3. OBJETIVOS
5.3. Resultado de Análise Estatística
Os resultados da análise estatística desse estudo levaram a submissão e publicação dos seguintes artigos científicos:
• Association study of the BIN1 and IL-6 genes on Alzheimer’s disease - publicado na revista Neuroscience Letter em Janeiro de 2016.
• Validating GWAS's variants from microglial genes implicated in Alzheimer
disease - manuscrito submetido em análise na revista Journal of Molecular Neuroscience.
• Epistasis interaction among BIN1, CLU and APOE genes in Alzheimer
disease - manuscrito submetido na revista Neuroscience Letter.
Os resultados estatísticos e a discussão da dissertação foram apresentados em formato de artigo. Para os artigos em submissão, foram incluídos os comprovantes de submissão e os artigos na íntegra.
Neuroscience
Letters
j o u r n a l h o m e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / n e u l e t
Researchpaper
Association
studyoftheBIN1and
IL-6genesonAlzheimer’sdisease
LígiaRamosdosSantosa,b,∗,LucianoBelcavelloa,DanielaCamporeza,b,
CaerêIamondeMacieldeMagalhãesc,ElianaZandonadec,RenatoLírioMorelatod,
FlaviaImbroisiValleErreraa,e,IuriDrumondLouroa,b,
MarialDoCarmoPimentelBatituccia,f,FlaviadePaulaa,b
aNúcleodeGenéticaHumanaeMolecular,DepartamentodeCiênciasBiológicas,CentrodeCiênciasHumanaseNaturais,UniversidadeFederaldoEspírito Santo,Vitória,ES,Brazil
bProgramadePós-Graduac¸ãoemBiotecnologia,UniversidadeFederaldoEspíritoSanto,Vitória,ES,Brazil
cLaboratóriodeEstatística(LESTAT),CentrodeCiênciasExatas,UniversidadeFederaldoEspíritoSanto,Vitória,ES,Brazil
dHospitaldaSantaCasadeMisericórdiadeVitória,EscolaSuperiordeCiênciasdaSantaCasadeMisericórdiadeVitória,Vitória,ES,Brazil eEscolaSuperiordeCiênciasdaSantaCasadeMisericórdiadeVitória,Vitória,ES,Brazil
fProgramadePós-Graduac¸ãoemBiologiaVegetal,UniversidadeFederaldoEspíritoSanto,Vitória,ES,Brazil
h i g h l i g h t s
•BIN1(rs744373)andIL-6(rs1800795)SNPswerestudiedinAlzheimer’sdisease(AD).
•NoassociationwithADforbothSNPsinalocatedsoutheastBrazilianpopulation.
•TheresultssuggestthattheywerenotimplicatedwithADforallpopulations.
a r t i c l e i n f o
Articlehistory: Received9October2015
Receivedinrevisedform7December2015 Accepted21December2015
Availableonline28December2015 Keywords: Alzheimer’sdisease Case-controlstudy −174G/CIL-6 Brazilianpopulation a b s t r a c t
Genome-wideassociationstudy(GWAS)hasidentifiedseveralnovelgenesassociatedwiththeriskof Alzheimer’sdisease(AD),whichisaprogressiveneurodegenerativediseaseinelders.However,mostof thenovelgeneshavenotbeenvalidatedthroughreplicationinseparatedpopulations.Amongthem,the BIN1geneisinvolvedinendocytosisandintracellulartraffickingaswellasintheformationofamyloid plaquesandneurofibrillarytangles,whicharethemainpathologicalhallmarksofAD.TheIL-6genehas alsobeenfrequentlyassociatedwithAD;however,consistentresultshavenotbeenfound.IL-6,acytokine fromtheimmunesystem,isimplicatedinthepathogenesisofseveraldegenerativediseases.Similarto BIN1,itissuggestedthatIL-6isalsoinvolvedintheformationofamyloidplaques.Inthiscase-control study,weaimedtoinvestigatewhethersinglenucleotidepolymorphismsintheBIN1(rs744373)and IL-6(rs1800795)genesareassociatedwithAD.GenotypefrequencieswereevaluatedviaPCR-RFLPin82 late-onsetADpatientsand159elderlyhealthycontrols,whowerematchedbyageandgender.Inthis study,noassociationwasfoundforeitherpolymorphism,suggestingthatthesegenesarenotimplicated intheaetiologyofADinallpopulations.
©2015ElsevierIrelandLtd.Allrightsreserved.
∗ Correspondingauthorat.LaboratóriodeGenéticaHumanaeMolecular,Depar- tamentodeCiênciasBiológicas,CCHN,UniversidadeFederaldoEspíritoSanto,Av. Mal.Campos1468,Maruípe29043-900,Vitória,ES,Brazil.Fax:+552733357250.
E-mailaddresses:ligia.ramos.santos@gmail.com(L.RamosdosSantos),
belcavello@gmail.com(L.Belcavello),danielacamporez@gmail.com(D.Camporez),
magalhaes.caere@gmail.com(C.IamondeMacieldeMagalhães),
elianazandonade@uol.com.br(E.Zandonade),renato.morelato@hotmail.com
(R.LírioMorelato),flavia.errera@gmail.com(F.ImbroisiValleErrera),
Iurilouro@yahoo.com(I.DrumondLouro),docarmobatitucci@yahoo.com.br
(M.DoCarmoPimentelBatitucci),flapvit@yahoo.com.br(F.dePaula).
1. Introduction
Dementiaisa neurocognitivedisorderinelderlypeoplethat affect memory, language and learning ability. In 2013, it was estimatedthattherewere44millionspeoplewithdementiaworld- wide.Anextremetype ofdementiaisAlzheimer’sdisease(AD), whichthisaccountsfor60to80%ofdementiacases[1].Thisneu- rodegenerativediseaseischaracterizedbymemoryandcognitive damage.Pathologically,itischaracterizedbytwochangesinside thebraincortex:neurofibrillarytanglesandplaquesofamyloid
http://dx.doi.org/10.1016/j.neulet.2015.12.046
(presenilin1)andPSEN2(presenilin2)accountsfor5%ofADcases withMendelianpatternsandleadtotheearlyonsetofthedisease. Insomecaseswherethediseaseappearsaftertheageof65,con- sideredlate-onsetAD(LOAD),therearemultifactorialcauses,with geneticandenvironmentalcauses[2].Currently,theonlygenewith asusceptibilityriskfactorthathasbeenwellestablishedforLOAD isAPOE(apolipoproteinE)fromthe4allele.Therefore,effortsare beingmadetoidentifymorevariantscorrespondingtotheriskof LOAD[3].
Thelargegenome-wideassociationstudy(GWAS)scannedthe entirehumangenometodiscovernovelgenesleadingtoADsus- ceptibility[4].Therecentlydiscovered,Bridgingintegrator1(BIN1) genewasoneofthenovelvariants.GWAS analysesdisclosed a singlenucleotidepolymorphism(SNP),rs744373,inBIN1asbeing associatedwithAD[5].TheBIN1geneisinvolvedinmanycellular processes,includingendocytosis,apoptosisandsignaltransduction
[2].TheproteinfromBIN1interactswiththeTauprotein,suggest- ingthatitplaysaroleintheformationofneurofibrillarytangles. Duetothisendocytosisactivity,BIN1mayplayaroleinamyloid plaqueformationinthepathogenesisofAD[6].
ThepathwaythatleadstothepathogenicelementsofAD is crossedwiththeinflammatoryprocess[7].ThehumanInterleukin-6 gene(IL-6)isamajorcytokineandisimplicatedinthepathogenesis ofmanyinflammatorydiseases,suchasAD[8].TheIL-6genehasa polymorphism,−174G/C(rs1800795),inthepromoterregion,and previousstudieshavefoundthatitwasassociatedwithAD[9–11]. Thisstudyattemptstoinvestigatetherelationshipofrs744373 BIN1 and −174G/C IL-6 with LOAD in samples from south- eastern Brazil. The validation of the polymorphisms though a case-controlstudywasimportanttoenhanceourknowledgeabout thepathwaysleadingtothedevelopmentofAD,andalsototrace disease-relatedriskfactorsindistinctpopulations.
2. Materialsandmethods
2.1. Subjects
In this case-control study, a total of 241 individuals were included82 patientswithadiagnosisofLOAD accordingtothe criteriaoftheNationalInstituteofNeurologicalandCommunica- tiveDisordersandStrokeandtheAlzheimer’sDiseaseandRelated DisordersAssociation(NINCDS-ADRDA),and159healthycontrols matched for sex and age. Informationon gender,age, ethnical background composition, age at onset of disease, APOE status, schooling,scale of Mini-MentalState Examination(MMSE) and ClinicalDementiaRatingScale(CDR)inpatientsandcontrolsis giveninTable1.ThisstudywasacceptedbytheEthicsCommittee ofHumanResearchofEscolaSuperiordeCiênciasdaSantaCasade MisericórdiadeVitória,ES,Brazil,andalloftheelderlyortheirrel- ativesgaveawritteninformedconsenttoparticipate.Theelderly selected(caseandcontrols)toparticipateinthisstudywerefrom Vitória,ES,inthesoutheasternpartofBrazil.
Theparticipants,includinghealthycontrolsandpatients,were assistedattheGeriatricUnityoftheHospitalSantaCasadeMis- ericórdiadeVitória(HSCMV)andattheCentrodeReferênciade AtendimentoaoIdoso(CRAI),both atES,Brazil.Allparticipants werediagnosedattheNeurogeriatricUnityatHSCMVandCRAI.
ThepatientswerediagnosedforprobableAD,withacompre- hensivediagnostic evaluationfor dementia, and fulfilled others criteria,suchas MMSE;and theClinical Dementia RatingScale (CDR). Totalsample 82(100%) 159(100%) Gender Woman 54(65.9%) 116(73.0%) Man 28(34.1%) 43(27.0%) Schooling Literati 44(53.7%) 101(63.5%) Illiterate 38(46.3%) 58(36.5%) Ethnicalbackground Caucasians* 47(57.3%) 88(55.3%) Afro-Brazilians 35(42.7%) 71(44.7%) APOEstatus 4+ 35(42.7%) 111(69.8%) 4− 47(57.3%) 48(30.2%)
ADmeanSD ControlmeanSD
Age 81.2±7.5 79.2±7.8
Onsetofdisease 4 –
MMSE 14–4 >28
CDR 15±4 –
NumberswithpercentagesinparenthesisshowtheproportionsofADpatients andhealthycontrols;*=BraziliandescendantsofCaucasians;4+=4 carries;
4−=4non-carries;SD=standarddeviation;MMSE=Mini-MentalStateExamina- tion(valueconsideringtheschoolinglevelofcontrolsandpatientsandtheaverage evolutiontimeofthediseaseinADpatients);CDR=DementiaRatingScale(mild andmoderatedementia);>=Greaterthan.
2.2. Bloodsampling
BloodwascollectedattheGeriatricUnityofHSCMVandCRAI, between2007and2008.GenomicDNAwasextractedfrom5mL ofperipheralbloodandcollectedintotubescontaining5%EDTA followingapreviousmethodology[12].
2.3. 174G/CIL-6polymorphismgenotyping
The polymorphism was determined by the conventional PolymeraseChainReaction-RestrictionFragmentLengthPolymor- phism(PCR-RFLP)technique,usingpreviouslypublishedprimers
[13].PCRwasperformedonaVeriti®96(AppliedBiosystems)with
thefollowingconditions:denaturationat94◦Cfor4min,followed by28cyclesat94◦Cfor30s,62◦Cfor30s,and72◦Cfor30s,and anextraextensionphaseat72◦C for10min.Then, 20Lofthe PCRproductsweregeneratedusingaMIXtubeof1XPCRBuffer, 1,5mMMgCl2,10pmolofeachprimer,0.2mMofdeoxynucleotide
triphosphates,0.05units/lofTaqDNApolymeraseand20ngof DNA.ThePCRproductswereelectrophoresedfor1hand50minat 200V,300mAand100Wona7%polyacrylamidegel.Thegelwas stainedwithsilvernitrate,andtheamplificationofthe305-base- pair(bp)fragmentwasvisualized.Afteramplification,10Lofthe PCRproductswassubjectedtodigestionat37◦Covernightwith 0.2unitsofNlaIII(NewEnglandBiolabs,USA).Thedigestionprod- uctswereelectrophoresedona7%polyacrylamidegelstainedwith silvernitratewiththesameprotocolusedontheelectrophorezed PCRproducts.
2.4. rs744373polymorphisminBIN1genotyping
The polymorphism was determined by the conventional PCR-RFLP technique using degenerated primers designed by Primer3 v.0.4.0 web software [14]. The forward primer (5- CACCAGGGACAGGCAGGTCTGAGAC-3) was degenerated at nucleotideAlocalizedattheendoftheprimer, andthereverse primer(5-CACATCTTAGCCACAGAACAGG-3)wasnotdegenerated.
created a site (GANTC) for the HinfI restriction enzyme, while theCnucleotidedidnotallowdigestion.ThePCRwasperformed witha Veriti® 96 (Applied Biosystems)using these conditions:
denaturationat94◦Cfor4min,followedby32cyclesat94◦Cfor 30s, 72◦Cfor 30s, 72◦C for 30s, andan extraextension phase at72◦Cfor10min.Then,10LofPCRproductsweregenerated usingaMIXtubeof1XPCRBuffer,1,5mMMgCl2,10pmolofeach
primer,0.2mMofdeoxynucleotidetriphosphates,0.05units/lof TaqDNA polymeraseand20ngofDNA.ThePCRproductswere electrophoresedfor2hand15minat200V,300mAand100Won a7%polyacrylamidegelandlaterstainedwithsilvernitrate.The amplificationofthe250-base-pairfragmentwasvisualized.After electrophoresis,5LoftheamplifiedPCRproductsweresubjected todigestionat37◦Covernightwith0.2unitsofHinfI(NewEngland Biolabs,USA).Thedigestionproductswereelectrophoresedona 7%polyacrylamide gelstained withsilver nitratefollowing the sameprotocolusedontheelectrophoresedPCRproducts. 2.5. Statisticalanalysis
Pearson’schi-squaretest wasperformedusingtheSPSSsoft- warev.20,forWindows[16]tocomparefrequencydifferencesfor eachgenotypebetweenthesamples.ThePearson’schi-squaretest isastatisticaltestappliedtosetsofcategoricaldatatoevaluatehow likelyitisthatanyobserveddifferencebetweenthesetsaroseby chance[17].TotesttheassociationbetweentheAPOE4statusand polymorphisms−174G/CIL-6andrs744373BIN1,alogisticregres- sionanalysisthroughSPSSsoftwarev.20,forWindowswasused
[16].Thelogisticregressionanalysishasabinarylogisticmodelto estimatetheprobabilityofabinaryresponsebasedononeormore independentvariables[18].Thepvalueinthelogicregressionwas adjustedusingage,gender,schoollevelandethnicalbackground asvariables,sincethosevariablesareriskfactorsforAlzheimer’s diseaseandcaninfluencethediagnoseofthedisease.Moreover, wecalculatedtheHardy-WeinbergEquilibrium(H-WE),performed with1degreeoffreedom.p≤0.05wasconsideredstatisticallysig- nificant.ThegenotypefrequencyoftheAPOE4statuswasinferred inourpreviouslystudy[19].
3. Results
rs744373 BIN1 did not deviate from the predicted Hardy–Weinberg equilibrium (HWE) in both healthy controls andADpatients.Therewasagenotypedistributionof−174G/C IL-6inHWEinADpatients,butnotinhealthycontrols.Therefore, werandomlyre-genotyped20(12.5%)individualcontrolsamples, andallofthemmatchedthepreviouslydeterminedgenotype.
ThegenotypefrequenciesinADpatientsandhealthycontrolsof −174G/CIL-6andrs744373BIN1areshowninTable2.TheIL-6GG genotypesweremostfrequentinADpatients(61%),andinhealthy controlsitsfrequencywasslightlyhigher(49.7%)thenGC(49.1%). ThegenotypeCCwastheleastfrequentedinADpatients(1.2%)and inhealthycontrols(1.3%).Therewasnostatisticallysignificantdif- ferenceintheIL-6GG,GCandCCgenotypesbetweenADpatient andhealthycontrolgroups.TheTCgenotypeforthers744373poly- morphismhadthegreatestfrequencyinADpatients(47.6%)and healthycontrols(49%)comparedwithothergenotypes,whilethe CCgenotypewaslowerinADpatients(8.5%)andhealthycontrols (10.1%).NostatisticalsignificancewasobservedfortheTT,TCand CCgenotypesintheADpatientandhealthycontrolgroups.
Thelogisticregressionanalysisresultsforrs744373BIN1are shown in Table 3. The −174G/C IL-6 polymorphism was not
AD Control pValue IL-6 GG 50(61%) 79(49.7%) 0.246 GC 31(37.8%) 78(49.1%) CC 1(1.2%) 2(1.3%) Total 82(100.0%) 159(100.0%) BIN1 TT 36(43.9%) 65(40.9%) 0.872 TC 39(47.6%) 78(49.0%) CC 7(8.5%) 16(10.1%) Total 82(100.0%) 159(100.0%) APOE 4+ 35(42.7%) 111(69.8%) 0.000 4− 47(57.3%) 48(30.2%) Total 82(100.0%) 159(100.0%)
NumberswithpercentagesinparenthesisshowtheproportionsofgenotypesinAD patientsandhealthycontrols;pvalue=ADpatientversushealthycontrol;pvalue considerer≤0.05;4+=4carries;4−=4non-carries.
Table3
LogisticregressionanalysesfortheassociationofADwithrs744373BIN1. AD(%) Control(%) OR(95%CI) pValue* OR(95%CI) pvalue** BIN1
TC 47.6 49.0 1(reference) – 1(reference) – TT 43.9 40.9 1.11(0.63–1.94) 0.720 1.20(0.66–2.19) 0.547 CC 8.5 10.1 0.86(0.33–2.30) 0.787 0.79(0.28–2.26) 0.660 OR=Oddsratio;CI=Confidenceinterval;pvalueconsiderer≤0.05;*=crudepvalue; **=pvalueadjustedbythevariablesage,gender,educationalattainment,ethnical backgroundandAPOE4status.
assessedforlogisticregressionanalysisbecauseitsgenotypedis- tributioninthecontrolgroupwasnotinHWE.Afteradjustingfor age,gender,educationalattainment,andAPOE4status,thelogis- ticregressionanalysisshowednoassociationsforADwiththeCC genotypeofBIN1(p=0.660;OR=0.79;95%CI=0.28–2.26)ortheTT genotype(p=0.547;OR=1.20;95%CI=0.66–2.19).
4. Discussion
Genome-wideassociationstudyidentifiednewvariantsassoci- atedwiththeriskforAlzheimer’sdisease;however,mostofthe associations need tobereplicated in separatedpopulations for validation. Inthis study,we investigatedwhetherthers744373 BIN1and−174G/CIL-6polymorphismswereassociatedwithAD. AlthoughtheBIN1geneisoneofthemostrelevantgenesinAD,our samplepopulationshowednoassociation.Wealsodidnotdetect anassociationbetweenthe−174G/CIL-6polymorphismandAD.
TheBIN1proteinisinvolvedinendocytosisandintracellular trafficking.Therefore,thisproteinmayplayaroleinthepathway leadingtoADbecauseitcouldaffectthetransportandprocessing ofAPPinsidecells,resultinginamyloidplaqueformation[2].This hypothesiswassupportedbystudiesthatfoundelevatedlevelsof theBIN1proteininADbrains,becauseinvitroandinvivostud- iesshowedmolecularinteractionsbetweentheBIN1proteinand Tau,themainproteininvolvedintheformationofneurofibrillary tangles[6].
The rs744373 BIN1 polymorphism showeda significantrisk association for AD in GWAS studiesas well as in LOAD meta- analyses studies from AlzGene (OR=1.17, 95% CI=1.13–1.20). Furthermore, it rankssecond in the top resultsfor genes with positive associations with AD (OR=1.166, 95% CI=1.13–1.20), suggestingastrongrelationshipwiththedisease[20].Thispoly-
tions.However,thesamestudydidnotfindanassociationforADin populationsfromFinland,whichwassimilartoourstudy.Lambert etal.[21]usedalogisticregressionwithanadditivegeneticmodel consideringgender,age and diseasestatus totest whetherthe SNPwasassociatedwithAD;ourstudyusedadifferentstatistical approachtotesttheassociation;withaChi-squaretest.Moreover, oursample(n=241)wasminorcomparedwiththeirstudy(Ital- ian,n=2725;Spanishn=1555;Finland,n=1092).Thesamplesize andstatisticalanalysesmayhaveinfluencedthedetectionlimitsof associationinoursamplepopulation.Additionally,anotherfactor toconsideristhattheBrazilianpopulationisamixture,withances- tryderivativesfromIberianCaucasians,WestAfricansandNative Americans[22,23].Thisethnicalbackgroundvariabilitycouldmask somegenemarkersforAD[24],suchasvariantsintheBIN1gene, intheBrazilianpopulation.Whenco-relatedwithAPOEε4carriers
[21],thestudyfoundnoassociationforthers744373BIN1poly- morphismandADinthethreesamplesstudied.Ourdataalsoused alogisticregressionadjustedforage,gender,educationalattain- ment,ethnicalbackgroundandAPOE4statusanddidnotfinda significantassociation.
InAlzheimer’sdisease,theformationoftheamyloidpeptide caninducemicrogliaandastrocytestosecreteimmunesystemele- ments,suchascytokineinterleukin6[25].TheIL-6geneislocalized onchromosome7p21[7]andisthoughttoincreasetheamountof proteininamyloidplaques.Moreover,elevatedlevelsofIL-6have beenfoundintheearlyformationofamyloidplaquesandhave onlybeendetectedonlyinbrainswithAD,notinthoseoftheelderly withoutAD[25,26].
Thepolymorphism−174G/CintheIL-6genealterstherateof genetranscriptionandserumlevelsinADpatientbrains,indicating itsroleinthedevelopmentofAD[27].However,theassociationof the−174G/CIL-6polymorphismwithADisnotconsensual.From 22case-controlstudiesofthe−174G/CIL-6riskassociationforAD listedonAlzGene[20]forCaucasians,12showedpositiveassoci- ations,9showednegativeassociationsand1indicatedatrendfor LOAD.Furthermore,themajorityofstudiesinEuropeanindividu- alsdidnotfindsignificantassociationforthispolymorphismwith AD[28–34].Studiesin Brazilianpopulations withthe−174G/C IL-6polymorphism are scarceand not consensual.In thestudy ofMoraisetal.[11],theCalleleof−174G/CIL-6wasassociated withADinadominantmodelanalysis,usingasamplelocatedin themidwestregionofBrazil(Brasilia)consistingof532individuals (120ADpatients;412controls).ThestudybyRasmussenetal.[35]
wentintheoppositedirection.Inasampleof365individuals(200 ADpatients;165controls)fromSãoPaulocity,noassociationwas foundfor−174G/CIL-6usingtheChi-Squaretest[33].Thereasons fortheconflictingresultsmaybeexplainedbytheinteractionof −174G/CIL-6withotherSNPs.Forinstance,Rasmussenetal.[35]
foundaprotectiveassociationforADintheIL-6genewhenallele Aof−597G/AwasassociatedwithalleleGof−174G/C,suggest- ingthatthesamegenemayinteractasahaplotypethatchanges theodds forAD.Additionally,in theMansoori etal.[36]study, theTalleleoftheSNPrs1801133inthemethylenetetrahydrofolate reductase(MTHFR)geneinteractedwiththeCalleleof−174G/CIL- 6toincreaseriskofAD(OR=2.5)inanItalianpopulation.MTHFR isanimportantgeneinthemetabolismoffolate(acidfolic)and homocysteine[37].In ourstudy,thispolymorphism showedno associationwithAD.WebelievethatIL-6playsaroleinthedisease thatmaybeunderpoweredalone,anditisnecessarytoconsider itsinteractionswithothersSNPsforassociationswithAD.
Case-controlstudiesareimportanttotracesusceptibilitygenes becauseeachpopulationmayhave differentmarkersfor LOAD. Associationstudiesareessentialtovalidatenovelgenesdisclosed
ofneurofibrillarytanglesandamyloidplaques,itisstillnotfully understoodhoweitherSNPdrivesADdevelopment[2,25].There- fore,bothpolymorphismsrequiremoreresearchtobetterelucidate theirroleintheaetiologyofAD.
5. Conclusion
Ourdatasuggeststhatthe−174G/CIL-6and rs744373BIN1 polymorphismsdonothaveanassociationwithsporadicADwhen stratifiedforallele4ofAPOEinasmallpopulationfromoneloca-