Association of genetic polymorphisms of CXCL12/SDF1 gene and its receptor, CXCR4, to the susceptibility and prognosis of non-small cell lung cancer.

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling or

licensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of the

article (e.g. in Word or Tex form) to their personal website or

institutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies are

encouraged to visit:

ContentslistsavailableatScienceDirect

Lung

Cancer

jo u rn al h om epa g e :w w w . e l s e v i e r . c o m / l o c a t e / l u n g c a n

Association

of

genetic

polymorphisms

of

CXCL12/SDF1

gene

and

its

receptor,

CXCR4,

to

the

susceptibility

and

prognosis

of

non-small

cell

lung

cancer

Yao-Ling

Lee

_,

_Wu-Hsien

_Kuo

_,

_Chiao-Wen

_Lin

_,

_Wei

_Chen

_,

_Wei-Erh

_Cheng

_,

Shuo-Chueh

Chen

_,

_Chuen-Ming

_Shih

a_{SchoolofMedicalLaboratoryandBiotechnology,ChungShanMedicalUniversity,No.110,Sec.1,Chien-KuoN.Road,Taichung402,Taiwan}

b_{DepartmentofMedicine,Armed-ForceTaichungGeneralHospital,Taichung404,Taiwan}

c_{InstituteofBiochemistryandBiotechnology,ChungShanMedicalUniversity,No.110,Sec.1,Chien-KuoN.Road,Taichung402,Taiwan}

d_{DivisionofPulmonaryandCriticalCareMedicine,DepartmentofInternalMedicine,ChinaMedicalUniversityHospital,No.2,Yuh-DerRoad,Taichung404,Taiwan}

e_{DepartmentofRespiratoryTherapy,ChinaMedicalUniversity,No.91Hsueh-ShihRoad,Taichung404,Taiwan}

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received29May2010

Receivedinrevisedform

14December2010

Accepted19December2010

Keywords:

Non-smallcellcancer

CXCL12 SDF1 CXCR4 Polymorphism PCR-RFLP

a

b

s

t

r

a

c

t

Background:TheaimofthisstudywastoevaluatetherelationsofchemokineCXCL12,previouslyknown asstromalcell-derivedfactor-1(SDF1),anditsreceptor,CXCR4,genevariantsonnon-smallcelllung cancer(NSCLC)riskanddiseaseseverity.

Methods:Throughacase–controlstudydesign,genomicDNAsamplesof247NSCLCpatientsand328 ageandsex-matchedcontrolsweresubjectedtopolymerasechainreaction-restrictionfragmentlength polymorphismanalysis.Thevalidityofthistechniquewasprovenbydirectsequencingofamplified prod-ucts.StatisticalanalyseswereconductedtoexplorethecontributionofpolymorphismoftheCXCL12/SDF1 geneandCXCR4,inthesusceptibilitytoandprognosisofNSCLC.

Results:Overall,thegenotypefrequenciesofCXCL12/SDF1geneandCXCR4,weresignificantlydifferent betweenlungcancerpatientsandcontrols(p<0.0001),andalsodifferentbetweenpatientswithlung cancersofvariousstages(p<0.0001).Logisticregressionanalysisrevealedthathigheroddsratios(ORs) forlungcancerwereseenforindividualswithCXCL12/SDF1AA(anORof1.95,95%CI1.08–3.50,p=0.018), orCXCR4TT(anORof4.71,95%CI1.99–11.2,p<0.0001),andforindividualswithbothCXCL12/SDF1AA andCXCR4TTgenotypes(anORof12.4,95%CI1.56–98.3,p=0.002).Thepatientscarryingahomologous AAgenotypeatCXCL12/SDF1,orahomologousTTgenotypeatCXCR4,hadatendencytoadvanceddisease andtowardpoorerprognosescomparedwithotherpatients.

Conclusion:AsignificantassociationbetweenthepolymorphismsofCXCL12/SDF1andCXCR4,andthe susceptibilitytoandprognosisofNSCLCwasdemonstrated.

© 2011 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Lungcancerisoneofthemostcommonmalignanciesworldwide [1] and the leading cause of cancerdeaths in industrial coun-tries, includingTaiwan[2],withtheworstoutcome [3].Cancer isa resultof multiplegene–environmentinteractions occurring overseveraldecades.Themajorriskfactorforlungcancerisan excessiveexposuretotobaccosmoke.However,onlyabout11%of tobaccosmokersultimatelydeveloplungcancer,suggestingthat geneticfactorsmayinfluencetheriskforlungcanceramongthose whoareexposedtocarcinogens.Epidemiologicalstudiesrevealed

∗ Correspondingauthorat:DivisionofPulmonaryandCriticalCareMedicine,

DepartmentofInternalMedicine,ChinaMedicalUniversityHospital,No.2,Yuh-Der

Road,Taichung404,Taiwan.Tel.:+886422077618;fax:+886422077618.

E-mailaddress:[email protected](C.-M.Shih).

anincreasedriskofapproximately14-foldforlungcanceramong regular tobaccosmokers.Aftertheeffectoftobacco smokewas stratified,anapproximately2.5-foldriskwasattributabletoa fam-ilyhistoryoflungcancer[4,5].Therefore,itisrationaltospeculate thatcertaincommongeneticvariantsorpolymorphismsmayhave animpactonlungcancerrisk.

Invasion, angiogenesis, migration, and metastasis are inter-twinedprocessesregulatedbyoverlappingmolecularpathways. Chemokinesandtheirreceptorscomposeonesuchpathwayand are involvedin cell trafficking,migration, and proliferation [6]. CXCL12/SDF1isanimportantalpha-chemokinethatbindstothe G-protein-coupled seven-transmembrane span CXCR4 [7]. The SDF-1-CXCR4axisregulatestrafficking ofnormal and malignant cells [8–10]. Thereare fourgroups of chemokine receptors: C, CC, CXC, and CX3C [11]. CXCR4 is expressed in dendritic cells, naïveTcells,NKcells,andmonocytesandisalsothechemokine receptor most commonly expressed in tumors. Within normal

0169-5002/$–seefrontmatter © 2011 Elsevier Ireland Ltd. All rights reserved.

148 Y.-L.Leeetal./LungCancer73 (2011) 147–152

cells, chemokine receptorsare important in immune cell func-tionandmigrationofstemcellstositesofinjury.Withintumor cells, chemokine receptorexpression is related tothe develop-ment of metastases preferentially to sites with expression of thecorrespondingchemokine [12]. Theligandfor CXCR4 isthe chemokineCXCL12/SDF1whichisexpressedinthelungandother sitesofmetastases.CXCR4/CXCL12/SDF1alsoindirectlypromotes tumor metastasis by mediating proliferation and migration of tumorcellsandenhancingtumor-associatedangiogenesis[13–16]. CXCL12/SDF1geneislocatedonchromosome10q11.1[17,18],and ithasbeenrevealedthatasinglenucleotidepolymorphism(SNP), aguaninetoadenine(G→A),atposition801ofthe3-untranslated generegionmayaffecttheexpressionofCXCL12/SDF1chemokine [19].TheCXCL12/SDF1 A/Ahomozygoteshad beensuggestedto altertheproductionofCXCL12/SDF1[20,21]andareassociatedwith theriskofcarcinogenesisofvariousorigins,includinglungcancer [22–24].Meanwhile,itsreceptor,CXCR4,islocatedonchromosome 2q2[25]andasilentSNPofCXCR4,acytosinetothymine(C→T), isfoundatcodon138[26].Sofar,onlytheeffectofgenetic poly-morphismsofCXCL12/SDF1hasbeenstudiedinIranianlungcancer patients[24].TheeffectofgeneticpolymorphismsofCXCL12/SDF1 andCXCR4 inNSCLC hasnotbeen studied.Toclarifythe influ-enceofgeneticpolymorphismsofCXCL12/SDF1andCXCR4onthe susceptibilityandclinicopathologicaldevelopmentofNSCLC,the relationshipbetweenSNPsofCXCL12/SDF1andCXCR4genesand NSCLCriskaswellastheclinicopathologicalcharacteristicswas investigatedinthisstudy.

2. Materialsandmethods 2.1. Studypopulation

A totalof 247 NSCLC patients, consisting of 178 males and 69 females with a median ageof 63.8, who were admitted to ChinaMedicalUniversityHospital,Taichung,Taiwanbetween2005 and 2009, were included in this study. Of them, 152 patients had adenocarcinomas(AD) and 95 patients had squamous car-cinomas (SQ). The histological determination, including tumor types and stages, was performed according to the WHO clas-sification (WHO,1982) and theTNMsystem (Mountain,1986), respectively.Meanwhile,328unrelatedcontrols,consistingof248 males and 80 females with a median age of 64.2, were ran-domlyselected from a pool of healthy volunteers who visited the general health check-up center of China Medical Univer-sityHospitalduringthesameperiod.Allcontrolshadnoknown medicalillnessorhereditarydisordersandweretakingno medi-cations.

A detailed questionnaire, included informationon the aver-agenumberofcigarettessmokeddailyandthenumberofyears thesubjects had been smoking,for each caseand controlwas completedbyatrainedinterviewer.Thisstudywasapprovedby theResearchEthicsCommitteeofChinaMedicalUniversity,and informedconsentwasobtainedfromeachparticipantpriortothe commencement.Basiccharacteristics,includingageandgender,of allparticipantsaresummarizedinTable1toshownosignificant differencesintheabove-mentionedfeaturesbetweenpatientsand controls.

2.2. SamplecollectionandgenomicDNAextraction

Venous blood from each subject wasdrawninto Vacutainer tubes containing EDTA and stored at 4◦C. Genomic DNA was extractedbyQIAampDNAbloodminikits(Qiagen,Valencia,USA) accordingtothemanufacture’sinstructions.DNAwasdissolvedin TEbuffer[10mMTris(pH7.8),1mMEDTA]andthenquantitated

Table1

Clinicalfeaturesofthestudypopulations.a

Variables NSCLC Control pvalue‡

Subjects 247 328 Gender(male/female) 178/69 248/80 0.34 Age(yr) 63.8±8.6 64.2±9.2 0.28 Brinkmanindexb _{589.4±58.2 586.8±56.8 0.45} Tumortype Adenocarcinomas(AD) 152 Squamouscarcinomas(SQ) 95 Tumorstage I+II 87 III+IV 160

a_{Dataarepresentedasno.ormean±SEM.}

b_{Brinkmanindex=dailycigarettenumbersmultipliedbysmokingyears.}

‡_{pvalueswerecalculatedusingtheMann–WhitneyUtest.}

byameasurementofOD260.Finalpreparationwasstoredat−20◦C andusedastemplatesforpolymerasechainreaction(PCR). 2.3. Polymerasechainreaction-restrictionfragmentlength polymorphism(PCR-RFLP)

The CXCL12/SDF1-3A and CXCR4 polymorphisms were determined by polymerase chain reaction-restriction frag-ment length polymorphism (PCR-RFLP) assay. The sequences of primers used to amplify the CXCL12/SDF1-3A genotype were 5-CAGTCAACCTGGGCAAA GCC-3 and 5 -CCTGAGAGTC-CTTTTGCGGG-3, and those used for the amplification of CXCR4 genotype were 5-AACTTCCTATG CAAGGCAGT-3 and 5-TATCTGTCATCTGCCTCACT-3 [22,27].PCRwasperformedina 10␮L reactionmixturecontaining 100ng DNA template, 1.0␮L of10×PCRbuffer(Invitrogen,Carslbad,CA),0.25UofTaqDNA polymerase(Invitrogen, Carslbad,CA), 0.2mMdNTPs (Promega, Madison,WI),and200nMofeachprimer(MDBio,Taipei,Taiwan). ThePCRcyclingstartedat94◦C for5minfollowedby35cycles of94◦Cfor1min,60◦Cfor1min,and72◦Cfor2min,withafinal stepat72◦Cfor20mintoallowacompleteextensionofallPCR fragments.TheresultsareshowninFig.1.

2.4. Statisticalanalysis

DifferencesinclinicaldatabetweentheNSCLCpatientsandthe controlsubjectswereexamined.Allcontinuousdataareexpressed as mean±standarddeviation and compared using a two-tailed Student’st-test.Categoricalvariables arereportedasa percent-age and compared using Chi-square (2_{)or Fisher’s exact test.} Hardy–Weinbergequilibriumwasassessedusinga goodness-of-fit 2 _{testfor biallelicmarkers.Thedistributionof CXCL12/SDF1} anditsreceptor,CXCR4,geneticpolymorphismbetweenhealthy subjectsandlungcancerpatientswasexaminedbythe2 _test. Significancewasacceptedatp<0.05.Oddsratios(ORs)and 95% confidenceintervals(CI)forlungcancerofeachspecificgenotype werecalculatedwithlogisticregressiontoquantitativelyassessthe degreeofassociationobserved.

TheKaplan–Meiermethodwasusedtoestimatetheprobability ofsurvivalasafunctionoftimeandmediansurvival.Thelogrank testwasusedtoassessthesignificanceofthedifferencebetween homozygousvariantgenotypeandothergenotypesofCXCL12/SDF1 anditsreceptor,CXCR4,amongpairsofsurvivalprobabilities. Sig-nificancewasacceptedatp<0.05.

3. Results

The associationof the CXCL12/SDF1 and its receptor, CXCR4, geneticpolymorphismwithcertainclinicopathologicalparameters oflungcancerpatientswasanalyzedandisshowninTables2and3.

Table2

TheassociationbetweentheSDF-1polymorphismandtheclinicopathologicparametersofthestudiedsubjects.

Characteristics Genotypes Total p-Value Oddsratio(95%CI)p-value†

GG(%) GA(%) AA(%) Non-cancercontrol 171(52) 136(42) 21(6) 328 1.00 Lungcancer 99(40) 112(45) 36(15) 247 <0.001* _{2.49(1.42–4.39)0.002} Tumortype AD 59(39) 72(47) 21(14) 152 0.004* _{2.34(1.42–4.44)0.009} SQ 40(42) 40(42) 15(16) 95 0.01* _{2.74(1.35–5.56)0.006} Tumorstage I+II 52(60) 33(38) 2(2) 87 <0.0001‡ _{11.47(2.68–49.0)<0.0001} III+IV 47(29) 79(50) 34(21) 160

*_{pvalue.Thefrequenciesofthegenotypesbetweenthecancerandnon-cancercontrolgroupswerecomparedwith}2_analysis.

‡ _{pvalue.Thefrequenciesofthegenotypesbetweenlungcancerswithdifferenttumorstageswerecomparedwith}2_analysis.

† _{OddsratiosandpvaluewerecalculatedusinglogisticregressiontomeasuretheassociationofthevarianthomozygousgenotypesAAwithlungcancerrisk,withthatof}

theGG/GAgenotypebeingreferredtoas1.

AD,adenocarcinoma;SQ,squamouscellcarcinoma.

Fig.1. Polymerasechainreaction-restrictionfragmentlengthpolymorphismof

CXCL12/SDF1andCXCR4gene.PCRproductsofCXCL12/SDF1andCXCR4genewere

subjectedtoenzymaticdigestionbyincubationwithHpaIIandBccIfor4hat37◦C

andthensubmittedtoelectrophoresisin3%agarosegels.ForCXCL12/SDF1,wildtype

homozygousalleles(G/G)yielded100and193-bpproducts,theheterozygousalleles

(G/A)yielded100-,193-and293-bpproducts,whilethemutatedtypehomozygous

alleles(A/A)yieldeda293-bpproduct.ForCXCR4gene,wild-typehomozygous

alle-les(C/C)yielded103and133-bpproducts,theheterozygousalleles(C/T)yielded

103-,133-and236-bpproducts,whilethemutatedtypehomozygousalleles(T/T)

yieldeda236-bpproduct.

Overall,genotypedistributionsoftheCXCL12/SDF1anditsreceptor, CXCR4 weresignificantlydifferentbetweennon-cancercontrols andlungcancerpatients(p<0.0001).ThefrequencyofCXCL12/SDF1 variantpolymorphic homozygote waslow withbeing 0.15and 0.06 in the cases and controls,respectively, while that of the wild-type allele was higher (0.71 for cases and 0.63 for con-trols).Thefrequency ofCXCR4variantpolymorphichomozygote was low with being 0.12 and 0.02 in the cases and control, respectively,while thatofthewild-typeallelewashigher(0.79 for cases and0.91for controls). Resultsof the2 goodness-of-fit test showedthatgenotype frequencieswereconsistentwith Hardy–Weinbergequilibriuminthepatientpopulation(2_=0.22; p>0.05)andthecontrolpopulation(2_{=0.77;p>0.05).Logistic} regressionanalysisrevealedthathigherORsforhavingNSCLC(an ORof4.02,95%CI2.39–6.76;p<0.0001),adenocarcinoma(anOR of 3.92, 95% CI 2.23–6.89; p<0.0001) and squamous cell carci-noma (anOR of 4.19, 95% CI 2.19–8.02;p<0.0001), were seen in patientshomozygous forCXCL12/SDF1variantallele(A/A),as comparedwithpatientswithatleastonewild-typeallele((G/G) or (G/A)). Also, higher ORs for having NSCLC (an OR of 4.02, 95% CI 2.39-6.76; p<0.0001), adenocarcinoma (an OR of 3.92, 95% CI 2.23–6.89;p<0.0001) and squamous cellcarcinoma(an ORof4.19, 95%CI 2.19–8.02;p<0.0001), wereseeninpatients homozygous for CXCR4 variant allele (T/T), as compared with patientswithat leastonewild-type allele((C/C) or(C/T)). Fur-thermore,patientscarryingavariantpolymorphichomozygoteof CXCL12/SDF1andCXCR4alsohadatendencyforadvanceddisease (p=0.001).

All patients were followed up during the study period. Those patients carrying a variant polymorphic homozygote of CXCL12/SDF1 andCXCR4 alsohada poorer prognosescompared

Table3

TheassociationbetweentheCXCR4polymorphismandtheclinicopathologicparametersofthestudiedsubjects.

Characteristics Genotypes Total p-Value Oddsratio(95%CI)p-value†

CC(%) CT(%) TT(%) Non-cancercontrol 274(84) 47(14) 7(2) 328 1.00 Lungcancer 172(70) 44(18) 31(12) 247 <0.0001* _{6.58(2.85–15.2)<0.0001} Tumortype AD 108(71) 29(19) 15(10) 152 <0.0001* _{5.02(2.00–12.6)<0.0001} SQ 64(67) 15(16) 16(17) 95 <0.0001* _{6.63(2.53–17.4)<0.0001} Tumorstage I+II 70(81) 15(17) 2(2) 87 0.017‡ _{6.42(1.47–28.1)0.005} III+IV 102(64) 29(18) 29(18) 160

*_{pvalue.Thefrequenciesofthegenotypesbetweenthecancerandnon-cancercontrolgroupswerecomparedwith}2_analysis.

‡ _{pvalue.Thefrequenciesofthegenotypesbetweenlungcancerswithdifferenttumorstageswerecomparedwith}2_analysis.

† _{OddsratiosandpvaluewerecalculatedusinglogisticregressiontomeasuretheassociationofthevarianthomozygousgenotypesTTwithlungcancerrisk,withthatof}

theCC/CTgenotypebeingreferredtoas1.

150 Y.-L.Leeetal./LungCancer73 (2011) 147–152

Fig.2. TheKaplan–MeiersurvivalcurveswithrespecttotheCXCL12/SDF1G→AandCXCR4C→Tgeneticpolymorphismsinrelationtonon-smallcelllungcancerpatients.

Thepvalueforeachanalysisisindicated.

withother patients (p<0.0001 and p<0.001,by log rank test) (Fig.2).

4. Discussion

Inthisstudy,weprovidednovelinformationabouttheeffectsof geneticpolymorphismsofCXCL12/SDF1anditsreceptor,CXCR4,on

thesusceptibilityandclinicopathologicaldevelopmentofNSCLC. It has been shown that over-expression of CXCL12/SDF1 and CXCR4 or interaction between both chemokines is associated with the development and metastasis of human lung cancer [15,28–30].ThebindingofCXCL12/SDF1toCXCR4induces intra-cellularsignaling throughseveral divergentpathwaysinitiating signals related tochemotaxis, cellsurvivaland/or proliferation,

increaseinintracellularcalcium,andgenetranscription. Homozy-gousCXCL12/SDF1-3A/Agenotypeisconsideredtobeassociated with an increased level of protein, which is available to bind its exclusive receptor CXCR4 that contributes to the protective effectofCXCL12/SDF1-3A/Ahomozygousalleles[21].Inthisstudy, significant difference in genotypic frequencies of CXCL12/SDF1 gene was found between controls and patients with NSCLC. Individualswith CXCL12/SDF1-3A/A homozygotes had a risk of 2.49-fold to have NSCLC compared with individuals with G/G homozygotes or G/A heterozygotes. As well, individuals with CXCR4 T/T homozygoteshad a risk of 6.58-fold tohave NSCLC comparedwithindividualswithC/ChomozygotesorC/T heterozy-gotes.

Tothebestofourknowledge,onlyRazmkhahetal.[24]have studiedtherelationshipbetweenCXCL12/SDF1G801Agene poly-morphism and Iranian lung cancer patients. They investigated 72patients, includedcarcinoidtumor,smallcell andnon-small cell lung cancer, and concluded that AA and AG genotypes of CXCL12/SDF1mightbeconsideredas factorsincreasingthe sus-ceptibilityofIranianpatientstolungcancer.Inthisstudyfocusing uponNSCLCpatients,bothCXCL12/SDF1anditsreceptor,CXCR4, genevariantswereevaluatedtorevealthatthegenetic polymor-phismsoftheCXCL12/SDF1andCXCR4increasedthesusceptibility toNSCLC.Significantdifferenceswerefoundinalleleandgenotype frequencydistributionofCXCL12/SDF1andCXCR4betweenNSCLC patientsandcontrols.

Tothebestofourknowledge,nostudyhadbeenconductedto investigatetheassociationbetweenCXCR4genepolymorphismand NSCLCrisk.Although,Hirataetal.[31]reportedthatA/AorA/G genotypeofCXCL12/SDF1wasassociatedwithprostatecancer.They alsodemonstratedthatA/AorA/GgenotypeofCXCL12/SDF1was significantlyassociatedwithhigherexpressionofCXCL12/SDF1and CXCR4.Subsequently,weinvestigatedtherelationshipsbetween polymorphisms and clinicopathological status of patients with NSCLC.TherewassignificantassociationbetweenCXCL12/SDF1and CXCR4genepolymorphismwithadvancedstagesofNSCLC.Survival analysiswasperformedandthosepatientscarryingavariant poly-morphichomozygoteofCXCL12/SDF1andCXCR4alsohadapoorer prognosis.

SincethesusceptibilitytoNSCLCisrecognizedtobeaffected bymultiplegeneticfactorsandgenotype,aswellasenvironment interactions,eachofwhichhasacertainextentofinfluenceonthe developmentofthisdisease,itisrationaltosaythatsucha rela-tionshipbetweengeneticpolymorphismsandNSCLCsusceptibility maybeethnic-dependent.

AlsoCXCL12/SDF1andCXCR4mayalsoplayatissue-specificrole, anddifferentregulatorymechanismsmayapplytodifferenttumors [31–35].Nevertheless,ourdatadiddemonstratethatthe polymor-phismsofCXCL12/SDF1andCXCR4genehadsignificantassociations withsusceptibilitytoandprognosisofNSCLCinourpopulation. However,inthisseries,theinterpretationofourresultswaslimited bysamplesize.Furtherstudyoflargerscaleonthesegenotypesand otherpolymorphismsisrequiredtoanalyzehaplotypesandtheir associationwithboththeonsetoflungcancerandthedevelopment ofmetastases.

In conclusion, by evaluating the polymorphic sites of CXCL12/SDF1 and CXCR4, a significant association of the poly-morphismswith NSCLCwasrevealed. Theresults ofthis study support a relation between CXCL12/SDF1 and CXCR4 genetic polymorphisms and the susceptibility to and prognosis of NSCLC.

Conflictofintereststatement Nonedeclared.

Acknowledgements

ThisstudywassupportedinpartbygrantDMR-96-031from ChinaMedicalUniversityandgrant DOH99-TD-C-111-005from theDepartmentofHealth(TheExecutiveYuan,RepublicofChina). References

[1] AlbergAJ,FordJG,SametJM.Epidemiologyoflungcancer:ACCP

evidence-basedclinicalpracticeguidelines(2nded.).Chest2007;132:29S–55S.

[2] ChiuHF,ChengMH,TsaiSS,WuTN,KuoHW,YangCY.Outdoorairpollution

andfemalelungcancerinTaiwan.InhalToxicol2006;18:1025–31.

[3] JemalA,SiegelR,WardE,HaoY,XuJ,ThunMJ.Cancerstatistics,2009.CA

CancerJClin2009;59:225–49.

[4]YoungRP,HopkinsRJ,HayBA,EptonMJ,MillsGD,BlackPN,etal.Agene-based

riskscoreforlungcancersusceptibilityinsmokersandex-smokers.Postgrad

MedJ2009;85:515–24.

[5]AmosCI,XuW,SpitzMR.Isthereageneticbasisforlungcancersusceptibility?

RecentResultsCancerRes1999;151:3–12.

[6]FultonAM.ThechemokinereceptorsCXCR4andCXCR3incancer.CurrOncol

Rep2009;11:125–31.

[7]Darash-YahanaM,PikarskyE,AbramovitchR,ZeiraE,PalB,KarplusR,etal.

RoleofhighexpressionlevelsofCXCR4intumorgrowth,vascularization,and

metastasis.FASEBJ2004;18:1240–2.

[8] MaksymRB,TarnowskiM,GrymulaK,TarnowskaJ,WysoczynskiM,LiuR,etal.

Theroleofstromal-derivedfactor-1–CXCR7axisindevelopmentandcancer.

EurJPharmacol2009;625:31–40.

[9]TarnowskiM,GrymulaK,RecaR,JankowskiK,MaksymR,TarnowskaJ,etal.

Regulationofexpressionofstromal-derivedfactor-1receptors:CXCR4and

CXCR7inhumanrhabdomyosarcomas.MolCancerRes2010;8:1–14.

[10]HassanS,FerrarioC,SaragoviU,QuennevilleL,GabouryL,BaccarelliA,etal.

Theinfluenceoftumor–hostinteractionsinthestromalcell-derived

factor-1/CXCR4ligand/receptoraxisindeterminingmetastaticriskinbreastcancer.

AmJPathol2009;175:66–73.

[11]VandercappellenJ,VanDammeJ,StruyfS.TheroleofCXCchemokinesand

theirreceptorsincancer.CancerLett2008;267:226–44.

[12] MullerA,HomeyB,SotoH,GeN,CatronD,BuchananME,etal.Involvementof

chemokinereceptorsinbreastcancermetastasis.Nature2001;410:50–6.

[13]WangJ,WangJ,SunY,SongW,NorJE,WangCY,etal.Diversesignaling

path-waysthroughtheSDF-1/CXCR4chemokineaxisinprostatecancercelllines

leadstoalteredpatternsofcytokinesecretionandangiogenesis.CellSignal

2005;17:1578–92.

[14]Chen Q, Wu MH, Zhou YH, Tang YL, Huang C, Li XL, et al. Inhibitory

effectofLRRC4onthemobilityandinvasionofglioblastomasthroughthe

SDF-1alpha/CXCR4axis.ZhongNanDaXueXueBaoYiXueBan2007;32:

735–41.

[15]OonakaharaK,MatsuyamaW,HigashimotoI,KawabataM,ArimuraK,Osame

M.Stromal-derivedfactor-1alpha/CXCL12-CXCR4axisisinvolvedinthe

dis-semination ofNSCLCcells into pleural space.AmJ RespirCell MolBiol

2004;30:671–7.

[16]SongN,HuangY,ShiH,YuanS,DingY,SongX,etal.Overexpressionof

platelet-derivedgrowthfactor-BBincreasestumorpericytecontentviastromal-derived

factor-1alpha/CXCR4axis.CancerRes2009;69:6057–64.

[17]RamachandranS,RamadasK,HariharanR,RejnishKumarR,RadhakrishnaPillai

M.SinglenucleotidepolymorphismsofDNArepairgenesXRCC1andXPDand

itsmolecularmappinginIndianoralcancer.OralOncol2006;42:350–62.

[18] ShirozuM,NakanoT,InazawaJ,TashiroK,TadaH,ShinoharaT,etal.

Struc-tureandchromosomallocalizationofthehumanstromalcell-derivedfactor1

(SDF1)gene.Genomics1995;28:495–500.

[19]ApostolakisS,BaritakiS,KochiadakisGE,IgoumenidisNE,PanutsopulosD,

SpandidosDA.Effectsofpolymorphismsinchemokineligandsandreceptors

onsusceptibilitytocoronaryarterydisease.ThrombRes2007;119:63–71.

[20]SorianoA,MartinezC,GarciaF,PlanaM,PalouE,LejeuneM,etal.Plasma

stro-malcell-derivedfactor(SDF)-1levels,SDF1-3_{Agenotype,andexpressionof}

CXCR4onTlymphocytes:theirimpactonresistancetohuman

immunodefi-ciencyvirustype1infectionanditsprogression.JInfectDis2002;186:922–31.

[21]WinklerC,ModiW,SmithMW,NelsonGW,WuX,CarringtonM,etal.Genetic

restrictionofAIDSpathogenesisbyanSDF-1chemokinegenevariant.ALIVE

Study,HemophiliaGrowthandDevelopmentStudy(HGDS),MulticenterAIDS

CohortStudy(MACS),MulticenterHemophiliaCohortStudy(MHCS),San

Fran-ciscoCityCohort(SFCC).Science1998;279:389–93.

[22]deOliveiraCE,CavassinGG,PerimAdeL,NasserTF,deOliveiraKB,FungaroMH,

etal.Stromalcell-derivedfactor-1chemokinegenevariantinblooddonorsand

chronicmyelogenousleukemiapatients.JClinLabAnal2007;21:49–54.

[23]RazmkhahM,TaleiAR,DoroudchiM,Khalili-AzadT,GhaderiA.Stromal

cell-derivedfactor-1(SDF-1)allelesandsusceptibilitytobreastcarcinoma.Cancer

Lett2005;225:261–6.

[24]RazmkhahM,DoroudchiM,GhayumiSM,ErfaniN,GhaderiA.Stromal

cell-derivedfactor-1(SDF-1)geneandsusceptibilityofIranianpatientswithlung

cancer.LungCancer2005;49:311–5.

[25] FedersppielB,MelhadoIG,DuncanAM,DelaneyA,SchappertK,Clark-LewisI,

etal.MolecularcloningofthecDNAandchromosomallocalizationofthegene

foraputativeseven-transmembranesegment(7-TMS)receptorisolatedfrom

152 Y.-L.Leeetal./LungCancer73 (2011) 147–152

[26]PetersenDC,GlashoffRH,ShresthaS,BergeronJ,LatenA,GoldB,etal.Risk

forHIV-1infectionassociatedwithacommonCXCL12(SDF1)polymorphism

andCXCR4variationinanAfricanpopulation.JAcquirImmuneDeficSyndr

2005;40:521–6.

[27]CheongJY,ChoSW,ChoiJY,LeeJA,KimMH,LeeJE,etal.RANTES,MCP-1,

CCR2,CCR5,CXCR1andCXCR4genepolymorphismsarenotassociatedwith

theoutcomeofhepatitisBvirusinfection:resultsfromalargescalesingle

ethnicpopulation.JKoreanMedSci2007;22:529–35.

[28]HuangYC,HsiaoYC,ChenYJ,WeiYY,LaiTH,TangCH.Stromalcell-derived

factor-1enhancesmotilityandintegrinup-regulationthroughCXCR4,ERKand

NF-kappaB-dependentpathwayinhumanlungcancercells.Biochem

Pharma-col2007;74:1702–12.

[29]BelperioJA,PhillipsRJ,BurdickMD,LutzM,KeaneM,StrieterR.The

SDF-1/CXCL12/CXCR4biologicalaxisinnon-smallcelllungcancermetastases.

Chest2004;125:156S.

[30] PhillipsRJ,BurdickMD,LutzM,BelperioJA,KeaneMP,Strieter RM.The

stromalderivedfactor-1/CXCL12-CXCchemokinereceptor4biologicalaxis

in non-small cell lung cancer metastases. Am J Respir Crit Care Med

2003;167:1676–86.

[31]HirataH,HinodaY,KikunoN,KawamotoK,DahiyaAV,SuehiroY,etal.CXCL12

G801Apolymorphismisariskfactorforsporadicprostatecancersusceptibility.

ClinCancerRes2007;13:5056–62.

[32]Poulain S, Ertault M, Leleu X, Coiteux V, Fernandes J, Stalnikiewicz L,

etal. SDF1/CXCL12(−801GA)polymorphism is a prognosticfactorafter

treatmentinitiationinWaldenstrommacroglobulinemia.LeukRes2009;33:

1204–7.

[33] LinGT,TsengHF,YangCH,HouMF,ChuangLY,TaiHT,etal.Combinational

polymorphismsofsevenCXCL12-relatedgenesareprotectiveagainstbreast

cancerinTaiwan.Omics2009;13:165–72.

[34]ChangSC,LinPC,YangSH,WangHS,LiAF,LinJK.SDF-1alphaG801A

polymor-phismpredictslymphnodemetastasisinstageT3colorectalcancer.AnnSurg

Oncol2009;16:2323–30.

[35] ChangCC,Chen SC,HsiehYH,ChenYC, ChenTY,ChuYH,etal. Stromal

cell-derivedfactor-1butnotitsreceptor,CXCR4,genevariantsincrease

suscep-tibilityandpathologicaldevelopmentofhepatocellularcarcinoma.ClinChem