Bioinformatic Analysis Suggests That Three Hub Genes May Be a Vital Prognostic Biomarker in Pancreatic Ductal Adenocarcinoma

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal malignancies worldwide due to its ineffective diagnosis and poor prognosis. It is essential to identify differentially expressed genes (DEGs) in PDAC to gain new insights into its underlying molecular mechanisms, as well as identify potential diagnostic and therapeutic targets. We screened 135 DEGs from the GSE15417, GSE16515, and GSE28735 PDAC and normal pancreatic tissue microarray data sets, and identified 16 DEGs that were correlated with PDAC prognosis through the Kaplan–Meier survival analysis and log-rank tests. The Cancer Genome Atlas and Oncomine databases validated the expression levels of 16 candidate genes (SLC6A14, GPRC5A, IFI27, ERP27, SDR16C5, SIDT2, TCN1, COL12A1, MMP1, CEACAM6, DKK1, ITGA2, KRT19, PLAU, ANO1, and GABRP). Weighted gene coexpression network analysis (WGCNA) and protein and protein interaction (PPI) analysis identified three hub genes—ERP27, ITGA2, and MMP1—that are likely important in PDAC prognosis. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis demonstrated that they were enriched in functions of extracellular matrix organization, extracellular structure organization, and positive regulation of cell migration. Taken together, we identified three pivotal genes for PDAC, which can improve our understanding of its pathogenesis, progression, and prognosis.

1. Introduction

Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal malignancies on account of its ineffective diagnosis and therapy (Oncology, 2014). The median survival duration of PDAC is less than 6 months, and the 5-year overall survival (OS) rate is a dismal 7.7% (Pierce, 2012; Oncology, 2014; andrew_aguirre@dfci.harvard.edu, 2017). Although early detection and surgical resection have increased the 5-year survival rate of stage I PDAC patients to 50%, most patients are already at the advanced stage when diagnosed (Hernandez and Lucas, 2016; Li et al., 2016; Taucher et al., 2016). Therefore, it is crucial to identify diagnostic or prognostic biomarkers and therapeutic targets to improve clinical outcomes.

Several mutations drive the development and progression of PDAC (Dickson, 2017), and understanding these genetic alterations in the context of biological pathways can help identify specific novel biomarkers. Zhou et al. (2015) demonstrated that p21-activated kinase 1 (PAK1) regulated the downstream signaling pathways of multiple growth factors, such as hepatocyte growth factor and MET receptor tyrosine kinase. Recent studies indicate that KRAS, SMAD4, CDK6, SOX9, GATA6, PIK3CA, and PIK3R3 are frequently mutated in PDAC (Fakhri and Lim, 2017). The mutagenic landscape of PDAC also includes genes related to histone modification, DNA damage repair, and mitotic cytokinesis, as well as drug targets such as ERBB2, MET, and FGFR1 (Dunne and Hezel, 2015; Long et al., 2016; Zhou et al., 2018).

High-throughput gene profiling and sequencing technologies have helped identify cancer-associated genes and biological pathways on the basis of histological tumor classification and clinical data. The Cancer Genome Atlas (TCGA) database includes the gene expression data of almost 30 human cancers (Jia et al., 2018), and the Gene Expression Omnibus (GEO) is a public database of microarrays or gene profiles of a wide range of diseases, including cancer (Edgar and Lash, 2002). The weighted gene coexpression network analysis (WGCNA) helps identify the correlation patterns among genes across microarray samples, and gene set enrichment analysis (GSEA) is a computational method that identifies the significance of a gene set vis-a-vis different physiological states (Holden et al., 2008). These bioinformatic resources have unearthed several novel biomarkers for the diagnosis, treatment, and prognosis of different cancer types, essential for the development of individualized treatment strategies.

The aim of this study was to systematically review all available gene expression data for PDAC, and perform a pooled analysis using the above bioinformatic tools to identify biomarkers and molecular targets. We identified ERP27, ITGA2, and MMP1 as novel potential diagnostic and prognostic biomarkers for PDAC, which need to be validated further for clinical application.

2. Materials and Methods

2.1. Patient samples

The study included two PDAC patients recruited from Shenzhen People's Hospital. Written informed consent was obtained from all patients before surgery, and the study was approved by our Institutional Ethics Committee.

2.2. Microarray data and identification of differentially expressed genes

The GSE15471, GSE16515, and GSE28735 microarray data sets—consisting of gene expression data of 117 PDAC and 97 adjacent nontumor tissues—were obtained from the GEO. The raw data of each microarray were normalized and analyzed by R package, limma, and RobustRankAggreg package. Differentially expressed genes (DEGs) were identified by classical t test with false discovery rate <0.05 and | logFC | >1 as the cutoff criterion.

2.3. Gene Ontology, Kyoto Encyclopedia of Genes and Genomes pathway enrichment, and GSEA

To indentify the potential functions of the identified DEGs, Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and GSEA were performed by R cluster Profiler package as previously described (Yu et al., 2012).

2.4. WGCNA and protein and protein interaction network analysis

The DEGs were qualified and then used to construct a WGCNA network by R WGCNA package. The primary protein and protein interaction (PPI) network was constructed by module node >0.6, and uploaded onto the Cytoscape software (www.cytoscape.org/), which was used to construct PPI subnetworks and obtain candidate genes by calculating node degree.

2.5. OS analysis

To analyze the prognostic value of the candidate genes, the PDAC patients were stratified into the high-expression and low-expression groups for each DEG, using the median expression level from TCGA (https://cancergenome.nih.gov/) as the cutoff. The two patient cohorts for each DEG were compared by Kaplan–Meier analysis and the log-rank test using the R survival package.

2.6. Analysis of candidate gene expression levels

The expression of candidate genes was analyzed by FirebrowseR package for PDAC from TCGA, which offers customizable functions such as tumor/normal tissue differential expression analysis, profiling according to cancer types or pathological stages, patient survival analysis, similar gene detection, correlation analysis, and dimensionality reduction analysis. The expression data were further validated using Oncomine (https://www.oncomine.org), a web application for translational bioinformatics that can compute gene expression signatures, clusters, and gene set modules, and automatically extracts biological insights from the raw data.

2.7. Identification of the candidate gene protein expression

The expression of the proteins encoded by the candidate genes was identified from the Human Protein Atlas (HPA; www.proteinatlas.org/), which includes quantitative transcriptomic data (RNA-Seq) and spatial proteomic data from immunohistochemistry (IHC) on tissue microarrays.

2.8. Immunohistochemistry

The expression of relative proteins can be analyzed by IHC. First, tissue slices were incubated overnight at 4°C with primary antibodies to ERP27, ITGA2 and MMP1. Then they were incubated with secondary antibodies at room temperature for 1 hour. Finally, slides were stained with DAB, counterstained with haematoxylin and photographed.

2.9. Statistical analysis

The data are presented as mean ± standard deviation. Dependent t test was used to compare the mean between two groups. Figure 5 shows the log-rank tests used.

3. Results

3.1. Identification of DEGs

Three microarray data sets of pancreatic cancer and adjacent nontumor tissues—GSE15471, GSE16515, and GSE28735—were downloaded from the GEO database and analyzed. GSE15471 and GSE28735 contained data from 36 PDAC and 45 paired adjacent nontumor tissues, and GSE1615 from 36 PDAC and 16 adjacent nontumor tissues, altogether comprising 117 PDAC and 97 nontumor tissues. A total of 135 DEGs were identified, of which 86 were upregulated and 49 downregulated in the PDAC compared with the normal tissues (Table 1). The top 20 DEGs are shown in Figure 1.

FIG. 1.

The DEGs in PDAC GSE15417, GSE16515, and GSE28735 were obtained and calculated by R package, and then indentified DEGs as FDR <0.05 and |FC| >1 as cutoff. The top 20 DEGs are shown in the figure, and red/blue were relative upregulation/downregulation comparing adjacent nontumor tissues with PDAC tissues. DEGs, differentially expressed genes; FDR, false discovery rate; PDAC, pancreatic ductal adenocarcinoma.

Table 1.

The Differentially Expressed Genes in Pancreatic Ductal Adenocarcinoma

Regulation	DEGs
SIDT2	GAMT	TCEA3	CCDC69	SLC30A2
CELA2B	TMED6	ERP27	CPA2	CTRL
RBPJL	CLPS	EGF	KLK1	PLA2G1B
PAIP2B	CELA3B	AOX1	ERO1B	CELA3A
PSAT1	REG1A	SLC43A1	P2RX1	BACE1
ABAT	GATM	HOMER2	MT1G	RGN
TMEM97	IL22RA1	FAM46C	GAS2	NUCB2
EPB41L4B	FAM129A	BTG2	SERPINI2	SYCN
Up-expression
KIAA1324	PDIA2	GNMT	LGALS2	SLC16A10
SLC4A4	C5	SLC39A5	PAK3	MT1M
GP2	FAM150B	PNLIP	NR5A2	BNIP3
RAB26	MT1F	PNLIPRP1	PNLIPRP2	CEL
CPA1	TRHDE	ACADL	ANPEP	PRSS3
EPHX1	NRCAM	CELP	AQP8	PM20D1
GPHA2	TPST2	MCOLN3	ALB	CELA2A
DPP10	GUCA1C	PRSS3P2	CTRC	AQP12B
EPHX2	DPEP1	IAPP	GSTA1	PDK4
FGL1
S100P	SLC6A14	CEACAM6	CTSE	CEACAM5
CLDN18	COL11A1	TFF1	SULF1	GABRP
LAMC2	POSTN	CST1	GJB2	KRT19
SDR16C5	MMP12	MMP1	LCN2	THBS2
Down-expression
TCN1	INHBA	DKK1	SLPI	ANXA10
CCL20	GPRC5A	ITGA2	MMP7	IFI27
GCNT3	KRT17	TMC5	COL8A1	PLAU
COL12A1	PLPP4	CAPG	ANO1	SERPINB3
EPYC	OLR1	CP	FOXQ1	COMP
COL10A1	TMPRSS4	CXCL5	FN1

DEGs, differentially expressed genes.

3.2. Functional enrichment analysis

The DEGs showed significant enrichment in the biological process (BP) terms of leukocyte migration, digestion, and extracellular matrix (ECM) organization, the cellular component (CC) terms proteinaceous ECM, endoplasmic reticulum lumen, and secretory granule lumen, and the molecular function (MF) terms endopeptidase activity, serine-type endopeptidase activity, serine-type peptidase activity, and serine hydrolase activity (Fig. 2A). It was displayed that most significant genes were involved in significant biological function, for instance, MMP1 may regulate digestion, multicellular organismal catabolic process, ECM organization, extracellular structure organization, and leukocyte (Fig. 2B). It was displayed that the gene network may regulate the biological function, such as FN1, ITGA2, LAMC2, CEACAM5, CXCL5, OLR1, and DKK1 regulating positive regulation of cell migration (Fig. 2C). Core enriched genes in metabolic pathways for KEGG pathway were discovered (Fig. 2D). Running enrichment score was the traditional method for visualizing GSEA. The results showed significant gene enrichment in pancreatic secretion (Fig. 2E).

FIG. 2.

GO and Pathway analysis of the DEGs. (A) Our gene ontology analysis showed that DEGs are significantly enriched in BP, CC and MF. (B) It was yielded that the genes with the most substantial difference were involved in significant biological function. (C) It was displayed that the gene networks may regulate the biological function, such as FN1, ITGA2, LAMC2, CEACAM5, CXCL5, OLR1, and DKK1 regulating positive regulation of cell migration. (D) KEGG pathway discovered core enriched genes in metabolic pathways. (E) GSEA showed significant gene enrichment in pancreatic secretion. BP, biological process; CC, cellular component; ECM, extracellular matrix; GO, gene ontology; GSEA, gene set enrichment analysis; ITGA2, integrin α2; KEGG, Kyoto Encyclopedia of Genes and Genomes; MF, molecular function; MMP1, matrix metalloproteinase 1.

3.3. Survival analysis for candidate genes

The PDAC patients were stratified into the high-expressing (n = 86) and the low-expressing (n = 49) groups for the 135 DEGs, using the median expression level as the cutoff, and the survival was analyzed by R package (data not shown). Sixteen DEGs were correlated with PDAC prognosis, with high expression of 16 of them—ITGA2, MMP1, ERP27, ANO1, COL12A1, DKK1, EPYC, GABRP, KRT19, PLAU, CEACAM6, GPRC5A, IFI27, SDR16C5, SLC6A14, and TCN1—significantly associated with shorter survival (p = 0.001, p = 0.027, p = 0.036, p = 0.002, p = 0.004, p = 0.014, p = 0.005, p = 0.027, p = 0.004, p = 0.002, p = 0.018, p < 0.001, p = 0.003, p = 0.0005, p = 0.005, and p = 0.021, respectively); in contrast, higher SIDT2 expression was associated with longer survival (p = 0.04) (Fig. 3). Therefore, these 17 DEGs were further analyzed.

FIG. 3.

Survival analysis for DEGs. The DEGs subjected to survival analysis by R package. Kaplan–Meier survival analysis and log-rank tests showed that higher ITGA2, MMP1, ERP27, ANO1, COL12A1, DKK1, EPYC, GABRP, KRT19, PLAU, CEACAM6, GPRC5A, IFI27, SDR16C5, SLC6A14, and TCN1 expression was associated with shorter survival time (p = 0.001, p = 0.027, p = 0.036, p = 0.002, p = 0.004, p = 0.014, p = 0.005, p = 0.027, p = 0.004, p = 0.002, p = 0.018, p < 0.001, p = 0.003, p = 0.0005, p = 0.005, and p = 0.021, respectively), however, the higher SIDT2 expression had longer survival time (p = 0.04).

3.4. DEG expression in PDAC

The mRNA expression data from 178 PDAC and 4 normal tissues were downloaded from TGCA, and the expression levels of the above candidate genes were compared using the R package. ERP27 and SID2 were downregulated (p = 0.03 and p < 0.01, respectively), and CEACAM6, GPRC5, IFI27, SDR16C5, SLC6A14, and TCN1 were upregulated in PDAC (p < 0.01) (Fig. 4A). ANO1, COL12A1, DKK1, GABRP, ITGA2, KRT19, MMP1, and PLAU were overexpressed in normal tissues (p < 0.01), while EPYC levels were similar in both PDAC and normal tissues (p > 0.05) (Fig. 4B). Oncomine database validated the findings of TGCA, and the results demonstrated that ERP27 and SID2 were underexpressed in PDAC (p < 0.01), but ITGA2, MMP1, CEACAM6, GPRC5, IFI27, SDR16C5, SLC6A14, TCN1, ANO1, COL12A1, DKK1, GABRP, KRT19, EPYC, and PLAU were overexpressed in PDAC (p < 0.01) (Fig. 5). Taken together, the expression of all candidate genes except egg yolk phosphatidylcholine (EYPC) in TCGA and Oncomine was consistent with the gene expression profiles from GEO.

FIG. 4.

The candidate gene expression in TCGA data. The candidate genes expression was performed by R package from TCGA data. (A) It was discovered that ERP27 and SID2 were downregulated in TP (p = 0.03 and p < 0.01, respectively), and CEACAM6, GPRC5, IFI27, SDR16C5, SLC6A14, and TCN1 were upregulated in TP (p < 0.01). (B) It was demonstrated that ANO1, COL12A1, DKK1, GABRP, ITGA2, KRT19, MPP1, and PLAU were overexpressed in TP (p < 0.01), but no significant difference for EPYC (p > 0.05). TCGA, The Cancer Genome Atlas.

FIG. 5.

The candidate gene expression in Oncomine data set. The results showed that ERP27 and SID2 were underexpressed in PDAC (p < 0.01); however, ITGA2, MMP1, CEACAM6, GPRC5, IFI27, SDR16C5, SLC6A14, TCN1, ANO1, COL12A1, DKK1, GABRP, KRT19, EPYC, and PLAU were overexpressed in PDAC (p < 0.01).

3.5. WGCNA and PPI network analysis and candidate gene identification

DEGs with similar expression patterns were clustered by WGCNA, which divided them into three modules: gray, turquoise, and blue (Fig. 6A) (Supplementary Table S1), and the PPI network was then constructed with 0.6 as the threshold. Further PPI subnetworks were constructed using Cytoscape, resulting in 39 nodes and 237 edges. Combined with the survival analysis, we obtained three core candidate genes—ITGA2, MMP1, and ERP27—on the basis of node degrees. The red and blue represent the up- and downregulation, respectively, and the degree is displayed with the size (Fig. 6B).

FIG. 6.

WGCNA and PPI analysis and candidate gene identification. (A) DEGs with similar expression patterns were clustered by WGCNA, which showed that DEGs can be divided into three modules, gray, turquoise, and blue. (B) PPI analysis suggested combining with survival analysis, and ITGA2, MMP1, and ERP27 were hub genes due to node degree. The red and blue represent the up- and downregulation, respectively, and the degree is displayed with the size. PPI, protein and protein interaction; WGCNA, weighted gene coexpression network analysis.

3.6. Candidate gene protein expression validation

The IHC and patient clinical data were extracted from HPA, which showed higher in situ expression of ITGA2 and MMP1 in PDAC tissues compared with normal pancreatic tissues, while ERP27 expression was lower in the PDAC tissues (Fig. 7). We have tested ITGA2, MMP1, and ERP27 protein expression in two paired PDAC tissues by IHC, and the results showed a higher ITGA2 and MMP1 expression in PDAC tissues compared with normal pancreatic tissues; yet, ERP27 expression was lower in the PDAC tissues (Supplementary Fig. S1).

FIG. 7.

Candidate gene protein expression validation. The IHC data were obtained from HPA. Staining yielded ITGA2, and MMP1 protein expression was higher in PDAC tissues than in normal pancreatic tissues, but ERP27 protein expression was lower in PDAC tissues than in normal pancreatic tissues. HPA, Human Protein Atlas; IHC, immunohistochemistry.

4. Discussion

PDAC has one of the worst survival rates among all cancers. Over the last decade, comparative genome-wide analyses has identified several loci associated with a high susceptibility to PDAC (Bhasin et al., 2015; Kong et al., 2016; Makohon-Moore et al., 2017; Ghosh et al., 2018). In the present study, we identified 135 DEGs from the GSE15471, GSE16515, and GSE28735 data sets of PDAC and normal pancreatic tissue microarrays, which minimized the statistical deviation (Zhou et al., 2018). To further extract the tumor-related genes from the DEGs with greater accuracy, we analyzed the survival trends of the PDAC patients relative to the DEG expression levels, and found that 16 genes were negatively correlated and only 1 gene was positively correlated with survival. Of these 17 genes (SLC6A14, GPRC5A, IFI27, ERP27, SDR16C5, SIDT2, TCN1, COL12A1, EPYC, MMP1, CEACAM6, DKK1, ITGA2, KRT19, PLAU, ANO1, and GABRP), all except EPYC were validated by TGCA as candidate genes likely playing a pivotal role in PDAC development. Seven of these candidate genes (SLC6A14, GPRC5A, ANO1, GABRP, ITGA2, KRT19, and PLAU) have been previously associated with PDAC.

In addition, ANO1, ITGA2, and KRT19 are well-established oncogenes (Ischenko et al., 2013; Liu et al., 2015; Nones et al., 2014; Sauter et al., 2015; Sauter et al., 2015a, 2015b; Yao et al., 2016), and genetic analysis has further reinforced the association between PLAU and PDAC (Bournet et al., 2012; Krug et al., 2014; Gutiérrez et al., 2015). A previous study showed that SLC6A14, GPRC5A, and GABRP were suitable therapeutic biomarkers for PDAC (Karunakaran et al., 2010; Mccann, 2015; Nakamura, 2008; Penheiter et al., 2015; Takehara et al. 2007). In this study, we identified some additional oncogenes—namely IFI27, ERP27, SDR16C5, SIDT2, TCN1, and COL12A1—that have not been previously linked to PDAC.

According to GO and KEGG pathway analyses, the most significantly enriched CC, MF, and BP terms among the oncogenes are proteinaceous ECM, endopeptidase, and serine-type endopeptidase activity, and positive regulation of leukocyte migration and ECM organization, respectively. This is highly significant since the ECM is a major factor influencing cancer metastasis, and PDAC in particular is highly invasive and metastatic (He et al., 2013). Furthermore, pancreatic secretion and digestion are now considered an important oncogenic function in the gastrointestinal cancers, and aberrant expression of genes that regulate ECM organization and pancreatic secretion has been associated with aggressive pancreatic tumors (Laklai et al., 2012; He et al., 2013; Biondani et al., 2018).

The major hub genes involved in PDAC were further screened by combining GSEA, WGCNA, PPI networks and survival analysis, which identified MMP1, ITGA2, and ERP27. All three genes were enriched in ECM organization and extracellular structure functions. MMP1 and ITGA2 were overexpressed, while ERP27 was inhibited in the PDAC tissues. Matrix metalloproteinase 1 (MMP1) gene is part of a cluster of MMP genes on chromosome 11 that encode enzymes such as collagenase-1, fibroblast collagenase, and interstitial collagenase. MMP1 is the key mediator in the breakdown of ECM, with major roles in embryogenesis, propagation, and tissue remodeling in various physiological processes. MMP1 also has a significant impact on cancer metastasis, since the degradation of basal membranes and ECM enables cancer cell migration and invasion. MMP1 overexpression is associated with the metastasis of various cancers, such as breast cancer, lung carcinoma, and colorectal cancer (Sunami et al., 2000; Hinoda et al., 2002; Zinzindohoué et al., 2005; Decock et al., 2008; Sauter et al., 2008).

Integrin α2 (ITGA2) gene is located on chromosome 5 and encodes the transmembrane receptor for collagens and related proteins. It is responsible for the adhesion of platelets and other cell types to the ECM (Ji et al., 2011; Heino, 2012). Nones et al. (2014) demonstrated upregulation of ITGA2 in 167 untreated resected PDAC compared with 29 adjacent nontransformed pancreatic tissues. In addition, ITGA2 mRNA levels are significantly high in gastric cancer tissues, and blocking ITGA2 not only inhibited cell migration but also induced apoptosis in gastric cancer cells (Chen et al., 2011). Xu et al. (2017) reported that ITGA2 stimulated yes-associated protein (YAP) activity and was associated with poor survival of PDAC patients. Furthermore, Zhao et al. (2015) found that ITGA2 was also a key gene involved in hepatocellular carcinoma (HCC) development.

ERP27 encodes a redox-inactive member of the protein-disulfide isomerase family, which assists almost one-third of all cellular proteins that pass and undergo oxidative folding in the endoplasmic reticulum, indicating its function in regulating pancreatic secretion (Alanen et al., 2006). Transcriptome analysis showed higher expression of ERP27 in the exocrine glandular cells compared with other tissues, which decreased significantly in an acute pancreatic model (Chen et al., 2010; Kober et al., 2013).

Taken together, we identified three pivotal genes that are potentially involved in PDAC prognosis through bioinformatic analysis, greatly contributing to our understanding of the pathogenesis, progression, and prognosis of PDAC. Further studies are needed to validate these genes functionally and elucidate their underlying molecular mechanisms.

Footnotes

Acknowledgment

The authors thank Prof. Guangchuang Yu for the use of R package.

Availability of Data and Materials

The data sets used during the present study are available from the corresponding authors on reasonable request. Study concept and design: L.-s.W., J.Y., and D.-f.L. Acquisition of data: X.C. and M.-f.Y. Analysis of data: D.-f.L. and M.-f.Y. Drafting of the article: D.-f.L. Statistical analysis: X.C. Obtained funding: J.Y. and D.-f.L. Revision of the article: W.F.

Ethics Approval and Consent to Participate

Shenzhen People's Hospital Institutional Ethics Committee.

Consent for Publication

The patients provided consent for data to be published. All authors gave approval for publication.

Author Disclosure Statement

The authors declare they have no conflicting financial interests.

Funding Information

This study was supported by the Natural Science Foundation of Hunan Province (No. 2017JJ3270); the Natural Science Foundation of Guangdong Province (No.2018A0303100024); the Three Engineering Training Funds in Shenzhen (No. SYLY201718 and ), and the Technical Research and Development Project of Shenzhen (No. JCYJ20150403101028164).

Supplementary Material

References

Alanen

H.I.

, Williamson

R.A.

, Howard

M.J.

, et al. 2006. ERp27, a new non-catalytic endoplasmic reticulum-located human protein disulfide isomerase family member, interacts with. ERp57. 281, 33727–33738.

Balachandran

V.P.

, Beatty

G.L.

, and Dougan

S.K.

2019. Broadening the impact of immonotherapy to pancreatic cancer: Challenges and opportunities. Gastroenterology. 156, 2056–2072.

Bhasin

, Bucur

, Ndebele

, et al. 2015. Abstract PR07: Transcriptome meta-analysis identifies new 5-gene classifier for early detection of pancreatic cancer. Cancer Res. 75, PR07-PR07.

Biondani

, Zeeberg

, Greco

M.R.

, et al. 2018. Extracellular Matrix composition modulates PDAC parenchymal and stem cell plasticity and behavior through the secretome. FEBS J. 85, 2104–2124.

Bournet

, Pointreau

, Souque

, et al. 2012. Gene expression signature of advanced pancreatic ductal adenocarcinoma using low density array on endoscopic ultrasound-guided fine needle aspiration samples. Pancreatology. 12, 27–34.

Chen

, Liu

N.N.

, Li

J.Q.

, et al. 2011. Association between ITGA2 C807T polymorphism and gastric cancer risk. World J. Gastroenterol. 17, 2860–2866.

Chen

, Sans

M.D.

, Strahler

J.R.

, et al. 2010. Quantitative organellar proteomics analysis of rough endoplasmic reticulum from normal and acute pancreatitis rat pancreas. J. Proteome Res. 9, 885.

Decock

, Hendrickx

, Vanleeuw

, et al. 2008. Plasma MMP1, MMP8 and MMP13 expression in breast cancer: Protective role of MMP8 against lymph node metastasis. Breast Cancer Res. 10, P31.

Dickson

2017. Pancreatic cancer: PDAC metastases show identical known driver gene mutations. Nat. Rev. Gastroenterol. Hepatol. 14, 138.

10.

Dunne

R.F.

, and Hezel

A.F.

2015. Genetics and biology of pancreatic ductal adenocarcinoma. Hematol. Oncol. Clin. North Am. 29, 595–608.

11.

Edgar

, and Lash

2002. The Gene Expression Omnibus (GEO): A gene expression and hybridization repository. Natl. Center Biotechnol. Inf. 207–210.

12.

Fakhri

, and Lim

K.H.

2017. Molecular landscape and sub-classification of gastrointestinal cancers: A review of literature. J. Gastrointest. Oncol. 8, 379.

13.

Ghosh

, Gunewardena

, Paul

, et al. 2018. Abstract LB-A26: Regulation of super-enhancers by GZ17–6.02 has therapeutic effect in pancreatic cancer (PDAC).

14.

Gutiérrez

M.L.

, Corchete

, Teodosio

, et al. 2015. Identification and characterization of the gene expression profiles for protein coding and non-coding RNAs of pancreatic ductal adenocarcinomas. Oncotarget. 6, 19070–19086.

15.

, Di

, Liang

, et al. 2013. The microRNA-218 and ROBO-1 signaling axis correlates with the lymphatic metastasis of pancreatic cancer. Oncol. Rep. 30, 651–658.

16.

Heino

2012. Integrin α2 (ITGA2). Encyclopedia of Signaling Molecules, 962–966.

17.

Hernandez

Y.G.

, and Lucas

A.L.

2016. MicroRNA in pancreatic ductal adenocarcinoma and itsprecursor lesions. World J. Gastrointest. Oncol. 8, 18–29.

18.

Hinoda

, Okayama

, Takano

, et al. 2002. Association of functional polymorphisms of matrix metalloproteinase (MMP)-1 and MMP-3 genes with colorectal cancer. Int. J. Cancer. 102, 526–529.

19.

Holden

, Deng

, Wojnowski

, et al. 2008. GSEA-SNP. Bioinformatics. 24, 2784–2785.

20.

Ischenko

, Zhi

, Moll

U.M.

, et al. 2013. Direct reprogramming by oncogenic Ras and Myc. Proc. Natl. Acad. Sci. U. S. A. 110, 3937–3942.

21.

, Long

, Briody

, et al. 2011. Progesterone modulates integrin {alpha}2 (ITGA2) and {alpha}11 (ITGA11) in the pregnant cervix. Reprod. Sci. 18, 156–163.

22.

Jia

, Li

, et al. 2018. Mining TCGA database for genes of prognostic value in glioblastoma microenvironment. Aging. 10, 592–605.

23.

Karunakaran

, Schoenlein

P.V.

, Thangaraju

, et al. 2010. Abstract 4838: Alpha-methyltryptophan, a blocker of the amino acid transporter SLC6A14, induces autophagy and apoptosis in SLC6A14-positive breast cancer and colon cancer cell lines. Cancer Res. 70, 4838.

24.

Kober

F.X.

, Koelmel

, Kuper

, et al. 2013. The crystal structure of the protein-disulfide isomerase family member ERp27 provides insights into its substrate binding capabilities. J. Biol. Chem. 288, 2029.

25.

Kong

, Wu

, Cheng

, et al. 2016. A subset of metastatic pancreatic ductal adenocarcinomas depends quantitatively on oncogenic Kras/Mek/Erk-induced hyperactive mTOR signalling. Gut. 65, 647.

26.

Krug

, Kühnemuth

, Griesmann

, et al. 2014. CUX1: A modulator of tumour aggressiveness in pancreatic neuroendocrine neoplasms. Endocr. Relat. Cancer. 21, 879–890.

27.

Laklai

, Kalluri

, Moses

, et al. 2012. Abstract A47: Reciprocal dialogue between cellular and extracellular matrix tension and tissue inflammation in pancreatic tumor progression. Cancer Res. 72, A47.

28.

, Yang

, Liao

, et al. 2018. Linc00483 as ceRNA regulates proliferation and apoptosis through activating MAPKs in gastric cancer. J. Cell. Mol. Med. 22, 3875–3886.

29.

, Hao

, Wang

, et al. 2016. Circular RNA expression profile of pancreatic ductal adenocarcinoma revealed by microarray. Cell. Physiol. Biochem. 40, 1334.

30.

Liu

P.F.

, Jiang

W.H.

, Han

Y.T.

, et al. 2015. Integrated microRNA-mRNA analysis of pancreatic ductal adenocarcinoma. Genet. Mol. Res. 14, 10288–10297.

31.

Long

, Zhang

, Liu

, et al. 2016. Identification of genes and pathways associated with pancreatic ductal adenocarcinoma by bioinformatics analyses. Oncol. Lett. 11, 1391–1397.

32.

Makohon-Moore

A.P.

, Zhang

, Reiter

JG.

, et al. 2017. Limited heterogeneity of known driver gene mutations among the metastases of individual patients with pancreatic cancer. Nat. Genet. 49, 358.

33.

McCann

2015. G protein-coupled receptor expression and function in cancer and the cancer associated microenvironment: A study of Chronic Lymphocytic Leukemia and Pancreatic Adenocarcinoma. PMID: 29872392.

34.

Nakamura

2008. Pancreatic Cancer Related Gene CST6 and GABRP (EP). The University of Tokyo. PMID: 30826798 and PMID: 17942900.

35.

Nones

, Waddell

, Song

, et al. 2014. Genome—Wide DNA methylation patterns in pancreatic ductal adenocarcinoma reveal epigenetic deregulation of SLIT—ROBO, ITGA2 and MET signaling. Int. J. Cancer. 135, 1110–1118.

36.

Oncology

T.L.

2014. Pancreatic cancer in the spotlight. Lancet Oncol. 15, 241.

37.

Penheiter

A.R.

, Erdogan

, Murphy

S.J.

, et al. 2015. Transcriptomic and immunohistochemical profiling of SLC6A14 in pancreatic ductal adenocarcinoma. Biomed. Res. Int. 2015, 593572.

38.

Pierce

B.A.

2012. A Cancer Journal for Clinicians, CA. Biomedical Market Newsletter. March, 21, 2012.

39.

Sauter

, Bonito

, Rapedius

, et al. 2015. Ca2+ and pH – Activated Channels in Pancreatic Ductal Adenocarcinoma (PDAC). PMID: 3180239.

40.

Sauter

, Sauter

D.R.P.

, and Drp

2015a. CIand KChannels in Pancreatic Ductal Adenocarcinoma (PDAC).

41.

Sauter

D.R.

, Novak

, Pedersen

S.F.

, et al. 2015b. ANO1 (TMEM16A) in pancreatic ductal adenocarcinoma (PDAC). Pflugers Arch. 467, 1495–1508.

42.

Sauter

, Rosenberger

, Beckmann

, et al. 2008. Matrix metalloproteinase 1 (MMP1) is associated with early-onset lung cancer. Cancer Epidemiol. Biomarkers Prev. 17, 1127.

43.

Sunami

, Tsuno

, Osada

, et al. 2000. MMP-1 is a prognostic marker for hematogenous metastasis of colorectal cancer. Oncologist. 5, 108–114.

44.

Takehara

, Hosokawa

, Eguchi

, et al. 2007. Gamma-Aminobutyric Acid (GABA) stimulates pancreatic cancer growth through overexpressing GABAA receptor P subunit. Cancer Res. 67, 9704–9712.

45.

Taucher

, Mangge

, and Haybaeck

2016. Non-coding RNAs in pancreatic cancer: Challenges and opportunities for clinical application. Cell. Oncol. 39, 295–318.

46.

The Cancer Genome Atlas Research Network. 2017. Integrated genomic characterization of pancreatic ductal adenocarcinoma. Cancer Cell. 32, 185–203.

47.

, Hao

, Zhao

, et al. 2017. Statins inhibit proliferation of pancreatic ductal adenocarcinoma cancer (PDAC) cells acting synergistically with metformin and inhibitors of MEK and MTOR. Gastroenterology. 152, S837.

48.

Yao

, Yang

, Liu

, et al. 2016. Glypican-3 and KRT19 are markers associating with metastasis and poor prognosis of pancreatic ductal adenocarcinoma. Cancer Biomarkers. 17, 1–8.

49.

, Wang

L.G.

, Han

, et al. 2012. clusterProfiler: An R package for comparing biological themes among gene clusters. OMICS. 16, 284–287.

50.

Zhao

, Wu

, and Lv

2015. miR-128 modulates hepatocellular carcinoma by inhibition of ITGA2 and ITGA5 expression. Am. J. Transl. Res. 7, 1564–1573.

51.

Zhou

, Jubb

A.M.

, Lyle

, et al. 2015. PAK1 mediates pancreatic cancer cell migration and resistance to MET inhibition. J. Pathol. 234, 502–513.

52.

Zhou

, Cheng

, Jiang

. 2018. Ten hub genes associated with progression and prognosis of pancreatic carcinoma identified by co-expression analysis. Int. J. Biol. Sci. 14, 124–136.

53.

Zinzindohoué

, Lecomte

, Ferraz

J.M.

, et al. 2005. Prognostic significance of MMP-1 and MMP-3 functional promoter polymorphisms in colorectal cancer. Clin. Cancer Res. 11, 594–599.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.18 MB

0.03 MB