Genetic Polymorphisms in the Polycomb Group Gene EZH2 and the Risk of Lung Cancer

GeneticPolymorphismsinthePolycombGroupGene

EZH2andtheRiskofLungCancer

Kyong-AhYoon,MD,HyeJinGil,MS,JihyeHan,MS,JaeheePark,MS,andJinSooLee,MD

Introduction:Polycombgroup(PcG)proteinsplayessentialrolesincellularmemorysystemsandascellcycleregulatorsbymain-taininghomeoticgenesintheirsilencedstates.EZH1andEZH2,thehumanhomologuesoftheDrosophilageneEnhancerofZeste(E(z)),aredefinedasPcGproteinsandcontainahighlyconservedmotif,calledtheSET(Su(var)3–9,EnhancerofZeste,Trithorax)domain,whichisrequiredforhistonemethyltransferaseactivity.IncreasedexpressionofthetranscriptionalrepressorEZH2hasbeenreportedtobeassociatedwithpoorprognosisinvariousmalignan-ciesincludingbreastcancerandprostatecancer.AlteredexpressionofEZH2wasalsodemonstratedinlungcancer,suggestinganinvolvementintheprogressionoflungcancer.

Methods:All41polymorphismsinEZH2weregenotypedin335patientswithlungcancerand335age-andgender-matchedhealthycontrols.Finally,26polymorphismswereselectedforthestatisticalanalysisbasedonminorallelefrequency(Ͼ0.05)andlinkagedisequilibrium.

Results:TwopolymorphismsofEZH2,rs6950683andrs3757441,showedastatisticallysignificantassociationwithreducedriskoflungcancer(adjustedOR͓aOR͔ϭ0.71,pϭ0.007;aORϭ0.73,pϭ0.015,respectively).Twocopiesofahaplotype(Ht2)ofEZH2alsoshowedasignificantassociationwithreducedlungcancerrisk(aORϭ0.45,95%confidenceintervalϭ0.23–0.87).

Conclusion:ThisisthefirststudytoshowasignificantassociationbetweenpolymorphismsofthePcGgeneEZH2andlungcancerrisk.ThisstudysuggestsacorrelationbetweenthegenotypevariantsinEZH2andreducedlungcancerriskandhencepresentsapossiblemarkerforlungcancersusceptibility.

KeyWords:EZH2,Singlenucleotidepolymorphism,Lungcancer,Polycombgroup.

(JThoracOncol.2010;5:10–16)

L

ungcanceristhemostcommoncancerandalsotheleadingcauseofcancerdeaththroughouttheworld.Al-thoughcigarettesmokingisawell-knownriskfactorforlung

ResearchInstituteandHospital,NationalCancerCenter,Goyang,Gyeonggi,Korea.

Disclosure:Theauthorsdeclarenoconflictsofinterest.

Addressforcorrespondence:Dr.JinSooLee,NationalCancerCenter,809Madu-dong,Ilsan-gu,Goyang,Gyeonggi411-769,SouthKorea.E-mail:jslee@ncc.re.kr

Copyright©2009bytheInternationalAssociationfortheStudyofLungCancer

ISSN:1556-0864/10/0501-0010

cancer,geneticdiversityplaysanimportantroleinindividualsusceptibilitytolungcancerfollowingexposuretotobaccocarcinogens.1,2CommonpolymorphismsofphaseIandIIenzymesandDNArepairenzymeshavebeenstudiedtoelucidatetheircorrelationswithlungcancersusceptibilitybecauseoftheabilityoftheseenzymestomodifytheeffectoftobaccocarcinogens.3–7Polymorphismsingenesencodingnicotineacetylcholinereceptorswererecentlyidentifiedasgeneticriskfactorsforlungcancer.8–11Inaddition,thepossibleassociationofcancersusceptibilityandgeneticvari-ationsinthegenesinvolvedinhistonemodificationhasbeeninvestigated.12–15Recently,anassociationwasreportedbe-tweencancerandapolymorphismofSMYD3,ahistoneH4lysine4-specificmethyltransferase.12Cebrianetal.13reportedapreliminaryobservationregardingtheassociationofbreastcancerwithvariantsofDNAmethyltransferaseandhistonemethyltransferases(HMTs).WealsoreportedthelungcancerriskassociatedwithpolymorphismsofHMTssuchasSUV39H2andRIZ.14,15EZH2isahumanhomologoftheDrosophilaenhancerofzestehomolog2(ezh2)andisalsoamemberofthefamilyofSET(Su(var)3–9,EnhancerofZeste,Trithorax)domainproteins.16,17TheEZH2geneislocatedonhumanchromo-some7q36andhastwoisoformswithdifferenttranscriptsizes.18EZH2isapolycombgroupproteinthatplaysanimportantroleintheregulationofgenerepressionatthechromatinlevelbymediatingthemethylationoflysine27inhistoneH3.19–21EZH2hasbeenreportedtobeinvolvedinoncogenesisbyinteractingwithregulatorsofcellprolifera-tionsuchasE2Fs,pRB,andcyclinA.22,23OverexpressionofEZH2wasreportedinmetastaticprostatecancerandwasfoundtobeassociatedwithapoorprognosisamongpatientswithprostatecancer.24Consistentwiththegrowth-stimulat-ingeffectofEZH2,knockdownofEZH2expressioninhib-itedtheproliferationofprostatecancercells.Similarresultswerealsoreportedinmultiplemyelomacells,andtheonco-genicactivityofEZH2requiredHMTactivity.25Recently,alteredexpressionlevelsofEZH2anditsroleindiseaseprogressionweredemonstratedinvariouscancers,suchasbreastcancer,pancreaticcancer,andgastriccancer.26–28TheroleofEZH2intumoraggressivenesswasrelatedtotherepressionofseveraltargetgenessuchasE-cadherin,RUNX3,andPSP94bytrimethylationofH3K27.29–31Basedonitsfunctionalsignificanceanditsdysregulatedexpressionpatterninmalignancies,thegeneticvariationsintheEZH2genecouldbeassociatedwithcancersusceptibility.

10

JournalofThoracicOncology•Volume5,Number1,January2010

JournalofThoracicOncology•Volume5,Number1,January2010EZH2PolymorphismsandLungCancer

TotestthehypothesisthatgeneticpolymorphismsoftheEZH2geneareassociatedwiththeriskoflungcancer,weanalyzedthepolymorphismsandhaplotypesofEZH2inaKoreanpopulation.

PATIENTSANDMETHODS

StudyPopulation

Thisstudywasahospital-basedcase-controlstudy.Threehundredthirty-fivecaseswererecruitedfrompatientswithhistologicallyconfirmedlungcancerwhovisitedtheNationalCancerCenterinKoreaandvoluntarilyparticipatedinaquestionnairesurveyconductedfromMay2002toJuly2003.Thesubjectsdonatedbloodforgenetictestsaftersigningtheinformedconsentform,whichhadbeenapprovedbytheinstitutionalreviewboard.Therewerenorecruitmentrestrictionswithregardtogenderorcancerstage,butonlysubjectswhowereyoungerthan70yearswererecruited.Noneofthesecancerpatientshadreceivedchemotherapyorradiotherapybeforetheirrecruitment.Asacontrolgroup,atotalof335controlsubjectswerematchedwithpatientswithlungcancerforage(Ϯ3years)andgender.Thesecontrolsubjectswithoutahistoryofcancerwererecruitedfrompatientswhovisitedourinstitutionforacancerscreeningprogram.Informationondemographiccharacteristicssuchasgender,age,smokinghabits,andfamilyhistoryofcancerwereobtainedfromself-administeredquestionnaires(forcontrols)orapersonalinterview(forcases)administeredbytrainedpersonnelafterwritteninformedconsentwasobtained.

polymorphismwasnotinHWE(PϽ0.01),itwasexcludedfromfurtheranalysis.

Weemployedawidelyusedmeasureoflinkagedis-equilibrium(LD)betweenallpairsofbiallelicloci,Le-wontin’sDЈ(DЈ),andr2.33LDblockswereidentifiedusingHaploviewsoftware.34Haplotypesofeachblockandindivid-ualwereinferredusingthealgorithmdevelopedbyStephensetal.(PHASE),35whichusesaBayesianapproachtoincor-porateaprioriexpectationsforhaplotypereconstruction.Phaseprobabilitiesforeachsitewerecalculatedforeachindividualbythissoftware,andthehaplotypewiththehigh-estprobabilityforeachsamplewasusedforfurtheranalysis.ThegeneticeffectsofinferredhaplotypeswereanalyzedinthesamewayasSNPs.TherelationshipbetweenEZH2polymorphismsandlungcancerriskwasanalyzedusingunconditionalmultiplelogisticregressionmodelswhilecon-trollingforfamilyhistory,pack-years,andsmokingstatusascovariates.WealsoanalyzedtheassociationbetweenEZH2polymorphismsandtheriskofnon-smallcelllungcancer(NSCLC)andsmallcelllungcancer(SCLC),thetwodistincthistologictypesinlungcancer.Analysesfortheassociationbetweenhaplotypesandlungcancerriskwereperformedusingunconditionallogisticregressionattheindividuallevel,inwhichthecovariatewasdefinedbythenumberofcopies(0,1,or2)ofeachhaplotypethatasubjectcarried,andthemodelalsoincludedfamilyhistoryofcancer,smokingstatus,andpack-yearsascovariates.Allreportedpvaluesaretwo-sided.ThestatisticalsoftwareStata/SEversion10wasusedforstatisticalanalyses(StataCorpLP,CollegeStation,TX).

GenotypingofEZH2Polymorphisms

Atotalof41singlenucleotidepolymorphisms(SNPs)oftheEZH2gene(NM_004456)wereselectedfromtheInternationalHapMapProjectdata(www.hapmap.org)forthisstudy.GenomicDNAwasextractedfromtheperipheralbloodusingaQIAampDNABloodMiniKit(Qiagen,Va-lencia,CA)inaccordancewiththemanufacturer’sinstruc-tions.ThehighlymultiplexedSNPGoldenGategenotypingassay(IlluminaInc.,SanDiego,CA)wasperformed,whichcombinesoligonucleotideligationwithanallele-specificex-tensionreaction.32RESULTS

ThedemographicfeaturesofpatientswithlungcancerandhealthycontrolsareshowninTable1.Lungcancerpatientsweremorelikelythanthecontrolstohaveeversmoked(pϽ0.001).Thedistributionoffamilyhistoryofcancerwassignificantlydifferentbetweenthecasesandcontrols(pϽ0.05).Ofthe335patientswithlungcancer,189(56%)hadadenocarcinomas,77(23%)hadsquamouscellcarcinomas,and29(9%)hadSCLC.

Atotalof41SNPsoftheEZH2gene(NM_012231)wereselectedfromtheInternationalHapMapProjectdataandanalyzedtomeasuretheLDaftergenotyping(Figure1A).ArepresentativeSNPwasonlyselectediftherewereabsoluteLDs(r2ϭ1);26SNPswerefinallyselectedforthestatisticalanalysisbasedonminorallelefrequencyandLDs.Thecommonhaplotypes(Ͼ6%)ofEZH2polymorphismsareindicatedinFigure1B.TheLDblockwasdeterminedfromtheresultsofLDanalysisusingtheHaploviewsoftware(Figure1C).

Table2presentstheminorallelefrequenciesofthe26SNPsamongpatientswithlungcancerandnormalcontrolsandtheestimatedoddsratiosoflungcancerriskbetweensubjectscarryingonevariantalleleandsubjectscarryingnovariantalleleofeachSNPbasedonamultiplelogisticregressionmodel(log-additivemodel)controllingforsmok-ingstatus,pack-years,andthefamilyhistoryofcancerascovariates.

StatisticalAnalysis

Totestforthedifferencesindemographiccharacteris-ticsbetweenlungcancercasesandcontrols,Pearson’s␹2testwasusedforcategoricalvariables,andtheWilcoxonranksumtestwasusedforcontinuousvariables.Withregardtosmokinghabits,smokingstatus,thenumberofcigarettessmokedperday,andthetimeofstartingandquittingwereinvestigated.Individualswhohadeitherformerlysmokedmorethan100cigarettesduringtheirlifetimeorcurrentlysmokedweredefinedaseversmokers.Asameasureofcumulativesmokingexposure,pack-yearsweredefinedastheaveragenumberofpacks(20cigarettes/pack)ofcigarettessmokedperdaymultipliedbythetotalnumberofyearsofsmoking.

TheHardy-Weinbergequilibrium(HWE)wastestedforthecontrolgenotypingresults.AplevelϽ0.01wasacceptedasstatisticallysignificantfortheHWEtest.Ifa

Copyright©2009bytheInternationalAssociationfortheStudyofLungCancer

11

Yoonetal.JournalofThoracicOncology•Volume5,Number1,January2010

TABLE1.DemographicCharacteristicsofPatientswithLungCancerandNormalControls

LungCancerCases

Phenotype

Age(median,range)Gender(n,%)MaleFemale

Smokingstatus(n,%)aNeversmokerEversmokerCurrentFormerUnknown

Pack-years(median,range)bFamilyhistory(n,%)aYesNo

Unknown

aNormalControls(n؍335)

58(28–73)226(67.5%)109(32.5%)139(41.5%)189(56.4%)86(25.7%)103(30.7%)

7

21(0.3–126)166(49.6%)167(49.9%)

2

All(n؍335)58(25–70)226(67.5%)109(32.5%)116(34.6%)218(65.1%)143(42.7%)75(22.4%)

1

34.5(0.08–135)128(38.3%)205(61.4%)

2

NSCLC(n؍306)

58(25–70)204(66.7%)102(33.3%)112(36.6%)193(63.1%)123(40.2%)70(22.9%)

1

34.75(0.08–135)119(38.9%)185(60.5%)

2

SCLC(n؍29)60(37–70)22(75.9%)7(24.1%)4(13.8%)25(86.2%)20(69.0%)5(17.2%)34(0.35–114)9(31.0%)20(69.0%)

pϽ0.05inPearson’s␹2testforthedifferencebetweenlungcancerpatientsandcontrols.bPack-yearsofsmokingwerecalculatedforeversmokersonly.pϽ0.05inWilcoxonranksumtestforthedifferencebetweenlungcancerpatientsandcontrol.

NSCLC,non-smallcelllungcancer;SCLC,smallcelllungcancer.

FIGURE1.Genemaps,haplotypes,andlinkagedisequilibriums(LDs)ofEZH2singlenucleotidepolymorphisms(SNPs).Codingexonsaremarkedbyblackblocksand5Јand3ЈUTRsbywhiteblocks.Thefirstbaseofthetranslationalstartsiteisdenotedasnucleotideϩ1.AsterisksindicatetheSNPsthatwereusedforthestatisticalanalysis.A,SNPsgenotypedinthestudysubjects.ThefrequenciesoftheSNPswerebasedonfindingsin670subjects(335patientswithlungcancerand335normalcontrols).B,FivecommonhaplotypesofEZH2withafrequencyofϾ0.06.C,LDblockinEZH2identi-fiedusingHaploview.

12

Copyright©2009bytheInternationalAssociationfortheStudyofLungCancer

JournalofThoracicOncology•Volume5,Number1,January2010EZH2PolymorphismsandLungCancer

TABLE2.SubgroupAnalysisofEZH2PolymorphismsAssociatedwiththeRiskofLungCancer

MinorAlleleFrequency

Lung

ControlsCancerNSCLCSCLC

Alleles(n؍335)(n؍335)(n؍306)(n؍29)G:AC:TG:CT:CT:CT:AA:CG:CC:TG:AC:TT:CG:CT:CA:CC:TT:GC:TG:AC:TT:GG:AT:CG:AG:CA:G

0.30.4790.1670.3540.3540.3540.4790.3540.3520.1690.3520.1690.0820.2960.1670.4780.1730.0570.1540.3540.0570.3520.3520.3540.3550.355

0.2460.5180.1730.3090.3100.3100.5160.3120.3120.1730.3100.1780.0790.2370.1730.5090.1810.0600.1630.3070.0660.3070.3090.3090.3090.310

0.2430.5160.1720.3120.3140.3140.5150.3150.3150.1720.3140.1760.080.2450.1720.5070.180.0570.160.310.0590.310.3120.3120.3120.314

0.1550.5340.1900.2760.2760.2760.5340.2760.2760.1900.2760.1900.0690.1550.1900.5340.1900.0860.1900.2760.1380.2760.2760.2760.2760.276

Lungcancervs.

ControlsOR(95%CI)

p

NSCLCvs.ControlsOR(95%CI)

p

SCLCvs.ControlsOR(95%CI)

p

rs#rs6950683rs2177567rs1880357rs1880358rs10233740rs1880355rs10952780rs6975291rs9691534rs28723387rs6959647rs10488070rs2302427rs3757441rs1061037rs6464926rs17171119rs17551792rs3735219rs2072407rs41277434rs740949rs734005rs734002rs734004rs887569

Position5ЈneargeneIntron1Intron1Intron1Intron1Intron1Intron1Intron1Intron1Intron2Intron3Intron4Exon6Intron6Intron8Intron8Intron8Intron10Intron12Intron15Intron18Intron19Intron19Intron19Intron19Intron19

0.71(0.55–0.91)0.0070.74(0.57–0.95)1.19(0.95–1.48)1.01(0.75–1.34)0.82(0.65–1.03)0.82(0.65–1.04)0.82(0.65–1.04)1.18(0.95–1.47)0.83(0.66–1.05)0.84(0.66–1.06)1.00(0.75–1.33)0.83(0.66–1.05)1.03(0.77–1.37)0.95(0.64–1.41)0.73(0.57–0.94)1.00(0.75–1.34)1.15(0.92–1.43)1.03(0.78–1.37)1.02(0.64–1.64)1.03(0.77–1.40)0.81(0.64–1.02)1.13(0.72–1.78)0.82(0.65–1.03)0.82(0.65–1.04)0.82(0.65–1.03)0.81(0.64–1.02)0.82(0.65–1.03)

0.1220.9600.0930.1050.1020.1330.1150.1320.9900.1180.8580.8060.0150.9740.2110.8260.9190.8290.0780.5910.0900.0970.0880.0780.089

1.17(0.94–1.47)1.01(0.75–1.35)0.83(0.65–1.05)0.84(0.66–1.06)0.83(0.66–1.06)1.17(0.94–1.47)0.84(0.66–1.07)0.85(0.67–1.08)1.00(0.74–1.34)0.84(0.66–1.07)1.03(0.77–1.38)0.97(0.65–1.45)0.77(0.59–0.99)1.00(0.75–1.35)1.13(0.91–1.42)1.04(0.78–1.39)0.99(0.61–1.60)1.03(0.75–1.39)0.82(0.65–1.04)1.01(0.63–1.64)0.83(0.65–1.05)0.83(0.66–1.06)0.83(0.65–1.05)0.82(90.65–1.04)0.83(0.65–1.05)

0.0210.40(0.19–0.83)0.0140.1540.9680.1240.1390.1360.1680.1530.1740.9840.1550.8510.8860.0380.9790.2670.8090.9640.8730.1040.9540.1180.1290.1180.1060.120

1.39(0.78–2.45)1.06(0.52–2.15)0.68(0.37–1.24)0.68(0.37–1.24)0.68(0.37–1.24)1.38(0.78–2.45)0.68(0.37–1.24)0.68(0.37–1.26)1.05(0.52–2.15)0.68(0.37–1.26)1.02(0.51–2.06)0.77(0.25–2.35)0.41(0.20–0.85)1.03(0.51–2.07)1.39(0.79–2.44)1.01(0.50–2.03)1.57(0.55–4.47)1.25(0.59–2.62)0.68(0.37–1.24)2.76(1.16–6.53)0.68(0.37–1.26)0.68(0.37–1.24)0.68(0.37–1.24)0.67(0.37–1.24)0.67(0.37–1.24)

0.2610.8820.2080.2080.2080.2610.2080.2200.8870.2200.9460.6480.0170.9340.2510.9790.3940.5590.2070.0210.2250.2090.2070.2040.204

TheORs(95%CI)andcorrespondingpvalueswereestimatedfromalog-additivemodelusingunconditionalmultiplelogisticregression,controllingforfamilyhistory,smokingstatus,andpack-yearsascovariates.pvaluesunder0.05areindicatedinboldfont.

NSCLC,non-smallcelllungcancer;SCLC,smallcelllungcancer;CI,confidenceinterval.

Among26SNPs,thevariantgenotypesofrs6950683andrs3757441showedasignificantassociationwithreducedlungcancerrisk(adjustedOR͓aOR͔ϭ0.71,95%confidenceinterval͓CI͔ϭ0.55–0.91,pϭ0.007andaORϭ0.73,95%CIϭ0.57–0.94,pϭ0.015,respectively).SubsetanalyseswereperformedfortheNSCLCgroupandtheSCLCgroupbasedonhistologiccelltype.Theassociationwithade-creasedriskoflungcancerwasmuchstrongerintheSCLCgroupthanintheNSCLCgroup(aORϭ0.40,95%CIϭ0.19–0.83,pϭ0.014andaORϭ0.41,95%CIϭ0.20–0.85,pϭ0.017).Specifically,rs41277434wasassociatedwithanincreasedriskofSCLCeventhoughthesamplesizeforSCLCwaslimited(aORϭ2.76,pϭ0.021).Theeffectsofrs6950683,rs3757441,andrs41277434werefurtherana-lyzedusingalternativemodelssuchascodominant,domi-nant,andrecessivemodels(Table3).Theeffectofrs6950683onlungcancerriskwasalsoanalyzedinadenocarcinomagroupandnever-smokersgroup(aORϭ0.81,pϭ0.16andaORϭ0.81,pϭ0.32,respectively).Lungcancerriskwasdecreasedinbothgroups,however,itwasnotstatisticallysignificant.The

genotypedistributionofrs6950683wasnotdifferentbetweeneversmokersandneversmokers(pϭ0.73).

From26SNPsoftheEZH2gene,fivecommonhaplo-types(Ͼ6%frequency)wereidentifiedwithacumulativefrequencyof91%inthecontrols(Figure1B).ThecommonhaplotypeswerelabeledandcategorizedasHt1toHt5basedontheircomputer-estimatedfrequency,andotherrarehaplo-typesweregroupedtogetherintheanalysesasshowninTable4.SubjectscarryingtwocopiesofHt2haplotypeshowedasignificantlydecreasedriskoflungcancer(aORϭ0.45,95%CIϭ0.23–0.87).

DISCUSSION

PolymorphismsoftheEZH2geneshowedasignificantassociationwithlungcancerinthiscase-controlstudy.EZH2isapolycombgroupproteincontainingaSETdomainthathasHMTactivity.IncreasedexpressionofEZH2wasre-portedalongwithagrowth-stimulatingeffectandincreasedtumoraggressivenessinvariouscancers.26–29Thefunctional

Copyright©2009bytheInternationalAssociationfortheStudyofLungCancer

13

Yoonetal.JournalofThoracicOncology•Volume5,Number1,January2010

TABLE3.AssociationofThreePolymorphismsofEZH2withLungCancerRiskAccordingtoHistologicTypeinPatientswithLungCancer

Frequency

SNP

Overall

rs6950683

GenotypeGGAGAATTCTCCTTGTGGGGAGAATTCTCCTTGTGGGGAGAATTCTCCTTGTGG

Controls166137321691343229836116613732169134322983611661373216913432298361

LC19412417193125172934021741151717311617271341209020902261

CodominantModelOR(95%CI)1

0.77(0.55–1.06)0.44(0.23–0.83)

1

0.81(0.58–1.12)0.45(0.24–0.85)

1

1.09(0.67–1.77)2.09(0.18–23.91)

1

0.79(0.57–1.10)0.49(0.26–0.93)

1

0.83(0.60–1.16)0.50(0.26–0.96)

1

1.00(0.60–1.66)1.32(0.08–21.58)

1

0.54(0.23–1.27)1

0.56(0.24–1.31)1

2.30(0.83–6.34)19.27(0.95–390.79)

p0.1080.0120.2020.0140.7330.5540.1620.030.2860.0360.9960.8460.182

DominantModelOR(95%CI)1

0.70(0.52–0.96)1

0.74(0.54–1.01)1

1.12(0.69–1.80)1

0.73(0.53–1.01)1

0.77(0.56–1.06)

p0.027

0.49(0.26–0.92)

1

0.058

0.49(0.26–0.92)

1

0.656

2.07(0.18–23.67)

1

0.055

0.54(0.29–1.01)

1

0.109

0.54(0.29–1.01)

1

1.01(0.61–1.66)1

0.41(0.18–0.95)1

0.42(0.18–0.98)1

2.66(1.01–6.98)

0.979

1.32(0.08–21.57)

1

0.038

1

0.044

1

0.047

16.89(0.85–333.86)

0.0630.8460.0560.0560.560.0260.026

RecessiveModelOR(95%CI)

1

p

rs3757441

rs41277434

NSCLCrs6950683

rs3757441

rs41277434

SCLCrs6950683

rs3757441

0.182

rs41277434

0.1090.054

Correctedpvalueswereobtainedusingthecorrectionformultipletests.TheORs(95%CI)andcorrespondingpvalueswerederivedfromalogisticanalysiscontrollingforsmokingstatus,pack-years,andfamilyhistoryascovariates.Oddsratiosthatarestatisticallysignificant(pϽ0.05)areindicatedinboldfont.

NSCLC,non-smallcelllungcancer;SCLC,smallcelllungcancer;CI,confidenceinterval;LC,lungcancercases.

significanceandalteredexpressionlevelofEZH2makeitfeasibletoevaluatetheassociationbetweenthegene’spoly-morphismsandcancersusceptibility.

Inthishospital-basedcase-controlstudy,41polymor-phismsofEZH2weregenotypedin335patientswithlungcancerand335healthycontrols.Controlsrecruitedamongthecancerscreeneesatourhospitalshowedahigherpreva-lenceofafamilyhistoryofcancerthandidthelungcancercases.Therefore,theassociationbetweentheSNPsandtheriskoflungcancerwasanalyzedbycontrollingforfamilyhistoryofcancer,aswellasotherpotentialconfoundingvariablessuchassmokingstatusandpack-years.Twopoly-morphismsofEZH2,rs6950683andrs3757441,showedaprotectiveeffectonlungcancerrisk.ThereducedlungcancerriskassociatedwithtwopolymorphismswassignificantintheNSCLCgroupaswellasintheSCLCgroup.AlthoughtheNSCLCandSCLCgroupspresentdifferentdemographicfeatures,asshowninTable1,theprotectiveeffectofEZH2polymorphismswasdemonstratedinbothofthesedistincthistologictypes.EventhoughthenumberofpatientswithSCLC(nϭ29)wasmuchsmallerthanthatofNSCLCpatients(nϭ306),theSCLCgrouphadmoremenandmorecurrentsmokersthantheNSCLCgroup.

Inthecaseofrs41277434,althoughanincreasedriskoflungcancerwasobservedintheSCLCgroup,thesamplesizewasnotsufficienttocomparewiththoseofthecontrolgroup.

Becausers6950683islocatednearexon1,itmayimpactgeneexpressionbyaffectingthepromoterregion.However,furtherbiologicand/orfunctionalevidenceisneededtoconfirmthegeneticeffectsofEZH2polymor-phismsonlungcancer.

BecauseEZH2isknownasapolycombgroupgenethatplaysanimportantroleincellcycleregulation,itcanbehypothesizedthatEZH2polymorphismshaveasignificantassociationwithothercancersbesideslungcancer.Recently,EZH2mediatedhistoneH3lysine27trimethylationwasreportedasamechanismoftumorsuppressorgenesilenc-ing.36OverexpressedEZH2wasassociatedwithpoorprog-nosisandEZH2knockdowninducedincreaseoftargetgenesinvolvedincellgrowthanddifferentiation.20,36BecauseofitsHMTactivitywithsubstratespecificityforhistoneH3lysine27,thegeneticvariationsintheEZH2genecouldbeassoci-

14

Copyright©2009bytheInternationalAssociationfortheStudyofLungCancer

JournalofThoracicOncology•Volume5,Number1,January2010EZH2PolymorphismsandLungCancer

TABLE4.AssociationAnalysisofHaplotypesofEZH2andLungCancerRiskAccordingtoHistologicCellTypeinPatientswithLungCancer

FrequencyN,(%)

HaplotypeHt1

0copy1copy2copiesHt2

0copy1copy2copiesHt3

0copy1copy2copiesHt4

0copy1copy2copiesHt5

0copy1copy2copiesOthers0copy1copy2copies

Controls(n؍335)1131635917213330283502287480300341279542

LungCancer(n؍335)

1111497519812116286463281522289451278516

NSCLC(n؍306)1011366917811216261423259452265401256455

SCLC(n؍29)1013620902540227024502261

Lungcancervs.

ControlsOR(95%CI)1

0.91(0.64–1.29)1.39(0.89–2.16)1

0.78(0.56–1.08)0.45(0.23–0.87)1

0.88(0.57–1.37)1.68(0.27–10.53)1

1.03(0.67–1.58)

p

NSCLCvs.ControlsOR(95%CI)1

0.91(0.64–1.30)1.38(0.88–2.17)1

0.80(0.57–1.12)0.50(0.26–0.97)1

0.89(0.57–1.40)1.80(0.29–11.23)1

0.98(0.63–1.53)

p

SCLCvs.ControlsOR(95%CI)1

0.95(0.39–2.32)1.44(0.47–4.35)1

0.56(0.24–1.31)

p

0.60.1460.610.1560.9140.522

0.1360.0180.1910.041

0.181

0.5750.5790.6190.527

1

0.78(0.25–2.42)

0.672

0.9050.923

1

1.86(0.73–4.80)

0.19

1

1.40(0.86–2.27)1.69(0.10–27.42)1

0.91(0.59–1.39)2.96(0.58–15.05)

0.1740.712

1

1.36(0.83–2.23)1.75(0.11–28.38)1

0.87(0.56–1.34)2.77(0.53–14.64)

0.2260.694

1

2.35(0.79–7.00)

0.123

0.6520.1910.5230.23

1

1.18(0.44–3.16)10.51(0.77–143.06)

0.7390.078

TheORs(95%CI)andcorrespondingpvalueswerederivedfromlogisticanalysiscontrollingforsmokingstatus,pack-years,andfamilyhistoryascovariates.NSCLC,non-smallcelllungcancer;SCLC,smallcelllungcancer;CI,confidenceinterval.

atedwithdiversityofenzymaticactivityandcancersuscep-tibility.Furtherepidemiologicstudiesinlargerpopulationsarerequiredtotestthishypothesis.

DespitetheimportantdiscoveryofaprotectiveeffectoftheEZH2polymorphisms,thisstudyonlyconsidersaKoreanpopulation,whichmaylimittheapplicationofthesefindingstootherethnicpopulations.Furthermore,thesubsetanalysisaccordingtohistologictypewaslimitedbysamplesize,particularlyforSCLC.

Inconclusion,thisisthefirststudytoshowasignificantassociationbetweenpolymorphismsoftheEZH2geneandlungcancerrisk.TheseresultssuggestthatthepresenceofthevariantalleleinEZH2maybeaprotectivefactorforthedevelopmentoflungcancerandcouldbeamarkerassociatedwithgeneticsusceptibilitytolungcancer.

SupportedbyaNationalCancerCenterResearchGrant(0710221)(toJ.S.L.).

REFERENCES

1.MillerMCIII,MohrenweiserHW,BellDA.Geneticvariabilityinsusceptibilityandresponsetotoxicants.ToxicolLett2001;120:269–280.

ACKNOWLEDGMENTS

2.Ingelman-SundbergM.Geneticvariabilityinsusceptibilityandresponsetotoxicants.ToxicolLett2001;120:259–268.

3.LarsenJE,ColosimoML,YangIA,BowmanR,ZimmermanPV,FongKM.CYP1A1Ile462ValandMPOG-463Ainteracttoincreaseriskofadenocarcinomabutnotsquamouscellcarcinomaofthelung.Carcino-genesis2006;27:525–532.

4.WenzlaffAS,CoteML,BockCH,etal.CYP1A1andCYP1B1polymorphismsandriskoflungcanceramongneversmokers:apopu-lation-basedstudy.Carcinogenesis2005;26:2207–2212.

5.WenzlaffAS,CoteML,BockCH,LandSJ,SchwartzAG.GSTM1,GSTT1andGSTP1polymorphisms,environmentaltobaccosmokeex-posureandriskoflungcanceramongneversmokers:apopulation-basedstudy.Carcinogenesis2005;26:395–401.

6.LeMarchandL,DonlonT,Lum-JonesA,SeifriedA,WilkensLR.AssociationofthehOGG1Ser326Cyspolymorphismwithlungcancerrisk.CancerEpidemiolBiomarkersPrev2002;11:409–412.

7.ItoH,MatsuoK,HamajimaN,etal.Gene-environmentinteractionsbetweenthesmokinghabitandpolymorphismsintheDNArepairgenes,APE1Asp148GluandXRCC1Arg399Gln,inJapaneselungcancerrisk.Carcinogenesis2004;25:1395–1401.

8.AmosCI,WuX,BroderickP,etal.Genome-wideassociationscanoftagSNPsidentifiesasusceptibilitylocusforlungcancerat15q25.1.NatGenet2008;40:616–622.

9.HungRJ,McKayJD,GaborieauV,etal.Asusceptibilitylocusforlungcancermapstonicotinicacetylcholinereceptorsubunitgeneson15q25.Nature2008;452:633–637.

10.ThorgeirssonTE,GellerF,SulemP,etal.Avariantassociatedwith

Copyright©2009bytheInternationalAssociationfortheStudyofLungCancer

15

Yoonetal.JournalofThoracicOncology•Volume5,Number1,January2010

11.12.

13.14.15.16.

17.

18.

19.20.21.22.23.

nicotinedependence,lungcancerandperipheralarterialdisease.Nature2008;452:638–642.

ShiraishiK,KohnoT,KunitohH,etal.Contributionofnicotineacetylcholinereceptorpolymorphismstolungcancerriskinasmoking-independentmannerintheJapanese.Carcinogenesis2009;30:65–70.TsugeM,HamamotoR,SilvaFP,etal.AvariablenumberoftandemrepeatspolymorphisminanE2F-1bindingelementinthe5ЈflankingregionofSMYD3isariskfactorforhumancancers.NatGenet2005;37:1104–1107.

CebrianA,PharoahPD,AhmedS,etal.Geneticvariantsinepigeneticgenesandbreastcancerrisk.Carcinogenesis2006;27:1661–1669.

YoonKA,HwangboB,KimIJ,etal.NovelpolymorphismsintheSUV39H2histonemethyltransferaseandtheriskoflungcancer.Car-cinogenesis2006;27:2217–2222.

YoonKA,ParkS,HwangboB,etal.GeneticpolymorphismsintheRb-bindingzincfingergeneRIZandtheriskoflungcancer.Carcino-genesis2007;28:1971–1977.

LaibleG,WolfA,DornR,etal.MammalianhomologuesofthePolycomb-groupgeneEnhancerofzestemediategenesilencinginDrosophilaheterochromatinandatS.cerevisiaetelomeres.EmboJ1997;16:3219–3232.

CardosoC,TimsitS,VillardL,KhrestchatiskyM,Fonte`sM,ColleauxL.SpecificinteractionbetweentheXNP/ATR-XgeneproductandtheSETdomainofthehumanEZH2protein.HumMolGenet1998;7:679–684.

CardosoC,MignonC,HetetG,GrandchampsB,FontesM,ColleauxL.ThehumanEZH2gene:genomicorganisationandrevisedmappingin7q35withinthecriticalregionformalignantmyeloiddisorders.EurJHumGenet2000;8:174–180.

CaoR,WangL,WangH,etal.RoleofhistoneH3lysine27methylationinPolycomb-groupsilencing.Science2002;298:1039–1043.

CaoR,ZhangY.ThefunctionsofE(Z)/EZH2-mediatedmethylationoflysine27inhistoneH3.CurrOpinGenetDev2004;14:155–164.

KirmizisA,BartleySM,KuzmichevA,etal.SilencingofhumanpolycombtargetgenesisassociatedwithmethylationofhistoneH3Lys27.GenesDev2004;18:1592–1605.

BrackenAP,PasiniD,CapraM,ProsperiniE,ColliE,HelinK.EZH2isdownstreamofthepRB-E2Fpathway,essentialforproliferationandamplifiedincancer.EmboJ2003;22:5323–5335.

ToniniT,BagellaL,D’AndrilliG,ClaudioPP,GiordanoA.Ezh2

24.25.26.27.28.29.30.31.32.33.34.35.36.

reducestheabilityofHDAC1-dependentpRb2/p130transcriptionalrepressionofcyclinA.Oncogene2004;23:4930–4937.

VaramballyS,DhanasekaranSM,ZhouM,etal.ThepolycombgroupproteinEZH2isinvolvedinprogressionofprostatecancer.Nature2002;419:624–629.

CroonquistPA,VanNessB.Thepolycombgroupproteinenhancerofzestehomolog2(EZH2)isanoncogenethatinfluencesmyelomacellgrowthandthemutantrasphenotype.Oncogene2005;24:6269–6280.FujiiS,OchiaiA.Enhancerofzestehomolog2downregulatesE-cadherinbymediatinghistoneH3methylationingastriccancercells.CancerSci2008;99:738–746.

GonzalezME,LiX,ToyK,etal.DownregulationofEZH2decreasesgrowthofestrogenreceptor-negativeinvasivebreastcarcinomaandrequiresBRCA1.Oncogene2009;28:843–853.

OugolkovAV,BilimVN,BilladeauDD.Regulationofpancreatictumorcellproliferationandchemoresistancebythehistonemethyltransferaseenhancerofzestehomologue2.ClinCancerRes2008;14:6790–6796.FujiiS,ItoK,ItoY,OchiaiA.Enhancerofzestehomologue2(EZH2)down-regulatesRUNX3byincreasinghistoneH3methylation.JBiolChem2008;283:17324–17332.

CaoQ,YuJ,DhanasekaranSM,etal.RepressionofE-cadherinbythepolycombgroupproteinEZH2incancer.Oncogene2008;27:7274–7284.

BekeL,NuyttenM,VanEyndeA,BeullensM,BollenM.ThegeneencodingtheprostatictumorsuppressorPSP94isatargetforrepressionbythePolycombgroupproteinEZH2.Oncogene2007;26:4590–4595.OliphantA,BarkerDL,StuelpnagelJR,CheeMS.BeadArraytechnol-ogy:enablinganaccurate,cost-effectiveapproachtohigh-throughputgenotyping.Biotechniques2002;Suppl:56–58,60–61.

HedrickPW.Gameticdisequilibriummeasures:proceedwithcaution.Genetics1987;117:331–341.

BarrettJC,FryB,MallerJ,DalyMJ.Haploview:analysisandvisual-izationofLDandhaplotypemaps.Bioinformatics2005;21:263–265.StephensM,SmithNJ,DonnellyP.Anewstatisticalmethodforhaplotypereconstructionfrompopulationdata.AmJHumGenet2001;68:978–989.

KondoY,ShenL,ChengAS,etal.GenesilencingincancerbyhistoneH3lysine27trimethylationindependentofpromoterDNAmethylation.NatGenet2008;40:741–750.

16

Copyright©2009bytheInternationalAssociationfortheStudyofLungCancer