identification of key genes and mirnas in osteosarcoma...

Research ArticleIdentification of Key Genes and miRNAs in OsteosarcomaPatients with Chemoresistance by Bioinformatics Analysis

Binbin Xie1 Yiran Li1 Rongjie Zhao 1 Yuzi Xu2 Yuhui Wu1 Ji Wang34

Dongdong Xia 5 Weidong Han 1 and Dake Chen 6

1Department of Medical Oncology Sir Run Run Shaw Hospital School of Medicine Zhejiang University Hangzhou China2Department of Sports Medicine School of Medicine Zhejiang University Zhejiang China3Department of Surgical Oncology Sir Run Run Shaw Hospital College of Medicine Zhejiang University HangzhouZhejiang 310016 China4Biomedical Research Center and Key Laboratory of Biotherapy of Zhejiang Province Hangzhou Zhejiang 310016 China5Orthopedic Department Ningbo First Hospital Ningbo 315000 China6Department of Urology Wenzhou Peoplersquos Hospital Wenzhou Zhejiang China

Correspondence should be addressed to Dongdong Xia xiadongdongqqcom Weidong Han hanwdzjueducnand Dake Chen 61502485qqcom

Received 23 November 2017 Revised 21 February 2018 Accepted 4 March 2018 Published 22 April 2018

Academic Editor Jialiang Yang

Copyright copy 2018 Binbin Xie et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Chemoresistance is a significant factor associated with poor outcomes of osteosarcoma patients The present study aims to identifyChemoresistance-regulated gene signatures and microRNAs (miRNAs) in Gene Expression Omnibus (GEO) database The resultsof Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) included positive regulation of transcriptionDNA-templated tryptophan metabolism and the like Then differentially expressed genes (DEGs) were uploaded to Search Toolfor the Retrieval of Interacting Genes (STRING) to construct protein-protein interaction (PPI) networks and 9 hub genes werescreened such as fucosyltransferase 3 (Lewis blood group) (FUT3) whose expression in chemoresistant samples was high but witha better prognosis in osteosarcoma patients Furthermore the connection between DEGs and differentially expressed miRNAs(DEMs) was explored GEO2R was utilized to screen out DEGs and DEMs A total of 668 DEGs and 5 DEMs were extracted fromGSE7437 and GSE30934 differentiating samples of poor and good chemotherapy reaction patients The Database for AnnotationVisualization and Integrated Discovery (DAVID) was used to perform GO and KEGG pathway enrichment analysis to identifypotential pathways and functional annotations linked with osteosarcoma chemoresistanceThe present studymay provide a deeperunderstanding about regulatory genes of osteosarcoma chemoresistance and identify potential therapeutic targets for osteosarcoma

1 Introduction

Osteosarcoma is one of themost common primarymalignantbone tumors in children and adolescents The worldwidemorbidity rates of osteosarcoma are approximately with anaverage incidence of 31 per million for each stage and 44per million for groups lt25 years old Additionally thereis a bimodal age distribution individuals aged 25ndash60 yearsand elderly individuals respectively In America the Annualage-standardized incidence of osteosarcoma has reachedstabilization from 1976 to 2005 [1ndash5]

Many factors are associatedwith tumor genesis includinghigh birth weight [6] pubertal hormones [7] and germlinegenetic variants [8 9]The common subtypes are osteoblasticchondroblastic and fibroblastic osteosarcomas which mayaccount for 70ndash80 of total cases [10] The standard therapyconsists of neoadjuvant chemotherapy (NACT) surgicalremoval of the primary tumor and adjuvant chemotherapyBefore the 1970s no more than 20 percent of patients werealive after 5 years when excision was major therapeuticmeasure for osteosarcoma [11 12] while it increased to60ndash70 percent for children and young adults with localized

HindawiBioMed Research InternationalVolume 2018 Article ID 4761064 10 pageshttpsdoiorg10115520184761064

2 BioMed Research International

disease after the chemotherapy was used as adjuvant therapyfor surgical resection [13 14] The current predicament ofosteosarcoma treatment is the five-year survival rate doesnot exceed 25 for patients aged 2ndash68 with poor initialresponse tending to have adverse outcomes [15]Therefore toimprove and modify chemotherapy regimens an increasingnumber of pharmacogenomics studies on osteosarcoma havebeen going on for some time such as drug reactions andtoxicity

Multidrug resistance protein 1 (MDR1) that is encodedby gene ATP-binding cassette subfamily B (MDRTAP)member 1 (ABCB1) has been shown to serve as a plausiblefactor in doxorubicin resistance which was validated to belinkedwith poor outcomes inmany osteosarcoma studies [1617] but whether there would be more valuable biomarkersremained to be explored In recent years microarray technol-ogy has substantially promoted the advance of understandingthe mechanisms underlying diseases Additionally the rapiddevelopment of bioinformatics enables us to comprehensivelyscreen out the hub genes associated with chemoresistanceby the process of high-throughput microarrays MicroRNAs(miRNAs) a group of highly conserved short noncodingsmall RNAs including generally 18ndash25 nucleotides in lengthcan suppress the translation of mRNA and cleave it by themodality of base-pairing to the target genesrsquo 31015840 untranslatedregion [18ndash20]

In the present study we analyzed the data of GSE87437and GSE30934 submitted by Serra Mand and Kobayashi E etal respectively to get 668 differentially expressed genes(DEGs) and 5 differentially expressed miRNA (DEMs)between samples of poor and good chemotherapy reactionpatients inGEO2R (httpwwwncbinlmnihgovgeogeo2r)To further understand the function of genes Gene Ontology(GO) Kyoto Encyclopedia of Genes and Genomes pathway(KEGG) Protein-Protein Interaction (PPI) networks andthe connections among DEGs and DEMs were performedin sequence We selected chemoresistance developmentrelated key genes and provided theoretical foundations formodifying and improving osteosarcoma treatment methods

2 Materials and Methods

21 Microarray Data The dataset of GSE87437 gene expres-sion array and GSE30934 miRNA expression array included10 and 8 samples from poor chemotherapy reaction patientsand 11 and 16 samples fromgood ones respectivelyMoreoverthe former dataset was based on GPL570 platform ([HG-U133 Plus 2] Affymetrix Human Genome U133 Plus 20Array) and the latter one was based on GPL10312 platform(3D-Gene Human miRNA Oligo chip v12-100)

22 Identification of DEGs GEO2R an R-associated webapplication was applied to filtrate DEGs between goodchemotherapy reaction samples and poor chemotherapyreaction samples In total 21 samples in GSE87437 and24 samples in GSE30934 were divided into two groupsrespectively and the concrete grouping schemes werealready shown in microarray data The 119875 lt 005 and| log FC| ⩾ 1 were considered as cutoff criterion All

results of DEGs were downloaded in text format hierarchi-cal clustering analysis being conducted later in Morpheus(httpssoftwarebroadinstituteorgmorpheus)

23 GO and Pathway Enrichment Analysis of DEGs Theonline tool Database for Annotation Visualization and Inte-grated Discovery (DAVID httpsdavidncifcrfgov) pro-vided comprehensive information for list of genes by GOand KEGG pathway analyses In addition GO enrichmentanalysis included three different aspects biological process(BP) molecular function (MF) and cellular component (CC)[21] KEGG enrichment analysis was associatedwith genomicinformationrsquos functional interpretation and practical appli-cation [22] The screened DEGs were uploaded to DAVIDV68 to perform GO and KEGG pathway analysis with thecriterion of 119875 lt 005 the results of which were downloadedin text format

24 PPI Networks Construction andModule Analysis To ana-lyze the connection among proteins DEGs were uploaded toSearch Tool for the Retrieval of Interacting Genes (STRINGhttpsstring-dborg) a database covering 9643763 pro-teins from 2031 organisms and the result whose minimuminteraction score was 04 was visualized in Cytoscape [23 24]Furthermore the Molecular Complex Detection (MCODE)was used to screen out significant modules based on theconstructed PPI networks with the criteria of degree cutoff= 2 node density cutoff = 01 node score cutoff = 02 119896-core= 2 and max depth = 100 and hub genes were exported Thefunctional enrichment analysis of genes in each module wasperformed inDAVID Besides the genes in eachmodulewereuploaded toDAVID andKEGGpathway enrichment analysiswas conducted with the condition of 119875 lt 005

25 Survival Analysis of Hub Genes The series matrix ofGSE21257 that contained osteosarcoma patientsrsquo prognos-tic information was downloaded from GEO database Thepatients were split into two groups high expression and lowexpression according to the expression level of a specific hubgeneThedatawas processed by graphpad prism software andthen exported the results

26 Prediction of miRNA Targets DEMs were acquired bythe parallel method of DEGs mentioned above miRWalk10(httpzmfummuni-heidelbergdeappszmfmirwalkindexhtml) an integrated miRNA target prediction platformincluding 10 databases (DIANAmT miRanda miRDBmiRWalk RNAhybrid PICTAR4 PICTAR5 PITA RNA22and Targetscan) was utilized to explore the correlationbetween DEMs and DEGs Besides different colors wereused to indicate the degrees of connections For example redcolor represented strong correlation

3 Results

31 Identification of DEGs A total of 668 DEGs wereobtained fromGSE87437 in the poor chemotherapy responsesamples and compared with those of good response with thecriteria of 119875 lt 005 and | log FC| ⩾ 10 comprising 422

BioMed Research International 3

Table 1 Key differentially expressed genes (DEGs) obtained fromGSE87374

Gene symbol Log FC 119875 valueZNRD1 14 001504282CDK1 131 003504221MYH7B 119 003527194GPR68 118 003724459CAT minus115 003367398FUT3 169 000007205IMPG2 111 004785566GPR180 minus103 000511768ANPEP 132 00354653

upregulated genes and 246 downregulated genes The keyDEGs are displayed in Table 1

32 Hierarchical Clustering Analysis of DEGs Hierarchicalclustering analysis was conducted throughMorpheus a web-based online tool with the series matrix data of the DEGsThe heat map is shown in Figure 1 (top 50 upregulated and 50downregulated genes)

33 GO Term Enrichment Analysis In order to under-stand the function of the identified DEGs deeply GO andKEGG analyses were performed in DAVID respectivelyThe result of GO analysis showed that DEGs were enrichedin biological process (BP) including positive regulation oftranscription and DNA-templated positive regulation ofsequence-specific DNA binding transcription factor activitynitric oxidemediated signal transduction positive regulationof transcription from RNA polymerase II promoter andregulation of phosphatidylinositol 3-kinase signaling Asfor molecular function (MF) the DEGs were enriched inestrogen response element binding Rac guanyl-nucleotideexchange factor activity calcium ion binding zinc ionbinding and phosphatidylinositol-45-bisphosphate 3-kinaseactivity Besides Cellular Component (CC) analysis showedthat the DEGs were enriched in proteinaceous extracellularmatrix cell surface P granule integral component of plasmamembrane and endocytic vesicle membrane as shown inFigure 2

34 KEGG Pathway Analysis KEGG pathway analysisshowed that DEGs were mainly involved in tryptophanmetabolism oxytocin signaling pathway glyoxylate anddicarboxylate metabolism cAMP signaling pathway anddopaminergic synapse (Figure 2)

35 PPI Networks and Modules Selection The PPI networksof DEGs were composed of 432 nodes and 428 edges(Figure 3) Then the networks were imported into Cytoscapesoftware analyzed by using plug-ins MCODE Eventually3 significant modules were selected (Figure 4) and theKEGGpathwaywasmainly associatedwith ribosome biogen-esis in eukaryotes calcium signaling pathway arachidonicacid metabolism proteoglycans in cancer and linoleic acidmetabolism (Figure 4)

ComparisonGoodPoor

Comparison Genesymbol

Row min Row max

MAKPRNTIL34KLHDC8AC3orf79SEMA6AMKXRASSF9PPEF2SFTA1PZFPM2ZNRF2P1LOC100996549ACAT1ASCL5KRBOX1FAM200ASOX11SNAPC3C5orf49MAGI2-AS2MAP2LRP12CLEC3ATMTC1RRS1-AS1ZNF226COL4A4PKDCCZNF7SERPINA5CTD-2350J171RAB3IPAK1KIAA0319LCDH12MEGF9BMPR1BAMIGO1LINC01442CCDC158WNT1ATOH7CENPICPOLOC652993BTBD8CXXC4LOXFPR1NDPBEX5BCL2LOC100128281TCF3PDCL2CXCL14ACOXLRHODIFNKFCRLALOC101927089DIO1FUT3HAR1AUTP6SLFN13PKD2L2FAM19A4ITIH3LOC101928152OR6B1C12orf56TMEM171ROR1-AS1PTPN22LOC100128325ELANECCSER1CHIT1CKBCD5LNLRC4VDRNPHS2LOC100288721 MARVELD2 RCAN3 NIPAL4LOC100129112 LOC100130642 SDHDIFNA1BRE-AS1PIANPMEP1BZNF557NKX2-3ST8SIA6-AS1LINC00970

GSM

2331

338

GSM

2331

340

GSM

2331

342

GSM

2331

345

GSM

2331

346

GSM

2331

348

GSM

2331

350

GSM

2331

351

GSM

2331

352

GSM

2331

339

GSM

2331

354

GSM

2331

335

GSM

2331

336

GSM

2331

337

GSM

2331

341

GSM

2331

343

GSM

2331

344

GSM

2331

347

GSM

2331

349

GSM

2331

353

GSM

2331

334

Figure 1 Heat map of the top 100 DEGs (50 upregulated genesand 50 downregulated genes) Red up-regulation Blue down-regulation

36 Hub Genes and Survival Analysis 9 hub genes werescreened out including zinc ribbon domain containing 1(ZNRD1) myosin heavy chain 7B (MYH7B) G protein-cou-pledreceptor 68 (GPR68) catalase (CAT) fucosyltransferase3 (Lewis blood group) (FUT3) interphotoreceptor matrix


Sig GO terms of DEGs-BP

Regulation of phosphatidylinositol 3-kinase signaling

Positive regulation of transcriptionfrom RNA polymerase II promoter

Nitric oxide mediated signal transduction

Positive regulation of sequence-specific DNA binding transcription factor activity

Positive regulation of transcription DNA-templated

1 2 3 40Enrichment score (minuslog10(P value))

(a)

Sig GO terms of DEGs-CC

Endocytic vesicle membraneIntegral component of plasma membrane

P granuleCell surface

Proteinaceous extracellular matrix

1 2 30Enrichment score (minuslog10(P value))

(b)

Sig GO terms of DEGs-MF

Phosphatidylinositol-4 5-bisphosphate 3-kinase activity

Zinc ion binding

Calcium ion binding

Rac guanyl-nucleotide exchange factor activity

Estrogen response element binding

05 10 15 20 2500Enrichment score (minuslog10(P value))

(c)

Sig pathways of DEGs

Signaling pathways regulating pluripotency of stem cells

Glyoxylate and dicarboxylate metabolism

Oxytocin signaling pathway

Tryptophan metabolism

Adrenergic signaling in cardiomyocytes


(d)

Figure 2 GO and KEGG pathway analysis of DEGs associated with osteosarcoma (a) Top 5 significantly enriched biological processes inDEGs (b) Top 5 significantly enriched cell component in DEGs (c) Top 5 significantly enriched molecular function in DEGs (d) Top 5significantly enriched KEGG pathway in DEGs

proteoglycan 2 (IMPG2) G protein-coupled receptor 180(GPR180) alanyl aminopeptidase membrane (ANPEP) andcyclin dependent kinase 1 (CDK1) (Table 1) Next survivalanalysis of these genes in GSE21257 which contained patientsrsquosurvival prognostic information showed that osteosarcomapatients with high mRNA expression of FUT3 meant a betteroverall survival (OS) despite its high expression in poorchemotherapy response samples (Figure 5) Additionally thesurvival prognostic information of GPR180 and CDK1 wasnot included in GSE21257

37 MiRNAndashDEG Pairs After the differentially expressedanalysis for the data of GSE30934 a total of 5 DEMs wereobtained between the poor chemotherapy response samplescompared with that of good response with the criteriaof 119875 lt 005 and | log FC| ⩾ 10 (Table 2) Next bas-ing onmiRWalk10 database (httpzmfummuni-heidelbergdeappszmfmirwalkindexhtml) the relationship betweenmiRNAs andDEGswas acquired and different kinds of colorswere on behalf of the number of miRNAndashDEG pairs indifferent database which stand for the degrees of connectionFor example red color represented to a strong correlation(Figure 6) After comparing the targets with hub genes wefound that ZNRD1 was the potential target of hsa-miR-543while CAT was the potential target of hsa-miR-518f Bothhsa-miR-543 and hsa-miR-518f matched the regulated genein expression trends

4 Discussion

In the present study we observed whether there weremore valuable genes like ABCB1 which could help improveand modify chemotherapy regimens in osteosarcoma Tofind out the specific chemotherapy response-associated


miRNA ID Log FC 119875 valuehsa-miR-543 minus3429933 000192hsa-miR-409-5p minus270517 000729hsa-miR-518f 1448711 002332hsa-miR-154 minus2619116 003838ebv-miR-BART1-3p minus1358071 004733

DEGs we analyzed the osteosarcoma gene expression arrayof GSE87437 in GEO2R where a total number of 668DEGs were obtained between good and poor chemother-apy response samples Besides to further understand thepotential biological functions we conducted GO KEGG andSTRING analyses Subsequently on the foundation of PPInetworks the selection of 9 hub genes and their survivalprognosis were completed In terms of the increasinglyprominent role of miRNA in cancer DEMs of osteosarcomamiRNAexpression array ofGSE30934was screened out in thesameway and criteria like DEGs andDEGs-miRNAnetworkwas constructed to show relationship between them [25]

Our results showed that many genes and miRNAsmay have functions in the development of chemoresis-tance in osteosarcoma and have the potential to becometreatment targets Here we exclusively focused on 9 hubgenes and two miRNAs Firstly 9 hub genes consisted ofZNRD1 MYH7B GPR68 CAT FUT3 IMPG2 GPR180ANPEP and CDK1 Our data showed that the expression ofZNRD1 was upregulated in chemoresistance osteosarcomasamples Previous studies demonstrated that methotrexate-resistant vincristine-resistant multidrug resistant pheno-types of gastric cancer cells could be regulated by the


Figure 3 PPI network constructed by STRING database for the DEGs

inhibition of ZNRD1Inosine monophosphate dehydroge-nase 2 (IMPDH2) upregulated DARPP-32downregulatedZNRD1 overexpressed miR-508-5pZNRD1ABCB1 activi-ties respectively [26ndash28] but further researches of ZNRD1in osteosarcoma chemoresistance remained to be conductedSimilar to ZNRD1 MYH7B was also found upregulated inchemoresistance osteosarcoma samples At present MYH7Bwas mainly involved in pathway of cardiomyocytes suchas mitochondrial apoptosis pathway [29] but studies aboutcancer were rare GPR68 a kind of pH-sensing protein

was associated with tumor cell biology such as tumoraggressiveness by triggering the intracellular signaling cas-cade to promote the development of microenvironment ofextracellular acidification [30 31] As shown in Table 1likewise we found the expression of GPR68 was upregu-lated in chemoresistance samples Daglioglu C validated thatpH-responsive Fe3O4SiO2(FITC)-BTNQURDOXmulti-functional nanoparticles could potentiate the chemothera-peutic efficacy of DOX against multidrug resistance as wellas counteract the survival ability of chemoresistant lung


Pathway ID Term Count P value

hsa03008 Ribosome biogenesis in eukaryotes 2 40 250E minus 02

Pathway ID Term Count

hsa04270 Vascular smooth muscle contraction 2 25 671E minus 02

hsa04022 cGMP-PKG signaling pathway 2 25 927E minus 02

hsa04020 Calcium signaling pathway 2 25 997E minus 02


hsa00590 Arachidonic acid metabolism 4 2667 143E minus 04

hsa05205 Proteoglycans in cancer 4 2667 433E minus 03

hsa00591 Linoleic acid metabolism 2 1333 493E minus 02

(a)

(b)

(c)

(d)

(e)

(f)

P value

P value

RCL1TFB2M

NOC4L

ZNRD1

UTP6

CYP2U1PLB1ALOX12B

CYP2C9

ESR1

RGS4

SSTR1FPR1

WNT1

FGF8LHX9

NTRK2 PTPN11 CAT MMP2

NR4A1RORA

VDR

CASR

ADRA1D

EDNRA

CPR68

KALRN

Figure 4 The top 3 modules from the PPI network (a) module 1 (b) module 2 (c) module 3 (d) the enriched pathways of module 1 (e) theenriched pathways of module 2 and (f) the enriched pathways of module 3

carcinoma A549DOX cell lines [32] GPR68 has becomean attractive target for drug development [33] Severalprevious studies demonstrated that decreased CAT washighly associated with chemoresistance for example Xu etal showed that intervention against miR-551bCATreactiveoxygen species (ROS)Mucin-1 (MUC1) pathway might helpovercome acquired chemoresistance [34] Tumor microenvi-ronment (TME) was characterized by hypoxia acidosis anddense extracellular matrix providing tumors with resistanceto various therapies which could be effectively changedby the intravenous injection of human serum albumin

(HAS)-chlorine e6 (Ce6)-CAT-paclitaxel (PTX) nanopar-ticles enzyme-loaded therapeutic albumin nanoparticlesMeanwhile H2O2 could relieve tumor hypoxia by generatingoxygen within TME triggered by CAT of those nanoparticleswhich made CAT a potential treatment target in varioustumors [35] Similarly our data showed that CAT wasdownregulated in chemoresistance samples which mightexacerbate local microenvironment to strengthen tumorchemoresistance through the way of hypoxia subsequentacidosis and the like High expression of FUT3 was provedto participate in the development of invasion metastasis and


ZNRD1

Survival time (days)

Expression score comparisonLow expression (n = 27) = 12 deathsHigh expression (n = 26) = 11 deaths

100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07770

(a)

MYH7B



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09012

(b)

IMPG2



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07076

(c)

FUT3



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 00301

(d)

CAT



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 06088

(e)

GPR68



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 04394

(f)

ANPEP



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09961

(g)

Figure 5 The survival prognostic value of hub gene in osteosarcoma from the GSE21257

resistance to therapy by increased fucosylation activity in oralsquamous cell carcinoma (OSCC) and the function couldbe blocked by inhibition of fucosylation [36] In our studyFUT3 was also observed upregulated in chemoresistancesamples but it was found to be associated with a bettersurvival prognosis (Figure 4) There were some reasons to

explain that In spite of its protumor action death pathwayswere proved to be relied on fucosylation and FUT3 wasdemonstrated to play an important role in natural killer-induced cytotoxicity after the recognition of sialyl Lewis Xwith the help of C-type lectin receptors [37 38] Thereforethe relationship between FUT3 and tumor was so complex


Figure 6 The network of miRNAndashDEG pairs

that little was known about its function in osteosarcomachemoresistance IMPG-2 a genemainly associatedwith reti-nal disease was upregulated in chemoresistant osteosarcomasamples in our study but more studies involving IMPG-2and cancer needed to be conducted [39] Furthermore ourdata showed the decreased expression of GPR180 in poorchemotherapy response samples Homoplastically Honda etal found that the methylation of GPR180 was probably toencode tumor suppressors and serves as a novel prognosticmarker and therapeutic target inHepatoblastoma [40] Basedon previous studies whether the gene GPR180 could havefunction in osteosarcoma by producing tumor suppressorsand its concrete role in chemoresistance remained to beexplored The gene ANPEP encoded aminopeptidase N(APN) A previous study [41] showed that ANPEP wasdownregulated in prostate cancer (PC) On the contrary ourstudy showed that ANPEP expression of good chemother-apy response samples was approximately two times that inchemoresistance osteosarcoma samples However the dif-ference caused by the types of tumors or chemoresistanceneeded to be further studied In urothelial carcinoma APN

could increase cytotoxicity of melphalan-flufenamide to playanticancer effect by amplifying the intracellular loading ofmelphalan [42] The studies of chemoresistance associatedwith ANPEP have not been conducted so far but Viktorssonet al [42] offered researchers a new way for treatmentwhich made ANPEP a significant therapeutic target Severalresearches in chemoresistance-associated fields have alreadydemonstrated that CDK1 participated in the developmentof chemoresistance in pathways such as GRP78CDC2[43] Likewise in our study the expression of CDK1 wasincreased Hayashi et al found increased DNA repair activityin the G2-M transition promoted temozolomide (TMZ)resistance and CDC2 inhibitor flavopiridol (FP) treatmentcould resensitize TMZ-resistant clones in a p53-independentmanner in glioma cells [44] Besides the combination of ERKinhibitor PD98059 and Taxol could improve the sensitivity oftaxol-resistant tumor cells with the decreased CDC2 activity[45]

Compared with that of good response 5 DEMs wereacquired in GSE30934 in poor chemotherapy response sam-ples (Table 2) Among them hsa-miR-543 and hsa-miR-518f


were found to have a relation to ZNRD1 and CAT respec-tively In our results hsa-miR-543 was downregulated inchemoresistant samples Previous study in this field was lim-ited In other aspects decreased expression of it was involvedin osteosarcoma angiogenesis which might be caused byconnective tissue growth factor (CTGF) in phospholipase C(PLC)protein kinase C (PKC120575) signaling pathway Besideshsa-miR-543 was also proved to be linked with tumor staging[46] cell proliferation and the glycolytic pathway [47] Thestudies between chemoresistance-promoted gene ZNRD1and hsa-miR-543 have not been conducted yet but biologicalfunctions mentioned above made hsa-miR-543 become animportant therapeutic target Moreover the role of hsa-miR-518f in chemoresistance or the development of tumor wasrarely known Hsa-miR-518e and hsa-miR-518b homologuesof hsa-miR-518f were demonstrated to be upregulated inhepatocellular carcinoma (HCC) [48] Besides a previousstudy showed that has-miR-518c-5p could regulate the growthand metastasis of oral cancer [49] Consequently furtherresearch to hsa-miR-518f was of great importance

In summary we identified 668 DEGs and 5 DEMs fromGEO2R between good chemotherapy response samples andpoor chemoresistance samples in osteosarcoma And manyof them such as ZNRD1 GPR68 CAT FUT3 ANPEPCDK1 and hsa-miR-543 might be key genes related toosteosarcoma chemoresistance These findings provided aseries of promising treatment targets and enlightened us onthe further investigations of the molecular mechanisms

Conflicts of Interest

The authors declare that there are no conflicts of interest

References

[1] L Mirabello R J Troisi and S A Savage ldquoInternationalosteosarcoma incidence patterns in children and adolescentsmiddle ages and elderly personsrdquo International Journal ofCancer vol 125 no 1 pp 229ndash234 2009

[2] L Mirabello R J Troisi and S A Savage ldquoOsteosarcomaincidence and survival rates from 1973 to 2004 data from thesurveillance epidemiology and end results programrdquo Cancervol 115 no 7 pp 1531ndash1543 2009

[3] HD Dorfman and B Czerniak ldquoBone cancersrdquoCancer vol 75no 1 supplement pp 203ndash210 1995

[4] K P Anfinsen S S Devesa F Bray et al ldquoAge-period-cohort analysis of primary bone cancer incidence rates in theUnited States (1976-2005)rdquo Cancer Epidemiology Biomarkers ampPrevention vol 20 no 8 pp 1770ndash1777 2011

[5] P C ValeryM Laversanne and F Bray ldquoBone cancer incidenceby morphological subtype a global assessmentrdquo Cancer Causesamp Control vol 26 no 8 pp 1127ndash1139 2015

[6] L Mirabello R Pfeiffer G Murphy et al ldquoHeight at diagnosisand birth-weight as risk factors for osteosarcomardquo CancerCauses amp Control vol 22 no 6 pp 899ndash908 2011

[7] J R B Musselman T L Bergemann J A Ross et al ldquoCase-parent analysis of variation in pubertal hormone genes andpediatric osteosarcoma a childrenrsquos oncology group (COG)studyrdquo International Journal of Molecular Epidemiology andGenetics vol 3 no 4 pp 286ndash293 2012

[8] M Kansara M W Teng M J Smyth and D M ThomasldquoTranslational biology of osteosarcomardquo Nature Reviews Can-cer vol 14 no 11 pp 722ndash735 2014

[9] L Mirabello K Yu S I Berndt et al ldquoA comprehensivecandidate gene approach identifies genetic variation associatedwith osteosarcomardquo BMC Cancer vol 11 Article ID 209 2011

[10] S H Orkin D E Fisher A T Look S Lux D Ginsburg andDG Nathan Oncology of Infancy and Childhood Elsevier HealthSciences 2009

[11] D C Dahlin andM B Coventry ldquoOsteogenic sarcoma a studyof six hundred casesrdquo The Journal of Bone amp Joint Surgery vol49 no 1 pp 101ndash110 1967

[12] R CMarcove VMike J VHajek A G Levin andRVHutterldquoOsteogenic sarcoma under the age of twenty-onerdquoThe Journalof Bone amp Joint Surgery vol 52 no 3 pp 411ndash423 1970

[13] P A Meyers G Heller J Healey et al ldquoChemotherapy for non-metastatic osteogenic sarcoma the memorial sloan-ketteringexperiencerdquo Journal of Clinical Oncology vol 10 no 1 pp 5ndash151992

[14] A M Goorin D J Schwartzentruber M Devidas et alldquoPresurgical chemotherapy compared with immediate surgeryand adjuvant chemotherapy for nonmetastatic osteosarcomaPediatric Oncology Group Study POG-8651rdquo Journal of ClinicalOncology vol 21 no 8 pp 1574ndash1580 2003

[15] B Kempf-Bielack S S Bielack H Jurgens et al ldquoOsteosarcomarelapse after combined modality therapy an analysis of unse-lected patients in the Cooperative Osteosarcoma Study Group(COSS)rdquo Journal of Clinical Oncology vol 23 no 3 pp 559ndash5682005

[16] H I Vos M J H Coenen H-J Guchelaar and D M WM te Loo ldquoThe role of pharmacogenetics in the treatment ofosteosarcomardquoDrug DiscoveryTherapy vol 21 no 11 pp 1775ndash1786 2016

[17] M Serra and C M Hattinger ldquoThe pharmacogenomics ofosteosarcomardquoThe Pharmacogenomics Journal vol 17 no 1 pp11ndash20 2017

[18] V N Kim J Han and M C Siomi ldquoBiogenesis of small RNAsin animalsrdquo Nature Reviews Molecular Cell Biology vol 10 no2 pp 126ndash139 2009

[19] D P Bartel ldquoMicroRNAs target recognition and regulatoryfunctionsrdquo Cell vol 136 no 2 pp 215ndash233 2009

[20] M A Valencia-Sanchez J Liu G J Hannon and R ParkerldquoControl of translation andmRNAdegradation bymiRNAs andsiRNAsrdquoGenesampDevelopment vol 20 no 5 pp 515ndash524 2006

[21] P Gaudet N Skunca J C Hu and C Dessimoz ldquoPrimer onthe Gene Ontologyrdquo inThe Gene Ontology Handbook vol 1446ofMethods in Molecular Biology pp 25ndash37 Springer New YorkNew York NY USA 2017

[22] M Kanehisa M Furumichi M Tanabe Y Sato and KMorishima ldquoKEGG new perspectives on genomes pathwaysdiseases and drugsrdquo Nucleic Acids Research vol 45 no 1 ppD353ndashD361 2017

[23] D Szklarczyk A Franceschini S Wyder et al ldquoSTRING v10protein-protein interaction networks integrated over the treeof liferdquo Nucleic Acids Research vol 43 pp D447ndashD452 2015

[24] P Shannon A Markiel O Ozier et al ldquoCytoscape a softwareenvironment for integrated models of biomolecular interactionnetworksrdquo Genome Research vol 13 no 11 pp 2498ndash25042003

[25] G A Calin C Sevignani C D Dumitru et al ldquoHumanmicroRNA genes are frequently located at fragile sites and


genomic regions involved in cancersrdquo Proceedings of theNational Acadamy of Sciences of the United States of Americavol 101 no 9 pp 2999ndash3004 2004

[26] L Hong T Qiao Y Han et al ldquoZNRD1 mediates resistance ofgastric cancer cells to methotrexate by regulation of IMPDH2and Bcl-2rdquo The International Journal of Biochemistry amp CellBiology vol 84 no 2 pp 199ndash206 2006

[27] L Hong Y Zhao J Wang et al ldquoReversal of multidrugresistance of adriamycin-resistant gastric adenocarcinoma cellsthrough the up-regulation ofDARPP-32rdquoDigestiveDiseases andSciences vol 53 no 1 pp 101ndash107 2008

[28] Y Shang B Feng L Zhou et al ldquoThe miR27b-CCNG1-P53-miR-508-5p axis regulates multidrug resistance of gastriccancerrdquo Oncotarget vol 7 no 1 pp 538ndash549 2016

[29] J Wang Z Jia C Zhang et al ldquomiR-499 protects cardiomy-ocytes from H

2O2-induced apoptosis via its effects on Pdcd4

and Pacs2rdquo RNA Biology vol 11 no 4 pp 339ndash350 2014[30] W-C Huang P Swietach R D Vaughan-Jones O Ansorge

and M D Glitsch ldquoExtracellular acidification elicits spatiallyand temporally distinct Ca2+ signalsrdquo Current Biology vol 18no 10 pp 781ndash785 2008

[31] H Saxena D A Deshpande B C Tiegs et al ldquoThe GPCROGR1 (GPR68) mediates diverse signalling and contractionof airway smooth muscle in response to small reductions inextracellular pHrdquo British Journal of Pharmacology vol 166 no3 pp 981ndash990 2012

[32] C Daglioglu ldquoEnhancing tumor cell response to multidrugresistance with pH-sensitive quercetin and doxorubicin conju-gated multifunctional nanoparticlesrdquo Colloids and Surfaces BBiointerfaces vol 156 pp 175ndash185 2017

[33] X-P Huang J Karpiak W K Kroeze et al ldquoAllosteric ligandsfor the pharmacologically dark receptors GPR68 and GPR65rdquoNature vol 527 no 7579 pp 477ndash483 2015

[34] X Xu A Wells M T Padilla K Kato K C H Kim andY Lin ldquoA signaling pathway consisting of miR-551b catalaseand MUC1 contributes to acquired apoptosis resistance andchemoresistancerdquo Carcinogenesis vol 35 no 11 pp 2457ndash24662014

[35] Q Chen J Chen C Liang et al ldquoDrug-induced co-assemblyof albumincatalase as smart nano-theranostics for deep intra-tumoral penetration hypoxia relieve and synergistic combina-tion therapyrdquo Journal of Controlled Release vol 263 pp 79ndash892017

[36] V Desiderio P Papagerakis V Tirino et al ldquoIncreased fucosy-lation has a pivotal role in invasive and metastatic properties ofhead and neck cancer stem cellsrdquo Oncotarget vol 6 no 1 pp71ndash84 2015

[37] K Higai A Ichikawa and K Matsumoto ldquoBinding of sialylLewis X antigen to lectin-like receptors on NK cells inducescytotoxicity and tyrosine phosphorylation of a 17-kDa proteinrdquoBiochimica et Biophysica Acta (BBA)mdashGeneral Subjects vol1760 no 9 pp 1355ndash1363 2006

[38] C Ohyama S Kanto K Kato et al ldquoNatural killer cells attacktumor cells expressing high levels of sialyl Lewis x oligosaccha-ridesrdquo Proceedings of the National Acadamy of Sciences of theUnited States of America vol 99 no 21 pp 13789ndash13794 2002

[39] D Bandah-Rozenfeld RW J Collin E Banin et al ldquoMutationsin IMPG2 Encoding interphotoreceptor matrix proteoglycan2 cause autosomal-recessive retinitis pigmentosardquo AmericanJournal of Human Genetics vol 87 no 2 pp 199ndash208 2010

[40] S Honda M Minato H Suzuki et al ldquoClinical prognosticvalue of DNAmethylation in hepatoblastoma four novel tumor

suppressor candidatesrdquo Cancer Science vol 107 no 6 pp 812ndash819 2016

[41] K D Soslashrensen M O Abildgaard C Haldrup et al ldquoProg-nostic significance of aberrantly silenced ANPEP expression inprostate cancerrdquo British Journal of Cancer vol 108 no 2 pp420ndash428 2013

[42] K Viktorsson C-H Shah T Juntti et al ldquoMelphalan-flufenamide is cytotoxic and potentiates treatment withchemotherapy and the Src inhibitor dasatinib in urothelialcarcinomardquo Molecular Oncology vol 10 no 5 pp 719ndash7342016

[43] W Li W Wang H Dong et al ldquoCisplatin-induced senescencein ovarian cancer cells is mediated by GRP78rdquo OncologyReports vol 31 no 6 pp 2525ndash2534 2014

[44] T Hayashi K Adachi S Ohba and Y Hirose ldquoThe Cdkinhibitor flavopiridol enhances temozolomide-induced cyto-toxicity in human glioma cellsrdquo Journal of Neuro-Oncology vol115 no 2 pp 169ndash178 2013

[45] X Lin Y Liao X Chen D Long T Yu and F Shen ldquoRegulationof oncoprotein 18Stathmin signaling by ERK concerns theresistance to Taxol in nonsmall cell lung cancer cellsrdquo CancerBiotherapy and Radiopharmaceuticals vol 31 no 2 pp 37ndash432016

[46] L-H Wang H-C Tsai Y-C Cheng et al ldquoCTGF promotesosteosarcoma angiogenesis by regulatingmiR-543angiopoietin2 signalingrdquo Cancer Letters vol 391 pp 28ndash37 2017

[47] H Zhang X Feng T Wang et al ldquoMiRNA-543 promotesosteosarcoma cell proliferation and glycolysis by partially sup-pressing PRMT9 and stabilizing HIF-1120572 proteinrdquo Oncotargetvol 8 no 2 pp 2342ndash2355 2017

[48] W Wang L-J Zhao Y-X Tan H Ren and Z-X Qi ldquoMiR-138induces cell cycle arrest by targeting cyclin D3 in hepatocellularcarcinomardquo Carcinogenesis vol 33 no 5 pp 1113ndash1120 2012

[49] M Kinouchi D Uchida N Kuribayashi et al ldquoInvolvementof miR-518c-5p to growth and metastasis in oral cancerrdquo PLoSONE vol 9 no 12 Article ID e115936 2014

Hindawiwwwhindawicom

International Journal of

Volume 2018

Zoology

Hindawiwwwhindawicom Volume 2018

Anatomy Research International

PeptidesInternational Journal of



Journal of Parasitology Research

GenomicsInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018


BioinformaticsAdvances in

Marine BiologyJournal of



Neuroscience Journal


BioMed Research International

Cell BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawiwwwhindawicom Volume 2018


Genetics Research International


Advances in

Virolog y Stem Cells International



Enzyme Research



MicrobiologyHindawiwwwhindawicom

Nucleic AcidsJournal of

Volume 2018

Submit your manuscripts atwwwhindawicom


disease after the chemotherapy was used as adjuvant therapyfor surgical resection [13 14] The current predicament ofosteosarcoma treatment is the five-year survival rate doesnot exceed 25 for patients aged 2ndash68 with poor initialresponse tending to have adverse outcomes [15]Therefore toimprove and modify chemotherapy regimens an increasingnumber of pharmacogenomics studies on osteosarcoma havebeen going on for some time such as drug reactions andtoxicity

Multidrug resistance protein 1 (MDR1) that is encodedby gene ATP-binding cassette subfamily B (MDRTAP)member 1 (ABCB1) has been shown to serve as a plausiblefactor in doxorubicin resistance which was validated to belinkedwith poor outcomes inmany osteosarcoma studies [1617] but whether there would be more valuable biomarkersremained to be explored In recent years microarray technol-ogy has substantially promoted the advance of understandingthe mechanisms underlying diseases Additionally the rapiddevelopment of bioinformatics enables us to comprehensivelyscreen out the hub genes associated with chemoresistanceby the process of high-throughput microarrays MicroRNAs(miRNAs) a group of highly conserved short noncodingsmall RNAs including generally 18ndash25 nucleotides in lengthcan suppress the translation of mRNA and cleave it by themodality of base-pairing to the target genesrsquo 31015840 untranslatedregion [18ndash20]

In the present study we analyzed the data of GSE87437and GSE30934 submitted by Serra Mand and Kobayashi E etal respectively to get 668 differentially expressed genes(DEGs) and 5 differentially expressed miRNA (DEMs)between samples of poor and good chemotherapy reactionpatients inGEO2R (httpwwwncbinlmnihgovgeogeo2r)To further understand the function of genes Gene Ontology(GO) Kyoto Encyclopedia of Genes and Genomes pathway(KEGG) Protein-Protein Interaction (PPI) networks andthe connections among DEGs and DEMs were performedin sequence We selected chemoresistance developmentrelated key genes and provided theoretical foundations formodifying and improving osteosarcoma treatment methods

2 Materials and Methods

21 Microarray Data The dataset of GSE87437 gene expres-sion array and GSE30934 miRNA expression array included10 and 8 samples from poor chemotherapy reaction patientsand 11 and 16 samples fromgood ones respectivelyMoreoverthe former dataset was based on GPL570 platform ([HG-U133 Plus 2] Affymetrix Human Genome U133 Plus 20Array) and the latter one was based on GPL10312 platform(3D-Gene Human miRNA Oligo chip v12-100)

22 Identification of DEGs GEO2R an R-associated webapplication was applied to filtrate DEGs between goodchemotherapy reaction samples and poor chemotherapyreaction samples In total 21 samples in GSE87437 and24 samples in GSE30934 were divided into two groupsrespectively and the concrete grouping schemes werealready shown in microarray data The 119875 lt 005 and| log FC| ⩾ 1 were considered as cutoff criterion All

results of DEGs were downloaded in text format hierarchi-cal clustering analysis being conducted later in Morpheus(httpssoftwarebroadinstituteorgmorpheus)

23 GO and Pathway Enrichment Analysis of DEGs Theonline tool Database for Annotation Visualization and Inte-grated Discovery (DAVID httpsdavidncifcrfgov) pro-vided comprehensive information for list of genes by GOand KEGG pathway analyses In addition GO enrichmentanalysis included three different aspects biological process(BP) molecular function (MF) and cellular component (CC)[21] KEGG enrichment analysis was associatedwith genomicinformationrsquos functional interpretation and practical appli-cation [22] The screened DEGs were uploaded to DAVIDV68 to perform GO and KEGG pathway analysis with thecriterion of 119875 lt 005 the results of which were downloadedin text format

24 PPI Networks Construction andModule Analysis To ana-lyze the connection among proteins DEGs were uploaded toSearch Tool for the Retrieval of Interacting Genes (STRINGhttpsstring-dborg) a database covering 9643763 pro-teins from 2031 organisms and the result whose minimuminteraction score was 04 was visualized in Cytoscape [23 24]Furthermore the Molecular Complex Detection (MCODE)was used to screen out significant modules based on theconstructed PPI networks with the criteria of degree cutoff= 2 node density cutoff = 01 node score cutoff = 02 119896-core= 2 and max depth = 100 and hub genes were exported Thefunctional enrichment analysis of genes in each module wasperformed inDAVID Besides the genes in eachmodulewereuploaded toDAVID andKEGGpathway enrichment analysiswas conducted with the condition of 119875 lt 005

25 Survival Analysis of Hub Genes The series matrix ofGSE21257 that contained osteosarcoma patientsrsquo prognos-tic information was downloaded from GEO database Thepatients were split into two groups high expression and lowexpression according to the expression level of a specific hubgeneThedatawas processed by graphpad prism software andthen exported the results

26 Prediction of miRNA Targets DEMs were acquired bythe parallel method of DEGs mentioned above miRWalk10(httpzmfummuni-heidelbergdeappszmfmirwalkindexhtml) an integrated miRNA target prediction platformincluding 10 databases (DIANAmT miRanda miRDBmiRWalk RNAhybrid PICTAR4 PICTAR5 PITA RNA22and Targetscan) was utilized to explore the correlationbetween DEMs and DEGs Besides different colors wereused to indicate the degrees of connections For example redcolor represented strong correlation

3 Results

31 Identification of DEGs A total of 668 DEGs wereobtained fromGSE87437 in the poor chemotherapy responsesamples and compared with those of good response with thecriteria of 119875 lt 005 and | log FC| ⩾ 10 comprising 422









ComparisonGoodPoor


Row min Row max


GSM

2331

338

GSM

2331

340

GSM

2331

342

GSM

2331

345

GSM

2331

346

GSM

2331

348

GSM

2331

350

GSM

2331

351

GSM

2331

352

GSM

2331

339

GSM

2331

354

GSM

2331

335

GSM

2331

336

GSM

2331

337

GSM

2331

341

GSM

2331

343

GSM

2331

344

GSM

2331

347

GSM

2331

349

GSM

2331

353

GSM

2331

334











(a)






(b)



Zinc ion binding

Calcium ion binding




(c)








(d)




4 Discussion





















(a)

(b)

(c)

(d)

(e)

(f)

P value

P value

RCL1TFB2M

NOC4L

ZNRD1

UTP6

CYP2U1PLB1ALOX12B

CYP2C9

ESR1

RGS4

SSTR1FPR1

WNT1

FGF8LHX9


NR4A1RORA

VDR

CASR

ADRA1D

EDNRA

CPR68

KALRN





ZNRD1



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07770

(a)

MYH7B



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09012

(b)

IMPG2



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07076

(c)

FUT3



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 00301

(d)

CAT



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 06088

(e)

GPR68



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 04394

(f)

ANPEP



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09961

(g)














References

























































Volume 2018

Zoology











Volume 2018

















Advances in




Enzyme Research





Volume 2018










ComparisonGoodPoor


Row min Row max


GSM

2331

338

GSM

2331

340

GSM

2331

342

GSM

2331

345

GSM

2331

346

GSM

2331

348

GSM

2331

350

GSM

2331

351

GSM

2331

352

GSM

2331

339

GSM

2331

354

GSM

2331

335

GSM

2331

336

GSM

2331

337

GSM

2331

341

GSM

2331

343

GSM

2331

344

GSM

2331

347

GSM

2331

349

GSM

2331

353

GSM

2331

334











(a)






(b)



Zinc ion binding

Calcium ion binding




(c)








(d)




4 Discussion





















(a)

(b)

(c)

(d)

(e)

(f)

P value

P value

RCL1TFB2M

NOC4L

ZNRD1

UTP6

CYP2U1PLB1ALOX12B

CYP2C9

ESR1

RGS4

SSTR1FPR1

WNT1

FGF8LHX9


NR4A1RORA

VDR

CASR

ADRA1D

EDNRA

CPR68

KALRN





ZNRD1



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07770

(a)

MYH7B



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09012

(b)

IMPG2



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07076

(c)

FUT3



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 00301

(d)

CAT



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 06088

(e)

GPR68



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 04394

(f)

ANPEP



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09961

(g)














References

























































Volume 2018

Zoology











Volume 2018

















Advances in




Enzyme Research





Volume 2018










(a)






(b)



Zinc ion binding

Calcium ion binding




(c)








(d)




4 Discussion





















(a)

(b)

(c)

(d)

(e)

(f)

P value

P value

RCL1TFB2M

NOC4L

ZNRD1

UTP6

CYP2U1PLB1ALOX12B

CYP2C9

ESR1

RGS4

SSTR1FPR1

WNT1

FGF8LHX9


NR4A1RORA

VDR

CASR

ADRA1D

EDNRA

CPR68

KALRN





ZNRD1



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07770

(a)

MYH7B



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09012

(b)

IMPG2



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07076

(c)

FUT3



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 00301

(d)

CAT



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 06088

(e)

GPR68



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 04394

(f)

ANPEP



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09961

(g)














References

























































Volume 2018

Zoology











Volume 2018

















Advances in




Enzyme Research





Volume 2018

















(a)

(b)

(c)

(d)

(e)

(f)

P value

P value

RCL1TFB2M

NOC4L

ZNRD1

UTP6

CYP2U1PLB1ALOX12B

CYP2C9

ESR1

RGS4

SSTR1FPR1

WNT1

FGF8LHX9


NR4A1RORA

VDR

CASR

ADRA1D

EDNRA

CPR68

KALRN





ZNRD1



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07770

(a)

MYH7B



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09012

(b)

IMPG2



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07076

(c)

FUT3



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 00301

(d)

CAT



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 06088

(e)

GPR68



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 04394

(f)

ANPEP



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09961

(g)














References

























































Volume 2018

Zoology











Volume 2018

















Advances in




Enzyme Research





Volume 2018



ZNRD1



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07770

(a)

MYH7B



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09012

(b)

IMPG2



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 07076

(c)

FUT3



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 00301

(d)

CAT



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 06088

(e)

GPR68



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 04394

(f)

ANPEP



100

80

60

40

20

0

Surv

ival

pro

babi

lity

0 2000 4000 6000 8000

n = 53

P = 09961

(g)














References

























































Volume 2018

Zoology











Volume 2018

















Advances in




Enzyme Research





Volume 2018












References

























































Volume 2018

Zoology











Volume 2018

















Advances in




Enzyme Research





Volume 2018

































Volume 2018

Zoology











Volume 2018

















Advances in




Enzyme Research





Volume 2018


identification of key genes and mirnas in osteosarcoma...

Documents