Screening and Survival Analysis of Hub Genes in Gastric Cancer Based on Bioinformatics

Abstract

Screening for hub genes associated with gastric cancer and elucidating possible molecular mechanisms of gastric cancer. Five gastric cancer-related gene expression profiles were extracted from the GEO database, and differentially expressed genes (DEGs) were obtained using GEO2R. Gene ontology (GO) enrichment analyses were performed by DAVID, and protein–protein interaction (PPI) network of the DEGs was constructed by STRING and Cytoscape software. Survival value for hub gene comes from the Kaplan–Meier plotter platform. In addition, potential miRNAs of hub genes were predicted by miRWalk. Four hundred seventy-six DEGs were identified in the five expression profiles, these genes are mainly involved in extracellular matrix (ECM)-receptor interaction, chemical carcinogenesis, gastric acid secretion, and PI3K-Akt signaling pathway. Combined with the results of the PPI network and CytoHubba, six hub genes were screened: SERPINH1, NPY, PTGDR, GPER, ADHFE1, and AKR1C1. These genes are highly expressed in gastric cancer tissues, and the overexpression level of these genes is associated with poor survival. A series of miRNAs such as hsa-miRNA-92a-1, hsa-miRNA-647, and hsa-miRNA-507 may play a key role in hub gene regulation. Our studies indicate that SERPINH1, NPY, PTGDR, GPER, ADHFE1, and AKR1C1 may be potential biomarkers and therapeutic targets for gastric cancer in the future.

1. Background

Gastric cancer is a common malignant tumor of the digestive tract, ranking fourth in the world for oncology (Jemal et al., 2010; Wong et al., 2011). The incidence of gastric cancer is concealed and difficult to detect in early stage. Previous studies have shown that Helicobacter pylori infection is one of the risk factors for gastric cancer (Yip and Teoh, 2018; Shi et al., 2019). Current epidemiology and related studies have found that lifestyle, infection factors, environmental factors, genetic factors, and abnormal gene expression are closely related to the occurrence of gastric cancer (Jia et al., 2010; Boland and Yurgelun, 2017). The pathogenesis of gastric cancer is complicated, and the specific causes and pathogenesis need further study. With the rapid development of molecular biology, especially the wide application of chip technology, more and more experimental platforms use molecular biology technology to screen genes closely related to tumorigenesis. It is easier to elucidate the pathogenesis of tumor by combining chip data with bioinformatics.

In this study, we reanalyzed the mRNA expression profile of gastric cancer in the National Center for Biotechnology Information (NCBI) database, and screened out the differentially expressed genes (DEGs) that were expressed in more than three expression profiles. The DEGs provided a theoretical basis for the further study of the diagnostic markers and pathogenesis of gastric cancer.

2. Methods

2.1. Microarray data source

Data sets of mRNA expression profiles associated with gastric cancer were downloaded from the GEO database of the NCBI. In this study, five different gastric cancer gene expression profiles (GSE79973, GSE103236, GSE54129, GSE26942, and GSE26899) were used for comparative analysis of gastric cancer and paracancerous samples.

2.2. Identification of differentially expressed genes

The DEGs in gastric cancer and paracancerous tissues were screened and analyzed by using the GEO2R tool of GEO database. GEO2R is a data set analysis tool based on t-test or ANOVA. The two groups of data under the same experimental conditions were compared, and the DEGs were screened between cancer and paracancerous tissues. In this study, the DEGs were analyzed using p < 0.01 and LogFc < −1 or LogFc >1 as screening conditions.

2.3. Gene ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of differentially expressed genes

Gene ontology (GO) analysis and KEGG pathway analysis were performed on DEGs by using the DAVID (https://david.ncifcrf.gov/home.jsp) online analysis tool. The research includes the biological processes (BP), molecular functions (MF), cellular components (CC), and KEGG signaling pathways. In addition, the cutoff value for screening was set at p < 0.01.

2.4. Protein–protein interaction network construction and hub gene identification

DEGs were uploaded to an online software STRING (http://string-db.org/) to produce a protein–protein interaction (PPI) network graph, and the interaction between these DEGs was analyzed. The protein interaction network of DEGs was constructed by Cytoscape software. Protein interaction patterns and hub genes were screened from protein interaction networks. Genes with 10 or more gene degrees in the PPI network were deemed as hub genes.

2.5. Survival analysis of hub genes

Hub genes were uploaded to the online software Kaplan–Meier Plotter (http://kmplot.com) to analyze the correlation between the hub genes and the survival time of gastric cancer patients. The hazard ratio (HR) is given with 95% confidence interval and the logarithmic rank p value is calculated.

2.6. MiRNA screening to regulate hub genes

Six hub genes associated with gastric cancer were imported into the miRWalk2.0 website for screening miRNAs that regulate target genes. MiRNA identified by miRWalk, miRanda, miRDB, RNA22, and TargetScan prediction tools is included as the potential regulatory miRNA of hub genes.

3. Results

3.1. Identification of differentially expressed genes

In this study, we analyzed 5 different gastric cancer gene expression profiles, involving 432 gastric cancer samples and 64 paracancerous cancers tissues (Table 1). Using GEO2R to analyze gene expression profiles, there were 611 DEGs in GSE26899, 436 DEGs in GSE26942, 5258 DEGs in GSE54129, 2570 DEGs in GSE79973, and 1769 DEGs in GSE103236 (Fig. 1). Sixty-six genes were differentially expressed in 5 gene expression profiles. In this study, genes with differential expression of more than three expression profiles were included as inclusion criteria, and 476 DEGs were screened (Table 2).

FIG. 1.

Venn diagram for the DEGs between gastric cancer and paracancer tissues in five different gene expression profiles. DEGs, differentially expressed genes.

Table 1.

Details for GEO Gastric Cancer Data

GEO	Platform	Sample	Tumor	Normal
GSE79973	GPL570	Stomach	10	10
GSE103236	GPL4133	Stomach	10	9
GSE54129	GPL570	Stomach	111	21
GSE26942	GPL6947	Stomach	205	12
GSE26899	GPL6947	Stomach	96	12

Table 2.

Screening for Differentially Expressed Genes in Gastric Cancer by Integrating Microarray Combinations

GEO	Gene terms	Total
GSE79973GSE103236 GSE54129 GSE26942 GSE26899	FNDC1 GKN1 IRX3 BGN ATP11B OSBPL7 GIF SULF1 SOSTDC1 KCNJ16 ATP4A CD36 MAL AQP4 CST1 ESRRG CPA2 ABCA8 HOXA10 CAPN13 FUT9 ATP4B ADH1A KRT6B SELENBP1 CLIC6 TRIM50 DGKD MFSD4A GC ALDH6A1 PDGFD THY1 SPP1 GCNT2 XYLT2 TIMP1 COL8A1 MMP11 CDH3 CEMIP RNASE1 CLDN1 SCNN1G FBXL13 COBLL1 FBP2 GSTA4 FOXC1 CHGA ADH7 SULT2A1 SFRP4 CCKBR HPGD TMEM158 FAP LIFR THBS2 GATA5 METTL7A FAM189A2 SERPINE1 KCNE2 COL10A1 PMEPA1	66
GSE79973 GSE103236 GSE54129 GSE26942	ITGA5 MAOA KIAA1324 CYSTM1 NFE2L2 KLF4 CYP3A5 NTN4 TNFRSF17 CA9 CYP4F12 SCNN1B ZBTB7C GGT6 ABCC5 CMTM4 RARRES1 FAM46C TMED6 RNASE4 GKN2 ST6 GALNAC1 CHPT1 GSTA1 AKR1B10 POU2AF1 MRAP2 FOLR1 SLC12A2 FA2H AMPD1 SMIM24 IRF4 LIPF MYRF NQO1 CD79A CYP2C18 ALDH1A1 FMO5 SLC26A9 PGC AKR1C3 C1orf116 PRDM16 SULT1C2 FCGBP GPRC5C AKR7A3 VSIG1 FAM3B SH3BGRL2 ADH1C PIK3C2G CXCL17 CAPN8 LYPD6B SLC22A23 CA4 ADAM28 KRT20 HRASLS2 FBP1 ARSD ADTRP RORC VSIG2 PSCA SPTSSB TFF1 LDHD PROM2 AADAC CAPN9 VILL TMEM171 RAB27B SCGB2A1 SLC7A8 TMPRSS2 SLC28A2 EPN3 FOXA1 ALDH3A1 PTGR1 ELOVL6 RBM47 CA2 LTF BCAS1 SIDT2 KLK11 DNER RASSF6 GCNT1 JCHAIN	96
GSE79973 GSE103236 GSE54129 GSE26899	SORBS2 CLDN3 SST TCEA3	4
GSE79973 GSE103236 GSE26942 GSE26899	ZNF385B C5 SLC9A2 MT1E ADHFE1 SLC25A4 SH3GL2 CKMT2 DRD5 CKB CLDN7 CAMK2N1 ETNPPL FGA GPX3 HDC GHRL APOBEC2 KCNJ13 KIT PLCXD3 TNFSF11 RPRM CKM HPN RGMB FGG IGFALS MT1M APLP1 SLC2A12 CHIA MYRIP SIGLEC11 PTGS1	35
GSE79973 GSE103236 GSE54129 GSE26942 GSE26899	EPB41L4B COL1A1 COL18A1 TNFAIP6 PDGFRB FCGR2A LRP8 COL11A1 COL1A2 APOC1 COL12A1 IGFBP7 SPHK1 PDPN SERPINE2 S100A10 COL6A3 CTHRC1 FJX1 HEYL IGF2BP3 PRRX1 COL5A2 PDK4 NNMT TEAD4 NR3C2 FSCN1 SRPX2 BMP1 PLA2G7 MFAP2 IL32 PLAU TGM2ADAMTS9 CXCL8 ANGPT2 ARHGEF37 SERPINH1 CTSL OLFML2B LY6E PPP1R3C	44
GSE79973 GSE103236 GSE54129	MZB1 UNC5CL IGFBP2 ANXA10 CHRM3 FAM3D HYAL1 CXCL14 ARL14 TFF2 ARHGAP18 TCN1 FAR1 FABP4 SNX24 MAMDC2 TRNP1 COLCA1 AKR1C4 GPT PDIA2 MUC6 CTSE SOWAHA REG3A SOX2 DHCR24 TSC22D3 PIGR ODAM SORD GPT2 CLDN18	33
GSE79973 GSE103236 GSE26942	MT1G MT1F KLRB1 PP7080 MST1 ARRDC4 CYB5R1 AZGP1 IRX2 MT1H NKX3–2 ALCAM HAS3 CYP2S1 CYFIP2 DERL3 RAB26 PDILT ADA DUOX1 ITM2A PKDCC	22
GSE79973 GSE103236 GSE26899	COL4A5 NPY ECHDC3 GPER1 SMIM11A TRIM15 CLCNKA KRT17 S100A3 TPD52L1 CGNL1 LAMC2 TUBB3 PTGDR2 PGA3	15
GSE79973 GSE54129 GSE26942	PAQR8 C1orf115 THBS1 HTRA1 HOXB7 PDLIM7 UGT2B15 FZD2 AXDND1 C10orf10 ENTPD5 NECTIN3 EFEMP2 COL5A1 GPAT3 CTSK FAM107B COL4A2 ISLR SYTL2 COL4A1 FCER1G LGALS1 SCIN SMPDL3A RAB31 GUCA2B SSTR1 ALDOB HOXC6 CHN1	31
GSE79973 GSE54129 GSE26899	OLR1 MMP3 ZBTB16 LILRB3 LIPG GSTM2 HAVCR2 C2orf40 RBPMS2 JAG2 TYROBP MDK	12
GSE79973 GSE26942 GSE26899	ITGA2 KRT80 ESM1 CDCA7 TNFRSF12A DPT ARPC1B ACACB PNPLA7 CKS1B ETV4 GAMT GJB2 UPP1	14
GSE103236 GSE54129 GSE26942	MECOM SLC44A4 FOXQ1 TMEM30B DSC2 ANG PLCXD1 UBL3 PELI2 CXADR DHRS7 PXMP2 ANKRD22 ATP2A3 IQGAP2 CLDN23 ITPKA 2-Mar PLLP	19
GSE54129 GSE26942 GSE26899	MMP7 XK GDPD5 APOE CMTM3 CDH11 GPD1L LRRC32 FAM19A5 TREM2 FXYD5 MYOC LINC01094 DIO2 WISP1 VCAN ERO1B AJUBA SLC26A7 RCN3 PTPRZ1 OSMR TNFRSF11B SNX10 FCGR1B LEF1 FKBP10 LZTS1 TPCN2 LOC643201 BUB1 SPINK2 CNTN3 COL3A1 ANOS1 HOXC9 LINC01105 ADGRL3 C16orf89 TNFRSF10B AMOTL1 KLK10 CLEC4A HSD11B1 SPARC CTSZ FGD6 C6 ADAM12 P4HA3 ADAMTS2 CTSB MSR1 CWH43 ECT2 C2 NID2 GPX8 ENO1 MYZAP LOX AKR1C1 NSG1 CLEC11A PSAPL1 EPHB2 MYO1B FPR3 ASPA NOX4 LINC00982 ADRB2 TNFSF4 INHBA LOXL2 PLXDC2 RDH12 SALL4 PAIP2B DTL FERMT1 PKM ELL2 KBTBD12 ADIPOQ	85

3.2. Gene ontology and KEGG pathway analysis of differentially expressed genes

Four hundred seventy-six DEGs in gastric cancer and paracancer tissues were uploaded to DAVID's website. Signal pathway enrichment analysis revealed that these DEGs are mainly involved in extracellular matrix (ECM)-receptor interaction, protein digestion and absorption, chemical carcinogenesis, gastric acid secretion, PI3K-Akt signaling pathway, and glycolysis/gluconeogenesis. GO analysis showed that DEGs were mainly involved in cell adhesion, endoderm cell differentiation, angiogenesis regulation, cell and cell adhesion, regulation of cell proliferation, and regulation of gene expression, and closely related to the cell receptor and ECM (Fig. 2).

FIG. 2.

Gene ontology and pathway enrichment analysis of DEGs.

3.3. Protein–protein interaction network construction and analysis of modules

Using MCODE to analyze the entire PPI network, three higher scoring modules were selected. The score of model 1 is 15.867, which contains 16 hub genes, including SERPINH1, COL5A2, COL18A1, COL4A2, COL4A1, COL8A1, COL6A3, P4HA3, COL11A1, COL10A1, COL3A1, COL4A5, COL1A1, COL5A1, COL1A2, and COL12A1 (Fig. 3A). The score of mode 2 was 6.0, which contained 14 hub genes, including ITGA2, NPY, PTGDR, GPER, MMP7, FGA, FGG, IL8, ISLR, PLAU, C5, FPR3, SSTR1, and SST (Fig. 3B). Mode 3 has a score of 5.143 and contains 29 hub genes as follows: CLDN7, CHGA, RDH12, MT1E, MT1H, ATP4A, MT1M, ADH1A, ALDH1A1, CCKBR, NQO1, ADHFE1, ACACB, AKR1C4, CLDN18, MT1F, DHCR24, CLDN1, ADH7, MT1G, CLDN3, HDC, CLDN23, AGXT2L1, AKR1C3, AKR1C1, FBP2, ALDOB, and FBP1 (Fig. 3C). KEGG analysis of 59 hub genes revealed that these genes are closely related to protein digestion and absorption, ECM-receptor interaction, focal adhesion, and PI3K-Akt signaling pathway (Table 3).

FIG. 3.

Top three modules from the protein–protein interaction network. (A) Module 1; (B) module 2; and (C) module 3.

Table 3.

KEGG Pathway Enrichment Analysis of the Genes in Module

Term	Count	p
Hsa04974:Protein digestion and absorption	13	5.6E-13
Hsa04512:ECM-receptor interaction	11	3.4E-10
Hsa05146:Amoebiasis	9	6.5E-7
Hsa04510:Focal adhesion	11	1.4E-6
Hsa04611:Platelet activation	9	3.1E-6
Hsa04151:PI3K-Akt signaling pathway	11	1.3E-4
Hsa04978:Mineral absorption	5	2.5E-4
Hsa00010:Glycolysis/gluconeogenesis	5	1.3E-3
Hsa04530:Tight junction	5	3.3E-3
Hsa04670:Leukocyte transendothelial migration	5	8.8E-3

ECM, extracellular matrix.

3.4. Survival analysis of hub gene

To further evaluate the prognostic value of hub genes, we analyzed the survival curves of each gene. On the Kaplan–Meier plotter analysis platform, 876 gastric cancer cases were analyzed for survival analysis of hub genes. Among the 59 genes, 22 genes were not significantly correlated with the prognosis of gastric cancer, 31 genes had been reported to be significantly correlated with the occurrence of gastric cancer, and 6 genes were associated with the prognosis of gastric cancer and had not been reported (Fig. 4).

FIG. 4.

Kaplan–Meier overall survival analysis for six hub genes expressed in GC patient samples. (A) SERPINH1; (B) NPY; (C) PTGDR; (D) GPER; (E) ADHFE1; (F) AKR1C1.

3.5. miRNA screening of regulatory hub genes

The potential miRNAs of the six hub genes were predicted using the miRWalk2.0 website. As showed in Table 4, top three predicted miRNAs of hub genes were displayed according to the prediction program numbers.

Table 4.

Hub Genes and Their Predicted miRNAs

Gene	Predicted miRNA	Total
SERPINH1	hsa-miRNA-92a-1, hsa-miRNA-30c-1, hsa-miRNA-1265	91
NPY	hsa-miRNA-647, hsa-miRNA-363, hsa-miRNA-586	24
PTGDR	hsa-miRNA-507, hsa-miRNA-1301, hsa-miRNA-20b	116
GPER	hsa-miRNA-644, hsa-miRNA-1226, hsa-miRNA-10a	37
ADHFE1	hsa-miRNA-630, hsa-miRNA-372, hsa-miRNA-106a	128
AKR1C1	hsa-miRNA-944, hsa-miRNA-1226, hsa-miRNA-597	23

4. Discussion

Gastric cancer is one of the most deadly tumors in the world, with 5-year survival rate of only 20%–40% (Shah and Kelsen, 2010). The occurrence and development of gastric cancer are a very complicated process, so it is particularly important to study the molecular mechanism of gastric cancer from the molecular level with genomics, proteomics, metabolomics, and other methods. Based on the rapid development of gene chip and sequencing platform in disease research, it provides a good technical method for us to study the pathogenesis of gastric cancer.

In this study, we screened out 59 hub genes from 5 different gene expression profiles. These genes mainly involved in protein digestion and absorption, ECM-receptor interaction, focal adhesion, PI3K Akt signaling pathway, and other signaling pathways. Among these DEGs, SERPINH1, NPY, PTGDR, GPER, ADHFE1, and AKR1C1 genes were significantly correlated with gastric cancer survival, which has not been reported in the literature.

The pathogenesis of cancer is multifactorial and involves multiple signaling cascades. SERPINH1, also known as Hsp47, is an important chaperone protein required for correctly folding and secreting collagen, and its expression level is closely related to the occurrence of tumors (Durate and Bonatto, 2018). Previous research found high expression in renal clear cell carcinoma, and lung cancer (Yamamoto et al., 2013; Kamikawaji et al., 2016; Qi et al., 2018). Yamamoto et al. (2013) believed that the high expression of SERPINH1 was closely related to the migration and invasion of cervical squamous cells. Zhao et al. (2014) used small interfering RNA to knock out the expression of HSP47 in glioma cells, and the growth, migration, and invasion of glioma cells were inhibited in vitro. Using predictive software, the differential expression of SERPINH1 may be regulated by hsa-miRNA-92a-1, hsa-miRNA-30c-1, and hsa-miRNA-1265.

NPY gene encodes neuropeptides, which widely express the central nervous system and affect the physiological processes such as stress response, circadian rhythm, and cardiovascular function. Jeppsson et al. (2017) believed that NPY promotes inflammatory tumorigenesis by enhancing epithelial cell proliferation. Abnormal expression of NPY and related receptors in prostate cancer is closely related to tumor development (Rasiah et al., 2006; Ruscica et al., 2006). At present, the correlation between NPY and gastric cancer has not been reported. Hsa-miRNA-647, hsa-miRNA-363, and hsa-miRNA-586 may bind to the 3′ UTR of genes to regulate NYP expression.

PTGDR encodes a superfamily of guanine nucleotide binding protein-coupled receptors (GPCR) that respond to extracellular signals and activate intracellular signaling pathways. Ooki et al. (2017) found that methylation of PTGDR promoter can be used as a biomarker for early detection of cancer in nonsmall-cell lung cancer. Methylation of promoter of PTGDR gene is also associated with colorectal carcinogenesis (Kalmár et al., 2015). PTGDR may be the target of hsa-miRNA-507, hsa-miRNA-1301, and hsa-miRNA-20b through prediction.

GPER gene encodes multiple membrane proteins, which are located in the endoplasmic reticulum. Current studies have found that GPER inhibits angiogenesis of breast cancer and is associated with the prognosis of breast cancer (Wang et al., 2018). As a mechanical regulator, GPER protein is involved in the regulation of pancreatic stellate cells and tumor microenvironment, as well as the regulation of adrenal cortical cancer and colorectal cancer (Casaburi et al., 2017; Liu et al., 2017; Cortes et al., 2019). The MiRWalk2.0 website predicted that the expression of GPER gene may be regulated by hsa-miRNA-644, hsa-miRNA-1226, and hsa-miRNA-10a.

ADHFE1 encodes hydroxyacid/oxoacid hydrogenase, which promotes the oxidation of 4-hydroxybutyric acid in mammalian tissues. ADHFE1 is a breast cancer oncogene that reduces patient survival (Mishra et al., 2018). The expression of ADHFE1 protein was negatively correlated with methylation of DNA promoter and negatively correlated with the degree of differentiation of colorectal cancer (Tae et al., 2013). The expression of ADHFE1 is regulated by hsa-miRNA-630, hsa-miRNA-372, and hsa-miRNA-106a.

AKR1C1 gene encodes the aldo/keto reductase superfamily member, which converts aldehydes and ketones to alcohols using NADH or NADPH as cofactors. Matsumoto et al. found that the AKR1C1 protein plays an important role in regulating bladder cancer metastasis and drug resistance (Matsumoto et al., 2016). High expression of AKR1C1 promoted the migration and proliferation of lung cancer cells (Tian et al., 2016; Zhu et al., 2018). In primary human breast cancer, AKR1C1 expression is an independent prognostic marker (Wenners et al., 2016). Numerous miRNAs such as hsa-miRNA-944, hsa-miRNA-1226, and hsa-miRNA-597 play a regulatory role in the expression of AKR1C1.

In conclusion, according to the biological information, we identified some hub genes in gastric cancer. These genes play an important role, but have not been reported before. This study provides a new biomarker for the study of gastric cancer.

Footnotes

Acknowledgment

This work was supported by the Science Study of Guangxi Health Commission (Z20180228).

Author Disclosure Statement

The authors declare they have no competing financial interests.

References

Boland

C.R.

, and Yurgelun

M.B.

2017. Historical perspective on familial gastric cancer. Cell. Mol. Gastroenterol. Hepatol. 3, 192–200.

Casaburi

, Avena

, De Luca

, et al. 2017. GPER-independent inhibition of adrenocortical cancer growth by G-1 involves ROS/Egr-1/BAX pathway. Oncotarget, 8, 115609–115619.

Cortes

, Sarper

, Robinson

, et al. 2019. GPER is a mechanoregulator of pancreatic stellate cells and the tumor microenvironment. EMBO Rep. 20. pii: e46556.

Durate

B.D.P.

, and Bonatto

2018. The heat shock protein 47 as a potential biomarker and a therapeutic agent in cancer research. J. Cancer Res. Clin. Oncol. 144, 2319–2328.

Jemal

, Center

M.M.

, DeSantis

, et al. 2010. Global patterns of cancer incidence and mortality rates and trends. Cancer Epidemiol. Biomarkers Prev. 19, 1893–1907.

Jeppsson

, Srinivasan

, and Chandrasekharan

2017. Neuropeptide Y (NPY) promotes inflammation- induced tumorigenesis by enhancing epithelial cell proliferation. Am. J. Physiol. Gastrointest. Liver Physiol. 312, G103–G111.

Jia

, Persson

, Hou

, et al. 2010. A comprehensive analysis of common genetic variation in MUC1, MUC5AC, MUC6 genes and risk of stomach cancer. Cancer Causes Control, 21, 313–321.

Kalmár

, Péterfia

, Hollósi

, et al. 2015. DNA hypermethylation and decreased mRNA expression of MAL, PRIMA1, PTGDR and SFRP1 in colorectal adenoma and cancer. BMC Cancer, 15, 736.

Kamikawaji

, Seki

, Watanabe

, et al. 2016. Regulation of LOXL2 and SERPINH1 by antitumor microRNA-29a in lung cancer with idiopathic pulmonary fibrosis. J. Hum. Genet. 61, 985–993.

10.

Liu

, Chen

, Jiang

, et al. 2017. Epigenetic down regulation of G protein-coupled estrogen receptor (GPER) functions as a tumor suppressor in colorectal cancer. Mol. Cancer, 16, 87.

11.

Matsumoto

, Tsuda

, Yoshida

, et al. 2016. Aldo-keto reductase 1C1 induced by interleukin-1β mediates the invasive potential and drug resistance of metastatic bladder cancer cells. Sci. Rep. 6, 34625.

12.

Mishra

, Tang

, Putluri

, et al. 2018. ADHFE1 is a breast cancer oncogene and induces metabolic reprogramming. J. Clin. Invest. 128, 323–340.

13.

Ooki

, Maleki

, Tsay

J.J.

, et al. 2017. A panel of novel detection and prognostic methylated DNA markers in primary non-small cell lung cancer and serum DNA. Clin. Cancer Res. 23, 7141–7152.

14.

, Zhang

, Peng

, et al. 2018. SERPINH1 overexpression in clear cell renal cell carcinoma: Association with poor clinical outcome and its potential as a novel prognostic marker. J. Cell. Mol. Med. 22, 1224–1235.

15.

Rasiah

K.K.

, Kench

J.G.

, Gardiner-Garden

, et al. 2006. Aberrant neuropeptide Y and macrophage inhibitory cytokine-1 expression are early events in prostate cancer development and are associated with poor prognosis. Cancer Epidemiol. Biomarkers Prev. 15, 711–716.

16.

Ruscica

, Dozio

, Boghossian

, et al. 2006. Activation of the Y1 receptor by neuropeptide Y regulates the growth of prostate cancer cells. Endocrinology, 147, 1466–1473.

17.

Shah

M.A.

, and Kelsen

D.P.

2010. Gastric cancer: A primer on the epidemiology and biology of the disease and an overview of the medical management of advanced disease. J. Natl. Compr. Canc. Netw. 8, 437–447.

18.

Shi

, Wang

, Guo

, et al. 2019. Helicobacter pylori-induced DNA damage is a potential driver for human gastric cancer AGS cells. DNA Cell. Biol. 38, 272–280.

19.

Tae

C.H.

, Ryu

K.J.

, Kim

S.H.

, et al. 2013. Alcohol dehydrogenase, iron containing, 1 promoter hypermethylation associated with colorectal cancer differentiation. BMC Cancer, 13, 142.

20.

Tian

, Li

, Jiang

, et al. 2016. High expression of AKR1C1 is associated with proliferation and migration of small-cell lung cancer cells. Lung Cancer (Auckl). 7, 53–61.

21.

Wang

, Li

, Ye

, et al. 2018. Oestrogen inhibits VEGF expression and angiogenesis in triple-negative breast cancer by activating GPER-1. J. Cancer. 9, 3802–3811.

22.

Wenners

, Hartmann

, Jochens

, et al. 2016. Stromal markers AKR1C1 and AKR1C2 are prognostic factors in primary human breast cancer. Int. J. Clin. Oncol. 21, 548–556.

23.

Wong

H.L.

, Rabkin

C.S.

, Shu

X.O.

, et al. 2011. Systemic cytokine levels and subsequent risk of gastric cancer in Chinese Women. Cancer Sci. 102, 1911–1915.

24.

Yamamoto

, Kinoshita

, Nohata

, et al. 2013. Tumor-suppressive microRNA-29a inhibits cancer cell migration and invasion via targeting HSP47 in cervical squamous cell carcinoma. Int. J. Oncol. 43, 1855–1863.

25.

Yip

H.C.

, and Teoh

A.Y.B.

, 2018. Importance of timely eradication of Helicobacter pylori to prevent peptic ulcer recurrence and gastric cancer. Gastrointest. Endosc. 88, 251–252.

26.

Zhao

, Jiang

, Yao

, et al. 2014. Heat shock protein 47 regulated by miR-29a to enhance glioma tumor growth and invasion. J. Neurooncol. 118, 39–47.

27.

Zhu

, Chang

L.L.

, Yan

F.J.

, et al. 2018. AKR1C1 activates STAT3 to promote the metastasis of non-small cell lung cancer. Theranostics, 8, 676–692.