes in CTShave been associatedwith CRC [34–37]. For example,MYOT,ASPA, and KIAA2022 were downregulated in CRC [34], thedownregulation of ARHGEF37 was associated with a poorprognosis inCRC [35], higher expression levels ofBCL-2werecorrelated with a better survival prognosis in CRC [36], andPPARGC1A was a negative predictor for CRC prognosis [37].

Altogether, a number of the abnormally expressed genesin CTS compared to colon normal stroma identified by themeta-analysis have been associated with CRC pathology andprognosis.

3.2. Identification of Pathways Significantly Associated with theDEGs. GSEA [15] identified 44 KEGG pathways that weresignificantly associated with the upregulated genes in CTS.These pathwaysweremainly involved in cellular development(p53 signaling, Wnt signaling, apoptosis, Notch signaling,focal adhesion, endocytosis, ECM-receptor interaction, celladhesion molecules, adherens junction, tight junction, gapjunction, and regulation of actin cytoskeleton), immuneregulation (leukocyte transendothelial migration, comple-ment and coagulation cascades, natural killer cell medi-ated cytotoxicity, Toll-like receptor, chemokine signaling,and cytokine-cytokine receptor interaction), andmetabolism(purine metabolism and pyrimidine metabolism) (Figure 2,Supplementary Table S5). Previous studies have shown thatsome of these pathways were significantly associated withcolon cancer [38–41]. For example, the Wnt and Notchpathways were associated with colon cancer development[38, 39]. The cytokine-cytokine receptor interaction pathwaywas significantly enriched in CRC [34]. The ECM and ECM-associated proteins [39], the glycosaminoglycan metabolism,and chondroitin sulfate/dermatan sulfate metabolism path-ways played key roles in mediating tumormicroenvironment[40, 41].

In addition, GSEA identified six KEGG pathways thatwere significantly associated with the downregulated genes inCTS (Supplementary Figure S2).Most of these pathways weremetabolism-related, including purine metabolism, histidine

4 Journal of Oncology

Regulation of actin cytoskeletonPyrimidine metabolism

Cytokine-cytokine receptor interactionGap junction

Notch signaling pathwayApoptosis

Tight junctionChemokine signaling pathway

Adherens junctionToll-like receptor signaling pathway

EndocytosisCell adhesion molecules (CAMs)

Natural killer cell mediated cytotoxicityPurine metabolism

Complement and coagulation cascadesp53 signaling pathway

Glycosaminoglycan biosynthesis-chondroitin sulfateWnt signaling pathway

Focal adhesionLeukocyte transendothelial migration

ECM-receptor interactionSmall cell lung cancer

Pathways in cancer

1.321.351.411.441.441.441.481.48

1.611.88

1.981.981.98

2.262.31

2.692.692.692.692.692.692.69

5.21

0 54321

- LoＡ10 (FDR)

Figure 2: Significantly upregulated KEGG pathways in CTS relativeto colon normal stroma identified by GSEA [15]. FDR: false discoveryrate.

metabolism, glycine, serine, and threonine metabolism, anddrug metabolism-cytochrome p450. These pathways havebeen associated with colon and other cancers [42–44]. Forexample, impaired purine metabolism was associated withthe progression of cancer [42]. Histidine metabolism couldboost cancer therapy [43]. Cytochrome P450 enzymes wereassociated with the metabolism of anticancer drugs and theirexpression was associated with a poor prognosis in CRCpatents [44].

3.3. Identification of Upstream TFs, Kinases, and MMTRsSignificantly Associated with the DEGs. We identified 11 sig-nificant upstreamTFs regulating the DEGs, including SUZ12,NFE2L2, RUNX1, ESR1, STAT3, TCF3, FOSL2, SALL4, AR,SMC3, and SOX2, of which the genes encoding RUNX1and SALL4 were upregulated in CTS (Figure 3(a)). Mostof these TFs have been associated with colon cancer [45–49]. For example, SUZ12 was the most significant upstreamTF which could contribute to the CRC development [45].RUNX1 mutations were associated with the CRC risk [46].TCF3 and FOSL2 were associated with the tumorigenesis ofCRC [47, 48]. The overexpression of SOX2 was associatedwith the progression and a poor prognosis in colon cancer[49].

Moreover, we identified 124 significant protein kinasesthat regulate the DEGs (Figure 3(b), Supplementary TableS6). These kinases mainly included cell cycle regula-tion kinases (CDKs), signaling MAP kinases (MAPKs,MAP2Ks, and MAP3Ks), and ribosomal kinases (RPS6KA1,RPS6KA3, and RPS6KA5). MAPK14 was the most sig-nificant upstream kinase negatively regulating the forma-tion of colitis-associated colon tumors [50]. Furthermore,we constructed a TF-kinase interaction network of theseTFs and kinases (Figure 3(c)). In the network, the mostconnected TFs included SUZ12, NFE2L2, RUNX1, STAT3,FOSL2, AR, SMC3, ESR1, and TCF3, and the most con-nected kinases included MAPK14, CDK1, CSNK2A1, CDK2,MAPK3, HIPK2, ERK1, and CDK4. It indicates that the cellcycle regulation may play a pivotal role in CTS.

MMTRs are interesting biomarkers and targets formetabolism-targeted cancer therapy [51]. We identified 9(HNF1A, NFKB1, ZBTB7A, ATF6, TEAD4, TFAP2B, JAZF1,

FNTB, and EP300) and 12 (PKNOX2, GATA2, MAPK10,TEAD1, TOX, MEF2A, GATA5, ELK1, MAZ, NHLH1, ATF1,and RAD21) MMTRs for the upregulated and the down-regulated genes in CTS, respectively (Supplementary TableS7), and built the regulatory networks associated with theseMMTRs (Figure 4). In the networks, ATF6 (activatingtranscription factor 6), a TF regulating unfolded proteinresponse during endoplasmic reticulum (ER) stress, targeted163 upregulated genes, and PKNOX2 (PBX/knotted 1 home-obox 2), which plays key roles in regulating cell proliferation,differentiation, and death, targeted 131 downregulated genes.Interestingly, two members of the GATA family of TFs(GATA2 and GATA5) were the MMTRs that regulated thedownregulated genes in CTS (Figure 4(b)).

Altogether, the identification of upstream TFs, kinases,and MMTRs significantly associated with the DEGs mayprovide insights into the TME thatmediates the developmentof colon cancer.

3.4. CD8+ T Cells Are More Enriched in CTS than in NormalStroma. We compared the enrichment levels of CD8+ T cellsbetweenCTS andnormal stroma and found thatCD8+T cellsshowed significantly higher enrichment levels in CTS thanin normal stroma (Student's t-test, p=0.016) (Figure 5). Thissuggests an antitumor immune response activity in the TMEof colon cancer.

3.5. Identification of Prognostic Factors in Colon Cancer Basedon the DEGs and�eir Upstream Regulators. We investigatedthe association between the transcriptional signatures of CTSand survival prognosis (overall survival (OS) and disease-free survival (DFS)) in the TCGA colon cancer dataset. Thetranscriptional signatures included the top 10 upregulatedand top 10 downregulated genes in CTS on the basis of ES, 45hub genes (≥3 degrees) from the zero-order PPI network ofthe DEGs (Supplementary Table S8), and the genes encoding11 TFs, 124 kinases, and 21 MMTRs regulating the DEGs. Wefound that the expression of many of these transcriptionalsignatures was significantly associated with the survival ofcolon cancer patients. For example, the expression of CEBPB,a gene significantly upregulated in CTS and a hub node in thePPI network, had a significant negative correlationwithOS incolon cancer (Figure 6(a)). The negative correlation betweenCEBPB expression and survival has also been demonstratedin other cancer types, such as high-grade serous ovariancancer [52]. PPARGC1 was significantly downregulated inCTS and was a hub node in the PPI network, while itsexpression had a significant positive correlation with OS incolon cancer (Figure 6(a)). PPARGC1A was indicated as atumor suppressor in colon cancer [53] and ovarian cancer[54], as well as a negative prognostic biomarker for CRC [37].Our data indicate that the deregulation of these genes in CTSis prognostic for colon cancer patients.

Among the upstream regulators (TFs, kinases, andMMTRs) of the DEGs, the expression of STAT3, RPS6KA5,IKBKE, ERBB2,MTOR, andNFKB1 had a positive correlationwith OS in colon cancer, while the expression of CDK1,CDK5, and BRD2 had a negative correlation with OS in coloncancer (Figure 6(a)).The deregulation of these genes has been

Journal of Oncology 5

SOX2

SMC3

AR

SALL4

FOSL2

TCF3

STAT3

ESR1

RUNX1

NFE2L2

SUZ12

0 4321

1.37

1.62

1.88

1.99

2.02

2.28

2.38

2.51

2.61

3.47

4.71

Transcription Factor

- LoＡ10 (p value)

(a)

DNAPKGSK3BETA

ABL1RPS6KA3

PKBALPHACK2ALPHA

JNK1ERK2

MAPK8AKT1ERK1

HIPK2GSK3BCDK2

MAPK3CDK4

MAPK1CSNK2A1

CDK1MAPK14

0 2015105

9.289.779.9910.0610.32

11.4711.5

13.6213.713.81

15.4715.6115.9616.52

17.118.06

20.3322.27

23.4823.03

Protein Kinase

- LoＡ10 (p value)

(b)

Phosphorylation PPIKinase Intermediate protein Transcription factor

(c)

Figure 3:�e significant upstream transcriptional factors (TFs) and kinases that regulate the differentially expressed genes (DEGs) between CTSand colon normal stroma identified by eXpression2Kinases [16]. (a) Significant upstream TFs regulating the DEGs. (b) Significant upstreamkinases regulating the DEGs. (c) A TF-kinase interaction network of the significant upstream TFs and kinases regulating the DEGs.

6 Journal of Oncology

CARD10

CRB3

CECR6

F12

CLK2

GSTO1SLC25A22

CHTF18RHOV

TMEM44

TMEM184B

SCO2

CCCCLDN1

CLK2CRPS6KA4

RAB3IP

SLC1A4SSSSSSCXADR

GADD45A

YY1SLC5A1

CBX3

EP300

TEAD4

HNF1A

IRAK1HDGF

INSIG1

SIX4

BAZ1AATP8B1 UNC5B BAZABBAZ1AABAZBAZABA EVI58B1B1BB18 1BB1BBB11BB1BBB18 1BPIP4K2B

ADAM12HIST1H1B

PPP1R14C

CHMP4C

EGR2

ASPH

DAP

NFKB1

PABPC1

44GADD4D4D4G DDD4ADGG 44TPBG

SERPINE1

SLC1A7TRAF2

FRMD5

EFNB2

FAM43ARCC2

VSNL1

SLC12A8

LOX

TMEM105

HAS2

FOXC1

SOX9TWIST1 BDNF

SERPINA1 C4orf32

TESC

GATA2

ETV5AGRN

USP13

FAM64A

NCOR2

CDK6FOXA2

TPM4

SFTA2

CTHRC1P4HB

PLCB4

CLDN4IER3

LMNB2

BGN

FGR

FZD2

SFXN1PLOD2

CHSY1

TNIP3

ZNF668

NME2TBC1D4

VDAC1

FFFOXTR

LAMC2

ARL5B

S100A12

ADAMTS6ST6GALNAC5

UACA

RAB2A

HEATR3

SLC24A3

TMEM158

CYP4F11

VEZT

SLC35E1

LYZ

NRAS

PRDX4

TMEM204

GPBAR1

GPRC5A

TM4SF1

SEMA3G

KDELR3

CEACAM6

DOPEY2

CEACAM3

TMED3

TMEM200A

NCEH1 FAM122B

CARSPECAM1

RCN3

TFPT

GZMB

SEC24B

NDFIP2

FARSB

FKBP1444FKBP14FKBP1444SEC61B

KDELR2

ATF6

GNPNAT1

SULF1

KDELR1

CHPF

BCL6B

PCDH17

SEMA5A

RCN1HSPA12B

TWF2GPR4

KCNQ1

FGFRL1

SALL4

MEOX1

CDR2LOTUB2

PODXLSULT2B1

GRP

S100A11LRRC15 MET

SPSB4MPZL2

NOS3

TIMP1AMMECR1NT5DC2FTSJ1AGFG1EIF2AK4EIF2AK4F2 K4K44IF2AK42 K4K4IF2AK4F2 K4K44

LEF1LCN2

DLC1

IL1RAPAPLN

SERPINB5INBINB5INB5SERP1ASAP1

NOD2NXT1

ICAM1

CDS1

NCF2CFNCCPLEK2

ITGA2

NFE2L3

FZD6RALA

REPIN1

BICD1

SPINK1

VIL1

AURKBAAAAURKBRAUAURKBURACCNB1IP1DCBLD1 P1PB1IPPP1IPB1IPIPP1PB1IPPESM1

UPF3A

SPSSPFTL

PYCR1 BACE2

LACTBBBB PHPT1PT1PT1PT1SPTBN5

SNX4

EHD4

YRDC

TCEA1

ARRDC1

CYP51A1

PVR

HS3ST3A1SLC29A2

TSPAN17

LRRC8E

NOTCH3

ICA1

INHA

UBE2INDUFA4L2

ZBTB7A

FNTB

CEBPB

CNN2

CRIP2

E2F7NCOA7

NUAK1

CKAP4

FUT1

LRP8CEP72

88SMAGP

GPC1GPC1PCPLVAP

TUBA1C

KRT17

LBX2 PTPRBBXBX2BXAP2S1

FBXO28

HS6ST2

VAMP5

TM7SF3ENPEPLZTS1

SMYD2S11S1

ZMYND19

FOXQ1CLEC11A

EIF4E2

TNFRSF11BHYAL2

PSMA5TFAP2B

PLXDC2

EFHD1

TRIB3

SLC39A4

EMID1

BAMBI

AFAP1L1

JAZF1

FADDPLAUR

ARHGAP18AAAAAAAAAAAAGUK1

POSTN

GOLGA2

FAM89A

CST1

(a)

ZNF772

CXCL12

LSM14A

CNR2

SLC22A7

CXCXCXIGSF10

KCNB1FLJ25758

ENDOU

GATA2

ZNF780B

GPLD1

MAOB

C1QTNF9

XK

SCUBE1

ZFP36L1IL9

OMA1SIK3

TXNIP

OMA1MMAA1MAMA1MAA1MAMA1MMAA1MATMEM220

SLC17A3

C10orf760oC10o0oTRIM3

PASD1PASRASGRP222 MYCBP2GNG7

ANKRD13A

USP4

RAD21

CLIP4G

CLGNG7G

CAICDA

ZPBP2

44 CD5L

MEF2A

MST1DCAF5

ZNF540PAN2

0PAN2PAN2PAN2

BHMT

LOC401052

ZNF304

SV2A

MAPK10

ICA1L

TOX

EPM2A

CHST9

FCRL4

MYOT

EFCAB1

UST

YTHDC1

BCL2APOBEC2APOBECPO 2C222AAPOBEC2AP ECC2C2APOBECAP 2CC22NEUROG3

SLITRK2

PIK3C2B

PDE6B

NPTX1NHLH1

ATF1 KLK2K2KLK2LK2MTMR10

ZNF704FAM9C

USP8

GATS

ELK1

ACHE

GKAP1KGKKSHBG

C16orf89

PXK

FAM120AOS

RGS9

FAM131B

PELI2

CA14

FBXO15CCDC136

FSIP2

H3F3B

ESYT3CTF1 CRY2

AKAP7

BRS3

TMEM116

RIT2

MTX3

ANTXR2

ARHGAP20

PCF11CR2

RSPO1

EPHB6B66B6LRRTM3

BAALC

33333 RCAN2SLC9A9

CRTC1

GAS8

ATP5L

P2RY1

PPARGC1AFAM124A

ADAMTS19

FAM102A

HVCN1

C7OSBPL11

ARHGEF37

SMAP2

MAZ

CLCN6

FNIP1

NDUFB8METTL7A

TMEM100SEMA6A

BCL6

LYST

TTBK2

TUBCASP8

SSBP2

BBBBBBBBBBBSCN4B

RABGAP1L

TEAD1

DNMT3A

CBFA2T2

GPR162

STAT6

DNAH5RASSF5

DDDNDCDC42EP4

CDH12

STEAP4 SNAP91

DISP1

MFN2

PTBP2

KIAA0513 ARHGAP15

ABCB7 MAP2K6

GATA5ANKS1B

GABRQA

SLC9A3

PTH1R IGSF11

SLC9ASLCSLC9ZNF407

KCNIP3PDE9A

MMMERICH1

PKNOX2

RFX2

SEMA3DFHIT

CRADD

TGM4M4M4M4ZNF268

VAT1L MT1B

ATRNL1

TRERF1TRTRRTRRERF1F1RERF1RERFERF1RRREEERF1ERFRF11PRKG2

FGF13

SMARCA2

FHL1

PDE4D CTNND2

ANK2

ZNF536

COLEC12

KCNIP4

NPVF

KCKKCNIKCCNRSPO2

NIP4PNIP4P4CTNNA2

(b)

Figure 4: Regulatory networks of the master metabolic transcriptional regulators (MMTRs) and their targeted differentially expressed genes(DEGs) between CTS and normal stroma identified by iRegulon [17]. (a) Regulatory network of the MMTRs and their targeted upregulatedgenes in CTS. (b) Regulatory network of the MMTRs and their targeted downregulated genes in CTS. The green color octagon indicatesMMTRs and purple color oval indicates DEGs.

Journal of Oncology 7

Tumor stroma Normal stroma

p = 0.016

−0.4

−0.2

0.0

0.2

0.4

0.6

CD8+

T ce

ll en

richm

ent l

evel

Figure 5: CD8+ T cells have significantly higher enrichment levels in CTS than in colon normal stroma. Student's t-test p value is shown.

associated with tumor progression in a wide variety of cancertypes [55–60].

In addition, we identified 18 transcriptional signaturesof CTS whose expression was significantly associated withDFS in colon cancer individually (Supplementary FigureS3). These genes included CEBPB, BCL2, PAN2, NOS3,FTL, ARHGEF37, SMC3, EP300, JAK2, RPS6KA3, RPS6KA1,PRKACA,HIPK1,HIPK2,MAPK8,GSK3A,CLK2, andCDK3.It indicates that these CTS transcriptional signatures could bebiomarkers for colon cancer relapse.

Furthermore, we used themultivariate analysis to validatethe association between the prognostic CTS transcriptionalsignatures and survival using the colon metabase data [10].For OS analysis, a total of 482 patients were split intotwo groups: high-risk group (N=241) versus low-risk group(N=241) based on the prognostic index (SupplementaryFigure S4A). As expected, the high-risk group had worseOS than the low-risk group (Figure 6(b)). Similarly, forDFS analysis, we divided patients into the high-risk group(N=272) and the low-risk group (N=273) based on theprognostic index (Supplementary Figure S4B) and found thatthe high-risk group had worse DFS compared to the low-riskgroup (Figure 6(c)).These results proved the prognostic valueof these CTS transcriptional signatures in colon cancer.

4. Discussion

The tumor stroma constitutes an important component ofthe TME that mediates tumor growth, immune evasion, andmetastasis [1]. Thus, it is important to identify molecularfeatures in the tumor stroma. To this end, we performeda meta-analysis of three CTS transcriptome datasets foridentifying CTS-associated transcriptional signatures. Weidentified a number of upregulated and downregulated genesin CTS compared to colon normal stroma. Furthermore,we identified upregulated and downregulated pathways sig-nificantly associated with these deregulated genes in CTS.The upregulated pathways were mainly involved in cellulardevelopment, immune regulation, and metabolism, and the

downregulated pathways were mostly metabolism-related.These results revealed the abnormal alterations of cellulardevelopment, immune regulation, and metabolism pathwaysin CTS. We found that CD8+ T cells were more enriched inCTS than in colon normal stroma, suggesting an immuneinfiltration microenvironment in CTS. Furthermore, weidentified numerous CTS transcriptional signatures whoseexpression was significantly associated with prognosis incolon cancer, such as CEBPB, PPARGC1, STAT3, MTOR,BCL2, JAK2, and CDK1. These transcriptional signaturesare mainly involved in immune regulation (CEBPB, STAT3,and JAK2), metabolism (PPARGC1 and MTOR), cell cycle(CDK1), and apoptosis (BCL2), suggesting that the deregu-lation of these pathways in CTSmay contribute to the alteredprognosis in colon cancer.

To verify the association of the identified transcriptionalsignatures with CTS, we analyzed the TCGA colon cancerdataset. We divided these cancers into high-stroma-contentand low-stroma-content groups on the basis of their intra-tumoral stromal content evaluated by ESTIMATE [19] andfound that 153 upregulated genes in CTS had significantlyhigher expression levels in the high-stroma-content groupthan in the low-stroma-content group. These genes included18 hub genes in the PPI network of DEGs and 6 TFs, 40kinases, and 12 MMTRs encoding genes that regulated theDEGs (Supplementary Figure S5, Table S9). We also found27 downregulated genes in CTS which had significantlylower expression levels in the high-stroma-content group,including 14 hub genes, and genes encoding 2 TFs, 18kinases, and 3 MMTRs (Supplementary Figure S5, Table S9).Interestingly, most of the downregulated hub genes in CTSwere also downregulated in the high-stroma-content coloncancers (Supplementary Figure S5). These results indicatethat many transcriptional signatures of CTS identified bythe meta-analysis are tumor stroma-specific. In addition, wefound that CD8+ T cells had significantly higher enrichmentlevels in CTS versus colon normal stroma (Student's t-test,p=0.016), as well as in the high-stroma-content colon can-cers versus the low-stroma-content colon cancers (Student's

8 Journal of Oncology

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

CEBPB

Prob

abili

ty o

f ove

rall

surv

ival

p = 0.036

PPARGC1A

p = 0.003

STAT3

p = 0.009

NFKB1

Prob

abili

ty o

f ove

rall

surv

ival

p = 0.03

CDK1

p = 0.008

CDK5

p = 0.04

BRD2

p = 0.047

RPS6KA5

Prob

abili

ty o

f ove

rall

surv

ival

p = 0.045

ERBB2

Overall survival time (days)

p = 0.018

IKBKE

p = 0.043

MTOR

p = 0.024

Lower expression level

Higher expression level

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

0 30001000

0.0

0.8

0.4

(a)

1.0

0.8

0.6

0.4

0.2

0.00 50 100 150 200

Overall survival time (months)

Prob

abili

ty o

f ove

rall

surv

ival

Concordance Index=63.4,Log-Rank Equal Curves p=1.105e-06,R^2=0.074/0.988Risk Groups Hazard Ratio=2.07 (conf.int.1.53~2.79),p=1.959e-06

Low-risk group

High-risk group

241,+:173,CI=55.8241,+:125,CI=57

241134172860100154241137224679134202

(b)

1.0

0.8

0.6

0.4

0.2

0.00 50 100 150 200

Disease-free survival time (months)

Prob

abili

ty o

f dise

ase f

ree s

urvi

val

Concordance Index=61.64,Log-Rank Equal Curves p=0.0001785,R^2=0.044/0.935Risk Groups Hazard Ratio=2(conf.int.1.38~2.89),p=0.0002425

Low-risk group

High-risk group

273,+:230,CI=57.6272,+:190,CI=56.9

273

145153363118190272

124153464132211

(c)

Figure 6: �e CTS gene signatures whose expression is associated with prognosis in colon cancer. (a) Kaplan-Meier survival curves show thegene signatures whose expression is significantly associated with overall survival (OS) in colon cancer in the TCGA colon cancer dataset (log-rank test, p<0.05). (b) Multivariate Cox regression analysis shows that the OS-associated CTS gene signatures are prognostic for OS in coloncancer in a SurvExpress built-in dataset (colon metabase) [10]. (c) Multivariate Cox regression analysis shows that the DFS-associated CTSgene signatures are prognostic for DFS in colon cancer in a SurvExpress built-in dataset (colon metabase) [10]. DFS: disease-free survival.

Journal of Oncology 9

0 1000 2000 3000 4000

1.0

High-stroma-content colon cancers

Disease-free survival time (days)

Prob

abili

ty o

f dise

ase-

free s

urvi

val

0.0

0.2

0.4

0.6

0.8

0 4000

0.0

0.2

0.4

0.6

0.8

1.0

Low-stroma-content colon cancers

1000 2000 3000

Higher CD8+ T cells enrichment levels Lower CD8+ T cells enrichment levels

p= 0.499 p= 0.031

Figure 7:�e higher enrichment levels of CD8+ T cells were associated with better disease-free survival in the low-stroma-content colon cancers,but not in the high-stroma-content colon cancers. ESTIMATE [19] was used to quantify the intratumoral stromal content (stroma score) ofTCGA colon cancer samples. High-stroma-content: stroma score >median; low-stroma-content: stroma score <median.

t-test, p=3.3∗10−8). It indicates that CD8+ T cells tend to haveelevated infiltration in the TME of colon cancer. Interestingly,we found that the higher enrichment levels of CD8+ T cellswere associated with better DFS in the low-stroma-contentcolon cancers, but not in the high-stroma-content coloncancers (Figure 7). It suggests that the immune cells exert anantitumor effect only when they have infiltrated into tumorcells and that the immune cells in the tumor stroma may nothave such a direct antitumor effect.

This study has identified a number of CTS-associatedtranscriptional signatures that could be biomarkers for coloncancer diagnosis and prognosis and may provide therapeutictargets for colon cancer. However, to translate these findingsinto clinical application, further experimental and clinicalvalidation would be necessary.

5. Conclusions

The identification of CTS-specific transcriptional featuresmay provide insights into the mechanism that mediates thedevelopment of colon cancer and thus has potential clinicalimplications for colon cancer diagnosis and treatment.

Abbreviations

CTS: Colon tumor stromaDEGs: Differentially expressed genesPPI: Protein-protein interactionES: Effect sizeTME: Tumor microenvironmentECM: Extracellular matrixTFs: Transcription factorsMMTRs: Master metabolic transcription factorsOS: Overall survivalDFS: Disease-free survivalCRC: Colorectal cancer

GEO: Gene Expression OmnibusTCGA: The Cancer Genome AtlasFDR: False discovery rateGSEA: Gene-set enrichment analysisFFM: Function-first moduleCFM: Connection-first module.

Data Availability

The datasets (GSE31279, GSE35602, and GSE46824) weredownloaded from the NCBI GEO database (https://www.ncbi.nlm.nih.gov/geo/), the TCGA colon cancer dataset wasdownloaded from the website https://portal.gdc.cancer.gov/,and the colon metabase dataset was from SurvExpress(http://bioinformatica.mty.itesm.mx/SurvExpress).

Conflicts of Interest

The authors declare that they have no conflicts of interestregarding the publication of this paper.

Authors’ Contributions

Md. Nazim Uddin performed data analyses and helpedwrite and prepare the manuscript. Mengyuan Li performeddata analyses and helped prepare the manuscript. XiaoshengWang conceived the research, designed analysis strategies,and wrote the manuscript. All the authors read and approvedthe final manuscript.

Acknowledgments

This work was supported by China Pharmaceutical Uni-versity (grant numbers 3150120001 and 2632018YX01 toXiaosheng Wang).

10 Journal of Oncology

Supplementary Materials

Supplementary Figure S1: illustration of PCA and densityplots as validation tools for batch effect removal. Plot ofprincipal components: (A) before batch effect removal and(B) after batch effect removal. Plot of density: (C) beforebatch effect removal and (D) after batch effect removal. Themultidimensional scaling of the datasets demonstrates that,before the batch effect adjustment, each dataset obviouslyseparated from all the others, whereas, after batch effectadjustment, samples from all the datasets are incorporatedclearly. Supplementary Figure S2: six KEEG pathways thatare significantly associated with the downregulated genesin colon tumor stroma relative to normal stroma. Supple-mentary Figure S3: the genes (DEGs and their upstreamregulators) whose expression is significantly associated withdisease-free survival in colon cancer (log-rank test, p<0.05).Supplementary Figure S4: patients divided into the high-risk group and the low-risk group based on the prognosticgene signatures identified. A.Overall survival. B. Disease-freesurvival. Supplementary Figure S5: numbers of overlappinggenes between the DEGs between colon tumor stroma andnormal stroma and their upstream regulators and the DEGsbetween high-stroma-content and low-stroma-content coloncancers. UP TCGA: upregulated differentially expressedgenes between high-stroma-content and low-stroma-contentTCGA colon cancer samples and DOWN TCGA: downreg-ulated differentially expressed genes between high-stroma-content and low-stroma-content TCGA colon cancer sam-ples. Supplementary Table S1: a summary of the datasetsused in this study. Supplementary Table S2: upregulatedgenes in colon tumor stroma versus colon normal stroma.Supplementary Table S3: downregulated genes in colontumor stroma versus colon normal stroma. SupplementaryTable S4: the top 10 upregulated and top 10 downregulatedgenes in colon tumor stroma. Supplementary Table S5: 44KEGG pathways that were significantly associated with theupregulated genes in colon tumor stroma (CTS). Supple-mentary Table S6: upstream transcription factors and kinasesregulating the differentially expressed genes between colontumor stroma and normal stroma. Supplementary TableS7: master metabolic transcriptional regulators (MMTRs)(iRegulon normalized enrichment score NES > 3.0) regulat-ing the differentially expressed genes between colon tumorstroma and normal stroma. Supplementary Table S8: hubgenes in the protein-protein interaction network of thedifferentially expressed genes between colon tumor stromaand normal stroma. Supplementary Table S9: overlappinggenes between the DEGs between colon tumor stroma andnormal stroma and their upstream regulators and the DEGsbetween high-stroma-content and low-stroma-content coloncancers. (Supplementary Materials)

References

[1] K. C. Valkenburg, A. E. de Groot, and K. J. Pienta, “Targetingthe tumour stroma to improve cancer therapy,” Nature ReviewsClinical Oncology, vol. 15, no. 6, pp. 366–381, 2018.

[2] K. M. Bussard, L. Mutkus, K. Stumpf, C. Gomez-Manzano,and F. C. Marini, “Tumor-associated stromal cells as keycontributors to the tumor microenvironment,” Breast CancerResearch, vol. 18, no. 1, p. 84, 2016.

[3] D. Hanahan and L. M. Coussens, “Accessories to the crime:functions of cells recruited to the tumor microenvironment,”Cancer Cell, vol. 21, no. 3, pp. 309–322, 2012.

[4] F. Bray, J. Ferlay, I. Soerjomataram, R. L. Siegel, L. A. Torre, andA. Jemal, “Global cancer statistics 2018: GLOBOCAN estimatesof incidence and mortality worldwide for 36 cancers in 185countries,” CA: A Cancer Journal for Clinicians, vol. 68, no. 6,pp. 394–424, 2018.

[5] A. Calon, E. Lonardo, and A. Berenguer-Llergo, “Stromalgene expression defines poor-prognosis subtypes in colorectalcancer,” Nature Genetics, vol. 47, pp. 320–329, 2015.

[6] C. Isella, A. Terrasi, S. E. Bellomo et al., “Stromal contributionto the colorectal cancer transcriptome,”Nature Genetics, vol. 47,no. 4, pp. 312–319, 2015.

[7] M. Abba, S. Laufs, M. Aghajany, B. Korn, A. Benner, andH. Allgayer, “Look who’s talking: Deregulated signaling incolorectal cancer,” Cancer Genomics & Proteomics, vol. 9, no. 1,pp. 15–25, 2012.

[8] N. Nishida, M. Nagahara, T. Sato et al., “Microarray analysisof colorectal cancer stromal tissue reveals upregulation of twooncogenic miRNA clusters,” Clinical Cancer Research, vol. 18,no. 11, pp. 3054–3070, 2012.

[9] “GEOAccession viewer,” https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE46824.

[10] R. Aguirre-Gamboa, H. Gomez-Rueda, E. Martınez-Ledesmaet al., “SurvExpress: an online biomarker validation tool anddatabase for cancer gene expression data using survival anal-ysis,” PLoS ONE, vol. 8, no. 9, p. e74250, 2013.

[11] J. Xia, E. E. Gill, and R. E. W. Hancock, “NetworkAnalystfor statistical, visual and network-based meta-analysis of geneexpression data,” Nature Protocols, vol. 10, no. 6, pp. 823–844,2015.

[12] W. E. Johnson, C. Li, and A. Rabinovic, “Adjusting batch effectsin microarray expression data using empirical Bayes methods,”Biostatistics, vol. 8, no. 1, pp. 118–127, 2007.

[13] W. G. Cochran, “The combination of estimates from differentexperiments,” Biometrics, vol. 10, no. 1, pp. 101–129, 1954.

[14] Y. Benjamini and Y. Hochberg, “Controlling the false discoveryrate: a practical and powerful approach to multiple testing,”Journal of the Royal Statistical Society B: Methodological, vol. 57,no. 1, pp. 289–300, 1995.

[15] A. Subramanian, P. Tamayo, V. K. Mootha et al., “Gene setenrichment analysis: a knowledge-based approach for inter-preting genome-wide expression profiles,” Proceedings of theNational Acadamy of Sciences of the United States of America,vol. 102, no. 43, pp. 15545–15550, 2005.

[16] D. J. B. Clarke, M. V. Kuleshov, B. M. Schilder et al., “EXpres-sion2Kinases (X2K) Web: Linking expression signatures toupstream cell signaling networks,” Nucleic Acids Research, vol.46, no. W1, pp. W171–W179, 2018.

[17] R. Janky, A. Verfaillie, H. Imrichova et al., “iRegulon: from agene list to a gene regulatory network using large motif andtrack collections,” PLoS Computational Biology, vol. 10, no. 7,Article ID e1003731, 2014.

[18] D. Petrochilos, A. Shojaie, J. Gennari, and N. Abernethy,“Using random walks to identify cancer-associated modules inexpression data,” BioData Mining, vol. 6, p. 17, 2013.

Journal of Oncology 11

[19] K. Yoshihara, M. Shahmoradgoli, E. Martınez et al., “Inferringtumour purity and stromal and immune cell admixture fromexpression data,” Nature Communications, vol. 4, p. 2612, 2013.

[20] X.-M. Yang, H.-Y. You, Q. Li et al., “CTHRC1 promoteshuman colorectal cancer cell proliferation and invasiveness byactivating Wnt/PCP signaling,” International Journal of Clinicaland Experimental Pathology, vol. 8, no. 10, pp. 12793–12801, 2015.

[21] B. Guo, H. Yan, L. Li, K. Yin, F. Ji, and S. Zhang, “Collagen triplehelix repeat containing 1 (CTHRC1) activates Integrin 𝛽3/FAKsignaling and promotes metastasis in ovarian cancer,” Journal ofOvarian Research, vol. 10, no. 1, p. 69, 2017.

[22] A. M. Chowdhury, H. Katoh, A. Hatanaka et al., “Multipleregulatory mechanisms of the biological function of NRF3(NFE2L3) control cancer cell proliferation,” Scientific Reports,vol. 7, no. 1, p. 12494, 2017.

[23] M. Palma, L. Lopez, M. Garcıa et al., “Detection of collagentriple helix repeat containing-1 and nuclear factor (erythroid-derived 2)-like 3 in colorectal cancer,” BMC Clinical Pathology,vol. 12, p. 2, 2012.

[24] R. M. S. Gill, A. Michael, L. Westley, H. M. Kocher, J. I. Murphy,and G. K. Dhoot, “SULF1/SULF2 splice variants differentiallyregulate pancreatic tumour growth progression,” ExperimentalCell Research, vol. 324, no. 2, pp. 157–171, 2014.

[25] H. Lee, B. Yeh, T. Chan et al., “Sulfatase-1 overexpressionindicates poor prognosis in urothelial carcinoma of the urinarybladder and upper tract,” Oncotarget, vol. 8, no. 29, pp. 47216–47229, 2017.

[26] B. Lu, Y. Fang, J. Xu et al., “Analysis of SOX9 expression incolorectal cancer,” American Journal of Clinical Pathology, vol.130, no. 6, pp. 897–904, 2008.

[27] S. Arikan, C. Cacina, E. Guler, S. Culcu, G. Tuna, and I. Yaylim-Eraltan, “The effects of NOS3 Glu298Asp variant on colorectalcancer risk and progression in Turkish population,” MolecularBiology Reports, vol. 39, no. 3, pp. 3245–3249, 2012.

[28] S. Chida, H. Okayama, M. Noda et al., “Stromal VCAN expres-sion as a potential prognostic biomarker for disease recurrencein stage II-III colon cancer,” Carcinogenesis, vol. 37, no. 9, pp.878–887, 2016.

[29] S. Demirkol, I. Gomceli, M. Isbilen et al., “A combined ULBP2and SEMA5A expression signature as a prognostic and predic-tive biomarker for colon cancer,” Journal of Cancer, vol. 8, no. 7,pp. 1113–1122, 2017.

[30] M. Fujita, Y. Furukawa, T. Tsunoda, T. Tanaka, M. Ogawa, andY. Nakamura, “Up-regulation of the ectodermal-neural cortex 1(ENC1) gene, a downstream target of the𝛽-catenin/T-cell factorcomplex, in colorectal carcinomas,”Cancer Research, vol. 61, no.21, pp. 7722–7726, 2001.

[31] Y. Li, J.Wei, C. Xu, Z. Zhao, and T. You, “Prognostic significanceof cyclin D1 expression in colorectal cancer: a meta-analysisof observational studies,” PLoS ONE, vol. 9, no. 4, Article IDe94508, 2014.

[32] X. Hu, X. Sui, L. Li et al., “Protocadherin 17 acts as a tumoursuppressor inducing tumour cell apoptosis and autophagy, andis frequently methylated in gastric and colorectal cancers,” �eJournal of Pathology, vol. 229, no. 1, pp. 62–73, 2013.

[33] Y. Gu, A. Li, H. Sun et al., “BCL6B suppresses proliferationand migration of colorectal carcinoma cells through inhibitionof the PI3K/AKT signaling pathway,” International Journal ofMolecular Medicine, vol. 41, no. 5, pp. 2660–2668, 2018.

[34] Y. Guo, Y. Bao, M. Ma, and W. Yang, “Identification of keycandidate genes and pathways in colorectal cancer by integrated

bioinformatical analysis,” International Journal of MolecularSciences, vol. 18, no. 4, 2017.

[35] Q. Sun, M. Li, and X. Wang, “The Cancer Omics Atlas:An integrative resource for cancer omics annotations,” BMCMedical Genomics, vol. 11, no. 1, p. 63, 2018.

[36] Q. Huang, S. Li, P. Cheng et al., “High expression of anti-apoptotic protein Bcl-2 is a good prognostic factor in colorectalcancer: result of a meta-analysis,” World Journal of Gastroen-terology, vol. 23, no. 27, pp. 5018–5033, 2017.

[37] S.-H. Yun, M.-S. Roh, J.-S. Jeong, and J.-I. Park, “Peroxisomeproliferator-activated receptor 𝛾 coactivator-1𝛼 is a predictor oflymph nodemetastasis and poor prognosis in human colorectalcancer,” Annals of Diagnostic Pathology, vol. 33, pp. 11–16, 2018.

[38] F. E. Bertrand, C. W. Angus, W. J. Partis, and G. Sigounas,“Developmental pathways in colon cancer,” Cell Cycle, vol. 11,no. 23, pp. 4344–4351, 2012.

[39] A. M. Socovich and A. Naba, “The cancer matrisome: Fromcomprehensive characterization to biomarker discovery,” Semi-nars in Cell & Developmental Biology, 2018.

[40] N. Afratis, C. Gialeli, D. Nikitovic et al., “Glycosaminoglycans:key players in cancer cell biology and treatment,” FEBS Journal,vol. 279, pp. 1177–1197, 2012.

[41] Basappa, K. S. Rangappa, and K. Sugahara, “Roles of gly-cosaminoglycans and glycanmimetics in tumor progression andmetastasis,” Glycoconjugate Journal, vol. 31, no. 6-7, pp. 461–467,2014.

[42] J. Yin, W. Ren, X. Huang, J. Deng, T. Li, and Y. Yin, “Potentialmechanisms connecting purine metabolism and cancer ther-apy,” Frontiers in Immunology, vol. 9, 2018.

[43] C. Frezza, “Histidine metabolism boosts cancer therapy,”Nature, vol. 559, no. 7715, pp. 484-485, 2018.

[44] M. Kumarakulasingham, P. H. Rooney, S. R. Dundas et al.,“Cytochrome P450 profile of colorectal cancer: Identification ofmarkers of prognosis,” Clinical Cancer Research, vol. 11, no. 10,pp. 3758–3765, 2005.

[45] Y.-L. Liu, X. Gao, Y. Jiang et al., “Expression and clinico-pathological significance of EED, SUZ12 and EZH2 mRNAin colorectal cancer,” Journal of Cancer Research and ClinicalOncology, vol. 141, no. 4, pp. 661–669, 2015.

[46] M. L. Slattery, A. Lundgreen, J. S. Herrick, B. J. Caan, J. D. Potter,and R. K. Wolff, “Associations between genetic variation inRUNX1, RUNX2, RUNX3, MAPK1 and eIF4E and risk of colonand rectal cancer: additional support for a TGF-𝛽-signalingpathway,” Carcinogenesis, vol. 32, no. 3, pp. 318–326, 2011.

[47] K. Taniue, A. Kurimoto, Y. Takeda et al., “ASBEL –TCF3complex is required for the tumorigenicity of colorectal cancercells ,” Proceedings of the National Acadamy of Sciences of theUnited States of America, vol. 113, no. 45, pp. 12739–12744, 2016.

[48] T. Bae, K. Rho, J. W. Choi, K. Horimoto, W. Kim, and S. Kim,“Identification of upstream regulators for prognostic expressionsignature genes in colorectal cancer,” BMC Systems Biology, vol.7, no. 1, p. 86, 2013.

[49] T. J. Miller, M. J. McCoy, C. Hemmings, M. K. Bulsara, B.Iacopetta, and C. F. Platell, “The prognostic value of cancerstem-like cellmarkers SOX2 andCD133 in stage III colon canceris modified by expression of the immune-related markersFoxP3, PD-L1 and CD3,” Pathology, vol. 49, no. 7, pp. 721–730,2017.

[50] J. Gupta, I. delBarcoBarrantes, A. Igea et al., “Dual function ofp38𝛼 MAPK in colon cancer: suppression of colitis-associatedtumor initiation but requirement for cancer cell survival,”Cancer Cell, vol. 25, no. 4, pp. 484–500, 2014.

12 Journal of Oncology

[51] S. R. Rosario, M. D. Long, H. C. Affronti, A. M. Rowsam, K. H.Eng, and D. J. Smiraglia, “Pan-cancer analysis of transcriptionalmetabolic dysregulation using the cancer genome atlas,”NatureCommunications, vol. 9, 2018.

[52] D. Liu, X.-X. Zhang,M.-C. Li et al., “C/EBP𝛽 enhances platinumresistance of ovarian cancer cells by reprogramming H3K79methylation,”Nature Communications, vol. 9, no. 1, p. 1739, 2018.

[53] J. Feilchenfeldt, M.-A. Brundler, C. Soravia, M. Totsch, and C.A. Meier, “Peroxisome proliferator-activated receptors (PPARs)and associated transcription factors in colon cancer: reducedexpression of PPARgamma-coactivator 1 (PGC-1),” Cancer Let-ters, vol. 203, no. 1, pp. 25–33, 2014.

[54] Y. Zhang, Y. Ba, C. Liu et al., “PGC-1alpha induces apoptosis inhuman epithelial ovarian cancer cells through a PPARgamma-dependent pathway,” Cell Research, vol. 17, no. 4, pp. 363–373,2007.

[55] J. Concetti andC. L.Wilson, “NFKB1 and cancer: friend or foe?”Cells, vol. 7, no. 9, 2018.

[56] U. Saini, S. Naidu, A. C. Elnaggar et al., “Elevated STAT3expression in ovarian cancer ascites promotes invasion andmetastasis: A potential therapeutic target,” Oncogene, vol. 36,no. 2, pp. 168–181, 2017.

[57] W. J. Yang, Y. Qu, B. X. Tan et al., “Prognostic significance ofpreoperative IKBKE expression in esophageal squamous cellcarcinoma,” OncoTargets and �erapy, vol. 11, pp. 1305–1314,2018.

[58] W. Sung, Y. Lin, P. Wu et al., “High nuclear/cytoplasmic ratioof Cdk1 expression predicts poor prognosis in colorectal cancerpatients,” BMC Cancer, vol. 14, p. 951, 2014.

[59] K. Zhuang, J. Zhang, M. Xiong et al., “CDK5 functions as atumor promoter in human colorectal cancer viamodulating theERK5–AP-1 axis,” Cell Death & Disease, vol. 7, no. 10, p. e2415,2016.

[60] Y. Hu, J. Zhou, F. Ye et al., “BRD4 inhibitor inhibits colorectalcancer growth and metastasis,” International Journal of Molecu-lar Sciences, vol. 16, no. 1, pp. 1928–1948, 2015.

Stem Cells International

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

MEDIATORSINFLAMMATION

of

EndocrinologyInternational Journal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Disease Markers

Hindawiwww.hindawi.com Volume 2018

BioMed Research International

OncologyJournal of

Hindawiwww.hindawi.com Volume 2013

Hindawiwww.hindawi.com Volume 2018

Oxidative Medicine and Cellular Longevity

Hindawiwww.hindawi.com Volume 2018

PPAR Research

Hindawi Publishing Corporation http://www.hindawi.com Volume 2013Hindawiwww.hindawi.com

The Scientific World Journal

Volume 2018

Immunology ResearchHindawiwww.hindawi.com Volume 2018

Journal of

ObesityJournal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Computational and Mathematical Methods in Medicine

Hindawiwww.hindawi.com Volume 2018

Behavioural Neurology

OphthalmologyJournal of

Hindawiwww.hindawi.com Volume 2018

Diabetes ResearchJournal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Research and TreatmentAIDS

Hindawiwww.hindawi.com Volume 2018

Gastroenterology Research and Practice

Hindawiwww.hindawi.com Volume 2018

Parkinson’s Disease

Evidence-Based Complementary andAlternative Medicine

Volume 2018Hindawiwww.hindawi.com

Submit your manuscripts atwww.hindawi.com