identification and verification of biomarker in clear cell
TRANSCRIPT
Research ArticleIdentification and Verification of Biomarker in Clear Cell RenalCell Carcinoma via Bioinformatics and Neural Network Model
Bin Liu,1 Yu Xiao,2 Hao Li,3 Ai-li Zhang ,1 Ling-bing Meng ,4 Lu Feng,5 Zhi-hong Zhao,1
Xiao-chen Ni,1 Bo Fan,1 Xiao-yu Zhang,1 Shi-bin Zhao,6 and Yi-bo Liu1
1Department of Urinary Surgery, The Fourth Hospital of Hebei Medical University, No. 12 Jiankang Road, 050000, China2School of Basic Medicine, Peking University, No. 38 Xueyuan Road, Haidian District, Beijing 100191, China3Department of Oncology, Dongzhimen Hospital, Beijing University of Chinese Medicine, Beijing, China4School of Basic Medical Sciences, Hebei Medical University, Shijiazhuang, Hebei, China5MOH Key Laboratory of Geriatrics, Beijing Hospital, National Center of Gerontology, Beijing, China6Department of Reproductive Medicine, The Fourth Hospital of Hebei Medical University, No. 12 Jiankang Road, 050000, China
Correspondence should be addressed to Ai-li Zhang; [email protected] and Ling-bing Meng; [email protected]
Received 20 December 2019; Revised 10 May 2020; Accepted 22 May 2020; Published 17 June 2020
Academic Editor: Paul Harrison
Copyright © 2020 Bin Liu et al. This is an open access article distributed under the Creative Commons Attribution License, whichpermits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background. Clear cell renal cell carcinoma (ccRCC) is the most common subtype of kidney cancer, which represents the 9th mostfrequently diagnosed cancer. However, the molecular mechanism of occurrence and development of ccRCC is indistinct. Therefore,the research aims to identify the hub biomarkers of ccRCC using numerous bioinformatics tools and functional experiments.Methods. The public data was downloaded from the Gene Expression Omnibus (GEO) database, and the differently expressedgenes (DEGs) between ccRCC and normal renal tissues were identified with GEO2R. Protein-protein interaction (PPI) networkof the DEGs was constructed, and hub genes were screened with cytoHubba. Then, ten ccRCC tumor samples and ten normalkidney tissues were obtained to verify the expression of hub genes with the RT-qPCR. Finally, the neural network model wasconstructed to verify the relationship among the genes. Results. A total of 251 DEGs and ten hub genes were identified. AURKB,CCNA2, TPX2, and NCAPG were highly expressed in ccRCC compared with renal tissue. With the increasing expression ofAURKB, CCNA2, TPX2, and NCAPG, the pathological stage of ccRCC increased gradually (P < 0:05). Patients with highexpression of AURKB, CCNA2, TPX2, and NCAPG have a poor overall survival. After the verification of RT-qPCR, theexpression of hub genes was same as the public data. And there were strong correlations between the AURKB, CCNA2, TPX2,and NCAPG with the verification of the neural network model. Conclusion. After the identification and verification, AURKB,CCNA2, TPX2, and NCAPG might be related to the occurrence and malignant progression of ccRCC.
1. Introduction
Worldwide, renal cell carcinoma (RCC) represents the 9thmost frequently diagnosed cancer in men and the 10th inwomen, accounting for 5% and 3% of all oncological diagno-ses, respectively, [1]. According to the most updated dataprovided by the World Health Organization, there are morethan 140 000 RCC-related deaths yearly, with RCC rankingas the 13th most common cause of cancer death worldwide.Age and gender factors are closely related to the risk ofRCC. Other potential risk factors include lifestyle, complica-tions, drugs, and environmental factors [2]. The diagnosis
and management of RCC have changed remarkably rapidlyin the past decades through the unremitting efforts of gener-ation after generation of researchers. Despite progression incancer control and survival, locally advanced disease and dis-tant metastases are still diagnosed in a notable proportion ofpatients. Nevertheless, uncertainties, controversies, andresearch questions remain [3]. Further advances are expectedfrom the diagnosis, treatment, and prognosis evaluation.
RCC is a group of heterogeneous tumors with differentgenetic and molecular changes, clear cell renal cell carcinoma(ccRCC), papillary RCC (type 1 and type 2), and chromo-phobe RCC are the most common solid RCC, accounting
HindawiBioMed Research InternationalVolume 2020, Article ID 6954793, 24 pageshttps://doi.org/10.1155/2020/6954793
for 85.90% of all malignant RCC [4]. Among them, ccRCC isthe most common subtype of kidney cancer. Both sporadicand inherited RCC are usually associated with structuralchanges in the short arm of chromosome 3 [5]. In addition,the occurrence of RCC is related to multiple gene alterations,such as VHL, PBRM1, BAP1, SETD2, TCEB1, and KDM5C[3]. Although our understanding of the biology of RCC hasimproved, surgery is still the main treatment method ofRCC. Drugs and comprehensive therapies, identification ofnew target pathways, and optimal sequencing and combina-tion of existing targeted drugs are areas that are worthresearching [6].
Bioinformatics tools can screen differentially expressedgenes (DEGs) between diseased and normal tissues [3, 7, 8].These DEGs are related to the pathological stage, lesiongrade, and prognosis of patients. Zou et al. used microarraytechnology to identify the hub genes between malignantglioblastoma and normal brain tissue and obtained theimportant targets related to brain glioma [9]. Through aseries of bioinformatics analysis, Meng et al. concluded thatTPM2 may be an important biomarker for the occurrenceand development of atherosclerosis [10].
Therefore, this study will use bioinformatics technologyto explore the gene molecular markers of abnormal expres-sion during the occurrence of ccRCC and discuss the relatedpotential mechanisms. These differentially expressed genesmay affect the initiation and malignant progression of ccRCCand can be used as targets for diagnosis and treatment.
2. Material and Methods
2.1. Download Public Data. The Gene Expression Omnibus(GEO) database (http://www.ncbi.nlm.nih.gov/geo) is thelargest, most comprehensive, and publicly available sourceof gene expression data.
On 20 December, 2019, we set key words “(clear cell renalcell carcinoma) AND (normal kidney)” to detect the datasets,using a filter of “expression profiling by array.” The inclusioncriteria includes a diagnosis of clear cell renal cell carcinoma(data from papillary renal cell carcinoma diagnoses wereexcluded), the dataset including the gene expression profileof normal kidney (datasets which were composed of onlytumor data were excluded), a sample number of more thanforty per dataset (samples of less than forty were excluded),data from Homo sapiens (data from other species wereexcluded), and a series entry type, expression profiling byarray (data using methylation profiling only by array wereexcluded).
Therefore, GSE105288 (GPL10558, Illumina HumanHT-12V4.0 expression beadchip) and GSE66272 (GPL570(HG-U133_Plus_2) Affymetrix Human Genome U133 Plus 2.0Array) were obtained from the GEO database. A total of 44samples, including 35 ccRCC tissues and 9 normal renal tis-sues, were selected from GSE105288. A total of 53 samples,including 26 ccRCC tissues and 27 normal renal tissues, wereselected from GSE66272.
2.2. Differentially Expressed Genes (DEGs) between Normaland PCRC. GEO2R (http://www.ncbi.nlm.nih.gov/geo/
geo2r) could import data of the GEO database into the R lan-guage and perform differential analysis, essentially throughthe following two R packages, including limma packagesand GEOquery. Therefore, through the GEO2R tool, DEGswere identified between the normal and ccRCC groups. TheP values < 0.001 was defined as significant. The gene symbolswere necessary. SangerBox (https://shengxin.ren), one opentool, was used to draw volcano maps. Venn diagrams weredelineated using FunRich software (http://www.funrich.org/),which would visualize common DEGs shared betweenGSE105288 and GSE66272.
2.3. GO and KEGG Analysis. One online tool, DAVID(https://david.ncifcrf.gov/home.jsp) (version 6.8, Maryland,America), was applied to carry out the functional annotationfor DEGs. Gene Ontology (GO) [11] generally performsenrichment analysis of genomes. And there are mainly cellu-lar components (CC), biological processes (BP), and molecu-lar functions (MF) in the GO analysis. Kyoto Encyclopedia ofGenes and Genomes (KEGG) (https://www.kegg.jp/) [12] is acomprehensive database of genomic, chemical, and systemicfunctional information. Therefore, DAVID was used to makethe analysis of GO and KEGG. The Biological NetworksGene Oncology tool (BiNGO) (version 3.0.3) was used toanalyze and visualize the DEGs’ cellular component, biolog-ical process, and molecular function [13].
2.4. Protein-Protein Interaction (PPI) Network. The commonDEGs, shared between GSE105288 and GSE66272, wereconverted into differently expressed proteins. The STRING(Search Tool for the Retrieval of Interacting Genes) online
Table 1: Primers and their sequences for PCR analysis.
Primer Sequence (5′–3′)VEGFA-hF GGCAACTTACTTAGCCTCTT
VEGFA-hR AGGACAGTCTGAGTATGGGT
AURKB-hF GTTCGCATTCAACCTACCT
AURKB-hR GACGCCCAATCTCAAAGT
CCNA2-hF AACTGGGATAAGGAAGCT
CCNA2-hR CAGAAAGTATTGGGTAAGAA
MCM2-hF TTCTCCCTCACTTGTCCC
MCM2-hR CCTGTAATCCCAGCACTTT
MCM7-hF GGGGTAGGCAGAACTCAA
MCM7-hR CATGGAAGCGGTCTCAAA
SMC4-hF AGTGGCGTAGCACAGTAA
SMC4-hR ATTCCAAGATGATCCCTC
TPX2-hF GCAATCCTTCTGCCTTAG
TPX2-hR AGACCATCCTGGCTAACA
SLC2A1-hF GAGACGGGAAACCATCAA
SLC2A1-hR CTGCTCCTTCTTCAAACCAC
MCM5-hF GGGTGCGAGGAGAACAGT
MCM5-hR TGAGTCTGAGCCAGGGAG
NCAPG-hF AGAGTATTGTTGGCTTCC
NCAPG-hR AACTTCTGGACCATCACA
2 BioMed Research International
database (http://string-db.org) could construct the PPI net-work, which was visualized by Cytoscape (version 2.8) [14].
2.5. Significant Module and Hub Genes. Molecular ComplexDetection tool (MCODE) (version 1.5.1) [15], an openplug-in of Cytoscape, was performed to identify tested mostsignificant module from the PPI network, and the criteriawas that the maximum depth = 100, MCODE scores > 5,
cut − off = 2, k − score = 2, and node score cut − off = 0:2.Then, cytoHubba [16], a free plug-in of Cytoscape, wasapplied to authorize the hub genes, when the degree ≥ 10.
2.6. Expression Analysis of Hub Genes. The clustering analysisof expression level of hub genes was performed using heatmaps based on the GSE105288 and GSE66272. Also, theexpression profiles of hub genes in the ccRCC and normal
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 X Y
The gene positions are based on GRCh38.p2(NCBI). 2961 genes.
The differentially expressed genes on chromosomes
Underexpressed genesOverexpressed genes
(a)
12.7
10.16
7.62
5.08
2.54
0–0.81 –0.57 –0.34 –0.1 0.14 0.37 0.61
log2 (FoldChange)
Volcano plot
–log
10 (P
val
ue)
(b)
12.93
10.34
7.76
5.17
2.59
0–10.69 –6.82 –2.96 0.9 4.77 8.63 12.5
log2 (FoldChange)
Volcano plot
–log
10 (P
val
ue)
(c)
Figure 1: (a) The differentially expressed genes on chromosomes between ccRCC and normal kidney tissue. (b) One volcano plot presents theDEGs in the GSE105288. (c) Another volcano plot presents the DEGs in the GSE66272.
3BioMed Research International
groups were analyzed using Gene Expression Profiling Inter-active Analysis (GEPIA, http://gepia.cancer-pku.cn/) [17].
2.7. Effect of Hub Gene Expression for Pathological Stage andOverall Survival. The effect of hub gene expression for path-ological stage and overall survival was analyzed by theGEPIA. Finally, the correlation and linear regression analysesbetween AURKB, CCNA2, TPX2, and NCAPG were per-formed. And the receiver operator characteristic (ROC)
curve analysis was performed to test the sensitivity and spec-ificity of the hub gene expression for the diagnosis of ccRCC.The SPSS software (version 21.0; IBM; New York; America)was used to conduct all the statistical analysis. A P value <0.05 was defined as statistically significant.
2.8. RT-qPCR Assay. A total of 10 ccRCC participates wererecruited. After surgery, 10 ccRCC tumor samples fromccRCC patients and 10 adjacent normal kidney tissues
GSE105288 GSE66272
251 19731082
(a)
COLGALT1
HEPACAM2
PNMA2TMPRSS2PACRG
LOX
CALU
STC2RAB24 P4HA2
DPP6
PRKCSHHIGD1A
COL23A1
SLC2A1
TMCC1
VEGFA
HK2
CA9
YBX3DCTN4
TYMPPDGFD
EMCNP4HB
HRGKNG1
PLG
P4HA1
BNIP3
SLC2A12
PFKFB2
NDUFA4L2
PFKFB4
PGK1COX4I2
NPHS1TPI1
ALDOA
SLC1A4
VIM
STX4UNC5B
CD40ANGPT2
EGLN3
B3GNTL1
COL1A2
TAP1TAPBP
BST2
MAP4K4
ST14CANX
CCNA2
PSMB8
GRIK5
PREX1
SLC48A1CPOX
SCN2A
TBC1D24
PARM1
GNB5
SUCNR1
NRP2
MCOLN3
BHLHE41
KCNJ10
PARP9
ACPP
IFI16DTX1
DTX3
SCAMP5PAPPA
SHMT2
AMT
SLC7A8
CACNA2D2
COQ6
CHN2
ATP1A1
APEH
CASZ1
SLC41A2
EIF3B
CTSZ
GTF3A
YIPF5
HECW2
PLEKHO1
TRIM8
UBE2D2
HACL1
KLHL3
MCM5
MCM2
SPTSSA
NEK6
TOPBP1
CDCA7
GSAP
SLC51B
GABRA2
TMEM213
VWA8
NXPH2SLCO2B1
HOXA4
ATP6V0A4
SMDT1
ATP6V1D
DNASE2
ATP6V1C2
TACSTD2
LAIR1SETBP1
ERMP1
EHBP1L1
DNASE1
MLKLTRADDFUT11
SLC4A1MCFD2
NMI
PTTG1
RNF123
BID
AURKB USP2
SVOPL
CLIC5
SLC25A35SLC29A4
RARB
RNF145
LRBA
MCM7
PALM
RFC2
HMMR
ZDHHC3
PUS7
KIF13B
SMC4
TBC1D9B
TPX2
ARL2
ENSG00000259288 CENPM
RUFY1
DLGAP5
SORT1
CKAP2
KIF20B
DCAF11
NCK1 MTDHTPP2GLDC
FSCN1PHKA2
SHC1
TBL1XR1
TNIP1
L2HGDH
BCR
TRIB3
INTS8
MBNL1
FAM167ANETO2
ASUN
PPP1CC
XPO1
SPC25 NCAPG
SND1
TLK2NCKAP1L
CNEP1R1
PRLR
ELF5
ABCA1
SPAG4
CORO1C
SYNE2
(b)
Figure 2: (a) The Venn diagrammanifested that a total of 251 DEGs exist in the two datasets (GSE105288 and GSE66272) simultaneously. (b)The PPI network of the common DEGs.
4 BioMed Research International
Canonical glycolysisGlycolytic process
Peptidyl−proline hydroxylation to 4−hydroxy−L−prolineAngiogenesis
Cell proliferationFructose metabolic process
Cell divisionDNA replication initiation
Regulation of insulin secretionMitotic nuclear division
Regulation of actin cytoskeleton organizationCarbohydrate phosphorylation
Glycine catabolic processGlycine decarboxylation via glycine cleavage system
Cellular response to hypoxiaPlatelet degranulation
Sister chromatid cohesion
0.2 0.4 0.6RichFactor
0.01
0.02
0.03
0.04
P valueTop 18 of pathway enrichment
GeneNumber2.55.07.5
10.012.5
(a)
Basolateral plasma membraneEndoplasmic reticulum lumen
MembraneExtracellular exosome
MelanosomeCytosol
MCM complexPlatelet alpha granule lumen
CentrosomeIntracellular membrane−bounded organelle
Endoplasmic reticulumProcollagen−proline 4−dioxygenase complex
Lysosomal membraneCytoplasmMicrospike
TAP complexMicrovillus
0.00 0.25 0.50 0.75 1.00RichFactor
Top 18 of Pathway Enrichment
0.01
0.03
P value
0.02
GeneNumber2040
6080
(b)
Figure 3: Continued.
5BioMed Research International
Protein binding
Procollagen−proline 4−dioxygenase activity
Identical protein binding
Anion transmembrane transporter activity
Apolipoprotein binding
L−ascorbic acid binding
Actin binding
GTPase activator activity
Enzyme binding
TAP2 binding
Peptide antigen−transporting ATPase activity
DNA helicase activity
ATP binding
Double−stranded RNA binding
0.0 0.2 0.4 0.6RichFactor
Top 15 of pathway enrichment
GeneNumber
50100150
0.01
0.02
0.03
0.04
P value
(c)
hsa00051:Fructose and mannose metabolism
hsa01200:Carbon metabolism
hsa04966:Collecting duct acid secretion
hsa03030:DNA replication
hsa04066:HIF-1 signaling pathway
hsa00630:Glyoxylate and dicarboxylate metabolism
hsa05211:Renal cell carcinoma
hsa00010:Glycolysis / Gluconeogenesis
hsa01230:Biosynthesis of amino acids
0.050 0.075 0.100 0.125 0.150RichFactor
GeneNumber345
67
Top 10 of Pathway Enrichment
0.025
0.02
0.075
0.050
P value
(d)
Figure 3: Continued.
6 BioMed Research International
samples were obtained. The research conformed to the Dec-laration of Helsinki and was authorized by the Human Ethicsand Research Ethics Committees of the Fourth Hospital ofHebei Medical University. The informed consents wereobtained from all participates.
Total RNA was extracted from 10 ccRCC tumor samplesand 10 adjacent normal kidney tissue samples by the RNAisoPlus (TRIzol) kit (Thermo Fisher, Massachusetts, Americaand reverse transcribed to cDNA. RT-qPCR was performedusing a Light Cycler® 4800 System with specific primers forgenes. Table 1 presents the primer sequences used in theexperiments. The RQ values (2−ΔΔCt, where Ct is the thresh-old cycle) of each sample were calculated and are presentedas fold change in the gene expression relative to the controlgroup. GAPDH was used as an endogenous control.
2.9. The Confirmation Using The Cancer Genome Atlas(TCGA) Data. The gene expression dataset of ccRCC in theTCGA was downloaded using the University of CaliforniaSanta Cruz (UCSC) Xena (https://xena.ucsc.edu/welcome-to-ucsc-xena/). There were a total of 944 samples including537 ccRCC samples and 407 normal renal samples. The
IlluminaHiSeq was selected as gene expression RNAseq inthe research. In addition, the gene expression levels ofVEGFA, AURKB, CCNA2, MCM2, MCM7, SMC4, TPX2,SLC2A1, MCM5, and NCAPG between ccRCC and normalrenal samples were compared using the one-way ANOVA.
Furthermore, the effect of the gene expression of VEGFA,AURKB, CCNA2, MCM2, MCM7, SMC4, TPX2, SLC2A1,MCM5, and NCAPG on overall survival was analyzed byusing the TCGA data.
2.10. The Construction of Neural Network Model. The train-ing group was randomly divided into the calibration dataand training data according to the proportion of 3 : 7. Therewere 6 samples in the calibration data, and 20 samples inthe training data. We used MATLAB (version 8.3) toaccomplish the normalization processing of variable values,network simulation, network training, and network initiali-zation. The number of input neurons in the input layer isthe same as the number of input variables, and the numberis two. The hidden layer is designed as 1 layer, and the out-put layer is also designed for 1 layer. One output variable isthe intima-media thickness. When training to 2000 steps
Amine metabolic process
Peptidyl-proline hydroxylation
Peptidyl-proline hydroxylation to 4-hydroxy-L-proline
Peptidyl-amino acid modification
Nitrogen compound metabolic process
Cellular process
Peptidyl-proline modification
Cellular amino acid derivative metabolic process
Biological_process
Cellular metabolic process
Heterocycle metabolic process
4-hydroxyproline metabolic process
Monocarboxylic acid metabolic process
Cellular ketone metabolic process
Carboxylic acid metabolic process
Oxoacid metabolic process
Protein modification process
Protein metabolic process
Cellular protein metabolic process
Macromolecule modification
Primary metabolic process
Macromolecule metabolic process
Cellular amino acid and derivative metabolic process
Metabolic processOrganic acid metabolic process
Small molecule metabolic process
Cellular macromolecule metabolic process
(e)
Figure 3: (a) Detailed information relating to changes in the biological processes (BP) of DEGs in ccRCC and normal kidney tissue. (b)Detailed information relating to changes in the cellular components (CC) of DEGs in ccRCC and normal kidney tissue. (c) Detailedinformation relating to changes in the molecular functions (MF) of DEGs in ccRCC and normal kidney tissue. (d) KEGG pathwayanalysis for DEGs. (e) The BP analysis for DEGs via the BiNGO software.
7BioMed Research International
Membrane-bounded vesicle
Organelle lumen
Organelle
Membrane-enclosed lumen
Pigment granule
Cytoplasmicmembrane-bounded vesicle
Vesicle
Melanosome Nonmembrane-boundedorganelle
Cytoplasmic vesicle
Cell part
Cell
Cellular_component
Cell projection
Membrane-bounded organelle
Intracellular organelle
Nuclear part
Nucleoplasm
Intracellularnonmembrane-bounded
organelle
Nucleus
Nuclear lumen
Intracellularmembrane-bounded organelle
Intracellular organelle lumen
Macromolecular complex
Chromosome
intracellular organelle part
Intracellular part
Organelle part
Cytoplasmic part
Microvillus membrane
Nuclear chromosome part
MHC class I peptide loading complex
Cell projection membrane
Subsynaptic reticulum
Nucleoplasm part
Protein complex
MCM complex
Microvillus
Chromosomal part
Cytoplasm
Endoplasmic reticulum
Cell projection part
Intracellular
Endoplasmic reticulum part
(a)
Figure 4: Continued.
8 BioMed Research International
after repeated training, the falling gradient is 0, and thetraining speed is uniform [10]. At the same time, the train-ing error ≤ 0:05, and the R (relativity) value reached 0.9906.
3. Results
3.1. DEGs between Normal Kidney and ccRCC Samples. Thereare plenty of DEGs on all chromosomes between the ccRCCand normal samples (Figure 1(a)). One volcano plot presentsthe DEGs in the GSE105288 (Figure 1(b)), and another vol-cano plot presents the DEGs in the GSE66272 (Figure 1(c)).The Venn diagram manifested that a total of 251 DEGs existin the two datasets (GSE105288 and GSE66272) simulta-neously (Figure 2(a)).
3.2. Construction of the PPI Network. After construction ofthe PPI network for the common DEGs, there are 189 nodesand 406 edges in the PPI network (Figure 2(b)).
3.3. The Functional Enrichment Analysis of DEGs via GO andKEGG. GO analysis manifested that variations in DEGsrelated with biological processes (BP) were significantlyenriched in canonical glycolysis, glycolytic process, peptidyl-proline hydroxylation to 4-hydroxy-L-proline, angiogenesis,cell proliferation, fructose metabolic process, cell division,DNA replication initiation, regulation of insulin secretion,mitotic nuclear division, regulation of actin cytoskeleton orga-nization, carbohydrate phosphorylation, glycine catabolic pro-cess, glycine decarboxylation via glycine cleavage system,cellular response to hypoxia, and so on (Figure 3(a)). The var-iations in DEGs related with cellular components (CC) weresignificantly enriched in the basolateral plasma membrane,endoplasmic reticulum lumen, membrane, extracellular exo-some, melanosome, cytosol, MCM complex, and so on(Figure 3(b)). The variations in the DEGs related with molec-ular functions (MF) were significantly enriched in proteinbinding, procollagen-proline 4-dioxygenase activity, identical
Carbohydrate kinase activity
Kinase activity
6-Phosphofructo-2-kinaseactivity
Phosphofructokinase activity
Transferase activity, transferring phosphorus-containing groups
Phosphotransferase activity, alcohol group as acceptor
Transferase activity
Oxidoreductase activity
Catalytic activity
Procollagen-prolinedioxygenase activity
Procollagen-proline4-dioxygenase activity
Peptidyl-proline 4-dioxygenase activity
Kinase binding
Substrate-specifictransmembrane transporter
activity
Transmembrane transporter activity
Transporter activity
Substrate-specific transporter activity
Ion transmembrane transporter activity
TAP2 binding
Peptide binding
Peptide antigen binding
Antigen binding
TAP1 binding
Protein binding Molecular_function
Binding Peptidyl-proline dioxygenase activityTAP binding
Protein complex binding
Enzyme binding
Intermediate filament binding
(b)
Figure 4: (a) The CC analysis for DEGs via the BiNGO software. (b) The MF analysis for DEGs via the BiNGO software.
9BioMed Research International
protein binding, anion transmembrane transporter activity,apolipoprotein binding, L-ascorbic acid binding, actin bind-ing, and so on (Figure 3(c)). The KEGG pathway enrichmentanalysis showed that the top pathways related with DEGswere fructose and mannose metabolism, carbon metabolism,collecting duct acid secretion, DNA replication, HIF-1 sig-naling pathway, and so on (Figure 3(d)).
The BP analysis for DEGs is presented in Figure 3(e)via the BiNGO software (Figure 3(e)). The CC analysis forDEGs is presented in Figure 4(a) via the BiNGO software(Figure 4(a)). The MF analysis for DEGs is presented inFigure 4(b) via the BiNGO software (Figure 4(b)).
3.4. Significant Module Network and Identification of HubGenes. A significant module was screened from the PPI
network, and one module network consisted of 14 nodesand 84 edges (Figure 5(a)). Another module network con-sisted of 15 nodes and 32 edges (Figure 5(b)). And ten hubgenes were identified, including VEGFA, AURKB, CCNA2,MCM2, MCM7, SMC4, TPX2, SLC2A1, MCM5, andNCAPG (Figure 5(c)).
3.5. Difference of Expression of Hub Genes between ccRCCand Normal Kidney Samples. Hierarchical clustering allowedfor simple differentiation of ccRCC tissues from normal colo-rectal tissues via the expression levels of hub genes in theGSE105288 and GSE66272 datasets. One heat map showedthat the expressions of all the hub genes were higher in theccRCC samples than the normal samples in the GSE105288(Figure 6(a)). Another heat map also showed that the
AURKB
SPC25
SMC4
KIF20B
HMMR
TPX2CKAP2
MCM2
CENPM
MCM7
NCAPGPTTG1
CCNA2
DLGAP5
(a)
P4HA1COL23A1
EGLN3
P4HA2
VEGFA
COL1A2
PRKCSHSTC2
CALU
HRG
BNIP3
CA9PLG
PGK1
SLC2A1
(b)
MCM5
MCM7 NCAPG
MCM2
SLC2A1
CCNA2
TPX2
AURKBSMC4
VEGFA
(c)
VEGFA
AURKB
CCNA2
MCM2
MCM7
SMC4
TPX2
SLC2A1
MCM5
NCAPG
Rank Node
(d)
Figure 5: (a) A significant module was screened from the PPI network, and one module network consisted of 14 nodes and 84 edges. (b)Another module network consisted of 15 nodes and 32 edges. (c) Ten hub genes were identified, including VEGFA, AURKB, CCNA2,MCM2, MCM7, SMC4, TPX2, SLC2A1, MCM5, and NCAPG.
10 BioMed Research International
expressions of all the hub genes were higher in the ccRCCsamples than the normal samples in the GSE66272(Figure 6(b)). Through the GEPIA analysis, the expressionsof hub genes in the ccRCC patients were higher than the nor-mal individuals (Figure 7(a)).
3.6. Association between Hub Gene Expression, PathologicalStage, and Overall Survival. The results of GEPIA manifestedthat the expressions of AURKB, CCNA2, TPX2, and NCAPGwere significantly positively related with pathological stage
(P < 0:05), while the expressions of VEGFA, MCM2,MCM7, SMC4, SLC2A1, and MCM5 were not(Figure 7(b)). The results showed that the expression levelof VEGFA was not related with the overall survival ofccRCC patients (P > 0:05, Figure 8(a)). The overall sur-vival analysis showed that ccRCC patients with highexpression levels of AURKB (Figure 8(b)) and CCNA2(Figure 8(c)) had poorer overall survival times than thosewith low expression levels (P < 0:05). The expression levelof MCM2 was not related with the overall survival of
GSM
2825
240
GSM
2825
242
GSM
2825
243
GSM
2825
252
GSM
2825
255
GSM
2825
263
GSM
2825
265
GSM
2825
275
GSM
2825
277
GSM
2825
239
GSM
2825
241
GSM
2825
244
GSM
2825
245
GSM
2825
246
GSM
2825
247
GSM
2825
248
GSM
2825
249
GSM
2825
250
GSM
2825
251
GSM
2825
253
GSM
2825
254
GSM
2825
256
GSM
2825
257
GSM
2825
258
GSM
2825
259
GSM
2825
260
GSM
2825
261
GSM
2825
262
GSM
2825
264
GSM
2825
266
GSM
2825
267
GSM
2825
268
GSM
2825
269
GSM
2825
270
GSM
2825
271
GSM
2825
272
GSM
2825
273
GSM
2825
274
GSM
2825
276
GSM
2825
278
GSM
2825
279
GSM
2825
280
GSM
2825
281
GSM
2825
282
VEGFA
SLC2A1
SMC4
MCM5
MCM7
MCM2
CCNA2
TPX2
AURKB
NCAPG−3
−2
−1
0
1
2
3
Normal ccRCC
(a)
GSM
1618
388
GSM
1618
390
GSM
1618
392
GSM
1618
394
GSM
1618
396
GSM
1618
398
GSM
1618
400
GSM
1618
402
GSM
1618
404
GSM
1618
406
GSM
1618
408
GSM
1618
410
GSM
1618
412
GSM
1618
414 G
SM1618416 G
SM16
1841
8G
SM16
1842
0G
SM16
1842
2G
SM16
1842
4G
SM16
1842
6G
SM16
1842
8G
SM16
1843
0G
SM16
1843
2G
SM16
1843
4G
SM16
1843
6G
SM16
1843
8G
SM16
1844
0G
SM16
1838
7G
SM16
1838
9G
SM16
1839
1G
SM16
1839
3G
SM16
1839
5G
SM16
1839
7G
SM16
1839
9G
SM16
1840
1G
SM16
1840
3G
SM16
1840
5G
SM16
1840
7G
SM16
1840
9G
SM16
1841
1G
SM16
1841
3G
SM16
1841
5G
SM16
1841
9G
SM16
1842
1G
SM16
1842
3G
SM16
1842
5G
SM16
1842
7G
SM16
1842
9G
SM16
1843
1G
SM16
1843
3G
SM16
1843
5G
SM16
1843
7G
SM16
1843
9
SMC4
VEGFA
SLC2A1
MCM5
TPX2
NCAPG
AURKB
MCM7
CCNA2
MCM2
Normal ccRCC
−4
–2
0
2
4
(b)
Figure 6: (a) One heat map showed that the expressions of all the hub genes were higher in the ccRCC samples than the normal samples in theGSE105288. (b) Another heat map also showed that the expressions of all the hub genes were higher in the ccRCC samples than the normalsamples in the GSE66272.
11BioMed Research International
ccRCC patients (P > 0:05, Figure 8(d)). The expression levelof MCM7 was not related with the overall survival of ccRCCpatients (P > 0:05, Figure 8(e)). The expression level of SMC4was not related with the overall survival of ccRCC patients(P > 0:05, Figure 8(f)). The overall survival analysis showedthat ccRCC patients with high expression levels of TPX2had poorer overall survival times than those with low expres-
sion levels (P < 0:05, Figure 9(a)). The expression levels ofSLC2A1 (Figure 9(b)) and MCM5 (Figure 9(c)) were notrelated with the overall survival of ccRCC patients (P > 0:05).The overall survival analysis showed that ccRCC patientswith high expression levels of NCAPG had poorer overallsurvival times than those with low expression levels(P < 0:05, Figure 9(d)).
0
2
4
6
8
10
12
0
1
2
3
4
5
6
0
1
2
3
4
5
0
1
2
3
4
5
6
0
2
4
6
0
1
2
3
4
5
6
0
1
2
3
4
5
6
7
0
2
4
6
8
10
0
1
2
3
4
5
6
7
0
1
2
3
4
5
VEGFA AURKB CCNA2 MCM2 MCM7
SMC4 TPX2 SLC2A1 MCM5 NCAPG
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
KIRC(num(T)=523; num(N)=100)
⁎ ⁎ ⁎
⁎ ⁎ ⁎
(a)
4
6
8
10
12
Stage I Stage II Stage III Stage IV0
1
2
3
4
5
6
7
Stage I Stage II Stage III Stage IV
1
2
3
4
5
6
Stage I Stage II Stage III Stage IV
1
2
3
4
5
Stage I Stage II Stage III Stage IV2
3
4
5
6
7
8
Stage I Stage II Stage III Stage IV
1
2
3
4
5
6
7
Stage I Stage II Stage III Stage IV0
1
2
3
4
5
6
7
Stage I Stage II Stage III Stage IV2
4
6
8
10
Stage I Stage II Stage III Stage IV
2
3
4
5
6
7
Stage I Stage II Stage III Stage IV0
1
2
3
4
5
6
Stage I Stage II Stage III Stage IV
VEGFA AURKB CCNA2 MCM2 MCM7
SMC4 TPX2 SLC2A1 MCM5 NCAPG
F value = 2.67Pr(>F) = 0.0471
F value = 20.8Pr(>F) = 1.05e−12
F value = 13Pr(>F) = 3.5e−086
F value = 2.19Pr(>F) = 0.0879
F value = 2.18Pr(>F) = 0.0895
F value = 14.3Pr(>F) = 5.85e−09
F value = 1.91Pr(>F) = 0.127
F value = 0.548Pr(>F) = 0.649
F value = 15.3Pr(>F) = 1.52e−09
F value = 3.26Pr(>F) = 0.0213
6
(b)
Figure 7: (a) The comparison of expressions of all hub genes between ccRCC and normal kidney samples. (b) The relationship between theexpression of hub genes and pathological stage.
12 BioMed Research International
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0Overall survival
Months
Perc
ent s
urvi
val
Log-rank P = 0.2 HR(high) = 0.82
P(HR) = 0.2n(high) = 258n(low) = 258
Low VEGFA TPMHigh VEGFA TPM
(a)
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0Overall survival
Months
Perc
ent s
urvi
val
Log-rank P = 2.8e−06 HR(high) = 2.1
P(HR) = 4.6e−06n(high) = 257n(low) = 257
Low AURKB TPMHigh AURKB TPM
(b)
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0Overall survival
Months
Perc
ent s
urvi
val
Log-rank P = 0.0012 HR(high) = 1.7P(HR) = 0.0014
n(high) = 257n(low) = 257
Low CCNA2 TPMHigh CCNA2 TPM
(c)
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0Overall survival
Months
Perc
ent s
urvi
val
Log-rank P = 0.61 HR(high) = 0.92
P(HR) = 0.61 n(high) = 257n(low) = 258
Low MCM2 TPMHigh MCM2 TPM
(d)
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0Overall survival
Months
Perc
ent s
urvi
val
Log-rank P = 0.24 HR(high) = 0.83
P(HR)= 0.24 n(high) = 258n(low) = 258
Low MCM7 TPMHigh MCM7 TPM
(e)
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0Overall survival
Months
Perc
ent s
urvi
val
Log-rank P = 0.64 HR(high) = 1.1P(HR) = 0.64
n(high) = 257n(low) = 258
Low SMC4 TPMHigh SMC4 TPM
(f)
Figure 8: The overall survival Kaplan-Meier of six hub genes. (a) VEGFA, (b) AURKB, (c) CCNA2, (d) MCM2, (e) MCM7, and (f) SMC4.
13BioMed Research International
3.7. The Interaction Analysis among the Hub Genes. Throughthe Pearson correlation test, heat maps manifested that therewere strong correlations among hub genes in the GSE105288(Figure 10(a)) and GSE66272 (Figure 10(b)) datasets. Thecorrelation between AURKB, CCNA2, TPX2, and NCAPGwas strong (Figures 10(c)–10(h)).
3.8. ROC Analysis. To identify accurate thresholds for hubgenes to predict ccRCC, we constructed ROC. The expressionof all hub genes was associated with a diagnosis of ccRCC.The ROC curve of AURKB in the GSE105288 was shownin Figure 11(a). The ROC curve of CCNA2 in theGSE105288 was shown in Figure 11(b). The ROC curve ofTPX2 in the GSE105288 was shown in Figure 11(c). TheROC curve of NCAPG in the GSE105288 was shown inFigure 11(d). The ROC curve of AURKB in the GSE66272was shown in Figure 11(e). The ROC curve of CCNA2 inthe GSE66272 was shown in Figure 11(f). The ROC curve
of TPX2 in the GSE66272 was shown in Figure 11(g). TheROC curve of NCAPG in the GSE66272 was shown inFigure 11(h).
3.9. Results of RT-qPCR Analysis. As presented in Figure 12,the relative expression levels of VEGFA, AURKB, CCNA2,MCM2, MCM7, SMC4, TPX2, SLC2A1, MCM5, andNCAPG were significantly higher in the ccRCC samples,compared with the normal kidney tissues groups. The resultdemonstrated that VEGFA, AURKB, CCNA2, MCM2,MCM7, SMC4, TPX2, SLC2A1, MCM5, and NCAPG mightbe considered biomarkers for ccRCC.
3.10. The Verification by TCGA. According to the aboveexpression analysis, VEGFA, AURKB, CCNA2, MCM2,MCM7, SMC4, TPX2, SLC2A1, MCM5, and NCAPG weremarkedly upregulated in ccRCC tumor samples comparedwith the normal renal samples. After confirmation using
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0Overall survival
Months
Perc
ent s
urvi
val
Log-rank P = 0.0041 HR(high)= 1.6P(HR) = 0.0044n(high) = 256n(low) = 257
Low TPX2 TPMHigh TPX2 TPM
(a)
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0Overall survival
Months
Perc
ent s
urvi
val
Logrank P = 0.49 HR(high) = 0.9P(HR) = 0.49n(high) = 258n(low) = 258
Low SLC2A1 TPMHigh SLC2A1 TPM
(b)
Overall survival
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0
Months
Perc
ent s
urvi
val
Log-rank P = 0.17 HR(high) = 0.81
P(HR) = 0.17n(high) = 258n(low) = 258
Low MCM5 TPMHigh MCM5 TPM
(c)
Overall survival
0 50 100 150
0.0
0.2
0.4
0.6
0.8
1.0
Months
Perc
ent s
urvi
val
Logrank P = 0.014 HR(high) = 1.5 P(HR) = 0.015n(high) = 258n(low) = 257
Low NCAPG TPMHigh NCAPG TPM
(d)
Figure 9: The overall survival Kaplan-Meier of another four hub genes. (a) TPX2, (b) SLC2A1, (c) MCM5, and (d) NCAPG.
14 BioMed Research International
the TCGA data, these gene expression levels in the ccRCCsamples were also significantly higher than the normal renalsamples (P < 0:05, Figure 13).
The results showed that the expression level of VEGFA(P = 0:4946), MCM2 (P = 0:5249), MCM7 (P = 0:092),SMC4 (P = 0:856), SLC2A1 (P = 0:209), and MCM5
VEG
FA
AU
RKB
CCN
A2
MCM
2
MCM
7
SMC4
TPX2
SLC2
A1
MCM
5
NCA
PG
VEGFA
AURKB
CCNA2
MCM2
MCM7
SMC4
TPX2
SLC2A1
MCM5
NCAPG0.2
0.4
0.6
0.8
1
(a)
VEG
FA
AU
RKB
CCN
A2
MCM
2
MCM
7
SMC4
TPX2
SLC2
A1
MCM
5
NCA
PG
VEGFA
AURKB
CCNA2
MCM2
MCM7
SMC4
TPX2
SLC2A1
MCM5
NCAPG 0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
(b)
0 1 2 3 4 5 60
1
2
3
4
5
log2 (AURKB TPM)
log2
(CCN
A2
TPM
)
P value = 0R = 0.82
(c)
0 1 2 3 4 5 60
1
2
3
4
5
6
log2 (AURKB TPM)
log2
(TPX
2 TP
M)
P value = 0R = 0.93
(d)
0 1 2 3 4 5 60
1
2
3
4
5
log2 (AURKB TPM)
P value = 0R = 0.83
log2
(NCA
PG T
PM)
(e)
0 1 2 3 4 50
1
2
3
4
5
6
log2 (CCNA2 TPM)
log2
(TPX
2 TP
M)
P value = 0R = 0.9
(f)
0 1 2 3 4 50
1
2
3
4
5
log2 (CCNA2 TPM)
log2
(NCA
PG T
PM)
P value = 0R = 0.95
(g)
0
1
2
3
4
5
log2
(NCA
PG T
PM)
0 1 2 3 4 5 6log2 (TPX2 TPM)
P value = 0R = 0.91
(h)
Figure 10: (a) Heat maps showing the correlations between hub genes in the GSE105288 datasets. (b) Heat maps showing the correlationsbetween hub genes in the GSE66272 datasets. (c) The correlation between AURKB and CCNA2. (d) The correlation between AURKB andTPX2. (e) The correlation between AURKB and NCAPG. (f) The correlation between CCNA2 and TPX2. (f) The correlation betweenAURKB and NCAPG. (g) The correlation between CCNA2 and NCAPG. (h) The correlation between TPX2 and NCAPG.
15BioMed Research International
Specificity
Sens
itivi
ty
1.0 0.8 0.6 0.4 0.2 0.0
0.0
0.2
0.4
0.6
0.8
1.06.829 (0.857, 1.000)
AUC: 0.937
AURKB
(a)
Specificity
Sens
itivi
ty
1.0 0.8 0.6 0.4 0.2 0.0
0.0
0.2
0.4
0.6
0.8
1.0 7.237 (0.829, 1.000)
AUC: 0.933
CCNA2
(b)
Specificity
Sens
itivi
ty
1.0 0.8 0.6 0.4 0.2 0.0
0.0
0.2
0.4
0.6
0.8
1.0
7.209 (0.743, 0.889)
AUC: 0.895
TPX2
(c)
Specificity
Sens
itivi
ty
1.0 0.8 0.6 0.4 0.2 0.0
0.0
0.2
0.4
0.6
0.8
1.07.256 (0.829, 1.000)
AUC: 0.940
NCAPG
(d)
Specificity
Sens
itivi
ty
1.0 0.8 0.6 0.4 0.2 0.0
0.0
0.2
0.4
0.6
0.8
1.0
0.083 (0.846, 0.926)
AUC: 0.930
AURKB
(e)
Specificity
Sens
itivi
ty
1.0 0.8 0.6 0.4 0.2 0.0
0.0
0.2
0.4
0.6
0.8
1.0
−0.389 (1.000, 0.852)
AUC: 0.983
CCNA2
(f)
Figure 11: Continued.
16 BioMed Research International
(P = 0:303) was not related with the overall survival of ccRCCpatients. The overall survival analysis showed that ccRCCpatients with high expression levels of AURKB (P = 0:000),CCNA2 (P = 0:000), TPX2 (P = 0:000), and NCAPG(P = 0:000) had poorer overall survival times than those withlow expression levels (Figure 14).
3.11. The Neural Network Prediction Model between AURKB,CCNA2, TPX2, and NCAPG. The mean squared error is<0.05 (Figure 15(a)). The relativity of training is 0.9906.The relativity of validation is 0.99768. The relativity of testis 0.93812. And the relativity of all procedure is 0.97977(Figure 15(b)). Through verifying the predicted value of thedata against the actual value, we found that there are onlysmall differences in the comparison chart of training results(Figure 15(c)) and error analysis diagram (Figure 15(d)).Based on the above result, we could speculate that there werestrong correlations between AURKB, CCNA2, TPX2, andNCAPG.
Through the cubic spline interpolation algorithm, wefind the high-risk warning indicator of TPX2: CCNA2 < 5:0and 5:2 < AURKB. The three-dimensional stereogram couldpresent the warning range well (Figure 15(e)). The planegraph is also shown (Figure 15(f)).
4. Discussion
RCC is a common disease in the urinary system. Accordingto the statistics of the World Health Organization in 2018,its incidence is second only to prostate cancer and bladdercancer and is increasing year by year [1]. Although manygenes are considered potential therapeutic targets and prog-nostic predictors of RCC, the molecular mechanism of theoccurrence and development of RCC remains controversial.
With the continuous progress of science, microarraytechnology, as a special data mining method, is very influen-tial at present. This revolutionary technology transformstraditional molecular research from a situation that relies
on personal experience and subjective guesses to a moreobjective science [18–20].
In this paper, bioinformatics tools are used to mine thetargeted biomarkers of ccRCC. The results showed thatAURKB, CCNA2, TPX2, and NCAPG were highly expressedin ccRCC compared with renal tissue. With the increasingexpression of AURKB, CCNA2, TPX2, and NCAPG, thepathological stage of ccRCC increased gradually. Comparedwith the individuals with low expression of AURKB,CCNA2, TPX2, and NCAPG, patients with high expressionof AURKB, CCNA2, TPX2, and NCAPG have a poor overallsurvival.
Aurora kinase B (AURKB) is a serine/threonine kinasethat participates in the regulation of chromosome arrange-ment and segregation by binding to microtubules [21].Numerous studies have found that the overexpression ofAURKB exists in a variety of cancer cell lines [22–24]. Sor-rentino et al. found that AURKB is highly expressed in thy-roid carcinoma, and its expression level is related tomalignant degree. The block of AURKB expression or byusing an inhibitor of Aurora kinase activity significantlyreduced the growth of thyroid carcinoma cells [23].Katayama et al. have similar findings in colorectal cancer[25], Smith et al. in lung cancer [22], and Chieffi et al. in pros-tate cancer [24]. Abnormal mitotic regulation can induce theproduction of aneuploid cells and act as a driving role in theprocess of malignant progression, while serine/theronineprotein kinases of the Aurora family genes play a critical rolein the regulation of key cell cycle processes. The abnormalexpression of AURKB can produce malignant and invasiveaneuploid cells. This further indicates that AURKB is relatedto tumorigenesis [26, 27]. With the discovery of abnormalexpression of AURKB in cancer cells, researchers realizedthat it may become a new target for cancer treatments. Atpresent, many AURKB inhibitors have been developed,including AZD1152, AT9283, VX-680/MK-0457, PHA-680632, AMG-900, PHA-739358, and CYC-116, and someof them have entered clinical trials [28].
Specificity
Sens
itivi
ty
1.0 0.8 0.6 0.4 0.2 0.0
0.0
0.2
0.4
0.6
0.8
1.0−0.002 (0.962, 0.926)
AUC: 0.977
TPX2
(g)
Specificity
Sens
itivi
ty
1.0 0.8 0.6 0.4 0.2 0.0
0.0
0.2
0.4
0.6
0.8
1.00.302 (0.846, 0.963)
AUC: 0.934
NCAPG
(h)
Figure 11: ROC curves of hub genes for ccRCC.
17BioMed Research International
Con ccRCC
0
1
2
3VEGFA
Rela
tive e
xpre
ssio
n of
VEG
FA
⁎
(a)
Con ccRCC
0
2
4
6
8AURKB
Rela
tive e
xpre
ssio
n of
AU
RKB ⁎
(b)
Con ccRCC
0
2
4
6
8
10CCNA2
Rela
tive e
xpre
ssio
n of
CCN
A2 ⁎
(c)
Con ccRCC
0
5
10
15
20MCM2
Rela
tive e
xpre
ssio
n of
MCM
2 ⁎
(d)
Con ccRCC
0
2
4
6
8
10MCM7
Rela
tive e
xpre
ssio
n of
MCM
7 ⁎
(e)
Con ccRCC
0
2
4
6
8
10SMC4
Rela
tive e
xpre
ssio
n of
SM
C4
⁎
(f)
Con ccRCC
0
5
10
15
20
25TPX2
Rela
tive e
xpre
ssio
n of
TPX
2
⁎
(g)
Con ccRCC
0
5
10
15
20SLC2A1
Rela
tive e
xpre
ssio
n of
SLC
2A1 ⁎
(h)
Con ccRCC
0
5
10
15
20MCM5
Rela
tive e
xpre
ssio
n of
MCM
5 ⁎
(i)
Con ccRCC
0
5
10
15
20NCAPG
Rela
tive e
xpre
ssio
n of
NCA
PG
⁎
(j)
Figure 12: Relative expression of VEGFA, AURKB, CCNA2, MCM2, MCM7, SMC4, TPX2, SLC2A1, MCM5, and NCAPG by RT-qPCRanalysis. ∗P < 0:05, compared with normal kidney tissues.
18 BioMed Research International
The proteins encoded by CyclinA2 (CCNA2) belong to ahighly conserved cyclin family, which promotes cell transfor-mation by binding and activating cyclin-dependent kinases(CDKs) through G1/S and G2/M [29]. Previous studies havefound that the overexpression of CyclinA2 occurs in lungcancer [30, 31], breast cancer [32, 33], colorectal cancer[34], and other tumors and related to poor prognosis ofcancer patients. Aaltomaa et al. found that CyclinA2 wasexpressed in the cytoplasm of RCC but not in the normaltissue near the tumor, and the overexpression of CyclinA2was related to the survival time of patients with RCC,suggesting that it may be a prognostic indicator of RCC[35]. The increase of the CyclinA2 expression is related tothe uncontrolled and accelerated cell cycle, which leads togene amplification and chromosome ectopia. Gopinathanet al. found that knockout CyclinA2 in mice can inhibittumorigenesis [36]. Liang et al. found that the increasedexpression of sclerostin domain-containing protein1(SOSTDC1) can inhibit CyclinA2, while SOSTDC1 caninhibit tumor growth [37]. CyclinA2 can not only be usedas a predictor of prognosis and survival in patients withRCC but also has great potential in cancer treatment.
TPX2 microtubule nucleation factor (TPX2) encodes amicrotubule-associated protein that activates cell cycle kinasecalled Aurora A and regulates mitotic spindles. The overex-pression of TPX2 is related to the genesis of different can-cers and is closely related to chromosome instability. Theuncontrolled expression of TPX2 may eventually becomethe driving force of cancer development by inducing aneu-ploidy [38]. Zhang et al. found that compared with humanbronchial epithelial cells (16HBE), TPX is overexpressed inmalignant transformed 16HBE cells(16HBE-C) throughanti-benzo[a]pyrene-trans-7,8-dihydrodiol-9,10-epoxide, inwhich TPX2 RNA interference (RNAi) can lead to S-phasearrest, inhibit cell proliferation, and induce cell apoptosis.
TPX2 is tyrosine phosphorylated in malignant transformed16HBE-C, and this phosphorylation may be involved in themalignant proliferation of cancer cells [39]. Ma et al. foundthat the level of the TPX2 protein in normal bronchial epithe-lium and alveoli was very low, while the level of TPX2 proteinincreased gradually in squamous metaplasia, dysplasia, andcarcinoma in situ and invasive tumor. The immunohisto-chemical labeling index of TPX2 was related to the degreeof differentiation, stage, and lymph node metastasis of lungsquamous cell carcinoma, and the overexpression of TPX2was significantly correlated with the decrease of 5-year sur-vival rate [40]. Similar results were found in a variety ofcancers, such as colorectal cancer [41], cervical cancer[42], and prostate cancer [43]. The expression of TPX2 inRCC was significantly higher than that in normal renal tis-sue, and it was related to tumor size, histological grade,tumor stage, and poor prognosis [44–47]. This may bedue to the significant upregulation of TPX2 in RCC tissues,thus increasing the proliferation and invasive ability ofrenal cancer cells. From this point of view, TPX2 can notonly become a target for RCC treatment but also play a roleas an independent prognostic factor of RCC.
Non-SMC condensin I complex subunit G (NCAPG)coding a condensed protein complex subunit is responsiblefor chromosome condensation and stabilization duringmitosis and meiosis [48]. In recent years, there are moreand more studies on the abnormal expression of NCAPG inprostate cancer [49], lung cancer [50], breast cancer [51],and other cancers. In the study of Liu et al., NCAPG wasfound to be overexpressed in hepatocellular carcinoma com-pared with the adjacent normal tissue, and high levels ofNCAPG expression were found to significantly correlate withrecurrence, the time of recurrence, metastasis, differentiation,and TNM stage. The knockdown of NCAPG expression alsoinhibited tumor cell migration and the cell invasive capacity
Gene symbol
ccRCC
Normal
Color: sample_type
Gen
e exp
ress
ion
RNA
seq
-Illu
min
aHiS
eqU
nit:
log2
(nor
m_c
ount
+1)
–2.5
0
2.5
5
7.5
10
12.5
15
17.5
20TCGA clear cell renal cell carcinoma (ccRCC)
VEG
FA
AU
RKB
CCN
A2
MCM
2
MCM
7
SMC4
TPX2
SLC2
A1
MCM
5
NCA
PG
⁎
⁎⁎
⁎⁎
⁎ ⁎
⁎
⁎
⁎
Figure 13: The confirmation of gene expression level using The Cancer Genome Atlas (TCGA) data. The gene expression levels of VEGFA,AURKB, CCNA2, MCM2, MCM7, SMC4, TPX2, SLC2A1, MCM5, and NCAPG in ccRCC samples were significantly higher than the normalrenal samples.
19BioMed Research International
0
0.25
0.5
0.75
1
0 1,000 2,000 3,000 4,0000
0
0.25
0.5
0.75
1
0
1 1
0
0.25
0.5
0.75
1 1
0
0.25
0.5
0.75
1
VEGFA AURKB CCNA2
MCM2 MCM7 SMC4
TPX2 SLC2A1 MCM5
NCAPG
Perc
ent s
urvi
val
Perc
ent s
urvi
val
Perc
ent s
urvi
val
0.25
0.5
0.75
1
0.25
0.5
0.75
0.25
0.5
0.75
Perc
ent s
urvi
val
Perc
ent s
urvi
val
Perc
ent s
urvi
val
0
0.25
0.5
0.75
1
0.25
0.5
0.75
0.25
0.5
0.75
Perc
ent s
urvi
val
Perc
ent s
urvi
val
Perc
ent s
urvi
val
Perc
ent s
urvi
val
Days0 1,000 2,000 3,000 4,000
Days0 1,000 2,000 3,000 4,000
Days
0 1,000 2,000 3,000 4,000Days
0 1,000 2,000 3,000 4,000Days
0 1,000 2,000 3,000 4,000Days
0 1,000 2,000 3,000 4,000Days
0 1,000 2,000 3,000 4,000Days
0 1,000 2,000 3,000 4,000Days
0 1,000 2,000 3,000 4,000Days
P = 0.4946Log-rank test statistics = 0.4666
P = 0.000Log-rank test statistics = 21.59
P = 0.000Log-rank test statistics = 23.30
P = 0.5249Log-rank test statistics = 0.4042 P = 0.092
Log-rank test statistics = 2.831P = 0.856Log-rank test statistics = 0.033
P = 0.000Log-rank test statistics = 15.16
P = 0.209Log-rank test statistics = 1.581
P = 0.303Log-rank test statistics = 1.059
P = 0.000Log-rank test statistics = 20.47
LowHigh
Figure 14: The effect of gene expression on overall survival by using the TCGA data. The expression level of VEGFA, MCM2, MCM7, SMC4,SLC2A1, and MCM5 was not related with the overall survival of ccRCC patients. The ccRCC patients with high expression levels of AURKB,CCNA2, TPX2, and NCAPG had poorer overall survival times than those with low expression levels.
20 BioMed Research International
0 50 150 200 250 30010–6
10–4
10–2
100
102
MSE
Number of iterations100
(a)
5 10 15468
1012141618
Out
put~
=0.9
7⁎Ta
rget
+0.2
8
468
1012141618
Out
put~
=0.9
3⁎Ta
rget
+–0.
1
468
1012141618
Out
put~
=0.9
3⁎Ta
rget
+0.3
5
468
1012141618
Out
put~
=0.8
9⁎Ta
rget
+–0.
36
Training: R = 0.9906 Validation: R = 0.99768
Target5 10 15
Target
5 10 15Target
DataFitY = T
5 10 15Target
All: R = 0.97977Test: R = 0.93812
DataFitY = T
DataFitY = T
FitFitY = T
(b)
0 4 8 12 16 200
4
8
12
16
20
TPX2
Sample number
Raw dataTraining results
(c)
0 3 6 9
–1.5
–1.0
–0.5
–2.0
0.0
0.5
1.0
Abso
lute
erro
r
Sample number21181512
(d)
Figure 15: Continued.
21BioMed Research International
in vitro [52]. Through genome-wide functional knockoutscreen, Wang et al. believe that NCAPG is a necessaryclinical-related target for the growth of hepatocellularcarcinoma cells [53]. Ai et al. found that microRNA-181c(miR-181c) inhibits cancer by downregulating the expressionof NCAPG, affecting the infiltration, migration, proliferation,and apoptosis of hepatoma cells [54]. In the study of Araiet al., microRNA-99a-3p downregulated the expression ofNCAPG, thereby inhibiting cancer cell invasion incastration-resistant prostate cancer [49]. In conclusion,NCAPG represents a promising novel target and a prognos-tic biomarker for clinical management.
However, this study also has some shortcomings.Although the core genes screened in this study may play animportant role in the occurrence of ccRCC, more clinicalsamples and patient prognosis information are still neededfor verification.
5. Conclusion
To sum up, 251 differentially expressed genes and 10 hubgenes (especially AURKB, CCNA2, TPX2, and NCAPG)were screened from ccRCC and normal renal tissues bymicroarray technology, which could be used as diagnosticand therapeutic biomarkers for ccRCC. AURKB, CCNA2,TPX2, and NCAPG which might be related to the occurrenceand malignant progression of ccRCC.
Data Availability
The datasets used and/or analyzed during the current studyare available from the corresponding author on reasonablerequest.
Disclosure
The authors ensure that questions related to the accuracy orintegrity of any part of the work are appropriately investi-gated and resolved.
Conflicts of Interest
The authors declare that they have no conflicts of interests.
Authors’ Contributions
Ai-li Zhang and Ling-bing Meng contributed equally to thiswork.
Acknowledgments
We gratefully thank the Department of Urinary Surgery inThe Fourth Hospital of Hebei Medical University for thetechnical assistance.
References
[1] R. L. Siegel, K. D. Miller, and A. Jemal, “Cancer statistics,2018,” CA: a Cancer Journal for Clinicians, vol. 68, no. 1,pp. 7–30, 2018.
[2] U. Capitanio, K. Bensalah, A. Bex et al., “Epidemiology of renalcell carcinoma,” European Urology, vol. 75, no. 1, pp. 74–84,2019.
[3] U. Capitanio and F. Montorsi, “Renal cancer,” The Lancet,vol. 387, no. 10021, pp. 894–906, 2016.
[4] B. Shuch, A. Amin, A. J. Armstrong et al., “Understandingpathologic variants of renal cell carcinoma: distilling therapeu-tic opportunities from biologic complexity,” European Urol-ogy, vol. 67, no. 1, pp. 85–97, 2015.
[5] R. Srinivasan, C. J. Ricketts, C. Sourbier, and W. M. Linehan,“New strategies in renal cell carcinoma: targeting the geneticand metabolic basis of disease,” Clinical Cancer Research,vol. 21, no. 1, pp. 10–17, 2015.
[6] P. Singh, N. Agarwal, and S. K. Pal, “Sequencing systemic ther-apies for metastatic kidney cancer,” Current TreatmentOptions in Oncology, vol. 16, article 3, 2015.
[7] H. M. Huang, X. Jiang, M. L. Hao et al., “Identification of bio-markers in macrophages of atherosclerosis by microarrayanalysis,” Lipids in Health and Disease, vol. 18, no. 1, article107, 2019.
1 2 3 4 5 6 7 82
34
56
70
5
10
15
20
25
AURKB
CCNA2
TPX2
1 2 3 4 5 63
45
67
AURKB
CCN
(e)
1 2 3 4 5 6 7 8
2
3
4
5
6
7
0.000
3.300
6.600
9.900
13.20
16.50
19.80
23.10
26.40
AURK
B
CCNA2
TPX2
(f)
Figure 15: The neural network models. (a) The best training performance. (b) The relativity of training, validation, test, and all procedure. (c)The comparison chart of training results. (d) The error analysis diagram. (e) The high-risk warning range at the level of the three-dimensionalstereogram. (f) The high-risk warning range at the level of the plane form.
22 BioMed Research International
[8] Y. Qiu, L. B. Meng, C. Y. Di et al., “Exploration of the differen-tially expressed long noncoding RNAs and genes of morphinetolerance via bioinformatic analysis,” Journal of Computa-tional Biology, vol. 26, no. 12, pp. 1379–1393, 2019.
[9] Y. F. Zou, L. B. Meng, Z. K. He et al., “Screening and authen-tication of molecular markers in malignant glioblastoma basedon gene expression profiles,” Oncology Letters, vol. 18,pp. 4593–4604, 2019.
[10] L. B. Meng, M. J. Shan, Y. Qiu et al., “TPM2 as a potential pre-dictive biomarker for atherosclerosis,” Aging, vol. 11, no. 17,pp. 6960–6982, 2019.
[11] M. Ashburner, C. A. Ball, J. A. Blake et al., “Gene ontology: toolfor the unification of biology,”Nature Genetics, vol. 25, pp. 25–29, 2000.
[12] M. Tanabe and M. Kanehisa, “Using the KEGG databaseresource,” Curr Protoc Bioinformatics, vol. 11, no. 1, 2005.
[13] S. Maere, K. Heymans, and M. Kuiper, “BiNGO: a Cytoscapeplugin to assess overrepresentation of gene ontology categoriesin biological networks,” Bioinformatics, vol. 21, no. 16,pp. 3448-3449, 2005.
[14] M. E. Smoot, K. Ono, J. Ruscheinski, P. L.Wang, and T. Ideker,“Cytoscape 2.8: new features for data integration and networkvisualization,” Bioinformatics, vol. 27, no. 3, pp. 431-432, 2011.
[15] G. D. Bader and C. W. Hogue, “An automated method forfinding molecular complexes in large protein interaction net-works,” BMC Bioinformatics, vol. 4, no. 1, p. 2, 2003.
[16] C. H. Chin, S. H. Chen, H. H. Wu, C. W. Ho, M. T. Ko, andC. Y. Lin, “cytoHubba: identifying hub objects and sub-networks from complex interactome,” BMC Systems Biology,vol. 8, article S11, 2014.
[17] Z. Tang, C. Li, B. Kang, G. Gao, C. Li, and Z. Zhang, “GEPIA: aweb server for cancer and normal gene expression profilingand interactive analyses,” Nucleic Acids Research, vol. 45,no. W1, pp. W98–W102, 2017.
[18] T. Flieder, M.Weiser, T. Eller et al., “Interference of DOACs indifferent DRVVT assays for diagnosis of lupus anticoagu-lants,” Thrombosis Research, vol. 165, pp. 101–106, 2018.
[19] G. Hussain, J. Wang, A. Rasul et al., “Role of cholesterol andsphingolipids in brain development and neurological dis-eases,” Lipids in Health and Disease, vol. 18, no. 1, p. 26, 2019.
[20] W. Wang, W. Lou, B. Ding et al., “A novel mRNA-miRNA-lncRNA competing endogenous RNA triple sub-network asso-ciated with prognosis of pancreatic cancer,” Aging, vol. 11,no. 9, pp. 2610–2627, 2019.
[21] F. L. Chan, B. Vinod, K. Novy et al., “Aurora kinase B, a novelregulator of TERF1 binding and telomeric integrity,” NucleicAcids Research, vol. 45, no. 21, pp. 12340–12353, 2017.
[22] S. L. Smith, N. L. Bowers, D. C. Betticher et al., “Overexpres-sion of aurora B kinase (AURKB) in primary non-small celllung carcinoma is frequent, generally driven from one allele,and correlates with the level of genetic instability,” British Jour-nal of Cancer, vol. 93, no. 6, pp. 719–729, 2005.
[23] R. Sorrentino, S. Libertini, P. L. Pallante et al., “Aurora B over-expression associates with the thyroid carcinoma undifferenti-ated phenotype and is required for thyroid carcinoma cellproliferation,” The Journal of Clinical Endocrinology andMetabolism, vol. 90, no. 2, pp. 928–935, 2005.
[24] P. Chieffi, L. Cozzolino, A. Kisslinger et al., “Aurora B expres-sion directly correlates with prostate cancer malignancy andinfluence prostate cell proliferation,” Prostate, vol. 66, no. 3,pp. 326–333, 2006.
[25] H. Katayama, T. Ota, F. Jisaki et al., “Mitotic kinase expressionand colorectal cancer progression,” Journal of the NationalCancer Institute, vol. 91, no. 13, pp. 1160–1162, 1999.
[26] M. Tatsuka, H. Katayama, T. Ota et al., “Multinuclearity andincreased ploidy caused by overexpression of the aurora- andIpl1-like midbody-associated protein mitotic kinase in humancancer cells,” Cancer Research, vol. 58, no. 21, pp. 4811–4816,1998.
[27] T. Ota, S. Suto, H. Katayama et al., “Increased mitotic phos-phorylation of histone H3 attributable to AIM-1/Aurora-Boverexpression contributes to chromosome number instabil-ity,” Cancer Research, vol. 62, no. 18, pp. 5168–5177, 2002.
[28] A. Tang, K. Gao, L. Chu, R. Zhang, J. Yang, and J. Zheng,“Aurora kinases: novel therapy targets in cancers,” Oncotarget,vol. 8, no. 14, pp. 23937–23954, 2017.
[29] D. Gong, J. R. Pomerening, J. W. Myers et al., “Cyclin A2 reg-ulates nuclear-envelope breakdown and the nuclear accumula-tion of cyclin B1,” Current Biology, vol. 17, no. 1, pp. 85–91,2007.
[30] M. Volm, R. Koomägi, J. Mattern, and G. Stammler, “Cyclin Ais associated with an unfavourable outcome in patients withnon- small-cell lung carcinomas,” British Journal of Cancer,vol. 75, no. 12, pp. 1774–1778, 1997.
[31] Y. Dobashi, M. Shoji, S. X. Jiang, M. Kobayashi, Y. Kawakubo,and T. Kameya, “Active cyclin A-CDK2 complex, a possiblecritical factor for cell proliferation in human primary lung car-cinomas,” The American Journal of Pathology, vol. 153, no. 3,pp. 963–972, 1998.
[32] I. R. Bukholm, G. Bukholm, and J. M. Nesland, “Over-expres-sion of cyclin A is highly associated with early relapse andreduced survival in patients with primary breast carcinomas,”International Journal of Cancer, vol. 93, no. 2, pp. 283–287,2001.
[33] R. Michalides, H. van Tinteren, A. Balkenende et al., “Cyclin Ais a prognostic indicator in early stage breast cancer with andwithout tamoxifen treatment,” British Journal of Cancer,vol. 86, no. 3, pp. 402–408, 2002.
[34] J. Q. Li, H. Miki, F. Wu et al., “Cyclin a correlates with carcino-genesis and metastasis, and p27kip1 correlates with lymphaticinvasion, in colorectal neoplasms,” Human Pathology,vol. 33, no. 10, pp. 1006–1015, 2002.
[35] S. Aaltomaa, P. Lipponen, M. Ala-Opas, M. Eskelinen,K. Syrjänen, and V. M. Kosma, “Expression of cyclins Aand D and p21(waf1/cip1) proteins in renal cell cancer and theirrelation to clinicopathological variables and patient survival,”British Journal of Cancer, vol. 80, no. 12, pp. 2001–2007,1999.
[36] L. Gopinathan, S. L. Tan, V. C. Padmakumar, V. Coppola,L. Tessarollo, and P. Kaldis, “Loss of Cdk2 and cyclin A2impairs cell proliferation and tumorigenesis,” CancerResearch, vol. 74, no. 14, pp. 3870–3879, 2014.
[37] W. Liang, H. Guan, X. He et al., “Down-regulation ofSOSTDC1 promotes thyroid cancer cell proliferation viaregulating cyclin A2 and cyclin E2,” Oncotarget, vol. 6,no. 31, pp. 31780–31791, 2015.
[38] C. Aguirre-Portolés, A. W. Bird, A. Hyman, M. Cañamero,I. Pérez de Castro, and M. Malumbres, “Tpx2 controls spindleintegrity, genome stability, and tumor development,” CancerResearch, vol. 72, no. 6, pp. 1518–1528, 2012.
[39] L. Zhang, H. Huang, L. Deng et al., “TPX2 in malignantlytransformed human bronchial epithelial cells by anti-
23BioMed Research International
benzo[a]pyrene-7,8-diol-9,10-epoxide,” Toxicology, vol. 252,no. 1-3, pp. 49–55, 2008.
[40] Y. Ma, D. Lin, W. Sun et al., “Expression of targeting proteinfor xklp2 associated with both malignant transformation ofrespiratory epithelium and progression of squamous cell lungcancer,” Clinical Cancer Research, vol. 12, no. 4, pp. 1121–1127, 2006.
[41] A. H. Sillars-Hardebol, B. Carvalho, M. Tijssen et al., “TPX2and AURKA promote 20q amplicon-driven colorectal ade-noma to carcinoma progression,” Gut, vol. 61, no. 11,pp. 1568–1575, 2012.
[42] H. Chang, J. Wang, Y. Tian, J. Xu, X. Gou, and J. Cheng, “TheTPX2 gene is a promising diagnostic and therapeutic target forcervical cancer,” Oncology Reports, vol. 27, no. 5, pp. 1353–1359, 2012.
[43] P. Vainio, J. P. Mpindi, P. Kohonen et al., “High-throughputtranscriptomic and RNAi analysis identifies AIM1, ERGIC1,TMED3 and TPX2 as potential drug targets in prostate can-cer,” PLoS One, vol. 7, no. 6, article e39801, 2012.
[44] Q. I. Chen, B. Cao, N. Nan et al., “TPX2 in human clearcell renal carcinoma: expression, function and prognosticsignificance,” Oncology Letters, vol. 11, no. 5, pp. 3515–3521, 2016.
[45] Z. A. Glaser, H. D. Love, S. Guo et al., “TPX2 as a prognosticindicator and potential therapeutic target in clear cell renal cellcarcinoma,” Urologic Oncology, vol. 35, no. 5, pp. 286–293,2017.
[46] Y. Gu, L. Lu, L. Wu, H. Chen, W. Zhu, and Y. He, “Identifica-tion of prognostic genes in kidney renal clear cell carcinoma byRNA-seq data analysis,” Molecular Medicine Reports, vol. 15,no. 4, pp. 1661–1667, 2017.
[47] W. Wei, Y. Lv, Z. Gan, Y. Zhang, X. Han, and Z. Xu, “Identi-fication of key genes involved in the metastasis of clear cellrenal cell carcinoma,” Oncology Letters, vol. 17, no. 5,pp. 4321–4328, 2019.
[48] A. Eberlein, A. Takasuga, K. Setoguchi et al., “Dissection ofgenetic factors modulating fetal growth in cattle indicates asubstantial role of the non-SMC condensin I complex, subunitG (NCAPG) gene,” Genetics, vol. 183, no. 3, pp. 951–964,2009.
[49] T. Arai, A. Okato, Y. Yamada et al., “Regulation of NCAPG bymiR-99a-3p (passenger strand) inhibits cancer cell aggressive-ness and is involved in CRPC,” Cancer Medicine, vol. 7, no. 5,pp. 1988–2002, 2018.
[50] S. Li, Y. Xuan, B. Gao et al., “Identification of an eight-geneprognostic signature for lung adenocarcinoma,” Cancer Man-agement and Research, vol. 2018, pp. 3383–3392, 2018.
[51] J. Chen, X. Qian, Y. He, X. Han, and Y. Pan, “Novel key genesin triple-negative breast cancer identified by weighted gene co-expression network analysis,” Journal of Cellular Biochemistry,vol. 120, no. 10, pp. 16900–16912, 2019.
[52] W. Liu, B. Liang, H. Liu et al., “Overexpression of non-SMCcondensin I complex subunit G serves as a promising prognos-tic marker and therapeutic target for hepatocellular carci-noma,” International Journal of Molecular Medicine, vol. 40,no. 3, pp. 731–738, 2017.
[53] Y. Wang, B. Gao, P. Y. Tan et al., “Genome-wide CRISPRknockout screens identify NCAPG as an essential oncogenefor hepatocellular carcinoma tumor growth,” The FASEBJournal, vol. 33, no. 8, pp. 8759–8770, 2019.
[54] J. Ai, C. Gong, J. Wu et al., “MicroRNA-181c suppressesgrowth and metastasis of hepatocellular carcinoma bymodulating NCAPG,” Cancer Management and Research,vol. 2019, pp. 3455–3467, 2019.
24 BioMed Research International