The association between the survivin-31G/C promoter polymorphism and hepatocellular carcinoma risk in a Turkish population

(1)

The

association

between

the

survivin

31G/C

promoter

polymorphism

and

hepatocellular

carcinoma

risk

in

a

Turkish

population

Su¨leyman

Bayram

a,b,

*

,

Hikmet

Akkız

b

,

Aynur

Bekar

b

,

Ersin

Akgo¨llu¨

b

a_Adıyaman_University,_Adıyaman_School_of_Health,_Department_of_Nursing,₀₂₀₄₀_Adıyaman,_Turkey b

C¸ukurovaUniversity,FacultyofMedicine,DepartmentofGastroenterology,01330Adana,Turkey

1. Introduction

Hepatocellular carcinoma (HCC) is the ﬁfth most common

cancer in theworldwide and thethird leading causeof cancer

death.Becauseofitshighfatalityrates,theincidenceandmortality

ratesareapproximatelyequal[1].Itisnowwellestablishedthat

multipleriskfactorscontributetohepatocarcinogenesis,including

chronic hepatitis B virus (HBV) or hepatitis C virus (HCV)

infections,cirrhosis,carcinogenexposure (suchasaﬂatoxinB1),

excessivealcoholdrinking,andanumberofgeneticandepigenetic

alterations[2,3].However,HBVandHCVinfectionsarethemajor

causeof HCC, only a fractionof infected patients develop HCC

during their lifetime; therefore the identiﬁcationof other risk

characteristics(suchasgeneticpolymorphisms)tostratifythose

individuals into high-risk populations is needed. As in many

cancers,geneticpolymorphismsinvolved inmultistageof

hepa-tocarcinogenesismaydetermineindividual’ssusceptibilitytothe

developmentofHCC[4,5].Theidentiﬁcationofsinglenucleotide

polymorphisms(SNPs)thataffectgenefunctionorexpressionand

contribute toHCCsusceptibility isimportant asit may helpto

predictindividualandpopulationriskandclarifypathophysiologic

mechanisms relevant to HCC. Moreover, identifying genetic

biomarkers of HCC susceptibility and their application in

conjunctionwithtraditionaldiagnosis,staging andprognosis,it

mightbepossibletoreduceHCCmortalitythroughearlydiagnosis,

patientscareandpersonalizedtherapy[6].

Dysregulationsof thebalance betweenproliferationand cell

deathareinvolvedincarcinogenesisincluding

hepatocarcinogen-esisthroughprolongingcellsurvival,promotingaccumulationof

transformingmutations,andenhancingresistancetotherapy[7,8].

Indeed, recent studies have suggested that some molecules

involved in counteracting apoptosis, such as Bcl-XL, Mcl-1,

c-IAP1,XIAPandsurvivinareoverexpressedinHCC[7,9].Survivin,a

member of the inhibitor of apoptosisproteins (IAPs)family, is

involvedininhibitionofapoptosisandregulationofcelldivision

[10].Survivin’s cell division regulatoryfunction is executed by

bindingseveralstructuralcomponentsofthemitoticapparatus,i.e.

spindlemicrotubules,centrosomesorkinetochoresofmetaphasic

chromosomes [11]. Survivin, discoveredin 1997, is a 16.3kDa

proteinconsistingof142aminoacids.Thegeneencodingsurvivin

is of 14.5kb, which is located at the telomeric region of the

chromosome17q25[10].Previousstudiesfoundthatsurvivinwas

undetectableinterminallydifferentiatedadulttissues.However,

growingevidenceindicatesthatsurvivinisexpressedinnormal

adultcells,andthatitplaysaroleincertainphysiologicalprocesses

inmanyhumancells[12,13].Oneofthemostsigniﬁcantfeatures

ARTICLE INFO Articlehistory:

Received12October2010

Receivedinrevisedform5January2011 Accepted10January2011

Availableonline5February2011 Keywords: Hepatocellularcarcinoma Survivin Survivin31G/Cpolymorphism Susceptibilty ABSTRACT

Background:Survivin,amemberoftheinhibitorofapoptosisproteinfamily,functionsasakeyregulator ofapoptosisandcellcycleregulation.Acommonsinglenucleotidepolymorphism(31G>C)atthe survivinpromoterhasbeenextensivelystudiedinvariouscancersandreportedtoinfluencesurvivin expression,butitsassociationwithhepatocellularcarinoma(HCC)hasyettobeinvestigated.Theaimof thepresentstudywastoinvestigatewhetherthispolymorphismcouldbeinvolvedintheriskofHCC susceptibilty.Methods:Thegenotypefrequencyofsurvivin31G>Cpolymorphismwasdeterminedby usingapolymerasechainreaction-restrictionfragmentlengthpolymorphism(PCR-RFLP)methodin160 subjectswithHCCand241cancer-freecontrolsubjectsmatchedonage,gender,smokingandalcohol status.Results:Nostatisticallysignificantdifferenceswerefoundinthegenotypedistributionsofthe survivin31G>CpolymorphismamongHCCandcancer-freecontrolsubjects(p=0.28).Conclusion:Our resultsdemonstrateforthefirsttimethatthesurvivin31G/Cpolymorphismhavenotbeenanymajor roleingeneticsusceptibiltytohepatocellularcarcinogenesis,atleastinthepopulationstudiedhere.

ß2011ElsevierLtd.Allrightsreserved.

Abbreviations:SNP,singlenucleotidepolymorphism;HCC,hepatocellular carcino-ma;CI,conﬁdenceinterval;OR,oddsratio;PCR-RFLP,polymerasechain reaction-restrictionfragmentlengthpolymorphism.

*Correspondingauthorat:AdıyamanUniversity,AdıyamanSchoolofHealth, DepartmentofNursing,02040Adıyaman,Turkey.

Tel.:+904162233003;fax:+904162233005.

E-mailaddress:slymnbyrm81@gmail.com(S.Bayram).

ContentslistsavailableatScienceDirect

Cancer

Epidemiology

The

International

Journal

of

Cancer

Epidemiology,

Detection,

and

Prevention

j o urn a lhom e pa g e : ww w . ca nc e re pi d e mi ol o gy . ne t

1877-7821/$–seefrontmatterß2011ElsevierLtd.Allrightsreserved.

(2)

ofsurvivinisitsdifferentexpressionincancercomparedtonormal

tissues.Invarioushumancancersincludinghepatocellularcancer,

survivinisstronglyoverexpressedandhasbeenestablishedasa

prognosticmarker [9]. Increased survivinexpression in human

cancersisconsideredtobeanimportantmarkerforaggressiveand

chemoresistantdisease,signalingapoorprognosis[14].Morever,

Ye et al. [15] have suggested that survivin overexpression is

correlatedwithhighriskofdiseaserecurrenceandpoorprognosis

inHCC.

Survivin isexpressedinacellcycle-dependentmanner,with

maximal expression during G2/M phase of the cell cycle and

exhibitsarapiddownregulationinG1phase[11].Thisislargely

controlledatthetranscriptionallevel,andmediatedbyseveralcell

cycle-dependent elements (CDEs) (GGCGG) and one cell cycle

homologyregion(CHR)(ATTTGAA)locatedintheproximalregion

ofthesurvivinpromoter[16].SeveralSNPswereidentiﬁedwithin

thepromoterregionofthesurvivingene,oneofwhichislocatedat

the CDE/CHR repressor binding site (31bp from the ﬁrst

nucleotide ofthe ATGstart codon) [17].The referencenumber

ofthisSNPinthedatabaseoftheNationalCenterforBiotechnology

Information(NCBI)isrs9904341.ThebestdocumentedofthisSNP

isatposition31ofthesurvivingenepromoterandinvolvesthe

substitutionofguanine(G)forcytosine(C)andthecreationoftwo

alleles(31G and 31C)and three genotypesGG, GC, and CC.

Survivin31G/Cpolymorphism hasbeenassociatedwith

over-expressionofsurvivinatbothmessengerRNA(mRNA)andprotein

levelsandabberantcellcycle-dependenttranscriptionmediated

throughfunctionaldisruptionofbindingattheCDE/CHRrepressor

motifsinanumberofcancercelllines[17].

Severalcase–controlstudieshaveinvestigatedtheassociation

between 31G/C polymorphism within theCDE/CHR repressor

elementofthesurvivingenepromoterandcancerriskincluding

lungcancer[18],sporadiccolorectalcancer[19],cervicalcancer

[20],gastriccancer[21],esophagealsquamouscellcarcinoma[22],

urothelial carcinoma [23], pancreatic cancer [24] and acute

myeloidleukemia [25]. However, conﬂicting resultshave been

published[18–25].

Basedontheimportantroleofsurvivinincarcinogenesis,with

importantbiological,prognosticandtherapeuticimplications.We

hypothesizedthat functional 31G/Cpolymorphismin survivin

genemayactasageneticmodiﬁerinindividualsusceptibilityto

HCC.According toourrecent knowledge,no researchhasbeen

conductedtoevaluatesurvivin31G/Cpolymorphismandriskof

HCCdevelopment.Totestthehypothesisthatthepolymorphismof

survivin31G/Cis associated withrisk of developingHCC, we

performedgenotypinganalysisusingpolymerasechain

reaction-restrictionfragmentlengthpolymorphism(PCR-RFLP)assayina

hospital-basedcase–controlstudyof160HCCpatients and241

age,gender,smokingandalcoholconsumptionmatched

cancer-freecontrolsinTurkishpopulation.

2. Patientsandmethods

2.1. Studypopulation

Thestudypopulation andsubjectcharacteristicswere

previ-ously described elsewhere [4,5]. This is an ongoing molecular

epidemiologicstudyofHCCconductedinAdana,Turkeyandthe

subjectsrecruitmentwasapprovedbytheCommitteeforEthicsof

Medical Experiment on Human Subjects, Faculty of Medicine,

C¸ukurovaUniversity.Brieﬂy,allsubjectsweregenetically

unrelat-edTurkishandwerefromC¸ukurovaandthesurroundingregionsof

southernTurkey.Submissionoftheindividualstothestudywas

conditioned by an obtained written informed consent form

regarding the use of their blood samples for research studies.

ThestudyproceededinagreementwiththeHelsinkideclaration

approvedontheWorldMedicalAssociationmeetinginEdinburgh.

Bloodsampleswerecollectedfrom160consecutivepatientswith

HCC seen in the department of gastroenterology and general

surgery between September 2005and March2010. During the

sametime,241unrelatedcommunityresidentswithnoevidence

of hepatocellular or other cancer who entered thehospital for

health check-ups were enrolled as the control group. The 241

cancer-freecontrolsubjectsdidnothaveahistoryofliverdisease

andhadnoserologicalevidenceofhepatitisBorCvirusinfection.

Eachcontrolwaspair-matchedbysex,age(3years),smokingand

alcoholconsumption to a patient withHCC. Thesecharacteristics

allowedusthechoiceofacontrolpopulationwithoutanypossible

risk bias forHCC. The HCC diagnostic criteria was based on the

guidelineproposedbyEuropeanAssociationfortheStudyoftheLiver

(EASL)[26].WegaveadiagnosisofHCCwhenapatienthadoneor

moreriskfactors(i.e.,HBVorHCVinfection,orcirrhosis)andoneof

the following: >400ng/ml

a

-fetoprotein (AFP) and at least one

positive ﬁnding following examination using spiral computed

tomography (CT), contrast-enhanced dynamic MRI, or hepatic

angiography;or<400ng/ml

a

-fetoproteinandatleasttwoﬁndings

following CT,MR,or hepatic angiography.A positiveHCC ﬁnding

using dynamic CT or MRI is indicative of arterial enhancement

followedbyvenouswashoutinthedelayedportal/venousphase.In

addition,weperformedhistopathologicaldiagnosisforcasesthatdid

notfulﬁllalloftheclinicalnon-invasivediagnosticcriteriaofHCC.

Cirrhosiswasdiagnosedwithliverbiopsy,abdominalsonography,

andbiochemicalevidenceofparenchymaldamageplusendoscopic

esophageal or gastric varices [27]. Patients with cirrhosis were

classiﬁedintothreeChild-Pughgradesbasedontheirclinicalstatus

[28].SerumHBsAgandAnti-HCVwereassessedusingan

immunoas-say (Abbott Laboratories, Abbott Park, IL, USA). Serum AFP

concentrationwasmeasuredbymicroparticleenzymeimmunoassay

(AbbottLaboratories,AXSYM,USA).Heavyalcoholintakewasdeﬁned

asadailyminimumconsumptionof160galcoholforatleasteight

years.

Allsubjectswereinterviewedusingastructuredquestionnaire

to obtain information on demographic factors and health

characteristics.Technicianswhoperformedthebloodtestswere

blinded to the identity and disease status of participants.

Peripheralbloodsamplesweretakenfrompatientsandcontrols,

andbloodspecimens,includingwhitebloodcellsandserum,were

frozenat208Cuntilanalysis.

2.2. DNAextraction

A5mlsampleofvenousbloodwascollectedfromeachsubject

intoatesttubecontainingEDTAasanticoagulant.GenomicDNA

wasextractedfromperipheralwholebloodusingHighPurePCR

TemplatePreperationKit(RocheDiagnostics.GmbH,Mannheim,

Germany)accordingtothemanufacturer’sprotocol.

2.3. Polymerasechainreaction-restrictionfragmentlength

polymorphismanalysis

Polymerase chain reaction-restriction fragment length

poly-morphism(PCR-RFLP)analysiswasperformedtodeterminethe

genotype of the 31G/C polymorphism of survivin gene, as

described previously [19]. The 341 base pair (bp) fragment

encompassingtheGtoCpolymorphicsiteinsurvivinpromoter

regionwasampliﬁedusingspeciﬁcprimers50_-GTT_CTT_TGA_AAG

CAGTCGAG-30_and₅0_-GCC_AGT_TCT_TGA_ATG_TAG_AG-30_._The₂₀

_m

_l

PCRmixturecontainedapproximately250ngDNA,with0.25

m

M

ofbothprimer,0.1mMofeachdNTP,1PCRbuffer,1.5mMMgCI2

and 1UTaqpolymerase. Thefollowingcyclingconditionswere

used:948Cfor5min,followedby35cyclesof948Cfor30s,578C

(3)

AfterconﬁrmationofsuccessfulPCRampliﬁcationby1.5%agarose

gelelectrophoresis,eachPCRproductwasdigestedovernightwith

5 units EcoO109I (from Escherichia coli strain that carries the

EcoO109I gene from Escherichia coli H709c, recognizing the

sequence50_-RG#_GNCCY-30₎_enzyme_at₃₇_8C_(New_England_Biolabs

Inc., Beverly, MA) and electrophoresed on 3% agarose gel

containing 0.5

m

g/ml ethidium bromide and visualized under

UVillumination.PCRproductswithGatthepolymorphicsitewere

digestedintotwofragments,236bpand105bp,whilethosewith

CwerenotbecauseoftheabsenceofaEcoO109Irestrictionsite.

Samplesyielding236bpand105bpfragmentswerescoredasGG,

thosewithsingle341bpfragmentsasCC,and341bp,236bpand

105bpasGC.Toensurequalitycontrol,genotypingwasperformed

withoutknowledgeofthesubjects’case/controlstatusanda15%

random sample of cases and controls was genotyped twice by

differentpersons;reproducibilitywas100%.

2.4. Statisticalanalysis

ThesamplesizewascalculatedusingtheQUANTO1.1program

(hydra.usc.edu/gxe).Thedesiredpowerofourstudywassetat80%.

Data analysis was performed using the computer software

StatisticalPackageforSocialSciences(SPSS)forWindows(version

10.0).Differencesinthedistributionsofdemographic

character-isticsbetween thecasesand controlswereevaluatedusing the

Student’s t-test (for continuous variables) and

x

2 _test _(for

categorical variables). Theobserved genotype frequencies were

comparedwithexpectedvaluescalculatedfromHardy–Weinberg

equilibriumtheory(p2₊_2pq₊_q2₌_1;_where_p_is_the_frequency_of

thewild-typealleleandqisthefrequencyofthevariantallele)by

usinga

x

2

testwithdegreeoffreedomequalto1amongcasesand

controls, respectively. Pearson’s

x

2 _test _was_used _to_determine

whethertherewasanysigniﬁcantdifferenceinalleleandgenotype

frequencies between patients and controls. The associations

between survivin31G/C genotypesand the risk of HCC were

estimated by computing the odds ratios (ORs) and their 95%

conﬁdenceintervals(CIs)frombinarylogisticregressionanalysis.

The homozygousgenotype for the G alleleof survivin 31G/C

polymorphismwasusedasthereferenceincalculatingORsand

95%CIs.Statisticalmodelingwasperformedontherelativeriskof

the CC genotype or theGC genotype against theGG genotype

independently. Furthermore,toestimatetherecessiveor

domi-nant effect of survivin 31G/C genotype on risk, statistical

modelingwasperformedontherelativeriskoftheCCgenotype

againsttheGC+GGgenotypeortheCC+GCgenotypeagainstthe

GG genotype. Probability levels less than 0.05 were used as a

criterionofsigniﬁcance.

3. Results

3.1. Generalcharacteristicofthesubjects

A total of 401Turkish subjectswere enrolled in our study.

GeneralcharacteristicofthesubjectsaresummarizedinTable1.As

expected, no signiﬁcant difference was found between case

patients and control subjects with regard to age and sex

(p=0.85and p=0.98,respectively)which impliedthat ageand

sex matched adequately. Similarly, there were no signiﬁcant

differencesinsmokingstatusandalcoholconsumptionbetween

case andcontrolgroup.Inadditiontothese,Table1showsthe

distribution of demographic variables such as AFP, marker of

hepatitis,cirrhosisandChild-Pughgradeofcases.

3.2. Genotypefrequencydistributionofsurvivin31G/C

polymorphism

The frequency distributions of the different genotypes for

survivin 31G/C polymorphism are shown in Table 2. The

genotypic frequencies of thecontrol (n=241;

x

2

=0.005df=1,

p=0.94) werein Hardy–Weinberg equilibrium, suggesting that

therewasnopopulationstratiﬁcationandnosamplingbias.The

patients’ frequencieswere alsoinHardy–Weinberg equilibrium

(n=160;

x

2₌_0.061_df₌_1,_p₌_0.8)._The_allelic_frequencies_of_case

subjects (G,0.71; C, 0.29)were notsigniﬁcantly differentfrom

those of thecontrol subjects(G, 0.66; C, 0.34)(p=0.13).Thus,

genotypic frequencies in the cases were similar to that of the

controls(

x

2₌_2.58,_df₌_2,_p₌_0.28).

Table1

Distributionofselectedcharacteristicsinpatientswithhepatocellularcarcinomaandcontrols.

Characteristic Patients(%)(n=160) Controls(%)(n=241) p-Valuea

Age(years)mean(SD)(range) 59.1111.04(20–81) 58.8811.62(20–84) 0.85

Malesex 129(80.6%) 194(80.5%) 0.98 Smokingstatus 0.97 Ever 72(45.0%) 108(44.8%) Never 88(55.0%) 133(55.2%) Alcoholstatus 0.98 Drinker 45(28.1%) 68(28.2%) Non-drinker 115(71.9%) 173(71.8%) Viralinfection HBsAgpositive 94(58.8%) – Anti-HCVAbpositive 36(22.5%) – Bothpositive 2(1.3%) – Bothnegative 28(17.5%) – Livercirrhosis Present 129(80.6%) – Absent 31(19.4%) – Child-Pughclassiﬁcation A 21(16.3%) – B 43(33.3%) – C 65(50.4%) – a-Fetoprotein(ng/ml) <100 75(46.9%) – 100–400 23(14.3%) – >400 62(38.8%) –

Abbreviations:NS=notsigniﬁcant,n=totalnumberofcasepatientsorcontrolsubjects.

a

p-ValueswerederivedfromPearsonx2

(4)

3.3. Survivin31G/Cpolymorphismandriskofhepatocellular carcinoma

ToevaluatetheriskofHCCaccordingtothesurvivin31G/C

genotype,logistic regression analysiswas conducted (Table2).

UsingtheGGgenotypeasthereferencegenotype,individualswith

oneor two copies of Cvariantallele exhibited with0.75(95%

CI=0.49–1.13, p=0.17) and 0.63-fold (95% CI=0.29–1.38,

p=0.25) decreased HCC risk, respectively, although neither

associationreached statisticalsigniﬁcance. Studysubjects

com-bined withthesurvivin31 GC and CC genotypeshad a

non-signiﬁcantlowerHCCrisk(OR=0.73;95%CI=0.49–1.10,p=0.12),

comparedthiswiththesurvivin31GGgenotype.Whenweused

GGandGCgenotypesasareference,wefoundthattheORoftheCC

genotypewas0.74(95%CI=0.35–1.56,p=0.42).

To observe whether the effectof survivin31G/C

polymor-phismwasmodiﬁed byepidemiologicfactors,HCC patientsand

controlswerestratiﬁedonthebasisofvarioushostcharacteristics

includingsex,hepatocellularcarcinomaetiology (viralinfection

status) and age. Subgroup analysis revealed that the effect of

genderandviralinfectionstatuswerenotsigniﬁcantlydifferent

amongsurvivin31G/Cgenotypes(datanotgiven).Inadditionto

this,wecomparedtheageatdiagnosisforHCCpatientswhohad

differentgenotypesthearithmeticmeanagesandSDforGG,GC

andCCwere59.4210.65,59.2111.10and56.1813.88years

old,respectively.Therewasnosigniﬁcantdifferencebetweenthem

byt-test(p=0.66).

4. Discussion

Itiswidelyacceptedthattheregulationofprogrammedcell

death is important for the prevention of tumorigenesis.

Im-pairmentofapoptosisfacilitatestheaccumulationofgeneticerrors

byprolonging thecell cycle, promotingresistance to

immune-based cytotoxicity, and a selective growth advantage for the

altered cells contributing to carcinogenesis [7,14]. Because

survivinfunctionsas an inhibitor ofapoptosis that is essential

for eliminating mutated or transformed cells from body, it is

possible that subjects with a lower production genotype for

survivin 31G/C polymorphism may have increased apoptotic

capacity to eliminate cells with DNA damage, thus may have

decreasedsusceptibilitytohepatocellularcancer.

Thismolecularepidemiologicalstudyinvestigatedwhetherthe

survivin 31G/C polymorphism could have an impact on

susceptibility to HCC. Survivin 31G/C polymorphism was

selectedasthecandidate polymorphism becausethis

polymor-phismcouldaffectthebindingofelementsthatregulatecellcycle

dependenttranscriptionofsurvivingene,therebyhavingarolein

thederegulationofsurvivininHCC.Incontrarytoourexpectation,

inthepresentcase–controlstudyinTurkishpopulation,

distribu-tionofsurvivin31G/CgenotypewasnotdifferentbetweenHCC

casesand controls. Nosigniﬁcantassociationemergedbetween

risk of HCC and survivin 31G/C polymorphism in overall

statisticalanalyses.However,variantgenotypereducedtheHCC

risk,althoughnotstatisticallysigniﬁcantlyso.Ourresultsinline

withpreviousﬁndingsshowingthattherewasannoassociation

betweenthesurvivin31G/Cpolymorphismandriskforvarious

cancers inluding cervical cancer [20], gastric cancer [21],

esophageal squamous cell carcinoma [22], pancreatic cancer

[24], acute myeloid leukemia [25]. Nonetheless, it should be

noted that some studies, including lung cancer [18], sporadic

colorectal cancer [19], and urothelial cancer [23], have found

association with survivin 31G/C polymorphism. A rational

explanationforthiscancer-dependentdifferenceinriskconferred

by the examined survivin 31G/C polymorphism may be

attributable to differences in the pathways of carcinogenesis

among thevarioustypesof humancancers.,However, usingin

vitropromoterassay,Janget al.[18]foundthatthe31Gallele

hadasigniﬁcantlylowertranscriptionalactivitythan31Callele.

Chenet al.[29]showedthatsurvivinmRNAwasoverexpressedin

gastriccancercasesbutnosigniﬁcantdifferenceingastriccancer

tissuessubgroupedby31G/Cgenotypes.Inadditiontothis,the

contributionofgeneticpolymorphismstotheriskforcancermay

bedependent on thestudiedpopulation, as wellas on several

environmental andotherfactorsthat inﬂuencethatpopulation.

Geographicorethnicdifferenceshavebeenreportedregardingthe

genotypefrequencyofseveralpolymorphisms.

Similarlytoourﬁndings,survivin31G/Cpromoter

polymor-phismhasbeenshowntohavenoroleinvariouscarcinogenesis

[20–22,24,25].Soitismorelikelythatothermolecularandcellular

mechanismsareinvolved in survivinoverexpressionin

hepato-cellularcarcinogenesis.Forinstance,survivinexpressioncanbe

deregulatedincancerbyseveralmechanisms,including

ampliﬁ-cationofthesurvivinlocusonchromosome17q25,demethylation

ofsurvivinexons,increasedpromoteractivity,increasedupstream

signaling in the phosphatidylinositol 3 kinase and epigenetics

alterations by promoter methylations and demethylation of

survivinexons[16,30–33].

Survivin deserves growing attention as an ideal target for

cancertherapyduetoitsdifferentialexpressionintumorsversus

normal tissues, and the evidence for a dual role in both cell

apoptosisanddivision[9,34].Thereareseverallinesofevidence

suggesting that survivin is involved in hepatocarcinogenesis

[9,15,34]. In HCC cells, overexpression of survivin accelerates

the G1 checkpoint by releasing p21WAF1/Cip1 from the p21/

cyclin-dependent kinase 4 (CDK4) complex and enforcing the

formationof acomplexbetween pro-caspase3 andp21,which

suppressescelldeathsignaling[9].Overexpressionofsurvivinwas

Table2

AllelesandgenotypesfrequencydistributionofSurvivin31G/Cpolymorphismamongcasesandcontrolsaswellastheassociationwithhepatocellularcancerrisk. Cases(%),n=160 Control(%),n=241 p-Valuea

OR(95%CI) Allelefrequency G 228(71.2%) 319(66.2%) 1.00(reference) C 92(28.8%) 163(33.8%) 0.13 0.85(0.69–1.05) Generalgenotype GG 79(49.4%) 100(41.5%) 1.00(reference) GC 70(43.7%) 119(49.4%) 0.17 0.75(0.49–1.13) CC 11(6.9%) 22(9.1%) 0.25 0.63(0.29–1.38) Dominantgenotype GG 79(49.4%) 100(41.5%) 1.00(reference) GC+CC 81(50.6%) 141(58.5%) 0.12 0.73(0.49–1.10) Recessivegenotype GG+GC 149(93.1%) 219(90.9%) 1.00(reference) CC 11(6.9%) 22(9.1%) 0.42 0.74(0.35–1.56) a

(5)

found in 70% of HCC patients, which correlated with poor

prognosticparameters(highnuclearandhistologicgrade,

micro-vascular invasion,increasedproliferation),local recurrence,and

shorter disease free survival [15,34]. But, recently published

studies reported that HCC cells have higher levels of survivin

mRNAbutlesssurvivinproteincomparedwithpairednon-tumor

andcirrhoticlivertissues[35,36].Thereisalsocontroversyinthe

literatureregardingtheprognosticsigniﬁcanceofsurvivininHCC.

Forinstance,Chauet al.[35]showedthatlowersurvivinprotein

levels in HCC cells correlated with more agrressive tumor

behaviour,with more advanced stage disease and withpoorer

patient survival. Further studies are necessary to clarify these

discrepancies.

Thelimitationsofourstudyareasfollows.Firstlimitationofthe

presentstudyisthatitwashospital-basedcase–controlstudy,and

patientswereselectedatasingleinstitution(C¸ukurovaUniversity,

Balcalı Hospital) and thus may have been unrepresentative of

hepatocellular carcinomapatients in thegeneral population. In

addition, it should be noted that the control subjects were

recruitedatthesamehospital.However,inthecontrolgroup,the

agreementbetweentheobserveddistributionofsurvivin31G/C

genotypefrequencieswiththeexpectedaccordingtotheHardy–

Weinbergequilibriummodelsuggestednoselectionbias.Second,

thisstudyislimitedbytherelativelysmallnumberofcasesand

controls.Thereforefurtherstudieswithalargernumberofsubjects

areneededtoclarifythisissue.Third,wealsolimitedourstudyto

Turkishpopulationduetovariation inallelefrequencybetween

differentethnic groupshas beenobserved forsurvivin 31G/C

polymorphism. Fourth, due to the lack of data on survivin

expressionaccordingtosurvivin31G/CgenotypeinHCC,future

workneedtobedoneinordertoexplorethecorrelationbetween

levelsofsurvivininnormalliverandHCCtissuesinthecontextof

differentgenotypesofsurvivin31G/Cpolymorphism.

Inconclusion,ourresultsdemonstratefortheﬁrsttimethatthe

survivin31G/Cpolymorphismhavenotbeenanymajorrolein

geneticsusceptibiltytohepatocellularcarcinogenesiswithinthe

studiedpopulation.Beacusethisistheﬁrstreportconcerningthe

survivin 31G/C polymorphism and the risk of HCC in the

literature,further independent studies are required to validate

ourﬁndingsinalargerseries,aswellasinpatientsofdifferent

ethnicoriginsandtobetterunderstandsurvivin31G/C

polymor-phismandsusceptibiltyHCC.

Conﬂictofinterest

Allauthorsdeclarethattherearenoconﬂictsofinterest.

Acknowledgements

The authors thank all thesubjects who participated in this

study.ThisworkwassupportedbyC¸ukurovaUniversityResearch

FundFEF2008D4.

References

[1]ParkinDM,BrayF,FerlayJ,PisaniP.Globalcancerstatistics,2002.CACancerJ Clin2005;55:74–108.

[2]FaraziPA,DePinhoRA.Hepatocellularcarcinomapathogenesis:fromgenesto environment.NatRevCancer2006;6:674–87.

[3]El-SeragHB,RudolphKL.Hepatocellularcarcinoma:epidemiologyand molec-ularcarcinogenesis.Gastroenterology2007;132:2557–76.

[4]AkkızH,BayramS,BekarA,Akgo¨llu¨ E,OzdilB.Relationshipbetweenfunctional polymorphism inthe AuroraAgene andsusceptibility of hepatocellular carcinoma.JViralHepat2010;17:668–74.

[5]AkkızH,BayramS,BekarA,Akgo¨llu¨ E,OzdilB.CyclinD1G870Apolymorphism isassociatedwithanincreasedriskofhepatocellularcarcinomaintheTurkish population:case–controlstudy.CancerEpidemiol2010;34:298–302.

[6]LudwigJA,WeinsteinJN.Biomarkersincancerstaging,prognosisand treat-mentselection.NatRevCancer2005;5:845–56.

[7]MeletA,SongK,BucurO,JaganiZ,GrassianAR,Khosravi-FarR.Apoptotic pathwaysintumorprogressionandtherapy.AdvExpMedBiol2008;615:47– 79.

[8]FabregatI.Dysregulationofapoptosisinhepatocellularcarcinomacells.World JGastroenterol2009;15:513–20.

[9]ItoT,ShirakiK,SugimotoK,YamanakaT,FujikawaK,ItoM,etal.Survivin promotescellproliferationinhumanhepatocellularcarcinoma.Hepatology 2000;31:1080–5.

[10]SahNK,KhanZ,KhanGJ,BisenPS.Structural,functionalandtherapeutic biologyofsurvivin.CancerLett2006;244:164–71.

[11]LiF,AmbrosiniG,ChuEY,PlesciaJ,TogninS,MarchisioPC,etal.Controlof apoptosisandmitoticspindlecheckpointbysurvivin.Nature1998;396:580– 4.

[12]LiF,BrattainMG.RoleoftheSurvivingeneinpathophysiology.AmJPathol 2006;169:1–11.

[13]FukudaS,PelusLM.Survivin,acancertargetwithanemergingroleinnormal adulttissues.MolCancerTher2006;5:1087–98.

[14]DuffyMJ,O’DonovanN,BrennanDJ,GallagherWM,RyanBM.Survivin:a promisingtumorbiomarker.CancerLett2007;249:49–60.

[15]YeCP,QiuCZ,HuangZX,SuQC,ZhuangW,WuRL,etal.Relationshipbetween survivinexpressionandrecurrence,andprognosisinhepatocellular carcino-ma.WorldJGastroenterol2007;13:6264–8.

[16]LiF,AltieriDC.Thecancerantiapoptosismousesurvivingene: characteriza-tionoflocusandtranscriptionalrequirementsofbasalandcell cycle-depen-dentexpression.CancerRes1999;59:3143–51.

[17]XuY,FangF,LudewigG,JonesG,JonesD.Amutationfoundinthepromoter regionofthehumansurvivingeneiscorrelatedtooverexpressionofsurvivin incancercellsDNA.CellBiol2004;23:419–29.

[18]JangJS,KimKM,KangKH,ChoiJE,LeeWK,KimCH,etal.Polymorphismsinthe survivingeneandtheriskoflungcancer.LungCancer2008;60:31–9. [19]Gazouli M,Tzanakis N,Rallis G, Theodoropoulos G, PapaconstantinouI,

KostakisA,etal.Survivin31G/Cpromoterpolymorphism andsporadic colorectalcancer.IntJColorectalDis2009;24:145–50.

[20]Borbe´lyAA,MurvaiM,SzarkaK,Ko´nyaJ,GergelyL,Herna´diZ,etal.Survivin promoter polymorphism and cervical carcinogenesis. J Clin Pathol 2007;60:303–6.

[21]YangL,ZhuH,ZhouB,GuH,YanH,TangN,etal.Theassociationbetweenthe survivinC-31GpolymorphismandgastriccancerriskinaChinesepopulation. DigDisSci2009;54:1021–8.

[22]YangX,XiongG,ChenX,XuX,WangK,FuY,etal.Polymorphismsofsurvivin promoterareassociatedwithriskofesophagealsquamouscellcarcinoma.J CancerResClinOncol2009;135:1341–9.

[23]WangYH,ChiouHY,LinCT,HsiehHY,WuCC,HsuCD,etal.Association betweensurvivin gene promoter 31 C/Gpolymorphism and urothelial carcinomariskinTaiwanesepopulation.Urology2009;73:670–4. [24]TheodoropoulosGE,MichalopoulosNV,PanoussopoulosSG,TakaS,GazouliM.

Effectsofcaspase-9andsurvivingenepolymorphismsinpancreaticcancer riskandtumorcharacteristics.Pancreas2010;39:976–80.

[25]WagnerM,SchmelzK,Do¨rkenB,TammI.Epigeneticandgeneticanalysisof thesurvivinpromoterinacutemyeloidleukemia.LeukRes2008;3:1054–60. [26]BruixJ,ShermanM,LlovetJM,BeaugrandM,LencioniR,BurroughsAK,etal. Clinicalmanagementofhepatocellularcarcinoma.Conclusionsofthe Barce-lona-2000EASLconference.EuropeanAssociationfortheStudyoftheLiver.J Hepatol2001;35:421–30.

[27]TsaiJF,ChangWY,JengJE,HoMS,LinZY,TsaiJH.HepatitisBandCvirus infectionasriskfactorsforlivercirrhosisandcirrhotichepatocellular carci-noma:acase–controlstudy.Liver1994;14:98–102.

[28]PughRN,Murray-LyonIM,DawsonJL,PietroniMC,WilliamsR.Transectionof theesophagusforbleedingesophagealvarices.BrJSurg1973;60:646–9. [29]ChengZJ,HuLH,HuangSJ.Correlationof31G/Cpolymorphismsofsurviving

promotertotumorigenesisofgastriccarcinoma.AiZheng2008;27:258–63. [30]IslamA,KageyamaH,TakadaN,KawamotoT,TakayasuH,IsogaiE,etal.High expressionofsurvivin,mappedto17q25,issigniﬁcantlyassociatedwithpoor prognostic factors and promotes cell survival in human neuroblastoma. Oncogene2000;19:617–23.

[31]HattoriM,SakamotoH,SatohK,YamamotoT.DNAdemethylaseisexpressed inovariancancersandtheexpressioncorrelateswithdemethylationofCpG sitesinthepromoterregionofc-erbB-2andsurvivingenes.CancerLett 2001;169:155–64.

[32]VairaV,LeeCW,GoelHL,BosariS,LanguinoLR,AltieriDC.Regulationof survivinexpressionbyIGF-1/mTORsignaling.Oncogene2007;26:2678–84. [33]ChoS,LeeJH,ChoSB,YoonKW,ParkSY,LeeWS,etal.Epigeneticmethylation

andexpressionofcaspase8andsurvivininhepatocellularcarcinoma.Pathol Int2010;60:203–11.

[34]AltieriDC.Targetedtherapybydisablingcrossroadsignalingnetworks:the survivinparadigm.MolCancerTher2006;5:478–82.

[35]ChauGY,LeeAF,TsaySH,KeYR,KaoHL,WongFH,etal.Clinicopathological signiﬁcanceofsurvivinexpressioninpatientswithhepatocellularcarcinoma. Histopathology2007;51:204–18.

[36]AugelloC,CarusoL,MaggioniM,DonadonM,MontorsiM,SantambrogioR, etal.Inhibitorsofapoptosisproteins(IAPs)expressionandtheirprognostic signiﬁcanceinhepatocellularcarcinoma.BMCCancer2009;9:125.