Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Research ArticleScreening and Identification of Key Biomarkers for BladderCancer: A Study Based on TCGA and GEO Data
Yingkun Xu,1 Guangzhen Wu ,2 Jianyi Li,1 Jiatong Li,3 Ningke Ruan,4 Liye Ma,5
Xiaoyang Han,6 Yanjun Wei,6 Liang Li,7 Hongge Zhang,8 Yougen Chen,1
and Qinghua Xia 1
1Department of Urology, Shandong Provincial Hospital Affiliated to Shandong University, Jinan, China2Department of Urology, e First Affiliated Hospital of Dalian Medical University, Dalian, China3Department of Nephrology, Shandong Provincial Hospital, Shandong First Medical University,Shandong Academy of Medical Sciences, Jinan, China4 e Nursing College of Zhengzhou University, Zhengzhou, China5Department of Liver Transplantation and Hepatobiliary Surgery,Shandong Provincial Hospital Affiliated to Shandong University, Jinan, China6Department of Oncology, Shandong Provincial Hospital Affiliated to Shandong University, Jinan, China7Shangdong Academy of Medical Sciences, Jinan University, Jinan, China8Department of Urology, Tengzhou Hospital of Traditional Chinese Medicine, Tengzhou, China
Correspondence should be addressed to Qinghua Xia; [email protected]
Received 16 September 2019; Revised 18 November 2019; Accepted 26 December 2019; Published 25 January 2020
Academic Editor: Maxim Golovkin
Copyright © 2020 Yingkun Xu et al. .is is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Bladder cancer (BLCA) is a common malignant cancer, and it is the most common genitourinary cancer in the world. .erecurrence rate is the highest of all cancers, and the treatment of BLCA has only slightly improved over the past 30 years. Geneticand environmental factors play an important role in the development and progression of BLCA. However, the mechanism ofcancer development remains to be proven. .erefore, the identification of potential oncogenes is urgent for developing newtherapeutic directions and designing novel biomarkers for the diagnosis and prognosis of BLCA. Based on this need, we screenedoverlapping differentially expressed genes (DEG) from the GSE7476, GSE13507, and TCGABLCA datasets. To identify the centralgenes from these DEGs, we performed a protein-protein interaction network analysis. To investigate the role of DEGs and theunderlying mechanisms in BLCA, we performed Gene Ontology (GO) and Kyoto Gene and Genomic Encyclopedia (KEGG)analysis; we identified the hub genes via different evaluation methods in cytoHubba and then selected the target genes byperforming survival analysis. Finally, the relationship between these target genes and tumour immunity was analysed to explorethe roles of these genes. In summary, our current studies indicate that both cell division cycle 20 (CDC20) and abnormal spindlemicrotubule assembly (ASPM) genes are potential prognostic biomarkers for BLCA. It may also be a potential immunotherapeutictarget with future clinical significance.
1. Introduction
Bladder cancer (BLCA) is a serious health problemworldwide, and it is the second most common malignanttumour of all genitourinary tract tumours [1]. Risk factorsfor BLCA are known to include tobacco, schistosomiasis,eating habits, and lifestyle. In 2012, approximately 430,000
new cases of BLCA were diagnosed [2]. In China, the rate ofBLCA occurrence increased rapidly during the five-yearperiod from 2003 to 2008, and the growth rate in women washigher than that in men [3]. Despite surgery, dissection, andvarious adjuvant treatments for BLCA, the five-year survivalrate is still low, and the risk of recurrence is high. Accordingto reports, 30–70% of tumours reoccur [4] and 30% of
HindawiBioMed Research InternationalVolume 2020, Article ID 8283401, 20 pageshttps://doi.org/10.1155/2020/8283401
tumours develop into muscle-invasive diseases [5]. .ere-fore, there is an urgent need to discover new and reliableBLCA biomarkers.
In recent years, immunotherapy has become the focus ofcancer treatment strategies. Immunotherapy-related drugshave been approved for marketing and became available re-cently for asymptomatic or very mildly symptomatic prostatecancer [6], unresectable or metastatic melanoma [7], advancedmelanoma, and acute lymphocytic leukaemia [8]..e immuneresponse to cystatin has been confirmed very early [9].Intravesical instillation of BCG can kill bladder tumour cells byinducing the infiltration of cytotoxic T lymphocytes (CTL) inNMIBC patients [10]. Recently, it was reported that anti-PD-1/PD-L1 antibodies affect the growth of tumour cells by actingon T cells [11]. At present, there is an increasing number ofstudies on the role of immune checkpoints and immune cellsin influencing tumour development. .erefore, it is necessaryto find new possible prognostic and immunotherapeuticbiomarkers for BLCA.
To identify potential biomarkers for BLCA, we per-formed a series of analyses based on high-throughput se-quencing data obtained from three data sets, GSE7476,GSE13507, and TCGA BLCA. We first identified the DEGsthat are common among the three databases, as the com-bination of multiple databases can provide more credibleresults. .en, we used the Metascape website and the onlinetool from the DAVID website to analyse GO and KEGGterms, explore the main pathway of DEG enrichment, andexplore the research progress on the pathway in bladdercancer. .e protein interaction network between DEGs wasconstructed by using the online tool from the STRINGwebsite and illustrated with Cytoscape software. .en, weused the cytoHubba plugin for Cytoscape to search for thehub gene. Here, we used four different models, DEGREE,MCC, DMNC, and MNC, to screen out the most significanthub genes. We then used the Gene Expression ProfilingInteractive Analysis (GEPIA) and Human Protein Atlasonline tools to explore genes in the hub gene network thatare associated with bladder cancer prognosis. Finally, weused the UALCAN, cBioPortal, STRING, Cytoscape, andTIMER tools to explore this single gene and its main bio-logical role. We demonstrate that CDC20 and ASPM arepossible biomarkers for BLCA. After further exploration, wewere pleasantly surprised to find that both CDC20 andASPM are associated with the prognosis and immuno-therapy response of patients with BLCA. In vitro, we in-terfered with the expression of ASPM and CDC20 and thenused the cell counting kit-8 experiment and clone formationexperiment to detect the effect on the proliferation ofbladder cancer T24 cell line. In summary, our study providesnew potential prognostic markers and potential immuno-therapeutic targets for BLCA.
2. Materials and Methods
2.1. Microarray Data. GEO (https://www.ncbi.nlm.nih.gov/geo/) is a database containing high-throughput gene expres-sion data, chips, and microarrays [12]. We downloaded a geneexpression dataset (GSE7476) from GEO (Affymetrix
GPL3111 platform). .is gene expression dataset was trans-lated into commonly used gene symbols by using annotationinformation from the platform. .e GSE7476 dataset con-tained 9 BLCA tissue samples and 3 noncancer samples.
2.2. Data Processing. GEO2R (http://www.ncbi.nlm.nih.gov/geo/geo2r/) is an analysis tool that comes with theGEO database and is used to compare two sets of data; it canbe used to analyse any GEO series. Since the GSE13507dataset does not contain CELL subfiles that can be analysedby R language, we chose to use GEO2R to screen differ-entially expressed mRNA between normal tissue samplesand cancer tissue samples in the GSE13507 dataset [13].P< 0.05 and logFC> 1 or< –1 were set as the cut-off criteria..e GSE13507 dataset contained 188 BLCA tissue samplesand 68 noncancer samples.
2.3. e DEGs in BLCA from TCGA. TCGA is a vast re-pository of high-throughput data on DNA, RNA, andproteins in a variety of human cancers, which enables acomplete analysis of the expression of these components invarious cancer types. In the current study, we obtainedmRNA expression profiles from BLCA and adjacent normaltissues from GEPIA (TCGA Data Online Analysis Tool)(http://gepia2.cancer-pku.cn/#index) [14]. .e TCGAdataset contained 404 BLCA tissue samples and 19 non-cancer samples.
2.4. Functional and Pathway Enrichment Analyses. First, weperformed Gene Ontology (GO) and Kyoto Encyclopedia ofGenes and Genomes (KEGG) analysis on DEGs by using theMetascape software (http://metascape.org/gp/index.html#/main/step1) [15]. Metascape is an online analysis toolwith integrated discovery and annotation capabilities. Toensure the credibility of the results, we also analysed the datawith online tools from the DAVID website and visualizedthe results via the R language. .e DAVID website (https://david.ncifcrf.gov/home.jsp) is a bioinformatics data re-source composed of a comprehensive biological knowledgebase and analytical tools [16]. A P value< 0.05 was set as thecut-off criterion.
2.5. Protein-Protein Interaction (PPI) Network Constructionand Module Analysis. STRING (https://string-db.org/) candraw PPI networks after importing common DEGs intosearch tools to retrieve interacting genes [17]. First, we drewthe PPI network diagram of DEGs by using the STRINGwebsite. Cytoscape, a free visualization software, was appliedto visualize PPI networks and find hub genes [18]. .en, thehub genes were identified by four methods: DEGREE, MCC,DMNC, and MNC in cytoHubba [19].
2.6. Association of Hub Genes Expression with the Survival ofPatients with BLCA. .e GEPIA website can provide fastand customizable functions based on TCGA data. We firstanalysed the expression of the target gene using the GEPIA
2 BioMed Research International
website and then analysed the prognosis of the target geneusing the Human Protein Atlas website (https://www.proteinatlas.org/) [20]. .e results from the two websiteswere used to find the target gene. P< 0.05 was consideredstatistically significant.
2.7. Analysis of Target Genes. First, through the TIMERwebsite, the expression of the target gene in various tu-mours was found. .en, the UALCAN website (http://ualcan.path.uab.edu/analysis.html) was used to determinewhich factors are related to the expression of the targetgene in BLCA. .en, the STRING website was queried forthe ten genes with the highest correlation with the targetgene. In addition, Cytoscape software was used to map theprotein interactions and the enrichment pathways. Finally,these data were imported into the cBioPortal website(https://www.cbioportal.org/) to query the variation inBLCA [21]. .e cBioPortal website was used to analysecoexpression of the target gene and other genes in theenrichment pathway.
2.8. Correlation between mRNA Expression and Immune CellInfiltration and Immune Checkpoints. TIMER (https://cistrome.shinyapps.io/timer/) is a website dedicated toanalysing tumour immune relatedness. We used TIMER toanalyse the mRNA expression data for CDC20 and ASPM inTCGA BLCA tumour samples and its correlation with tu-mour infiltration of 6 immune cell types (B cells, CD4+T cells, CD8+ T cells, neutrophils, macrophages, and den-dritic cells) and 5 immunological checkpoints (PDCD1,CD274, PDCD1LG2, TOX, and CTLA4) [22].
2.9. Cell Culture and Reagents. .e human bladder cancercell line T24 was purchased from the Chinese Academy ofSciences cell bank. Cells were incubated at 37°C with 5% CO2in RPMI 1640 medium (Gibco BRL, Rockville, MD) sup-plemented with 10% heated-inactivated fetal bovine serum(FBS, Biological Industries, Kibbutz Beit HaEmek, Israel)and 1% penicillin-streptomycin (Macgene, Beijing, People’sRepublic of China).
2.10. Proliferation Analysis. Bladder cancer cell line (T24)was transfected with siASPM, siCDC20, or siControl in 24-well plates. After 24 h, the cells were seeded into 96-wellplates. Cell viability was then measured using Cell CountingKit-8 (CCK8) every 24 h. For the clone formation assay, cellswere plated in a six-well plate (500 cells per well). After 2weeks, the cells were fixed with 4% paraformaldehyde for2 h, stained with 1% crystal violet. All assays were conductedmore than three times.
2.11. Statistical Analysis. .e data are expressed as themean± S.E.M. Statistical analysis was performed usingPrism software (GraphPad, CA, USA). Statistical signifi-cance of differences between and among groups was assessed
using the t-test. Significant differences are indicated asfollows: ∗P< 0.05; ∗∗P< 0.01; ∗∗∗P< 0.001.
3. Results
3.1. Identification of DEGs in BLCA. We found the differ-entially expressed genes on chromosomes of BLCA cellsthrough the GEPIA website (Figure 1(a)). R studio was usedto investigate the DEGs via mining of the GEO (GSE7476)database (https://www.ncbi.nlm.nih.gov/geo/). We analysedthe DEGs in the database and showed them in a heat map(Figure 1(b)) and a volcano map (Figure 1(c)). In addition,we used GEO2R to compare cancer and normal tissues in theGEO (GSE13507) and identify genes that were differentiallyexpressed in this dataset. .en, we explored BLAC DEGs viathe GEPIA website based on the TCGA database. .e datawere filtered by logFC> 1 or< –1 and P< 0.05. .e over-lapping DEGs among the 3 datasets were identified, and 50upregulated genes (Figure 1(d)) and 241 downregulatedgenes (Figure 1(e)) were selected and presented using a Venndiagram. Fifty upregulated and 241 downregulated DEGs arelisted in Table 1.
3.2.GOandKEGGPathwayAnalysis. To further analyse thepotential functions of DEGs, GO analysis was performedon the DEGs by using Metascape online tools; we foundthat the DEGs were mostly enriched in the NABA corematrisome cellular component, muscle contraction, su-pramolecular fibre organization, muscle structure devel-opment, and tissue morphogenesis (Figures 2(a)–2(e))..en, we performed GO analysis through the DAVIDwebsite and visualized the data with the R language.Concerning biological processes (BPs), the DEGs wereenriched in response to steroid hormone stimuli, hor-mone stimuli, endogenous stimuli, and oestrogen stimuliand in cytoskeleton organization (Figure 2(g)). .echanges in cellular components (CCs) were significantlyenriched in the extracellular region, contractile fibre,extracellular region part, contractile fibre part, and actincytoskeleton (Figure 2(h)). .e changes in molecularfunction (MF) were significantly enriched in the cyto-skeletal protein binding, structural constituent of muscle,pattern binding, polysaccharide binding, and glycos-aminoglycan binding (Figure 2(i)). We further analysedthe DEG-enriched REACTOME and KEGG pathwaysthrough the DAVID online tool and visualized it with theR language. We found that the DEG genes were mainlyenriched in REACTOME pathways such as muscle con-traction, haemostasis, phase 1 functionalization, biolog-ical oxidation, signalling by PDGF, and signalling in theimmune system (Figure 2(k)). KEGG pathway analysisrevealed that the hub gene was mainly enriched in vas-cular smooth muscle contraction, hypertrophic cardio-myopathy (HCM), dilated cardiomyopathy (DCM), focaladhesion, arachidonic acid metabolism, and histidinemetabolism (Figure 2(j)). .en, we used the Clugo pluginfor Cytoscape to illustrate the results of KEGG pathanalysis (Figure 2(f )).
BioMed Research International 3
3.3. Protein-Protein Interaction and Screening of Hub Genes.To better understand the relationship between DEGs, weused the STRING online tool to study the relationshipbetween various DEGs (Figure 3(a)). .en, we identified 14
hub genes based on the DEGREE (Figure 3(b)), MCC(Figure 3(c)), DMNC (Figure 3(d)), and MNC (Figure 3(e))plugins for the Cytoscape software. .e 14 hub genes wereASPM, CCNB2, CDC20, CENPF, CEP55, HJURP, KIF20A,
�e differentially expressed genes on chromosomes
�e gene positions are based on GRCh38.p2(NCBI). 2876 genes
(a)
(b)
(c)
(d) (e)
1
6
Volcano
4
2
0
LogF
C
–2
–4
–6
0 2 4Adj. P value
GSE7467
18420
61
70
241
482
1173
TCGA
GSE13507GSE7467 GSE13507
3
5
518
332
7074
50
TCGA
6 8
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 X Y
Overexpressed genesUnderexpressed genes
Figure 1: Identification of DEGs shared between the three databases. (a) .e differentially expressed genes on chromosomes. (b) .e heatmap of GSE7467. (c).e volcano map of GSE7467. (d) A Venn diagram used to identify 50 promising upregulated target genes in BLCA. (e)A Venn diagram used to identify 241 promising downregulated target genes in BLCA.
4 BioMed Research International
NCAPG, NUSAP1, SPAG5, TOP2A, TRIP13, TROAP, andTTK (Figure 3(f )).
3.4. Differential Expression Analysis of Hub Genes in BLCAand Normal Bladder Tissues. To verify the differential ex-pression of the hub genes between BLCA and normalbladder tissues, we analysed the 14 hub genes using theGEPIA website-based TCGA database. We found thatASPM (Figure 4(a)), CCNB2 (Figure 4(b)), CDC20(Figure 4(c)), CENPF (Figure 4(d)), CEP55 (Figure 4(e)),HJURP (Figure 4(f)), KIF20A (Figure 4(g)), NCAPG(Figure 4(h)), NUSAP1 (Figure 4(i)), SPAG5 (Figure 4(j)),TOP2A (Figure 4(k)), TRIP1 3(Figure 4(l)), TROAP(Figure 4(m)), and TTK (Figure 4(n)) were significantlyupregulated in BLCA tissue compared with normal bladdertissue, and the differences were statistically significant.
3.5.DeterminationofCDC20andASPMas theTargetGenesbySurvival Analysis. To investigate the relevance of the hubgenes in BLCA patient survival, we performed a survivalanalysis of the hub genes using the Human Protein Atlasonline tool for differential analysis (Figure 5(a)–5(n)). Wefound that the analysis of CDC20 and ASPM expression andsurvival was statistically significant in BLCA. .e knot andthe GEPIA websites give the expression of each hub gene; wechose CDC20 and ASPM as our target genes, and all of thegenes with high expression status predicted poor prognosis.
3.6. e Biological Role of CDC20 in Tumours. To investigatewhether the CDC20 gene acts as an oncogene in other
tumours, we analysed the differential expression of CDC20in different tumours and normal tissues through the TIMERwebsite. We found that CDC20 is upregulated in a variety oftumours, including BLCA, BRCA, CHOL, COAD, ESCA,HNSC, KICH, KIRC, KIRP, LIHC, LUAD, LUSC, PPAD,READ, STAD, THCA, and UCEC (Figure 6(a)). .is sug-gests that the role of CDC20 in regulating the underlyingmechanisms of tumorigenesis and progression is identical indifferent tumours.We found that CDC20 is highly expressedin BLCA through the UALCAN website (Figure 6(b)). Wealso found that CDC20 has differential expression in patientswith different smoking habits (Figure 6(c)), histologicalsubtypes (Figure 6(d)), and molecular subtypes(Figure 6(e)). Moreover, CDC20 is related to the promotermethylation level in BLCA (Figure 6(f )). To explore theunderlying molecular mechanisms of CDC20, we firstidentified genes that have a protein-protein interaction withCDC20 via the STRING website (Figure 6(h)), and thenthrough KEGG pathway analysis of these related genes, wefound that CDC20-related genes are mainly enriched in thecell cycle, oocyte meiosis, and progesterone-mediated oocytematuration (Figure 6(i)). We found that CDC20 has a strongcoexpression relationship with BUB1, BUB3, BUB1B,CCNA2, CCNB1,MAD2L1, PLK1, and PTTG1 (Figure 6(j)).
3.7. e Biological Role of ASPM in Tumours. To investigatewhether the ASPM gene acts as an oncogene in other tu-mours, we analysed the differential expression of ASPM indifferent tumours and normal tissues via the TIMERwebsite.We found that ASPM is upregulated in a variety of tumours,
Table 1: A total of 291 DEGs were identified from the TCGA and GEO datasets, including 50 upregulated and 241 downregulated genes inthe comparison of BLAC tissues with normal tissues.
DEGs Genes name
Upregulated genes
MTFP1, IQGAP3, ESM1, FASN, CDC20, HILPDA, PAFAH1B3, ETV4, TTK, PODXL2, NUSAP1, TPX2,CENPF, CDT1, AURKB, KIF20A, SAPCD2, RAD54 L, KIF2C, HJURP, DTL, TROAP, TOP2A, NCAPG,ASPM, AURKA, PRC1, TK1, SYNE4, CDCA5, CA9, CDCA3, PFKFB4, SPAG5, TRIP13, ASF1B,
CELSR3, UHRF1, TMEM74B, CCNB2, POLQ, CEP55, IGSF9, TACC3,WDR72, ISG15, PRSS8, TNNT1,MMP1, TCN1
Downregulatedgenes
MOXD1, NDNF, ABCA8, SRPX, FGF9, FAM107A, MFAP4, FOXF1, OLFM1, TCF21, CFD, SCARA5,PRAC1, PAMR1, FCER1A, CPED1, ADAMTS8, COL16A1, ASPA, SPON1, OLFML3, DCN, FGL2,COLEC12, TMEM119, PDGFC, DIXDC1, GLT8D2, DPT, RERGL, TCEAL2, MYH11, MRGPRF,
SLC9A9, SDPR, GFRA1, BMP5, SMOC2, ALDH1A1, SPARCL1, ABI3BP, CNRIP1, EVA1C, PDGFD,ITM2A, CRISPLD2, GHR, CDH11, ADAMTS1, RERG, COX7A1, FLNC, HSPB6, PLAC9, TMOD1,OLFML1, HSD17B6, CYBRD1, SLIT2, JAM3, EMILIN1, CNN1, FHL1, ACTG2, LUM, BIN1, EGR2,PELI2, MAMDC2, CLIP3, ENPP2, ZEB2, RASL12, ITGA8, ACTA2, SORBS2, STON1, PDLIM3,CXCL12, LTBP4, C2orf40, NBEA, GPR183, GATA5, RNASE4, ANTXR2, SGCE, PLA2G4C, PAM,
ZNF521, TSHZ3, PALLD, PGM5, ACOX2, NR2F1, EDNRA, PTGS1, ROR2, GYPC, TGFBR2, LHFP,PARM1, C1S, RGL1, ALDH2, ATP1A2, RGS1, WLS, TAGLN, CRYAB, KCNMB1, FZD7, PRUNE2,SERPINF1, SORBS1, MSRB3, SYNPO2, LMOD1, LPP, PTGIS, MAOB, DPYSL2, BOC, EMP3, SELM,PLSCR4, KLF9, DKK3, VIM, DES, RGS2, SYNM, PCP4, PRICKLE2, GAS6, JAZF1, PLA2G4A, CALD1,PDK4, HAND2-AS1, COL6A2, C7, ZCCHC24, ACTC1, GSTM5, AEBP1, TGFB3, FILIP1L, P2RX1,FXYD6, DDR2, RNF150, TIMP2, SH3GL2, WFDC1, BNC2, A2M, TPM2, CASQ2, DACT3, TCF4,TPM1, RARRES2, GLIPR2, DACT1, AXL, CAV1, MAPRE2, NDN, FERMT2, PRICKLE1, RBPMS2,PTRF, CPVL, CTGF, TNFAIP8L3, PTGDS, DOCK11, ACKR1, PCOLCE2, PRRT2, FAM162 B, HDC,
CPE, EGR1, ZBTB16, EPDR1, SBSPON, DPYSL3, MYL9, CPXM2, COL6A3, CSRP1, HOXA13,MYOM1, FOS, MGP, DUSP1, ANGPTL2, AQP1, IGFBP6, NFIA, MFAP5, CCL19, EPB41L3, SFRP2,MAP1B, GAS1, CAP2, CCL2, FLNA, DKK1, C8orf4, C3, NFIB, CLIC4, AGR3, RASD1, ZFP36, NEXN,TGFB1I1, FOSB, FBLN2, PPP1R14A, PRKCDBP, TUBB6, REEP1, C11orf96, FAM129A, SMTN, APOD,
CYR61, SERPINA3, HSPB8, CKB, IGFBP2, SFRP1, ITGA5, PTGS2, NUPR1, AHNAK2
BioMed Research International 5
M5884: naba core matrisomeGO:0006936: muscle contractionGO:0097435: supramolecular fibre organization
R-HSA-1566948: elastic fibre formation
GO:0061061: muscle structure developmentGO:0048729: tissue morphogenesisGO:0031589: cell-substrate adhesion
0.0 2.5 5.0 7.5 10.0–Log10(P)
12.5 15.0 17.5 20.0
GO:0070848: response to growth factorGO:0051301: cell divisionGO:0000070: mitotic sister chromatid segregationGO:0001568: blood vessel developmentGO:0048545: response to steroid hormoneGO:0042692: muscle cell differentiationGO:0009611: response to woundingGO:0006937: regulation of muscle contractionGO:0016055: Wnt signaling pathwayGO:0055123: digestive system developmentGO:0061448: connective tissue developmentGO:0007507: heart development
GO:0060322: head development
(a)
NABA core matrisomemuscle contractionsupramolecular fiber organizationmuscle structure developmenttissue morphogenesiscell-substrate adhesionresponse to growth factorcell divisionmitotic sister chromatid segregationblood vessel developmentresponse to steroid hormonemuscle cell differentiationresponse to woundingregulation of muscle contractionWnt signaling pathwaydigestive system developmentconnective tissue developmentheart development
head developmentElastic fibre formation
(b)
10–2
10–3
10–4
10–6
10–10
10–20
(c)
MCODE1MCODE2MCODE3MCODE4MCODE5
(d)
MCODE1MCODE2MCODE3MCODE4MCODE5
(e) (f )
Figure 2: Continued.
6 BioMed Research International
including BLCA, BRCA, CHOL, COAD, ESCA, HNSC,KICH, KIRC, KIRP, LIHC, LUAD, LUSC, PPAD, READ,STAD, THCA, and UCEC (Figure 7(a)). .is suggests thatthe role of ASPM in regulating the underlying mechanismsof tumorigenesis and progression is identical in differenttumours. We found that ASPM is highly expressed in BLCAvia the UALCAN website (Figure 7(b)). We also found thatASPM has differential expression in patients with differentraces (Figure 7(c)), weights (Figure 7(d)), smoking habits(Figure 7(e)), and histological subtypes (Figure 7(f)). Toexplore the underlying molecular mechanisms of ASPM, we
first identified genes that have a protein-protein interactionwith ASPM via the STRING website (Figure 7(h)). .en,through the KEGG pathway analysis of these related genes,we found that ASPM-related genes are mainly enriched inthe cell cycle (Figure 7(i)). We found that ASPM has a strongcoexpression relationship with BUB1, CCNA2, CDC20,CDK1, and TTK (Figure 7(j)).
3.8. CDC20 and ASPM Act as Immune-Related Genes inBLCA. To investigate the relationship between CDC20,
GO: 0048545~response to steroid hormone stimulus
GO: 0043627~response to estrogen stimulus
GO: 0031960~response to corticosteroid stimulus
GO: 0022610~biological adhesion
GO: 0010033~response to organic substance
GO: 0009725~response to hormone stimulus
GO: 0009719~response to endogenous stimulus
GO: 0007155~cell adhesion
GO: 0007010~cytoskeleton organization
GO: 0000279~M phase
0e+
00
1e–
06
2e–
06
Pathway enrichment
Count15202530
9
8
7
6
–log10(P value)
P value
3e–
06
Path
way
nam
e
(g)
Pathway enrichment
Count
204060
910
876
–log10(P value)
0.0e
+00
2.5e
–06
5.0e
–06
P value
7.5e
–06
Path
way
nam
e
G0: 0044421~extracellular region part
G0: 0044449~contractile fiber part
GO: 0031012~extracellular matrix
GO: 0043292~contractile fiber
GO: 0030017~sarcomere
GO: 0030016~myofibril
GO: 0015629~actin cytoskeleton
GO: 0005856~cytoskeleton
GO: 0005578~proteinaceous extracellular matrix
GO: 0005576~extracellular region
(h)Pathway enrichment
Count
1020
6
5
4
3
–log10(P value)
0.00
000
0.00
025
0.00
050
0.00
075
0.00
100
0.00
125
Path
way
nam
e
GO: 0030247~polysaccharide binding
GO: 0019838~growth factor binding
GO: 0008307~structural constituent of muscle
GO: 0008092~cytoskeletal protein binding
GO: 0005539~glycosaminoglycan binding
GO: 0005523~tropomyosin binding
GO: 0005520~insulin-like growth factor binding
GO: 0005198~structural molecule activity
GO: 0003779~actin binding
GO: 0001871~pattern binding
P value
(i)
Pathway enrichmentCount
45678910
–log10(P value)
0.00
0.01
0.02
0.03
0.04
3.0
2.5
2.0
1.5
Path
way
nam
e
hsa05414: dilated cardiomyopathy
hsa05410: hypertrophic cardiomyopathy (HCM)
hsa04610: complement and coagulation cascades
hsa04510: focal adhesion
hsa04270: vascular smooth muscle contraction
hsa04260: cardiac muscle contraction
hsa00590: arachidonic acid metabolism
hsa00340: histidine metabolism
P value
(j)Pathway enrichment
Count
46810
3
2
–log10(P value)
0.00
0
0.02
5
0.05
0
0.07
5
P value
Path
way
nam
e
REACT_6900: signaling in immune system
REACT_649: Phase 1 functionalization
REACT_604: hemostasis
REACT_17044: muscle contraction
REACT_16888: signaling by PDGF
REACT_13433: biological oxidations
(k)
Figure 2: GO analysis and KEGG pathway analysis of DEGs. ((a), (b), (c), (d), (e)) GO analysis and KEGG pathway analysis from theMetascape website. (f ) Illustration of the results of KEGG pathway analysis with the Clugo plugin in Cytoscape software. ((g), (h), (i) (j), (k))Bubble diagram of BP, CC,MF, KEGG, and REACTOME analysis for BLCA. Significant pathways with P values< 0.01 were plotted by the Rlanguage.
BioMed Research International 7
ASPM, and tumour immunity, we analysed the relationshipbetween CDC20, ASPM, and immune cell infiltration via theTIMER website. We found that CDC20 is involved in theinfiltration of B cells, CD8+ T cells, and dendritic cells inBLCA (Figure 8(a)). ASPM is involved in the infiltration ofCD8+ T cells, neutrophils, and dendritic cells in BLCA(Figure 8(c)). Since immunotherapy is currently focused on
immunological checkpoint inhibitors such as PDCD1,CD274, PDCD1LG2, TOX, and CTLA4, we further analysedthe coexpression relationship of CDC20, ASPM, and im-mune checkpoint-related genes PDCD1, CD274,PDCD1LG2, TOX, and CTLA4. We were surprised to findthat CDC20 has a significant coexpression relationship withPDCD1, CD274, PDCD1LG2, TOX, and CTLA4
(a)
TTK
TOP2A TROAP
NCAPG
KIF20A
CEP55
HJURP
KIF2C
TK1NUSAP1SPAG5
CENPF
CCNB2
TRIP13
AURKA TPX2
CDC20
ASPMPRC1
AURKB
(b)
TUBB6
ALDH1A1
TROAPTACC3
POLQ
CENPF
PRC1
TTK
NUSAP1
CDCA5
RAD54L
SAPCD2
CENB2
MAPRE2TGFBR2
TK1
IQGAP3EGR1
FHL1
DTL
AURKB
VIM
CDCA3
ASPM
CDC20
KIF2C
TPX2
NCAPG
TRIP13
KIF20A
TOP2A
HJURP
ZBTB16SH3GL2
UHRF1
CEP55
ASF1B
SPAG5
AURKA
DES
CDT1
TPM2
TPM1
GLIPR2
(c)
ASPM
CENPF HJURP
CDCA5 TRIP13
NCAPGUHRF1
CCNB2
KIF20A
TTK
SPAG5
CDC20
TROAP
TOP2ACDCA3
CEP55
RAD54L
NUSAP1
CDT1
ASF1B
(d)
TRIP13CDC20
NCAPG
TROAP
KIF20A
CEP55
AURKB
NUSAP1
HJURPTK1ASPM
KIF2C
SPAG5
CCNB2
CENPF
AURKA
TTK
TOP2A
TPX2
PRC1
(e)
DMNCMCC
DEGREE MNC
0
0
0
0
0
1
0
14
5
0
0
1
5
0
0
(f )
Figure 3: Determination of the hub genes. (a) PPI network of 291 promising target genes in BLCA. ((b), (c), (d), (e)) Four different metrics:DEGREE, MCC, DMNC, and MNC. (f) A Venn diagram was used to identify 14 hub genes in BLCA.
8 BioMed Research International
Expr
essio
n –
log 2 (
TPM
+ 1
)
0
1
2
3
4
5
6∗
BLCA(num(T) = 404; num(N) = 28)
(a)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
8
0
2
4
6
∗
(b)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
∗
2
4
6
8
(c)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
0
2
4
6
8∗
(d)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
0
1
2
3
4
5
6∗
(e)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
0
1
2
3
4
5
6 ∗
(f )
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
0
1
2
3
4
5
6∗
(g)
Expr
essio
n –
log 2 (
TPM
+ 1
)BLCA
(num(T) = 404; num(N) = 28)
0
1
2
3
4
5
6 ∗
(h)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
2
4
6
8 ∗
(i)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
2
4
6
8∗
(j)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
0
2
4
6
8
∗
(k)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
2
4
6
8∗
(l)
Figure 4: Continued.
BioMed Research International 9
(Figure 8(b)). ASPM has a significant coexpression rela-tionship with CD274, PDCD1LG2, and TOX (Figure 8(d)).TOX is a newly discovered gene, and three consecutivearticles published in Nature recently introduced the role ofthe TOX gene in tumour immunotherapy. Fortunately, ourstudy found that CDC20, ASPM, and TOX also have strongcoexpression relationships; the above points lay a very solidfoundation for our future research.
3.9. InVitroCell Experiment. In vitro, we interfered with theexpression of ASPM and CDC20 and then used the cellcounting kit-8 experiment and clone formation experimentto detect the effect on the proliferation of bladder cancer T24cell line. .e above two experimental results show that theproliferation rate of the siASPM and siCDC20 group issignificantly lower than that of the siControl group(Figures 9(a) and 9(b)).
4. Discussion
Bladder cancer (BLCA) is a serious health problemworldwide and the second most common malignant tumourof all genitourinary tract tumours [1]. Unfortunately, untilrecently, the treatment of bladder cancer has progressed verylittle. For 30 years, clinicians have consistently used similar,limited treatments to serve patients. At present, transure-thral resection of bladder tumours is the most commonsurgical procedure for noninvasive bladder cancer, but therecurrence rate is higher [23]. .erefore, there is a veryurgent need to find new therapeutic strategies andbiomarkers.
In this study, we found 291 integrated DEGs in BLCA by acomprehensive analysis of GEO (GSE7476, GSE13507) andTCGA BLCA datasets. .e 291 integrated DEGs were then
subjected to GO (BP, CC and MF) analysis. .e DEG en-richment analysis produced the following terms: steroidhormone stimulus, hormone stimulus, cytoskeleton organi-zation, endogenous stimulus, and oestrogen stimulus (BP);extracellular region, contractile fibre, extracellular region part,contractile fibre part, and actin cytoskeleton (CC); and cy-toskeletal protein binding, structural constituent of muscle,pattern binding, polysaccharide binding, and glycosamino-glycan binding (MF). .ese results indicate that these DEGsare involved in the mitotic process and in the invasion andmetastasis of bladder cancer cells. .e REACTOME pathwayanalysis showed that DEGs are mainly enriched in the sixpathways: muscle contraction, haemostasis, phase 1 func-tionalization, biological oxidation, signalling by PDGF, andsignalling in the immune system..e KEGG pathway analysisshowed that DEGs are mainly enriched in the following sixpathways: vascular smooth muscle contraction, hypertrophiccardiomyopathy (HCM), dilated cardiomyopathy (DCM),focal adhesion, arachidonic acid metabolism, and histidinemetabolism. Two different pathway enrichment algorithmsshowed that DEGs are involved in the process of musclecontraction, and biomechanics play a key role in the devel-opment of bladder cancer. Biomechanics is related to thedeformability of cancer cells, which is involved in cell sig-nalling, cell adhesion, migration, invasion, and metastaticpotential. [24] Biomechanics is a discipline that uses me-chanical theory to study the movement of matter in livingorganisms. .erefore, studying these pathways will helpelucidate the underlying mechanisms of bladder cancer pro-liferation and invasion and help predict cancer progression..is is also related to immune system signalling, furtherconfirming that decreased immune system function is closelyrelated to tumorigenesis [25].
We constructed a PPI network with 291 integrated DEGsand identified the following 14 hub genes: ASPM, CCNB2,
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
0
2
4
6
∗
(m)
Expr
essio
n –
log 2 (
TPM
+ 1
)
BLCA(num(T) = 404; num(N) = 28)
0
1
2
3
4
5
∗
(n)
Figure 4: Expression analysis of 14 hub genes in BLCA based on GEPIA. (a) ASPM, (b) CCNB2, (c) CDC20, (d) CENPF, (e) CEP55, (f )HJURP, (g) KIF20A, (h) NCAPG, (i) NUSAP1, (j) SPAG5, (k) TOP2A, (l) TRIP13, (m) TROAP, (n), and TTK; P< 0.05 was consideredstatistically significant.
10 BioMed Research International
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 274)High expression (n = 132)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(a)
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 322)High expression (n = 84)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(b)
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 143)High expression (n = 263)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(c)
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 193)High expression (n = 213)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(d)1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 225)High expression (n = 181)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(e)
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 184)High expression (n = 222)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(f )
Figure 5: Continued.
BioMed Research International 11
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 271)High expression (n = 135)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(g)
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 261)High expression (n = 145)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(h)1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 281)High expression (n = 125)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(i)
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 301)High expression (n = 105)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(j)1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 183)High expression (n = 223)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(k)
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 126)High expression (n = 280)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(l)
Figure 5: Continued.
12 BioMed Research International
CDC20, CENPF, CEP55, HJURP, KIF20A, NCAPG,NUSAP1, SPAG5, TOP2A, TRIP13, TROAP, and TTK.Most of these factors affect the occurrence and developmentof cancer mainly by affecting the cell cycle. .ese hub genescan be used as therapeutic targets for bladder cancer. Wethen performed a prognostic analysis of these 14 hub genesusing the GEPIA and Human Protein Atlas websites. Sur-prisingly, the expression levels of CDC20 and ASPM areassociated with the prognosis of patients with bladdercancer. .erefore, we chose CDC20 and ASPM as targetgenes for this study.
CDC20 is associated with the meiotic cell cycle in oo-cytes, and APC is associated with the regulation of cell cycle.Human CDC20 is a homologue of the CDC20 protein and isgenerally considered to be one of the major regulators ofmitosis. It is responsible for activating the APC/C complex,an E3 ubiquitin ligase that targets mitotic proteins such assecurin and cyclin B for 26S proteasome degradation,allowing cells to exit mitosis [26]. Recent literature hasreported that CDC20 is highly expressed in pancreaticcancer [27], colon cancer [28], osteosarcoma cancer [29],lung adenocarcinoma [30], oral squamous cell carcinoma,and hepatocellular carcinoma [31, 32]. Consistent with ourresearch, increased expression of CDC20 in bladder cancerpatients is associated with poor prognosis [33]. It has beenreported that CDC20 may act as an oncoprotein to promotethe progression and development of human cancer and thatit is a promising therapeutic target [34]. We found thatCDC20 is highly expressed in BLCA via the UALCANwebsite. We also found that CDC20 is differentiallyexpressed in patients with different smoking habits, histo-logical subtypes, and molecular subtypes. In addition,CDC20 is associated with promoter methylation levels inBLCA. To investigate the relationship between CDC20 and
tumour immunity, we analysed the relationship betweenCDC20 and immune cell infiltration via the TIMER website,and we found that CDC20 is involved in the infiltration ofB cells, CD8+ T cells, and dendritic cells in BLCA. We weresurprised to find that CDC20 has a significant coexpressionrelationship with PDCD1, CD274, PDCD1LG2, TOX, andCTLA4.
ASPM is involved in the spindle tissue, spindle locali-zation, and cytokinesis of all dividing cells, and the extremeC-terminus of the protein is required for ASPM localizationand function. In addition, it may have a role in regulatingneurogenesis [35]..e purpose of this study was to assess thecritical role that ASPM plays in the development of cancerand determine whether it can be used as a biomarker forbladder cancer. Recent literature reports that ASPM ex-pression disorders are associated with the progression ofepithelial ovarian cancer [36], colorectal cancer [37],prostate cancer [38], and hepatocellular carcinoma [39].Consistent with our research, increased expression of ASPMin bladder cancer patients is associated with poor prognosis[40]. We found that ASPM is highly expressed in BLCA viathe UALCAN website. We also found that ASPM is dif-ferentially expressed in patients with different races, weights,smoking habits, and histological subtypes. To investigate therelationship between ASPM and tumour immunity, weanalysed the relationship between ASPM and immune cellinfiltration via the TIMER website, and we found that ASPMis involved in the infiltration of CD8+ T cells, neutrophils,and dendritic cells in BLCA. We were surprised to find thatASPM has a significant coexpression relationship withCD274, PDCD1LG2, and TOX.
Among these immune checkpoints, it is worth notingthat a molecule called TOX is a newly discovered gene, andthree consecutive articles published in Nature recently
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 115)High expression (n = 291)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(m)
1.0
0.90.80.70.60.50.40.30.20.1
0
Surv
ival
pro
babi
lity
0 1 2 3 4 5 6Time (years)
7 8 9 10 11 12 13
Low expression (n = 277)High expression (n = 129)
Dead (n = 179)Alive (n = 227)Density-deadDensity-aliveDensity-dead under cutoffDensity-dead over cutoff
(n)
Figure 5: Survival analysis of the 14 hub genes in BLCA based on the Human Protein Atlas. (a) ASPM, (b) CCNB2, (c) CDC20, (d) CENPF,(e) CEP55, (f ) HJURP, (g) KIF20A, (h) NCAPG, (i) NUSAP1, (j) SPAG5, (k) TOP2A, (l) TRIP13, (m) TROAP, (n), and TTK; P< 0.05 wasconsidered statistically significant.
BioMed Research International 13
∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗∗
0
5
10
ACC
.Tum
orBL
CA.T
umor
BLCA
.Nor
mal
BRCA
.Tum
orBR
CA.N
orm
alBR
CA-B
asal
.Tum
orBR
CA-H
er2.
Tum
orBR
CA-L
umin
al.T
umor
CESC
.Tum
orCH
OL.
Tum
orCH
OL.
Nor
mal
COA
D.T
umor
COA
D.N
orm
alD
LBC.
Tum
orES
CA.T
umor
ESCA
.Nor
mal
GBM
.Tum
orH
NSC
.Tum
orH
NSC
.Nor
mal
HN
SC-H
PVpo
s.Tum
orH
NSC
-HPV
neg.
Tum
orKI
CH.T
umor
KICH
.Nor
mal
KIRC
.Tum
orKI
RC.N
orm
alKI
RP.T
umor
KIRP
.Nor
mal
LAM
L.Tu
mor
LGG
.Tum
orLI
HC.
Tum
orLI
HC.
Nor
mal
LUA
D.T
umor
LUA
D.N
orm
alLU
SC.T
umor
LUSC
.Nor
mal
MES
O.T
umor
OV
.Tum
orPA
AD
.Tum
orPC
PG.T
umor
PRA
D.T
umor
PRA
D.N
orm
alRE
AD
.Tum
orRE
AD
.Nor
mal
SARC
.Tum
orSK
CM.T
umor
SKCM
.Met
asta
sisST
AD
.Tum
orST
AD
.Nor
mal
TGCT
.Tum
orTH
CA.T
umor
THCA
.Nor
mal
THYM
.Tum
orU
CEC.
Tum
orU
CEC.
Nor
mal
UCS
.Tum
orU
VM
.Tum
or
CDC2
0 ex
pres
sion
leve
l (lo
g 2 RSE
M)
(a)
∗∗∗
250
200
150
100
50
0
–50
Tran
scrip
t per
mill
ion
Normal (n = 19)TCGA samples
Primary tumor (n = 408)
(b)
∗∗∗∗∗∗
∗∗∗
∗∗∗
250
200
150
100
50
0
–50
Tran
scrip
t per
mill
ion
Normal(n = 19)
Non smoker(n = 106)
Smoker(n = 89)
Reformedsmoker
(<15 years)(n = 113)
Reformedsmoker
(>15 years)(n = 69)
TCGA samples
(c)
∗∗∗
∗∗∗
250
200
150
100
50
0
–50
Tran
scrip
t per
mill
ion
Normal (n = 19) Papillary tumor (n = 132)
TCGA samples
Non papillarytumor (n = 271)
(d)
∗∗∗
∗∗∗
∗∗∗
∗∗∗
∗∗∗
250
300
200
150
100
50
0
–50
Tran
scrip
t per
mill
ion
Normal(n = 19)
Neuronal(n = 20)
Luminal(n = 26)
LuminalInfiltrated(n = 78)
LuminalPapillary(n = 142)
Basalsquamous(n = 142)
TCGA samples
(e)
∗∗∗
0.0275
0.05
0.0225
0.02
0.0175
0.015
0.0125
0.01
Beta
val
ue
Normal (n = 21)TCGA samples
Primary tumor (n = 418)
(f )
Figure 6: Continued.
14 BioMed Research International
CDC20BUB1BCDC16CCNB1PLK1BUB3NEK2CCNA2PTTG1BUB1MAD2L1
4%4%5%
1.6%5%
0.8%4%
1.6%1.6%
8%2.4%
Genetic alterationMissense mutation (unknown significance)Truncating mutation (unknown significance)Amplification
Deep deletionNo alterations
5%
Blad
der U
roth
elia
lCa
rcin
oma
10%
15%
Alte
ratio
n fre
quen
cy
20%
25%
MutationAmplificationMultiple alterations
(g)
NEK2 PTTG1
CDC16
PLK1
MAD2L1
BUB3CCNA2
BUB1B
CCNB1
BUB1
CDC20
(h) (i)CDC20 vs. BUB1
mRNA expression (RNA Seq V2 RSEM):CDC20 (log2)
mRNA expression (RNA Seq V2 RSEM):CDC20 (log2)
mRNA expression (RNA Seq V2 RSEM):CDC20 (log2)
mRNA expression (RNA Seq V2 RSEM):CDC20 (log2)
mRNA expression (RNA Seq V2 RSEM):CDC20 (log2)
mRNA expression (RNA Seq V2 RSEM):CDC20 (log2)
mRNA expression (RNA Seq V2 RSEM):CDC20 (log2)
mRNA expression (RNA Seq V2 RSEM):CDC20 (log2)
8 9 10 11 12 13 14
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)m
RNA
expr
essio
n (R
NA
Seq
V2
RSEM
):BU
B1 (l
og2)
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)m
RNA
expr
essio
n (R
NA
Seq
V2
RSEM
):BU
B1 (l
og2)
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)m
RNA
expr
essio
n (R
NA
Seq
V2
RSEM
):BU
B1 (l
og2)
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)m
RNA
expr
essio
n (R
NA
Seq
V2
RSEM
):BU
B1 (l
og2)
77.5
88.5
99.510
10.511
11.512
y = 0.64x + 2.85R2 = 0.54
Spearman: 0.68(p = 1.80e – 18)Pearson: 0.73
(p = 2.12e – 22)
CDC20 vs. BUB3
8 9 10 11 12 13 149.5
10
10.5
11
11.5
12
12.5
13
13.5
y = 0.11x + 9.97R2 = 0.06
Spearman: 0.28(p = 1.465e – 3)
Pearson: 0.25(p = 5.323e – 3)
CDC20 vs. BUB1B
8 9 10 11 12 13 14
6
7
8
9
10
11y = 0.57x + 3.12
R2 = 0.54
Spearman: 0.67(p = 8.83e – 18)Pearson: 0.73
(p = 2.11e – 22)
CDC20 vs. CCNA2
8 9 10 11 12 13 14
7
8
9
10
11
12y = 0.69x + 2.33
R2 = 0.55
Spearman: 0.69(p = 3.50e – 19)Pearson: 0.74
(p = 4.91e – 23)
CDC20 vs. CCNB1
8 9 10 11 12 13 14
88.5
99.510
10.511
11.512
12.513
y = 0.65x + 3.64R2 = 0.56
Spearman: 0.69(p = 2.93e – 19)
Pearson: 0.75(p = 5.52e – 24)
CDC20 vs. MAD2L1
8 9 10 11 12 13 14
6
7
8
9
10
11 y = 0.65x + 2.16R2 = 0.44
Spearman: 0.63(p = 3.46e – 15)
Pearson: 0.66(p = 3.93e – 17)
CDC20 vs. PLK1
8 9 10 11 12 13 14
7
8
9
10
11
12
13
y = 0.78x + 1.77R2 = 0.6
Spearman: 0.75(p = 2.74e – 24)
Pearson: 0.78(p = 1.53e – 26)
CDC20 vs. PTTG1
8 9 10 11 12 13 14
7
8
9
10
11
12y = 0.69x + 2.21
R2 = 0.57
Spearman: 0.72(p = 1.85e – 21)
Pearson: 0.75(p = 2.74e – 24)
(j)
Figure 6: .e biological role of CDC20 in tumours. (a) Expression of CDC20 in various tumours. (b) Expression of CDC20 in BLCA basedon sample type. (c) Expression of CDC20 in BLCA based on patient smoking habits. (d) Expression of CDC20 in BLCA based on histologicalsubtype. (e) Expression of CDC20 based on molecular subtypes of BLCA. (f ) .e promoter methylation level of CDC20 in BLCA.(g) Variation of CDC20-related genes in BLCA. (h) Interacting proteins for the CDC20 gene STRING interaction network preview (showingtop 10 STRING interactants). (i) Illustration of the results of KEGG pathway analysis with the Clugo plugin in Cytoscape software. (j) Ascatter plot showing the correlation between CDC20 expression and the 8 hub gene signature. ∗P< 0.05, ∗∗P< 0.01, ∗∗∗P< 0.001.
BioMed Research International 15
ACC
.Tum
orBL
CA.T
umor
BLCA
.Nor
mal
BRCA
.Tum
orBR
CA.N
orm
alBR
CA-B
asal
.Tum
orBR
CA-H
er2.
Tum
orBR
CA-L
umin
al.T
umor
CESC
.Tum
orCH
OL.
Tum
orCH
OL.
Nor
mal
COA
D.T
umor
COA
D.N
orm
alD
LBC.
Tum
orES
CA.T
umor
ESCA
.Nor
mal
GBM
.Tum
orH
NSC
.Tum
orH
NSC
.Nor
mal
HN
SC-H
PVpo
s.Tum
orH
NSC
-HPV
neg.
Tum
orKI
CH.T
umor
KICH
.Nor
mal
KIRC
.Tum
orKI
RC.N
orm
alKI
RP.T
umor
KIRP
.Nor
mal
LAM
L.Tu
mor
LGG
.Tum
orLI
HC.
Tum
orLI
HC.
Nor
mal
LUA
D.T
umor
LUA
D.N
orm
alLU
SC.T
umor
LUSC
.Nor
mal
MES
O.T
umor
OV
.Tum
orPA
AD
.Tum
orPC
PG.T
umor
PRA
D.T
umor
PRA
D.N
orm
alRE
AD
.Tum
orRE
AD
.Nor
mal
SARC
.Tum
orSK
CM.T
umor
SKCM
.Met
asta
sisST
AD
.Tum
orST
AD
.Nor
mal
TGCT
.Tum
orTH
CA.T
umor
THCA
.Nor
mal
THYM
.Tum
orU
CEC.
Tum
orU
CEC.
Nor
mal
UCS
.Tum
orU
VM
.Tum
or
0
5
10
ASP
M ex
pres
sion
leve
l (lo
g 2 RSE
M)
∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗
(a)
∗∗∗
Normal (n = 19)TCGA samples
Primary tumor (n = 408)
25
20
15
10
5
0
–5
Tran
scrip
t per
mill
ion
(b)
Normal(n = 19)
Caucasian(n = 320)
African-american(n = 22)
TCGA samples
Asian(n = 44)
25
20
15
10
5
0
–5
Tran
scrip
t per
mill
ion
∗∗∗
∗∗∗
∗∗
(c)
Normal(n = 19)
Normal weight(n = 136)
Extreme weight(n = 124)
Extreme obese(n = 10)
Obese(n = 75)
TCGA samples
25
20
15
10
5
0
–5
Tran
scrip
t per
mill
ion
∗∗∗
∗∗∗
∗∗∗
∗
(d)25
20
15
10
5
0
–5
Tran
scrip
t per
mill
ion ∗∗∗
∗∗∗
∗∗∗
∗∗∗
Normal(n = 19)
Non smoker(n = 106)
Smoker(n = 89)
Reformedsmoker
(<15 years)(n = 113)
Reformedsmoker
(>15 years)(n = 69)
TCGA samples
(e)
Normal(n = 19)
25
20
15
10
5
0
–5
Tran
scrip
t per
mill
ion
∗∗∗
∗∗∗
TCGA samples
Papillary tumor(n = 132)
Non Papillary tumor(n = 271)
(f )
Figure 7: Continued.
16 BioMed Research International
DLGAP5
CDC20
ASPM
CDK1
KIF23
KIF11
NCAPG
TTK
CCNA2
BUB1
CENPE
4%
4%
9%
3%
4%
2.4%
2.4%
4%
1.6%
8%
4%
Genetic alterationMissense mutation (unknown significance)Truncating mutation (unknown significance)Amplification
Deep deletionNo alterations
Blad
der U
roth
elia
lCa
rcin
oma
MutationAmplificationDeep deletionMultiple alterations
10%
Alte
ratio
n fre
quen
cy
20%
30%
(g)
CCNA2 CENPE
CCD20
KIF23
DLGAP5
ASPM
CDK1NCAPG
BUB1
KIF11
TTK
(h) (i)ASPM vs. BUB1 ASPM vs. CCNA2
1
ASPM vs. CDC20 ASPM vs. CDK1 ASPM vs. TTK
y = 0.68x + 3.18R2 = 0.31
y = 0.6x + 3.92R2 = 0.54
y = 0.68x + 4.09R2 = 0.52
y = 0.56x + 4.89R2 = 0.51
y = 0.6x + 3.09R2 = 0.53
6 7 8 9 10 11 12 13mRNA expression (RNA Seq V2 RSEM):
ASPM (log2)
6 7 8 9 10 11 12 13mRNA expression (RNA Seq V2 RSEM):
ASPM (log2)
7
8
9
10
11
12
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)
7
8
9
10
11
12
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)
5
6
7
8
9
10
11
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)
8
9
10
11
12
13
14
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)
7
8
9
10
11
12
mRN
A ex
pres
sion
(RN
A S
eq V
2 RS
EM):
BUB1
(log
2)
Spearman: 0.66(p = 4.54e – 17)Pearson: 0.74
(p = 1.12e – 22)
ASPM mutatedBUB1 mutatedNeither mutated
6 7 8 9 10 11 12 13mRNA expression (RNA Seq V2 RSEM):
ASPM (log2)
ASPM mutatedTTK mutatedNeither mutated
ASPM mutatedNeither mutated
6 7 8 9 10 11 12 13mRNA expression (RNA Seq V2 RSEM):
ASPM (log2)ASPM mutatedNeither mutated
6 7 8 9 10 11 12 13mRNA expression (RNA Seq V2 RSEM):
ASPM (log2)ASPM mutatedNeither mutated
Spearman: 0.74(p = 5.61e – 23)Pearson: 0.78
(p = 5.75e – 27)
Spearman: 0.65(p = 1.36e – 16)Pearson: 0.72
(p = 1.41e – 21)
Spearman: 0.67(p = 8.68e – 18)
Pearson: 0.72(p = 5.13e – 21)
Spearman: 0.72(p = 3.10e – 21)
Pearson: 0.73(p = 5.72e – 22)
(j)
Figure 7: .e biological role of ASPM in tumours. (a) Expression of ASPM in various tumours. (b) Expression of ASPM in BLCA based onsample type. (c) Expression of ASPM in BLCA based on the patient race. (d) Expression of ASPM in BLCA based on patient weight.(e) Expression of ASPM in BLCA based on patient smoking habits. (f ) Expression of CDC20 in BLCA based on histological subtype.(g) Variation of ASPM-related genes in BLCA. (h) Interacting proteins for the ASPM gene STRING interaction network preview (showingtop 10 STRING interactants). (i) Illustration of the results of KEGG pathway analysis with the Clugo plugin in Cytoscape software. (j) Ascatter plot showing the correlation between ASPM expression and the 5 hub gene signature. ∗P< 0.05, ∗∗P< 0.01, ∗∗∗P< 0.001.
BioMed Research International 17
p = 7.19e – 02 p = 4.32e – 03 p = 1.65e – 04 p = 6.55e – 01 p = 2.83e – 01 p = 6.41e – 02 p = 2.66e – 10partial.cor = –0.149cor = –0.094 partial.cor = 0.916 partial.cor = –0.023 partial.cor = –0.056 partial.cor = 0.323partial.cor = 0.097
Purity B Cell CD8 + T Cell CD4 + T Cell Macrophage Neutrophil Dendritic Cell
0.25 0.50 0.75 1.000.25 0.50 0.75 1.00 1.250.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.1 0.2 0.30.40.0 0.2 0.40.5Infiltration Level
12
10
8
6
CDC2
0 Ex
pres
sion
Leve
l (lo
g 2 RSE
M)
(a)
p = 1.32e – 11 p = 1.53e – 04p = 1.95e – 02p = 8.84e – 11p = 5.81e – 04
Expr
essio
n Le
vel (
log 2 R
SEM
)
Expr
essio
n Le
vel (
log 2 R
SEM
)
Expr
essio
n Le
vel (
log 2 R
SEM
)
Expr
essio
n Le
vel (
log 2 R
SEM
)10.0
7.5
5.0
2.5
0.0 Expr
essio
n Le
vel (
log 2 R
SEM
) 10.0
7.5
5.0
2.5
0.06 8 10 12
Expression Level (log2 RSEM)6 8 10 12
Expression Level (log2 RSEM)6 8 10 12
Expression Level (log2 RSEM)6 8 10 12
Expression Level (log2 RSEM)6 8 10 12
Expression Level (log2 RSEM)
PDCD
1
CDC20 CDC20 CDC20 CDC20 CDC20
CD27
4
PDCD
1LG
2
TOX
CTLA
4
0
3
6
9
0
4
8
0
3
6
9
cor = 0.17 cor = 0.327 cor = 0.314 cor = 0.116 cor = 0.186
(b)
Purity B Cell CD8 + T Cell CD4 + T Cell Macrophage Neurophil Dendritie Cell
0.25 0.50 0.75 1.000.25 0.50 0.75 1.00 1.250.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.1 0.2 0.30.40.0 0.2 0.40.5Infiltration Level
12
10
8
6
4ASP
M E
xpre
ssio
n Le
vel (
log 2 R
SEM
)
cor = 0.046p = 3.75e − 01
partial.cor = −0.069p = 1.88e − 01
partial.cor = 0.337p = 3.43e − 11
partial.cor = 0.014p = 7.95e − 01
partial.cor = 0.037p = 4.76e − 01
partial.cor = 0.211p = 5.05e − 05
partial.cor = 0.339p = 2.73e − 11
BLCA
(c)
Expr
essio
n Le
vel (
log 2 R
SEM
)
64 8 10 12Expression Level (log2 RSEM)
64 8 10 12Expression Level (log2 RSEM)
64 8 10 12Expression Level (log2 RSEM)
cor = 0.334p = 4.22e − 12
cor = 0.181p = 2.34e − 04
cor = 0.21p = 1.82e − 05
0
3
6
9
Expr
essio
n Le
vel (
log 2 R
ESEM
)
0
3
6
9
Expr
essio
n Le
vel (
log 2 R
ESEM
)
0
3
6
9
ASPM ASPM ASPM
CD27
4
PDCD
1LG
2
TOX
(d)
Figure 8: Tumour immune correlation analysis based on the TIMER website. (a) Relationship between CDC20 expression and immunecells. (b) Relationship between CDC20 expression and immune checkpoints. (c) Relationship between ASPM expression and immune cells.(d) Relationship between ASPM expression and immune checkpoints.
0 24 48 72 96Time (h)
2.0
1.5
1.0
0.5
0.0
OD
(450
nm)
NCSi-ASPMSi-CDC20
(a)
NC Si-ASPM Si-CDC20
(b)
Figure 9: Experimental validation of ASPM and CDC20. (a) Cell Counting Kit-8 (CCK8) assay. (b) Clone formation assay. ∗∗∗P< 0.001.Student’s t-tests were used to evaluate the statistical significance of differences.
18 BioMed Research International
introduced the role of the TOX gene in tumour immuno-therapy [41–43]. TOX is a key regulator of Tcell dysfunction.TOX is specifically required for T cell differentiation in anenvironment of chronic antigen stimulation (e.g., tumoursand chronic infections) [41]. Manipulation of TOX ex-pression is thought to calibrate T cells to maintain theireffector function in cell differentiation and ultimatelyachieve long-lasting therapeutic outcomes [42, 43]. Webelieve that CDC20 and ASPM may affect the developmentof bladder cancer by affecting TOX molecules.
In addition, we interfered with the expression of ASPMand CDC20 in vitro and then used the cell counting kit-8experiment and clone formation experiment to detect theeffect on the proliferation of bladder cancer T24 cell line..eabove two experimental results show that the proliferationrate of the siASPM and siCDC20 group is significantly lowerthan that of the siControl group. In vitro experimentsfurther support our conclusions.
5. Conclusions
In summary, our current study reveals two potential bio-markers for BLCA, CDC20, and ASPM. Both of these genesmay be new immunotherapeutic targets for BLCA. In ad-dition, the biological role of the ASPM and CDC20 mole-cules in the development of bladder cancer was confirmed byin vitro experiments. We continue to study the underlyingmechanisms by using bioinformatics. .e findings of thisstudy may pave the way for the identification of otherfunctions in vitro and in vivo by our group and others.
Data Availability
.e data used to support the findings of this study areavailable from the corresponding author upon request.
Conflicts of Interest
.e authors declare that they have no conflicts of interest.
Authors’ Contributions
Yingkun Xu and Guangzhen Wu contributed equally to thiswork. Qinghua Xia, Yingkun Xu, and Guangzhen Wuconceived and designed the study; Yingkun Xu performedthe experiments; Yingkun Xu, Jianyi Li, Jiatong Li, Liye Ma,and Liang Li collected and analysed the data; Xiaoyang Han,Yanjun Wei, Hongge Zhang, and Yougen Chen contributedthe analysis tools; Yingkun Xu, Jianyi Li, and Ningke Ruanwrote the draft. .e final draft was read and approved by allauthors.
Acknowledgments
.e authors would like to thank the National Institutes ofHealth (NIH) for developing the TCGA and GEO databases..is work was funded by a grant from the National NaturalScience Foundation of China (grant serial number:81672553). In addition, Yingkun Xu wants to thank, in
particular, the patience, care, and support from Jiayi Li overthe passed time.
References
[1] A. Jemal, F. Bray, M. M. Center, J. Ferlay, E. Ward, andD. Forman, “Global cancer statistics,” CA: A Cancer Journalfor Clinicians, vol. 61, no. 2, pp. 69–90, 2011.
[2] S. Antoni, J. Ferlay, I. Soerjomataram, A. Znaor, A. Jemal, andF. Bray, “Bladder cancer incidence and mortality: a globaloverview and recent trends,” European Urology, vol. 71, no. 1,pp. 96–108, 2017.
[3] W. Li, Y. Huang, D. Sargsyan et al., “Epigenetic alterations inTRAMP mice: epigenome DNA methylation profiling usingMeDIP-seq,” Cell & Bioscience, vol. 8, no. 1, 2018.
[4] M. Babjuk, M. Burger, R. Zigeuner et al., “EAU guidelines onnon-muscle-invasive urothelial carcinoma of the bladder:update 2013,” European Urology, vol. 64, no. 4, pp. 639–653,2013.
[5] F. Millan-Rodrıguez, G. Chechile-Toniolo, J. Salvador-Bayarri, J. Palou, F. Algaba, and J. Vicente-Rodrıguez, “Pri-mary superficial bladder cancer risk groups according toprogression, mortality and recurrence,” Journal of Urology,vol. 164, no. 3 Part 1, pp. 680–684, 2000.
[6] P. W. Kantoff, C. S. Higano, N. D. Shore et al., “Sipuleucel-Timmunotherapy for castration-resistant prostate cancer,” eNew England Journal of Medicine, vol. 363, no. 5, pp. 411–422,2010.
[7] F. Stephen Hodi, S. J. O’Day, D. F. McDermott et al., “Im-proved survival with ipilimumab in patients with metastaticmelanoma,” e New England Journal of Medicine, vol. 363,no. 8, pp. 711–723, 2010.
[8] J. W. Kim, Y. Tomita, J. Trepel, and A. B. Apolo, “Emergingimmunotherapies for bladder cancer,” Current Opinion inOncology, vol. 27, no. 3, pp. 191–200, 2015.
[9] S. Brandau and H. Suttmann, “.irty years of BCG immu-notherapy for non-muscle invasive bladder cancer: a successstory with room for improvement,” Biomedicine & Phar-macotherapy, vol. 61, no. 6, pp. 299–305, 2007.
[10] D. S. Chen and I. Mellman, “Oncology meets immunology:the cancer-immunity cycle,” Immunity, vol. 39, no. 1, pp. 1–10,2013.
[11] P. A. Ott, F. S. Hodi, and C. Robert, “CTLA-4 and PD-1/PD-L1 blockade: new immunotherapeutic modalities with durableclinical benefit in melanoma patients,” Clinical Cancer Re-search: An Official Journal of the American Association forCancer Research, vol. 19, no. 19, pp. 5300–5309, 2013.
[12] T. Barrett, S. E. Wilhite, P. Ledoux et al., “NCBI GEO: archivefor functional genomics data sets—update,” Nucleic AcidsResearch, vol. 41, no. D1, pp. D991–D995, 2012.
[13] W.-J. Kim, E.-J. Kim, S.-K. Kim et al., “Predictive value ofprogression-related gene classifier in primary non-muscleinvasive bladder cancer,” Molecular Cancer, vol. 9, no. 1,3 pages, 2010.
[14] 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.
[15] Y. Zhou, B. Zhou, L. Pache et al., “Metascape provides abiologist-oriented resource for the analysis of systems-leveldatasets,” Nature Communications, vol. 10, no. 1, 2019.
[16] D. W. Huang, B. T. Sherman, and R. A. Lempicki, “Bio-informatics enrichment tools: paths toward the
BioMed Research International 19
comprehensive functional analysis of large gene lists,” NucleicAcids Research, vol. 37, no. 1, pp. 1–13, 2009.
[17] D. Szklarczyk, A. L. Gable, D. Lyon et al., “STRING v11:protein-protein association networks with increased coverage,supporting functional discovery in genome-wide experi-mental datasets,” Nucleic Acids Research, vol. 47, no. D1,pp. D607–D613, 2019.
[18] P. Shannon, A. Markiel, O. Ozier et al., “Cytoscape: a softwareenvironment for integrated models of biomolecular interac-tion networks,” Genome Research, vol. 13, no. 11,pp. 2498–2504, 2003.
[19] 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, no. Suppl 4, S11 pages, 2014.
[20] M. Uhlen, C. Zhang, S. Lee et al., “A pathology atlas of thehuman cancer transcriptome,” Science, vol. 357, no. 6352,Article ID eaan2507, 2017.
[21] E. Cerami, J. Gao, U. Dogrusoz et al., “.e cBio cancer ge-nomics portal: an open platform for exploring multidimen-sional cancer genomics data,” Cancer Discovery, vol. 2, no. 5,pp. 401–404, 2012.
[22] T. Li, J. Fan, B. Wang et al., “TIMER: a web server forcomprehensive analysis of tumor-infiltrating immune cells,”Cancer Research, vol. 77, no. 21, pp. e108–e110, 2017.
[23] A. J. Lightfoot, B. N. Breyer, H. M. Rosevear, B. A. Erickson,B. R. Konety, and M. A. O’Donnell, “Multi-institutionalanalysis of sequential intravesical gemcitabine and mitomycinC chemotherapy for non–muscle invasive bladder cancer,”Urologic Oncology: Seminars and Original Investigations,vol. 32, no. 1, pp. 35.e15–35.e19, 2014.
[24] S. Suresh, “Biomechanics and biophysics of cancer cells,” ActaBiomaterialia, vol. 3, no. 4, pp. 413–438, 2007.
[25] A. B. Bigley, G. Spielmann, E. C. P. LaVoy, and R. J. Simpson,“Can exercise-related improvements in immunity influencecancer prevention and prognosis in the elderly?” Maturitas,vol. 76, no. 1, pp. 51–56, 2013.
[26] J.-M. Peters, “.e anaphase-promoting complex,” MolecularCell, vol. 9, no. 5, pp. 931–943, 2002.
[27] D. Z. Chang, Y. Ma, B. Ji et al., “Increased CDC20 expressionis associated with pancreatic ductal adenocarcinoma differ-entiation and progression,” Journal of Hematology & On-cology, vol. 5, no. 1, p. 15, 2012.
[28] W.-J. Wu, K.-S. Hu, D.-S. Wang et al., “CDC20 over-expression predicts a poor prognosis for patients with colo-rectal cancer,” Journal of Translational Medicine, vol. 11, no. 1,p. 142, 2013.
[29] G. Shang, X. Ma, and G. Lv, “Cell division cycle 20 promotescell proliferation and invasion and inhibits apoptosis in os-teosarcoma cells,” Cell Cycle, vol. 17, no. 1, pp. 43–52, 2018.
[30] R. Shi, Q. Sun, J. Sun et al., “Cell division cycle 20 over-expression predicts poor prognosis for patients with lungadenocarcinoma,” Tumor Biology, vol. 39, no. 3, 2017.
[31] I. M. B. Moura, M. L. Delgado, P. M. A. Silva et al., “HighCDC20 expression is associated with poor prognosis in oralsquamous cell carcinoma,” Journal of Oral Pathology &Medicine, vol. 43, no. 3, pp. 225–231, 2014.
[32] J. Li, J.-Z. Gao, J.-L. Du, Z.-X. Huang, and L.-X. Wei, “In-creased CDC20 expression is associated with developmentand progression of hepatocellular carcinoma,” InternationalJournal of Oncology, vol. 45, no. 4, pp. 1547–1555, 2014.
[33] J.-W. Choi, Y. Kim, J.-H. Lee, and Y.-S. Kim, “High expressionof spindle assembly checkpoint proteins CDC20 andMAD2 is
associated with poor prognosis in urothelial bladder cancer,”Virchows Archiv, vol. 463, no. 5, pp. 681–687, 2013.
[34] Z. Wang, L. Wan, J. Zhong et al., “CDC20: a potential noveltherapeutic target for cancer treatment,” Current Pharma-ceutical Design, vol. 19, no. 18, pp. 3210–3214, 2013.
[35] J. Higgins, C. Midgley, A.-M. Bergh et al., “Human ASPMparticipates in spindle organisation, spindle orientation andcytokinesis,” BMC Cell Biology, vol. 11, no. 1, p. 85, 2010.
[36] A. Bruning-Richardson, J. Bond, R. Alsiary et al., “ASPM andmicrocephalin expression in epithelial ovarian cancer cor-relates with tumour grade and survival,” British Journal OfCancer, vol. 104, no. 10, pp. 1602–1610, 2011.
[37] N. Ertych, A. Stolz, A. Stenzinger et al., “Increased micro-tubule assembly rates influence chromosomal instability incolorectal cancer cells,” Nature Cell Biology, vol. 16, no. 8,pp. 779–791, 2014.
[38] J.-J. Xie, Y.-J. Zhuo, Y. Zheng et al., “High expression ofASPM correlates with tumor progression and predicts pooroutcome in patients with prostate cancer,” InternationalUrology and Nephrology, vol. 49, no. 5, pp. 817–823, 2017.
[39] S.-Y. Lin, H.-W. Pan, S.-H. Liu et al., “ASPM is a novel markerfor vascular invasion, early recurrence, and poor prognosis ofhepatocellular carcinoma,” Clinical Cancer Research, vol. 14,no. 15, pp. 4814–4820, 2008.
[40] Z. Xu, Q. Zhang, F. Luh, B. Jin, and X. Liu, “Overexpression ofthe ASPM gene is associated with aggressiveness and pooroutcome in bladder cancer,” Oncology Letters, vol. 17, no. 2,pp. 1865–1876, 2019.
[41] A. C. Scott, F. Dundar, P. Zumbo et al., “TOX is a criticalregulator of tumour-specific T cell differentiation,” Nature,vol. 571, no. 7764, pp. 270–274, 2019.
[42] F. Alfei, K. Kanev, M. Hofmann et al., “TOX reinforces thephenotype and longevity of exhausted T cells in chronic viralinfection,” Nature, vol. 571, no. 7764, pp. 265–269, 2019.
[43] O. Khan, J. R. Giles, S. McDonald et al., “TOX transcrip-tionally and epigenetically programs CD8+ Tcell exhaustion,”Nature, vol. 571, no. 7764, pp. 211–218, 2019.
20 BioMed Research International