Genetic polymorphisms of glutathione S-transferase genes GSTM1, GSTT1 and risk of CHD

Mutagenesispp.1–5,2010Mutagenesis Advance Access published March 30, 2010 doi:10.1093/mutage/geq014GeneticpolymorphismsofglutathioneS-transferasegenesGSTM1,GSTT1andriskofcoronaryheartdisease

JunWang,LiangJianZou,ShengDongHuang,

FangLinLu,XiLongLang,LinHan,ZhiGangSongand

ZhiYunXu*

DepartmentofCardiothoracicSurgery,ChanghaiHospital,SecondMilitary

MedicalUniversity,Shanghai,China.

Towhomcorrespondenceshouldbeaddressed.ZhiYunXu,Departmentof

CardiothoracicSurgery,ChanghaiHospital,SecondMilitaryMedicalUniversity,

Shanghai200433,China.Tel:þ86-21-81873436;Fax:þ86-21-65490979;

Email:zhiyunxu001@*pathogenesisofatherosclerosissuggestthattheoccurrenceofDNAalterationscontributetothemultifacetedpathogenesisoftheatherogenicprocess.Inparticular,deletionsormutationsofgenecodingforenzymesinvolvedinthemetabolismofhazardouscompoundsmayberesponsibleforindividualsusceptibilitytogenotoxicfactors,predisposingtothede-velopmentofDNAinsults(3,4).TheglutathioneS-transferases(GSTs)areagenesuperfam-ilyofphaseIImetabolicenzymesthatdetoxifyfreeradicals,particularlyintobaccosmoke,productsofoxidativestressandDownloaded from

carcinogenssuchasbenzopyreneandotherpolycyclic

ReceivedonFebruary24,2009;revisedonFebruary24,2010;aromatichydrocarbons(5).InadditiontotheirroleinphaseacceptedonMarch2,2010IIdetoxi cation,GSTsalsoplayanimportantroleinToclarifytheroleofglutathioneS-transferases(GSTs;modulatingtheinductionofotherenzymesandproteinsforGSTM1andGSTT1)statusinsusceptibilitytocoronarycellularfunctions,suchasDNArepair(6).GSTM1andheartdisease(CHD),ameta-analysisofpublishedstudiesGSTT1arethemostextensivelystudiedgenesintheGSTgenewasperformed.Atotalof19studiesincluding8020casessuperfamily.PolymorphicdeletionvariantsintheGSTM1andand11501controlswereincludedinthismeta-analysis.InGSTT1genesproduceeitherafunctionalenzyme(non-acombinedanalysis,therelativerisksforCHDofthedeletionallelesorheterozygousdeletion,GSTM1-1andGSTM1nullandGSTT1nullpolymorphismswere1.47GSTT1-1)orresultinthecompleteabsenceoftheenzyme

[95%con denceinterval(CI):1.08–2.01]and1.26(95%(homozygousdeletionalleles,GSTM1-nullandGSTT1-null)CI:0.90–1.75),respectively.Threepotentialsourcesof(7).Therefore,theseenzymesmayberelatedtotheriskforheterogeneityincludingethnicity,sourceofcontrolandatherosclerosisandCHD(8).

samplesizeofstudywerealsoassessed.However,noIndeed,severalstudieshaveinvestigatedtheassociationssigni cantassociationwasfoundinstrati edanalyses.BybetweentheGSTT1andGSTM1nullgenotypesandCHDpoolingdatafromeightstudies(2909casesand3745susceptibility.However,thesestudieshaveyieldedcontradic-controls)thatconsideredcombinationsofGSTT1andtoryresults,withsomestudiesshowingasigni cantassociation,GSTM1genotypes,astatisticallysigni cantincreasedriskwhileothersshowingnosuchassociation.SuchinconsistencyforCHD[oddsratio(OR52.38,95%CI:1.03–5.48)]wascouldbeduetothesmalleffectofthepolymorphismonCHDdetectedforindividualswithcombineddeletionmutationsriskandtherelativelysmallsamplesizeineachofthepublishedinbothgenescomparedwithpositivegenotypes.Resultsstudies.Wethereforeperformedameta-analysisofthefromthemeta-analysisof vestudiesonGSTsstrati edpublishedstudiestoclarifythisinconsistencyandtoestablishaccordingtosmokingstatusshowedanincreasedriskforacomprehensivepictureoftherelationshipbetweenGSTM1,individualswithnullgenotype(OR52.21,95%CI:GSTT1andCHD.

1.24–3.92forGSTM1andOR53.29,95%CI:1.49–7.26

forGSTT1)versusnon-nullgenotypes.Thismeta-analysisMaterialsandmethodssuggeststhattheGSTM1nullgenotypemayslightly

increasetheriskofCHDandthatinteractionbetweenLiteraturesearchanddataextraction

unfavourableGSTsgenotypesmayexist.PaperspublishedbeforetheendofNovember2009wereidenti edthrough

asearchofPubmedandEmbaseusingthefollowingterms‘glutathione

S-transferases’or‘GST’and‘coronaryheartdisease’or‘CHD’,without

restrictiononlanguage.Allreferencescitedinthesestudiesandpreviously

publishedreviewarticleswerereviewedtoidentifyadditionalworknotindexed

IntroductionbyMEDLINE.OnlythosestudiesassessingtheassociationbetweentheCHDand

theGSTsgenepolymorphismswereincluded.Theinclusioncriteriawere(i)

Coronaryheartdisease(CHD),includingitsmostsevereoriginalpaperscontainingindependentdata,(ii)identi cationofCHDwascomplication,myocardialinfarction,istheleadingcauseofcon rmedpathologically,(iii)suf cientdatatocalculatetheoddsratio(OR)or

P-valueand(iv)case–controlorcohortstudies.Themajorreasonsforexclusion

deathintheindustrializedworld(1).Traditionalrisksuchasofstudieswere(i)overlappingdataand(ii)case-onlystudies,family-basedhypertension,diabetesmellitus,dyslipidemiaandsmokingcanstudiesandreviewarticles.

onlyexplainapproximatelytwo-thirdsoftheobservedclinicalForeachstudy,thefollowinginformationwasextractedindependentlybyevents.Thishasmaintainedinterestinotherbiochemicalandtwoinvestigators: rstauthor’ssurname,publicationdate,gender,ethnicity,

genotypingmethod,cigarettesmokingstatus,clinicalcharacteristics,con r-geneticfactorsthatmightcontributetotheunderlyingmationofdiagnosis,totalnumberofcasesandcontrols.Theresultswere

pathophysiologyofvasculardisease(2).Recentstudiescomparedanddisagreementswerediscussedandresolvedwithconsensus.addressingtheissueofacquiredDNAmutationsintheWhereessentialinformationwasnotpresentedinarticles,everyeffortwas

madetocontacttheauthors.

ÓTheAuthor2010.PublishedbyOxfordUniversityPressonbehalfoftheUKEnvironmentalMutagenSociety.

Allrightsreserved.Forpermissions,pleasee-mail:journals.permissions@.1 at The 3rd Military Medical University on April 1, 2010

J.Wangetal.

Statisticalanalysis

FortheGSTM1andGSTT1gene,weestimatedtherisksofthenullgenotype

onCHDcomparedwiththenon-nullgenotypesintherecessivemodel(null

versesheterozygousþwildtype).Thestrengthoftheassociationbetweenthe

GSTM1andGSTT1geneandCHDriskwasmeasuredbyORswith95%

con denceintervals(CIs).

Cochran’sv2basedQ-statistictest(9,10)andI2-test(11)wereperformedto

assesspossibleheterogeneityinthecombinedstudies.Ifheterogeneityexisted,

therandomeffectsmodel(theDerSimonianandLairdmethod)(12),which

yieldswiderCIs,wasadoptedtocalculatetheoverallORvalue.Otherwise,the

xedeffectsmodel(theMantel–Haenszelmethod)wasused(13).Inaddition,

sourcesofheterogeneitywereinvestigatedbystrati edmeta-analysesbasedon

ethnicity(Caucasianandnon-Caucasianpopulation),sourceofcontrols

(populationandhospitalbased)andsamplesize(numberofcases.250or

,250).ThemethodofWoolf(14)wasusedtocontruct95%CIs.The

signi canceoftheoverallORwasdeterminedbytheZ-test.Funnelplotsand

Egger’slinearregressiontestwereusedtoassessevidenceforpotential

publicationbias(15).Inordertoassessthestabilityoftheresult,sensitivity

analyseswereperformed,eachstudyinturnwasremovedfromthetotal,and

theremainingwerereanalysed.TheanalysiswasconductedusingReview

Manager(version5.0,TheCochraneCollaboration,Oxford,UK).ThetypeI

errorratewassetat0.05.AllP-valuesweretwotailed.inpopulation-basedcontrolscomparedto1.81(95%CI:0.77–4.26;I2597%,Pheterogeneity50.00)inhospitalcontrols.Afterstrati cationforsamplesize,westillobservedpositiveresultsinbigstudies(datanotshown).GSTT1.Themeta-analysisresultedinastatisticallynon-signi cantassociationbetweenGSTT1de ciencyandCHD.TheoverallORwas1.26(95%CI:0.90–1.75;I2593%,Pheterogeneity50.00)(Figure2).Nosigni cantassociationwasfoundinstrati edanalysesaccordingtoethnicity,sourceofcontrolsandGSTT1genotypes.TheORwas1.22(95%CI:0.96–1.54;I2549%,Pheterogeneity50.06)inCaucasiansand1.20(95%CI:0.68–2.10;I2596%,Pheterogeneity50.00)innon-Caucasians.Whenstratifyingforsourceofcontrols,anORof0.97(95%CI:0.82–1.14;I2555%,P50.01)and2.09(95%CI:0.90–4.86;I2596%,Pheterogeneity50.00)resultedfornullgenotype,amongpopulation-andhospital-basedcontrols,respectively.Inthestrati edanalysisbysamplesize,nosigni cantassociationswere

foundinbigstudiesorsmallstudies(datanotshown).Downloaded from

ResultsGene–geneinteraction.TheeffectofeachgenotypeofGSTsCharacteristicsofstudieswasindependentlyassessed.NoassociationwasestablishedAtotalof19studieswereretrievedbasedonthesearchcriteriabetweenboththenullgenotypeofGSTM1,GSTT1andCHD.forCHDsusceptibilityrelatedtotheGSTpolymorphisms.TheThedataonbothnullgenotypeofGSTsamongcasesandmainstudycharacteristicsweresummarizedinTableI.Therecontrolswereavailableineightstudies,whichincluded2909are19studieswith8020CHDcasesand11501controlscasesand3745controls.TheinteractionbetweenGSTM1nullconcerningGSTM1polymorphismand17studieswithandGSTT1null,forwhichanORof2.38(95%CI:1.03–5.48;7318CHDcasesand10361controlsconcerningGSTT1I2596%,P50.00)forCHDappearedincomparedwithpolymorphism.individualswiththepositivegenotypes.

Meta-analysisresults

GSTM1.ForCHDriskandthenullgenotypeofGSTM1,ourGene–environmentinteraction.Thedataongenotypesofthemeta-analysisGSTM1andGSTT1amongcasesandcontrolsstrati edby2gaveanoverallORof1.47(95%CI:1.08–2.01;I595%,Psmokingstatuswereavailablein vestudies.Amongsmokers

heterogeneity50.00)withstatisticallysigni cant

between-studyheterogeneity(Figure1).Thisanalysisisbasedinall vestudies,individualswiththenullgenotypeofonpoolingofdatafromanumberofdifferentethnicGSTM1orGSTT1hadasigni cantlyincreasedCHDriskpopulations.Whenstratifyingforethnicity,anORof1.43comparedtothenon-nullgenotypeswithanORof2.21(95%(95%CI:0.92–2.22;I2593%,PCI:1.24–3.92;I2584%,P

heterogeneity50.00)and1.50

(95%CI:0.93–2.43;I2596%,PCI:1.49–7.26;I2590%,Pheterogeneity50.00)and3.29(95%

heterogeneity50.00),respectively.

fornullgenotype,amongCaucasiansheterogeneity50.00)resulted

andnon-Caucasians,

respectively.Byconsideringcontrolsourcesubgroups,theORSensitivityanalysesandpublicationbias.Sensitivityanalyseswas1.31(95%CI:0.99–1.72;I2591%,PindicatedthatthestudiesbyAbu-Ameroetal.werethemain

heterogeneity50.00)originofheterogeneityinoverallOR.Theheterogeneitywas

TableI.Demographyofthecombinedstudies

ReferenceYearCountry(ethnicity)GenotypingmethodNo.ofcasesNo.ofcontrolsControlssourceEvansetal.(16)1996SaudiArabiaPCR90884PopulationWilsonetal.(17)2000UKPCR356187PopulationLietal.(5)2000USAPCR400890PopulationWangetal.(18)2001USAPCR612256HospitalSalamaetal.(19)2002USAPCR12090PopulationWilsonetal.(20)2003SouthAsianPCR170203PopulationPalmeretal.(21)2003UKPCR5157HospitalOlshanetal.(22)2003USAPCR526868PopulationMasettietal.(23)2003ItalyPCR308122HospitalGirishaetal.(24)2004IndiaPCR197198PopulationTameretal.(25)2004TurkeyRT-PCR148247PopulationHayeketal.(26)2006UKPCR1932399PopulationAbu-Ameroetal.(27)2006SaudiArabiaPCR1054762HospitalCornelisetal.(28)2007CanadaPCR20422042PopulationKimetal.(29)2008KoreaPCR356336HospitalWangetal.(30)2008ChinesePCR277277HospitalMartinetal.(31)2009USAPCR6763PopulationManfredietal.(32)2009ItalyPCR18447HospitalMacieletal.(33)2009BrazilPCR8711577Population2 at The 3rd Military Medical University on April 1, 2010

Coronaryheartdiseasesusceptibility

effectivelydecreasedafterexclusionofthestudy(GSTM1:

I2560%;GSTT1:I2589%).Inaddition,noothersingle

studyin uencedthepooledORqualitatively,asindicatedby

sensitivityanalyses,suggestingthattheresultsofthismeta-

analysisarestable(datanotshown).

Begg’sfunnelplotandEgger’stestwereperformedto

evaluatethepublicationbiasofliteratures.AsshowninFigures3

and4,theshapeofthefunnelplotsseemedsymmetricalforbothgenes,suggestingnopublicationbiasamongthestudiesincluded.Thestatisticalresultsstilldidnotshowpublicationbias(P.0.05,forall).DiscussionLargesampleandunbiasedepidemiologicalstudiesofpre-dispositiongenespolymorphismscouldprovideinsightinto

the

Downloaded from

Fig.1.Meta-analysisofGSTM1nullgenotypeassociatedwithCHD.EachboxrepresentstheORpointestimate,anditsareaisproportionaltotheweightofthestudy.Thediamondrepresentstheoverallsummaryestimate,withCIrepresentedbyitswidth.

Fig.2.EffectoftheGSTT1nullversuspresentgenotypetheriskof

CHD.

3 at The 3rd Military Medical University on April 1, 2010

J.Wangetal.

0SE(log[OR])

0.1

0.2

0.3

0.4

0.5

0.10.20.512510distantcountries,sotheresultmustbeinterpretedwithcaution.Nosigni cantassociationbetweenvariantgenotypesandCHDriskwasobservedwhentheincludedstudieswerestrati edbycontrolsource.Suchresultcouldbeduetolimitednumberofstudiesthathadinsuf cientstatisticalpowertodetectaslighteffectormayhavegenerateda uctuatedriskestimate.IfgeneticsusceptibilitytoCHDis,inpart,mediatedthroughmetabolicgenepolymorphisms,itispossiblethatthecombinationsofcertaingenotypesmaybemorediscriminatingasriskfactorsforCHDthanasinglelocusgenotype.Amongthe19studiesincludedinthepresentmeta-analysis,eightstudiesinvestigatedtheinteractionbetweenGSTM1andGSTT1polymorphism.BypoolingthecollecteddataonGSTM1andGSTT1genotypes,astatisticallysigni cant2.38-foldincreasedriskforCHDappearedforindividualswithcombineddeletionmutationsinGSTT1andGSTM1genesincomparisonwithindividualswiththepositivegenotypes.Thisresultsuggeststhatinthepresenceofbothofthetworiskfactors,animportantnumberofCHDcaseswouldoccur.

Cigarettesmokingisapro-in ammatorystimulusandanDownloaded from Fig.3.FunnelplotofassociationbetweenGSTM1polymorphismandCHD

overallindividuals.FormalstatisticalcriteriabyEgger’stestwasalsoperformedtoinvestigatethesymmetryofthefunnelplot(P50.142).0SE(log[OR])

0.1

0.2

0.3

0.4

0.5OR

0.10.20.512510Fig.4.FunnelplotofassociationbetweenGSTT1polymorphismandCHDoverallindividuals.FormalstatisticalcriteriabyEgger’stestwasalsoperformedtoinvestigatethesymmetryofthefunnelplot(P50.763).

invivorelationshipbetweencandidategenesandcomplexdiseases.Thismeta-analysis,involvingatotalof8020CHDcasesand10501controlsfrom19case–controlstudies,examinedtheassociationoftwocommonlystudiedpoly-morphismsofGST(M1andT1)withCHDrisk.Resultsindicatedasigni cantassociationbetweennullpolymorphismofGSTM1andCHDrisk,whereastheGSTT1polymorphism(nullversusnon-deleted)seemsunrelatedtoCHDrisk.

Inmeta-analysis,heterogeneityevaluationwasalwaysconductedinstatisticalanalysis.Thus,severalsubgroupmeta-analyseswereperformedaccordingtoethnicitycontrolsource.Inracialsubgroups,nostatisticallysigni cantassoci-ationbetweenGSTM1orGSTT1polymorphismandCHDappearedinCaucasiansandnon-Caucasians.However,theCaucasianandnon-Caucasianpopulationreportsinthesubgroupanalysisincludeamixtureofpopulationsfromvery4importantriskfactorforCHD.Unfortunately,almostallthestudiesdidnotexploretheinteractionbetweenGSTsgenotypeandsmokinghabits.Thiswasprobablyduetothelowstatisticalpoweroftheindividualstudiestodetectinteractions.Ourresultsshowasigni cantassociationamongsmokerssubgroupbetweenGSTM1,GSTT1polymorphismandCHDrisk.TheresultssuggestthattherecouldbeaninteractionbetweencigarettesmokingandGSTs.Onepossiblecauseisthattobaccosmoke-inducedDNAdamagecausessmoothmusclecellproliferationintheintimaofarteries,therebycontributingtoatheroscleroticplaqueformation(8).Somelimitationsofthismeta-analysisshouldbeaddressed.Firstly,heterogeneityisapotentialproblemwheninterpretingalltheresultsofmeta-analysis.Althoughweminimizedthelikeli-hoodbyperformingacarefulsearchforpublishedstudies,usingtheexplicitcriteriaforstudyinclusion,thesigni cantbetween-studyheterogeneitystillexistedinmostofcomparison.Thepresenceofheterogeneitycanresultfromdifferencesintheagedistribution,selectionofcontrols,prevalencelifestylefactorsandsoon.Secondly,thethreesubgroupmeta-analysesconsideringinteractionsbetweenGSTM1,GSTT1nullgenotypeandcigarettesmoking,aswellasbetweenGSTT1nullandGSTM1nullgenotypes,wereperformedonthebasisofafractionofallthepossibledatatobepooled,soselectionbiasmayhaveoccurredandourresultsmaybeoverin ated.Inthiscontext,morereliableresultscanbeexpectedifindividualdataareavailableforapooledanalysis.Thirdly,onlypublishedstudieswereincludedinthismeta-analysis.Therefore,publicationbiasmayhaveoccurred,eventhoughtheuseofastatisticaltestdidnotshowit.Despitetheselimitations,thismeta-analysissuggeststhatGSTM1polymorphismsmayincreasetheriskofCHD,butnosigni canteffectforGSTT1polymorphisms.Inaddition,rgerstudiesofdifferentethnicpopulations,especiallywithdetailedindividualinformation,areneededtocon rmour ndings.AcknowledgementsWethankDrAlexandrePereira,DrJose´GeraldoMill,DrSteveHumphries,DrAndyOlshanandDrjosekriegerfortheircollaborationandformakingtheirdataavailableforthepresentmeta-analysis.Con ictofintereststatement:Nonedeclared. at The 3rd Military Medical University on April 1, 2010

Coronaryheartdiseasesusceptibility

References

1.AmericanHeartAssociation(1998)HeartandStrokeFacts.AmericanHeartAssociation,Dallas,TX,USA.

2.Sing,C.F.,Haviland,M.B.andReilly,S.L.(1996)Geneticarchitectureofcommonmultifactorialdiseases.CibaFound.Symp.,197,211–229;discussion,229–232.

3.DeFlora,S.,Izzotti,A.,Walsh,D.,Degan,P.,Petrilli,G.L.andLewtas,J.(1997)Molecularepidemiologyofatherosclerosis.FASEBJ.,11,1021–1031.

4.Martinet,W.,Knaapen,M.W.,DeMeyerv,G.R.,Herman,A.G.andKockx,M.M.(2002)ElevatedlevelsofoxidativeDNAdamageandDNArepairenzymesinhumanatheroscleroticplaques.Circulation,106,927–932.

5.Li,R.,Boerwinkle,E.,Olshan,A.F.etal.(2000)GlutathioneS-transferasegenotypeasasusceptibilityfactorinsmoking-relatedcoronaryheartdisease.Atherosclerosis,149,451–462.

6.Hayes,J.andPulford,D.(1995)TheglutathioneS-transferasesupergenefamily:regulationofGSTandthecontributionoftheisoenzymestocancerchemopreventionanddrugresistance.Crit.Rev.Biochem.Mol.Biol.,30,445–600.

7.Hayes,J.D.,Flanagan,J.U.andJowsey,I.R.(2005)Glutathionetransferases.Annu.Rev.Pharmacol.Toxicol.,45,51–88.

8.Bridges,B.A.,Bowyer,D.E.andHansen,E.S.(1990)Thepossibleinvolvementofsomaticmutationsinthedevelopmentofatheroscleroticplaques.ReportofICPEMCSubcommittee7/1.Conclusionsandmorphismasasusceptibilityfactorinsmoking-relatedcoronaryarterydisease.BasicRes.Cardiol.,99,223–229.26.Hayek,T.,Stephens,J.W.,Hubbart,C.S.,Acharya,J.,Caslake,M.J.,Hawe,E.,Miller,G.J.,Hurel,S.J.andHumphries,S.E.(2006)AcommonvariantintheglutathioneStransferasegeneisassociatedwithelevatedmarkersofin ammationandlipidperoxidationinsubjectswithdiabetesmellitus.Atherosclerosis,184,404–412.27.Abu-Amero,K.K.,Al-Boudari,O.M.,Mohamed,G.H.andDzimiri,N.(2006)TnullandMnullgenotypesoftheglutathioneS-transferasegeneareriskfactorforCADindependentofsmoking.BMCMedGenet.,7,38.28.Cornelis,M.C.,El-Sohemy,A.andCampos,H.(2007)GSTT1genotypemodi estheassociationbetweencruciferousvegetableintakeandtheriskofmyocardialinfarction.Am.J.Clin.Nutr.,86,752–8.29.Kim,S.J.,Kim,M.G.,Kim,K.S.,Song,J.S.,Yim,S.V.andChung,J.H.(2008)ImpactofglutathioneS-transferaseM1andT1genepolymorphismsonthesmoking-relatedcoronaryarterydisease.J.KoreanMed.Sci.,23,365–372.30.Wang,L.S.,Tang,J.J.,Tang,N.P.,Wang,M.W.,Yan,J.J.,Wang,Q.M.,Yang,Z.J.andWang,B.(2008)b.Med.,46,1720–1725.31.Martin,N.J.,Collier,A.C.,Bowen,L.D.,Pritsos,K.L.,Goodrich,G.G.,

Arger,K.,Cutter,G.andPritsos,C.A.(2009)PolymorphismsintheDownloaded from recommendations.Mutat.Res.,239,143–148.

9.Cochran,W.G.(1954)Thecombinationofestimatesfromdifferentexperiments.Biometrics,10,101–129.

10.Colditz,G.A.,Burdick,E.andMosteller,F.(1995)Heterogeneityinmeta-

analysisofdatafromepidemiologicstudies:acommentary.Am.J.Epidemiol.,142,371–382.

11.Higgins,J.P.,Thompson,S.G.,Deeks,J.J.andAltman,D.G.(2003)

Measuringinconsistencyinmeta-analyses.BMJ,327,557–560.

12.DerSimonian,R.andLaird,N.(1986)Meta-analysisinclinicaltrials.

ControlClin.Trials,7,177–188.

13.Mantel,N.andHaenszel,W.(1959)Statisticalaspectsoftheanalysisof

datafromretrospectivestudiesofdisease.J.NatlCancerInst.,22,719–748.

14.Woolf,B.(1955)Onestimatingtherelationbetweenbloodgroupand

disease.Ann.Hum.Genet.,19,251–253.

15.Egger,M.,DaveySmith,G.,Schneider,M.andMinder,C.(1997)Biasin

meta-analysisdetectedbyasimple,graphicaltest.BMJ,315,629–634.

16.Evans,D.A.,Seidega rd,J.andNarayanan,N.(1996)TheGSTM1genetic

polymorphisminhealthySaudiArabiansandFilipinos,andSaudiArabianswithcoronaryatherosclerosis.Pharmacogenetics,6,365–367.

17.Wilson,M.H.,Grant,P.J.,Hardie,L.J.andWild,C.P.(2000)

GlutathioneS-transferaseM1nullgenotypeisassociatedwithadecreasedriskofmyocardialinfarction.FASEBJ.,14,791–796.

18.Wang,X.L.,Greco,M.,Sim,A.S.,Duarte,N.,Wang,J.and

Wilcken,D.E.(2002)GlutathioneS-transferasemu1de ciency,cigarettesmokingandcoronaryarterydisease.J.Cardiovasc.Risk,9,25–31.

19.Salama,S.A.,Au,W.W.,Hunter,G.C.,Sheahan,R.G.,Badary,O.A.,

Abdel-Naim,A.B.andHamada,F.M.(2002)Polymorphicmetabolizinggenesandsusceptibilitytoatherosclerosisamongcigarettesmokers.Environ.Mol.Mutagen.,40,153–160.

20.Wilson,M.H.,Grant,P.J.,Kain,K.,Warner,D.P.andWild,C.P.(2003)

AssociationbetweentheriskofcoronaryarterydiseaseinSouthAsiansandadeletionpolymorphisminglutathioneS-transferaseM1.Biomarkers,8,43–50.

21.Palmer,C.N.,Young,V.,Ho,M.,Doney,A.andBelch,J.J.(2003)

AssociationofcommonvariationinglutathioneS-transferasegeneswithprematuredevelopmentofcardiovasculardiseaseinpatientswithsystemicsclerosis.ArthritisRheum.,48,854–855.

22.Olshan,A.F.,Li,R.,Pankow,J.S.,Bray,M.,Tyroler,H.A.,

Chambless,L.E.,Boerwinkle,E.,Pittman,G.S.andBell,D.A.(2003)Riskofatherosclerosis:interactionofsmokingandglutathioneS-transferasegenes.Epidemiology,14,321–327.

23.Masetti,S.,Botto,N.,Manfredi,S.,Colombo,M.G.,Rizza,A.,

Vassalle,C.,Clerico,A.,Biagini,A.andAndreassi,M.G.(2003)InteractiveeffectoftheglutathioneS-transferasegenesandcigarettesmokingonoccurrenceandseverityofcoronaryarteryrisk.J.Mol.Med.,81,488–494.

24.Girisha,K.M.,Gilmour,A.,Mastana,S.,Singh,V.P.,Sinha,N.,

Tewari,S.,Ramesh,V.,Sankar,V.H.andAgrawal,S.(2004)T1andM1polymorphisminglutathioneS-transferasegeneandcoronaryarterydiseaseinNorthIndianpopulation.IndianJ.Med.Sci.,58,520–526.

25.Tamer,L.,Ercan,B.,Camsari,A.,Yildirim,H.,Cicxek,D.,Sucu,N.,

Atesx,N.A.andAtik,U.(2004)GlutathioneS-transferasegenepoly-NQO1,GSTTandGSTMgenesareassociatedwithcoronaryheartdiseaseandbiomarkersofoxidativestress.Mutat.Res.,674,93–100.32.Manfredi,S.,Calvi,D.,delFiandra,M.,Botto,N.,Biagini,A.andAndreassi,M.G.(2009)GlutathioneS-transferaseT1-andM1-nullgenotypesandcoronaryarterydiseaseriskinpatientswithType2diabetesmellitus.Pharmacogenomics,10,29–34.33.Maciel,S.S.,PereiraAda,C.,Silva,G.J.,Rodrigues,M.V.,Mill,J.G.andKrieger,J.E.(2009)AssociationbetweenglutathioneS-transferasepolymorphismsandtriglyceridesandHDL-cholesterol.Atherosclerosis,206,204–208.

5 at The 3rd Military Medical University on April 1, 2010

本文来源：https://www.bwwdw.com/article/4271.html