ADNIincludesmorethan800participantsranginginagefrom55to90.Alltheseindi - PDF document

ADNIincludesmorethan800participantsrangi
ADNIincludesmorethan800participantsranginginagefrom55to90.Alltheseindividualswererecruitedfromover50sitesacrosstheUnitedStatesandCanada,includingapproximately200healthycontrols(HC),400patientsdiagnosedwithmildcognitiveimpairment(MCI)and200patientsdiagnosedwithearlyAD.TheADpatientswerefollowedfor2yearsandotherswerefollowedfor3years.Structural1.5-Tmagneticreson-anceimaging(MRI)collectedthefullsample.PIBandFDGpositronemissiontomography(PET)imagingofasubset,someotherbiologicalmarkers,andperformanceonneuropsychologicalorclinicalassessmentswerecollectedatbaselineandatfollow-upvisitsin6-to12-monthintervals.TheCSFBACEandgenome-widegenotypinginthisstudywererespectivelyavailableonapproximatelyhalfofthecohortandthefullADNIsample.FurtherinformationaboutADNIcanbefoundinpreviouspublicationsandatwww.adni-info.org[15].Standardprotocolapprovals,registrations,andpatientconsentsThisstudywasapprovedbyinstitutionalreviewboardsofallparticipatinginstitutionsandwritteninformedconsentwasobtainedfromallparticipantsorauthorizedrepresentatives.ParticipantsOurADNIcohortincludedallhealthycontrols(HC),MCIgroupandADgroupparticipantswithavailablebaselineCSFBACEsamplesandgenotypedata.Tore-ducethelikelihoodofpopulationstratificationeffectsintheGWAS,alltheparticipantswererestrictedtonon-HispanicCaucasians.Wealsotestedamultidimensionalscaling(MDS)plotandfoundsomegeneticoutliers(Additionalfile1).Afterqualitycontrol(QC)oftheCSFBACEdataandremovaloftheoutliers,therewere340participants(AD=86,MCI=163,HC=91)withCSFBACEdataleft.DetailedQCstepsforCSFandgenotypedatahavebeenpreviouslyreportedandarebrieflydescribedbelow.CSFBACEmeasurementandqualitycontrolSampleswereobtainedfrom382ADNIsubjects,enrolledat56participatingcentersusingpreviouslyreportedmethodsforCSFmeasurements[13].Formostsamples,thetimefromcollectiontofreezingwaswithin60min.Sampleswereprocessed,aliquoted,andstoredat80°CaccordingtotheADNIBiomarkerCoreLaboratoryStand-ardOperatingProcedures[16].CSFBACEproteinsofallthesamplesweretestedbyasolution-basedBACEenzymaticassaywhichhasbeenregardedasbestassayformatforBACEenzymaticactivityinpreviousstudies[13].Thisassayformatusesabiotinla-beled15aminoacidpeptidebiotin-KTEEISEVNFEVEFR(NFEV)astheBACEsubstrateandusesabaculovirusexpressedc-terminallytruncatedBACE(bBACE)astheBACEenzymestandard.AsourceofBACEusingeitherpurifiedrecombinanttruncatedBACE,humanorrhesusmonkeyCSFwasco-incubatedwiththisBACEsubstrate.TheBACEcleavageproductwasthendetectedusingan“anti-NF”neo-epitopespecificrabbitpolyclonalantibodyandanindirectanti-rabbithorseradishperoxidase(HRP)oralkalinephosphatase(AP)developmentofthereac-tion[13].TheBACEactivityassayincludestwosteps(1Enzyme+substrate,2ELISAtomeasureproduct)[13].LuminescencefromassayplateswasreadonEnVision(PerkinElmer,model2104).ThecountsfromindividualCSFsampleswereconvertedtoBACEen-zymaticactivityusingcoefficientsdeterminedbyaquadraticfittothebBACEstandardcurve[13].Meanandstandarddeviations(SD)baselineofCSFBACEmeasureswerecalculatedbyobserversblindtodiagnosticinformationandsubjectswhohadavaluegreaterorsmallerthan3-foldSDfromthemeanvaluewereregardedasextremeoutliersandremovedfromtheanalysis.GenotypingandqualitycontrolSinglenucleotidepolymorphism(SNP)genotypingformorethan620,000targetSNPswascompletedonallADNIparticipantsusingthefollowingprotocol.Atotalof7mLbloodofeachparticipantwastakeninEDTAcontainingVacutainertubesandgenomicDNAwasextractedusingtheQIAampDNABloodMaxiKit(Qiagen,Inc.,Valencia,CA)followingthemanufacturer’sprotocol.EBV-transformedBlymphoblastoidcelllineswereestablished.GenomicDNAsampleswereanalyzedusingtheHuman610-QuadBea

dChip(Illumina,Inc.,SanDiego,CA)accordin
dChip(Illumina,Inc.,SanDiego,CA)accordingtothemanufacturer’sproto-cols(InfiniumHDAssay;SuperProtocolGuide;rev.A,May2008).StringentQCassessmentwasperformedusingthePLINKsoftwarepackage(http://pngu.mgh.harvard.edu/purcell/plink/),releasev1.07,15asdescribedpreviously.StringentQCassessmentfollowedthesecriteria:mini-mumcallrateforSNPsandindividuals�95%,minimumminorallelefrequencies(MAF)�0.05,Hardy-WeinbergequilibriumtestP�0.001.TherestrictiontoSNPswithaMAFgreaterthan5%servedtoreducethelikelihoodoffalse-positiveresultsinthecontextofmodestsamplesizetoenhancestatisticalpower[14].What’smore,eliminationofrelativelyraremarkersreducedthesever-ityofthemultiplecomparisoncorrectionwhichinturnenhancedstatisticalpower.AftertheQCprocedure,all340participantsremainedintheanalysisandonly519,442outof620,901SNPsremainedintheanalysis.Theoverallgenotypingratefortheremainingdatasetwas99.5%.Huetal.BMCMedicalGenetics (2018) 19:75 Page2of8StatisticalanalysesTofindtheassociationofCSFBACEwiththegeneticpoly-morphism,aseparateGWASwasperformedusingPLINKsoftwareunderanadditivegeneticmodel.ThethresholdsofP05andP08wereusedforsuggestiveandgenome-widesignificantassociationsrespectively[17].Theanalysisincludedatotalof519,442genotypedvariants.Age,genderandAPOE4statuswereincludedascovari-ates.One-wayanalysisofvariance(ANOVA)andTukey’smultiplecomparisonstestwereusedtodeterminethedifferenceofCSFBACEenzymaticactivityindifferentdiag-nosticgroups.TheeffectsofgenotypesonCSFBACEwereexaminedwithamultiplelinearregressionmodel.Therela-tionsbetweentopSNPandAD-relatedphenotypeswerealsoperformedusingPLINKsoftware.Genome-wideasso-ciationswerevisualizedbyasoftwareprogram(R,version3.4.0;TheRFoundation).Regionalassociationswerevisual-izedwiththeLocusZoomwebtool(http://locuszoom.org/).ResultsDemographic,clinical,andcognitivecharacteristicsThedemographicinformationonthefinalsetof340non-HispanicCaucasianparticipantsandsummarysta-tisticsontheCSFBACEareshowedinTable1.Thisstudyincludes86ADsubjects,163MCIsubjectsand91HCsubjects.AsshownintheAdditionalfile2,therewasnosignificantdifferenceingender(ADP=0.32470,MCIP=0.26220,HCP=0.82720,TotalP=0.81770)amongthethreediagnosticgroups.Forage(Additionalfile3),thoughwedidnotfindthelinearcorrelationintheAD,MCIandtotalsamples(ADP=0.9007,MCIP=0.4659,TotalP=0.1447),theresultsshowedalinearcorrelationbetweenageandCSFBACEactivityinHCgroups(P=0.0063).Asexpected,the4alleleofAPOEgenesubstantiallyincreasedtheriskofADwithdoseeffects.Furthermore,CSFBACEenzymaticactivitydidnothavestatisticaldifference(P=0.2762,afterANOVAandTukey’smultiplecomparisonstest)amongthreediagnosticgroups.CharacteristicsofSNPsassociatedwithCSFBACEactivityAfteradjustingforage,genderandAPOE4status,oneSNP(rs1481950,NM_015941.3:c.871-5558Gᆐ&#x.399;餀T)wasidentifiedassociatedwithCSFBACEenzymaticactivityatgenome-widesignificancelevelsofP08.AsshownintheFig.1,oneSNP(rs1481950)intheregionofATP6V1Hgenereachedgenome-widesignificance(P=4.88×109).Moreover,allSNPswhosePvaluesreachedthelevelofP05forCSFBACEenzymaticactivityandtheirannotationinformationarelistedinTable2.AlthoughtheseSNPsdidnotreachthegenome-widesignificancelevelinthislimitedsample,theymaystillhavethepotentialsignificanceinotherlargersamples.SoweregardedtheseSNPswhosePvaluesreachedthelevelofP05assuggestiveSNPs.Sevenindependentpotentialcandidates(uncorrectedP05)wereincluded(rs2507780,rs1866007,rs7903757,rs6473895,rs4075903,rs9693226,rs1900511).TheQuantile-Quantileplot(QQplot)isshowninFig.2andtheinflationfactoris1.Thelinkagedisequilibrium(LD)patternbetween

rs1481950andnearbySNPsisshowninFig.3a.Th
rs1481950andnearbySNPsisshowninFig.3a.ThesenearbySNPsshowedassociationwithCSFBACElevelswiththeP.01.AstheFig.3bshows,whenthetopSNP(rs1481950)wascontrolled,thesenearbySNPsalsodisappeared,whichindicatedthatthesenearbySNPsweredrivenbythetopSNP.CSFBACEactivityandtopSNPAlltheseeightSNPswereanalyzedfurthertoexam-inepossibleinteractionsamongbaselinediagnosisandgenotypesonassociatedCSFBACEenzymaticactivitybyamultiplelinearregressiontestconsider-ingage,genderandAPOE4statusascovariates.ForthetopSNP(rs1481950),asindicatedinFig.4,CSFBACEactivityinthewholesamplesshowedpositivelinearassociationwiththeGalleleinamultiplelin-earregressionmodelafteradjustingforage,gender,andAPOE4status(GG+GT:63.86±22.64pM,TT:45.49±17.99pM,P=4.88×109,Beta:22.48,95%CI:15.22~29.75).FurtheranalysisfoundthislinearassociationbetweengenotypeandCSFBACEactivityinbothAD(GG+GT:60.14±19.13pM,TT:43±16.85pM,P=0.02720Beta:0.23820,95%CI:0.02137~0.43360)andMCIgroups(GG+GT:69.36±23.99pM,TT:46.99±18.72pM,P=0.00030,Beta:0.28280,95%CI:0.13040~0.42210).However,inHCgroup,resultsdidnotshowsignificantdifferencesinCSFBACEactivityamongthreegenotypesubgroupsinbothANOVA(P=0.06370)Table1DemographicinformationBaselinediagnosisADMCIHCTotaln8616391340Age(years),mean±SD(range)75.1±7.8(56.4–89.1)74.2±7.7(54.4–88.8)75.5±5.2(62.0–89.6)74.8±7.1(54.4–89.6)Gender,male/female47/39109/5449/42205/135APOE4carrier(%)45.333.113.230.9CSFBACE(pM),mean±SD44.5±17.548.8±20.146.1±18.047.0±18.9ADAlzheimer’sdisease,MCImildcognitiveimpairment,HChealthycontrol,SDstandarddeviationHuetal.BMCMedicalGenetics (2018) 19:75 Page3of8andmultiplelinearregressionanalysis(P=0.16480,Beta:0.14690,95%CI:-0.06708~0.34790).TopSNPandAD-relatedphenotypesThisstudyalsotestedrelationsbetweenrs1481950andAD-relatedphenotypes.Inallsamples,resultsshowedthatmutationsofrs1481950hadsignificantcorrelationswiththeatrophyofmiddletemporalgyrus(Beta:-0.01425,95%CI:-0.02803~0.00046,P=0.04389)andparahippocampalgyrus(Beta:-0.12170,95%CI:-0.22010~0.02327,P=0.01570).Similarlytothisresult,inADpatients,resultsalsoshowedthatmutationsofrs1481950hadsignificantcorrelationswiththeatro-phyofmiddletemporalgyrus(Beta:-0.13140,95%CI:-0.19900~0.06380,P=0.01894)andparahip-pocampalgyrus(Beta:-0.15100,95%CI:-0.25480~0.04708,P=0.04648).InMCIpatients,resultsshowedthatmutationsofrs1481950hadcorrelationswiththeatrophyofhippocampus(Beta:-0.04772,95%CI:-0.09203~0.00342,P=0.03619),entorhinalcortex(Beta:-0.19020,95%CI:-0.36850~0.01194,P=0.03728)andparahippocampalgyrus(Beta:-0.15180,95%CI:-0.28830~0.01528,P=0.03002)(Fig.5).ForHCgroups,thoughwedidnotfindsomecorrelationsbetweenrs1481950andbrainatrophy,theresultsshowedthatmu-tationsofrs1481950hadsignificantcorrelationswiththeCSFphosphorylatedtauprotein(p-tau)(Beta:8.52600,95%CI:2.74700~14.31000,P=0.00426).DiscussionInthismulticenterstudy,weidentifiedagenome-widesignificantassociationofaSNP(rs1481950)inthegeneATP6V1HregionwithCSFBACEactivityandfoundsevenadditionalsuggestiveassociationloci.Interestingly,Fig.1ManhattanplotsforassociationswithCSFBACE.IntheManhattanplot,theblueandredlinesrepresentthe-log10(105)and-log10(3.10×108)thresholdlevelsTable2SuggestiveSNPsSNPChrPositionSNPTypeClosestRefSeqgeneGenenamePvaluers25077808100588738intronSNX31sortingnexin313.15×106rs18660071560526051intronRORA-AS1RORARARrelatedorphanreceptorA6.02×106rs79037571071552620intronCDH23LOC105378355cadherinrelated239.21×106rs6473895853850344upstreamvariant2KBRGS20regulatorofG-proteinsignaling208.42×106rs40759031140186373intronLRRC4Cleucinerichrepeatcontain

ing4C9.34×106rs9693226853540570MAPK
ing4C9.34×106rs9693226853540570MAPK6PS1mitogen-activatedproteinkinase6pseudogene17.38×106rs19005111071559814intronLOC1053783559.32×106SNPsinglenucleotidepolymorphism,ChrchromosomeHuetal.BMCMedicalGenetics (2018) 19:75 Page4of8thisstudyisthefirsttoshowthatthers1481950riskvariantinATP6V1HsignificantlyaffectsCSFBACEac-tivity.WefoundstatisticallysignificantdifferencesinCSFBACEactivitybetweenthetwogenotypegroups(GG+GTandTT).ItisworthnotingthatthisstatisticaldifferencesbetweengenepolymorphismofATP6V1HandCSFBACEactivityexistedinbothAD(P=0.0027)andMCIgroups(P=0.0003).Theseresultsindicatedthatthers1481950riskvariant(G)inATP6V1HmightsignificantlyincreasetheCSFBACEactivityespeciallyintheADandMCIindividuals.Rs1481950isinanintronicregion.Currentstudieshaven’tfoundtheexactmechanismsofimpairingtheexpressionofATP6V1H.However,therolesofintronicregionswhichhavebeenignoredallthetimehaveattractedtheattentionofscientists.Andmoreandmorestudieshaveprovedthatintronicregionmayplaysomeimportantrolesincontrolingtheinitiationandtermin-ationoftranscription.Moreover,commonmolecularmechanismsforanintronicSNPtoaltermRNAlevelsaretoaffecttranscription,RNAelongation,splicing,ormaturation[18–22].TheATPase,H+transporting,lysosomal50/57kDa,V1subunitHgene(ATP6V1H)atChr8q11.2encodesfortheV1HsubunitofvacuolarATPase(V-ATPase)[23,24].ATP6V1Hgenewasmainlystudiedanddiscussedaboutitsrolesindiabetesinpreviousresearchesandthedatashowedthatthedown-regulationofitsgeneexpressioncorrelateswiththepresenceoftype-2diabetes[25].Thoughthereisstillnoresearchpointoutadirectrela-tionbetweenATP6V1HgeneandAD,somestudiesaboutencodedproteinandmetabolicprocessofBACEindicatethatmutationsofATP6V1HgenemaycontributetotheincreasedBACEactivity.V-ATPasebelongstotherotaryATPasefamilyandisamultiproteinmembranecomplex.ThemostimportantfunctionofV-ATPaseistoacidifyintracellularcompartmentsbyusingtheenergygatheredfromATPhydrolysistopumpprotons[23].OneofthemostimportantinfluencefactorsofBACEactivityisPH.AcidicIntracellularenvironment(PH=4.5)isoptimalforBACEactivity.SothedysfunctionofV-ATPasemayleadtothechangeoftheacidicIntracellularenvironmentandtheninfluencetheBACEactivity.BACEactivityisalsoas-sociatedwithmatureprocessing.ThematureprocessofBACErequirestheformationofdisulfidebonds,glycosyla-tionandsomeothermodificationprocesses[26].Thesestepsoccursintheendoplasmicreticulumandgolgibodyandmaybeinfluencedbythechangesoftheintracellularenvironment.Moreover,V-ATPasealsoplaysanimport-antroleinlysosomalacidification.LysosomalpathwayisFig.2Quantile-QuantileplotFig.3RegionalplotsforassociationswithCSFBACE.aRegionalassociationresultsfortheATP6V1Hregionofchromosome8.bAssociationresultsforchromosome8:54290285–55090285controllingforrs1481950Huetal.BMCMedicalGenetics (2018) 19:75 Page5of8animportantdegradationpathwayofBACEprotein[27,28].Tosumup,V-ATPasemayplaysomeimport-antrolesinbothBACEproteinlevelsandactivity.Theresultsalsoshowedthatmutationsofrs1481950hadsignificantcorrelationswiththeatrophyofAD-relatedencephalicregionsincludingmiddletemporalgyrus,parahippocampalgyrus,hippocampusandento-rhinalcortex.Moreover,theresultsshowedthatmuta-tionsofrs1481950hadsignificantcorrelationswiththeCSFp-tau.Theseresultsindicatedthatmutationsofrs1481950mayrelatetoADandmayinfluencethevol-umeofAD-relatedencephalicregionsbychangingthemetabolismofsomeAD-relatedproteinssuchastau.Inthisstudy,wedidnotfindanysignificantdifferencesinCSFBACEactivityamongthethreediagnosticgroups.ThisaspectofourresearchisinlinewithtwopreviousstudiesincludinganADNIcohortstudy[29,30].Somepreviousstudiesshowdifferentresultsinc

ludingincreasedactivityinMCIandAD[31],an
ludingincreasedactivityinMCIandAD[31],andincreasedBACE1activ-ityinMCIbutnotinAD[13,32].Theseinconsistentre-sultsmaybeexplainedbythecharacteristicsofthestudysamples,thewiderangeofBACE1activitymeasurements,andthelargeoverlapbetweenthegroups.Thereareseveralpotentiallimitationsofthisstudy.First,theADNI-1samplewaslimitedinsamplesizewhenCSFBACEdata,differentgenotypesanddiagnosissubgroupsweretakenintoconsideration,whichmakeseffectivesamplesizesmallforsometests.Thus,itwillstillbenecessarytoreplicatethesefindingsinalargerdataset.Second,recentstudiesalsoindicatedthatexceptCSFBACE,plasmaticBACEwasalsoassociatedwithAD[33].Theassociationsamonggenepolymorphism,CSFBACE,plasmaticBACEandADstillneedfurtherstudy.Third,oursamplewasrestrictedtoCaucasianstoavoidgeneticsstratificationacrossethnicities,whileallthegenesmayshowdifferentfrequenciesandpolymor-phismsindifferentpopulations.Therelationshipsbe-tweenthesegenesandADneedtobetestedinmorepopulations.Fourth,theADNIdatabasedidnotcoverthedetailedclassificationoftheADpatients(sporadicandfamilial;earlyonsetandlateonset).Sofurtherstud-ieswithabetterprofileofADpatientsshouldbetestedforcomparativepurposeswithotherstudies.ConclusionsInsummary,afteraseparateGWASofCSFBACE,wefoundatopSNP(rs1481950)inATP6V1HgenewiththePvaluereachinggenome-widesignificanceandsevensuggestiveSNPswiththePvaluelowerthan105.Rs1481950riskvariant(G)inATP6V1HmayincreasetheCSFBACEactivity.Sevengenes(SNX31,RORA,CDH23,RGS20,LRRC4C,MAPK6PS1,LOC105378355)wereregardedascandidategenes.TheseresultsprovideFig.5Forestplotofrs1481950andAD-relatedencephalicregionsinADNI.24M:Thefollow-upperiodis24months;BL:Baseline;CI:credibilityintervalFig.4MeanCSFBACEactivityasafunctionofbaselinediagnosisandgenotype.MeanandstandarderrorsofCSFBACEactivityareshownforgroupsdefinedbybaselinediagnosis.TheCSFBACEactivityinthetotalsamplesshowedpositivelinearassociationwiththeGalleleinamultiplelinearregressionmodelafteradjustingforage,gender,andAPOE4status(GG+GT:63.86±22.64pM,TT:45.49±17.99pM,P=4.88×109).ThePvalueofamultiplelinearregressionmodelineachdiagnosisgroupwasdisplayedinthefig.(AD:P=0.027MCI:P=0.0003HC:P=0.165)Huetal.BMCMedicalGenetics (2018) 19:75 Page6of8cluestosomenovelpathogenicgenesassociatedwithsomeBACErelateddiseases,suchasAD.Thein-depthdiscussionandstudyoftheseassociationscanhelpustofindtheexactmechanismsofAD,whichmayindi-catesomenewdiagnosticmethodsandtherapeuticdirections.AdditionalfilesAdditionalfile1:TheMDSplotofsamples.(TIF8989kb)Additionalfile2:ThecorrelationbetweengenderandCSFBACEactivity.(TIF2480kb)Additionalfile3:ThecorrelationbetweenageandCSFBACEactivity.(TIF2436kb)AbbreviationsAD:Alzheimerdisease;ADNI:Alzheimer’sDiseaseNeuroimagingInitiative;ANOVA:One-wayanalysisofvariance;AP:Alkalinephosphatase;APP:Amyloidprecursorprotein;A
:Amyloid-beta;BACE:
-siteAPPcleavingenzyme;bBACE:BaculovirusexpressedcterminallytruncatedBACE;CSF:Cerebrospinalfluid;GWAS:Genome-wideassociationstudy;HC:Healthycontrols;HRP:Horseradishperoxidase;LD:Linkagedisequilibrium;MAF:Minorallelefrequencies;MCI:Mildcognitiveimpairment;MRI:Magneticresonanceimaging;PET:Positronemissiontomography;QC:Qualitycontrol;SD:Standarddeviations;SNPs:SinglenucleotidepolymorphismsAcknowledgementsDatacollectionandsharingforthisprojectwasfundedbytheAlzheimer’sDiseaseNeuroimagingInitiative(ADNI)(NationalInstitutesofHealthGrantU01AG024904)andDODADNI(DepartmentofDefenseawardnumberW81XWH-12-2-0012).ADNIisfundedbytheNationalInstituteonAging,theNationalInstituteofBiomedicalImagingandBioengineering,andthroughgenerouscontributionsfromthefollowing:AbbVie,Alzheimer’s

Association;Alzheimer’sDrugDiscoveryFoun
Association;Alzheimer’sDrugDiscoveryFoundation;AraclonBiotech;BioClinica,Inc.;Biogen;Bristol-MyersSquibbCompany;CereSpir,Inc.;Cogstate;EisaiInc.;ElanPharmaceuticals,Inc.;EliLillyandCompany;EuroImmun;F.Hoffmann-LaRocheLtd.anditsaffiliatedcompanyGenentech,Inc.;Fujirebio;GEHealthcare;IXICOLtd.;JanssenAlzheimerImmunotherapyResearch&Development,LLC.;Johnson&JohnsonPharmaceuticalResearch&DevelopmentLLC.;Lumosity;Lundbeck;Merck&Co.,Inc.;MesoScaleDiagnostics,LLC.;NeuroRxResearch;NeurotrackTechnologies;NovartisPharmaceuticalsCorporation;PfizerInc.;PiramalImaging;Servier;TakedaPharmaceuticalCompany;andTransitionTherapeutics.TheCanadianInstitutesofHealthResearchisprovidingfundstosupportADNIclinicalsitesinCanada.PrivatesectorcontributionsarefacilitatedbytheFoundationfortheNationalInstitutesofHealth(www.fnih.org).ThegranteeorganizationistheNorthernCaliforniaInstituteforResearchandEducation,andthestudyiscoordinatedbytheAlzheimer’sTherapeuticResearchInstituteattheUniversityofSouthernCalifornia.ADNIdataaredisseminatedbytheLaboratoryforNeuroImagingattheUniversityofSouthernCalifornia.DatausedinpreparationofthisarticlewereobtainedfromtheAlzheimer’sDiseaseNeuroimagingInitiative(ADNI)database(adni.loni.usc.edu).Assuch,theinvestigatorswithintheADNIcontributedtothedesignandimplementationofADNIand/orprovideddatabutdidnotparticipateinanalysisorwritingofthisreport.AcompletelistingofADNIinvestigatorscanbefoundat:http://adni.loni.usc.edu/wp-content/uploads/how_to_apply/ADNI_Acknowledgement_List.pdf.FundingThisworkwassupportedbygrantsfromtheTaishanScholarsProgramofShandongProvince(ts201511109andtsqn20161079),QingdaoKeyHealthDisciplineDevelopmentFund,QingdaoOutstandingHealthProfessionalDevelopmentFund,andQingdaoInnovationandEntrepreneurshipLeadingTalentProgram.AvailabilityofdataandmaterialsDataareavailabletoresearchersbyapplyingtotheADNIconsortia.Applicationisrequiredtoprotectparticipantconfidentiality.TheADNIdataareavailableat(http://adni.loni.usc.edu/).Authors’contributionsLTandJTYledandsupervisedtheresearch.JTYdesignedtheresearch.HH,HYLandJQLpreparedthegenotypingdata,performedthegeneanalysis,andperformedqualitycontrolandGWAS.HHdraftedthemanuscript.Alltheauthorsreviewed,commentedandapprovedthemanuscript.EthicsapprovalandconsenttoparticipateThestudyprocedureswereapprovedbytheinstitutionalreviewboardsofallparticipatingcenters(https://adni.loni.usc.edu/wp-content/uploads/how_to_apply/ADNI_Acknowledgement_List.pdf),andwritteninformedconsentwasobtainedfromallparticipantsortheirauthorizedrepresentatives.Ethicsapprovalwasobtainedfromtheinstitutionalreviewboardsofeachinstitutioninvolved:OregonHealthandScienceUniversity;UniversityofSouthernCalifornia;UniversityofCalifornia—SanDiego;UniversityofMichigan;MayoClinic,Rochester;BaylorCollegeofMedicine;ColumbiaUniversityMedicalCenter;WashingtonUniversity,St.Louis;UniversityofAlabamaatBirmingham;MountSinaiSchoolofMedicine;RushUniversityMedicalCenter;WienCenter;JohnsHopkinsUniversity;NewYorkUniversity;DukeUniversityMedicalCenter;UniversityofPennsylvania;UniversityofKentucky;UniversityofPittsburgh;UniversityofRochesterMedicalCenter;UniversityofCalifornia,Irvine;UniversityofTexasSouthwesternMedicalSchool;EmoryUniversity;UniversityofKansas,MedicalCenter;UniversityofCalifornia,LosAngeles;MayoClinic,Jacksonville;IndianaUniversity;YaleUniversitySchoolofMedicine;McGillUniversity,Montreal-JewishGeneralHospital;SunnybrookHealthSciences,Ontario;U.B.C.ClinicforAD&RelatedDisorders;CognitiveNeurology—St.Joseph’s,Ontario;ClevelandClinicLouRuvoCenterforBrainHealth;NorthwesternUniversity;PremiereResearchInst(PalmBeac

hNeurology);GeorgetownUniversityMedicalC
hNeurology);GeorgetownUniversityMedicalCenter;BrighamandWomen’sHospital;StanfordUniversity;BannerSunHealthResearchInstitute;BostonUniversity;HowardUniversity;CaseWesternReserveUniversity;UniversityofCalifornia,Davis—Sacramento;NeurologicalCareofCNY;ParkwoodHospital;UniversityofWisconsin;UniversityofCalifornia,Irvine—BIC;BannerAlzheimer’sInstitute;DentNeurologicInstitute;OhioStateUniversity;AlbanyMedicalCollege;HartfordHospital,OlinNeuropsychiatryResearchCenter;Dartmouth-HitchcockMedicalCenter;WakeForestUniversityHealthSciences;RhodeIslandHospital;ButlerHospital;UCSanFrancisco;MedicalUniversitySouthCarolina;St.Joseph’sHealthCareNathanKlineInstitute;UniversityofIowaCollegeofMedicine;CornellUniversity;andUniversityofSouthFlorida:USFHealthByrdAlzheimer’sInstitute.CompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.Publisher’sNoteSpringerNatureremainsneutralwithregardtojurisdictionalclaimsinpublishedmapsandinstitutionalaffiliations.Authordetails1DepartmentofNeurology,QingdaoMunicipalHospital,SchoolofMedicine,QingdaoUniversity,No.5DonghaiMiddleRoad,Qingdao266071,ShandongProvince,China.2DepartmentofNeurology,WeihaiWeiPeople’sHospital,Weihai,China.3ClinicalResearchCenter,QingdaoMunicipalHospital,QingdaoUniversity,Qingdao,China.4DepartmentofNeurology,UniversityofCalifornia,SanFrancisco,675NelsonRisingLane,Suite190,Box1207,SanFrancisco,CA94158,USA.Received:11January2018Accepted:2May2018References1.ScheltensP,etal.Alzheimer’sdisease.Lancet.2016;388(10043):505–17.2.VassarR,CitronM.Abeta-generatingenzymes:recentadvancesinbeta-andgamma-secretaseresearch.Neuron.2000;27(3):419–22.3.ZhongZ,etal.Levelsofbeta-secretase(BACE1)incerebrospinalfluidasapredictorofriskinmildcognitiveimpairment.ArchGenPsychiatry.2007;64(6):718–26.Huetal.BMCMedicalGenetics (2018) 19:75 Page7of84.FukumotoH,etal.Beta-secretaseproteinandactivityareincreasedintheneocortexinAlzheimerdisease.ArchNeurol.2002;59(9):1381–9.5.YangLB,etal.Elevatedbeta-secretaseexpressionandenzymaticactivitydetectedinsporadicAlzheimerdisease.NatMed.2003;9(1):3–4.6.LiR,etal.Amyloidbetapeptideloadiscorrelatedwithincreasedbeta-secretaseactivityinsporadicAlzheimer’sdiseasepatients.ProcNatlAcadSciUSA.2004;101(10):3632–7.7.JohnstonJA,etal.Expressionandactivityofbeta-siteamyloidprecursorproteincleavingenzymeinAlzheimer’sdisease.BiochemSocTrans.2005;33(Pt5):1096–100.8.AhmedRR,etal.BACE1andBACE2enzymaticactivitiesinAlzheimer’sdisease.JNeurochem.2010;112(4):1045–53.9.HolsingerRM,etal.Increasedbeta-secretaseactivityincerebrospinalfluidofAlzheimer’sdiseasesubjects.AnnNeurol.2004;55(6):898–9.10.VerheijenJH,etal.DetectionofasolubleformofBACE-1inhumancerebrospinalfluidbyasensitiveactivityassay.ClinChem.2006;52(6):1168–74.11.ZetterbergH,etal.ElevatedcerebrospinalfluidBACE1activityinincipientAlzheimerdisease.ArchNeurol.2008;65(8):1102–7.12.EwersM,etal.IncreasedCSF-BACE1activityisassociatedwithApoE-epsilon4genotypeinsubjectswithmildcognitiveimpairmentandAlzheimer’sdisease.Brain.2008;131(Pt5):1252–8.13.WuG,etal.Decreaseinage-adjustedcerebrospinalfluidbeta-secretaseactivityinAlzheimer’ssubjects.ClinBiochem.2008;41(12):986–96.14.ChenJ,etal.Genome-wideassociationstudyidentifiesMAPTlocusinfluencinghumanplasmataulevels.Neurology.2017;88(7):669–76.15.PetersenRC,etal.Alzheimer’sDiseaseNeuroimagingInitiative(ADNI):clinicalcharacterization.Neurology.2010;74(3):201–9.16.ShawLM,etal.CerebrospinalfluidbiomarkersignatureinAlzheimer’sdiseaseneuroimaginginitiativesubjects.AnnNeurol.2009;65(4):403–13.17.KimS,etal.Genome-wideassociationstudyofCSFbiomarkersAbeta1-42,t-tau,andp-tau181pintheADNIcoho

rt.Neurology.2011;76(1):69–79.18.Giacope
rt.Neurology.2011;76(1):69–79.18.GiacopelliF,etal.ThefirstintronofthehumanosteopontingenecontainsaC/EBP-beta-responsiveenhancer.GeneExpr.2003;11(2):95–104.19.HugoH,etal.MutationsintheMYBintronIregulatorysequenceincreasetranscriptionincoloncancers.GenesChromosomesCancer.2006;45(12):1143–54.20.WangD,etal.RegulatorypolymorphisminvitaminKepoxidereductasecomplexsubunit1(VKORC1)affectsgeneexpressionandwarfarindoserequirement.Blood.2008;112(4):1013–21.21.BaralleM,etal.InfluenceofFriedreichataxiaGAAnoncodingrepeatexpansionsonpre-mRNAprocessing.AmJHumGenet.2008;83(1):77–88.22.BurattiE,etal.NuclearfactorTDP-43bindstothepolymorphicTGrepeatsinCFTRintron8andcausesskippingofexon9:afunctionallinkwithdiseasepenetrance.AmJHumGenet.2004;74(6):1322–5.23.MarshanskyV,RubinsteinJL,GruberG.EukaryoticV-ATPase:novelstructuralfindingsandfunctionalinsights.BiochimBiophysActa.2014;1837(6):857–79.24.ColacurcioDJ,NixonRA.Disordersoflysosomalacidification-theemergingroleofv-ATPaseinagingandneurodegenerativedisease.AgeingResRev.2016;32:75–88.25.MolinaMF,etal.DecreasedexpressionofATP6V1Hintype2diabetes:apilotreportonthediabetesriskstudyinMexicanAmericans.BiochemBiophysResCommun.2011;412(4):728–31.26.HoglS,etal.Determinationoftheproteolyticcleavagesitesoftheamyloidprecursor-likeprotein2bytheproteasesADAM10,BACE1andgamma-secretase.PLoSOne.2011;6(6):e21337.27.QingH,etal.DegradationofBACEbytheubiquitin-proteasomepathway.FASEBJ.2004;18(13):1571–3.28.KohYH,etal.BACEisdegradedviathelysosomalpathway.JBiolChem.2005;280(37):32499–504.29.PerneczkyR,AlexopoulosP,Alzheimer'sDiseaseeuroimagingInitiative.CerebrospinalfluidBACE1activityandmarkersofamyloidprecursorproteinmetabolismandaxonaldegenerationinAlzheimer’sdisease.AlzheimersDement.2014;10(5Suppl):S425–9.e130.SavageMJ,etal.SolubleBACE-1activityandsAbetaPPbetaconcentrationsinAlzheimer’sdiseaseandage-matchedhealthycontrolcerebrospinalfluidfromtheAlzheimer’sDiseaseNeuroimagingInitiative-1baselinecohort.JAlzheimersDis.2015;46(2):431–40.31.RosenC,etal.Cerebrospinalfluidprofilesofamyloidbeta-relatedbiomarkersinAlzheimer’sdisease.NeuroMolecularMed.2012;14(1):65–73.32.WuG,etal.Decreaseinbrainsolubleamyloidprecursorproteinbeta(sAPPbeta)inAlzheimer’sdiseasecortex.JNeurosciRes.2011;89(6):822–32.33.ShenY,etal.IncreasedPlasmaBeta-Secretase1MayPredictConversiontoAlzheimer'sDiseaseDementiainIndividualsWithMildCognitiveImpairment.BiolPsychiatry.2018;83(5):447-55.Huetal.BMCMedicalGenetics (2018) 19:75 Page8of8RESEARCHARTICLEOpenAccessGenome-wideassociationstudyidentifiedATP6V1HlocusinfluencingcerebrospinalfluidBACEactivityHaoHu1,HaiyanLi2,JieqiongLi1,JintaiYu1,3,4*,LanTan1,3*andAlzheimer’sDiseaseNeuroimagingInitiativeAbstractBackground:Theactivityofcerebrospinalfluid(CSF)*Correspondence:yu-jintai@163.com;dr.tanlan@163.com1DepartmentofNeurology,QingdaoMunicipalHospital,SchoolofMedicine,QingdaoUniversity,No.5DonghaiMiddleRoad,Qingdao266071,ShandongProvince,ChinaFulllistofauthorinformationisavailableattheendofthearticle©TheAuthor(s).2018OpenAccessThisarticleisdistributedunderthetermsoftheCreativeCommonsAttribution4.0InternationalLicense(http://creativecommons.org/licenses/by/4.0/),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedyougiveappropriatecredittotheoriginalauthor(s)andthesource,providealinktotheCreativeCommonslicense,andindicateifchangesweremade.TheCreativeCommonsPublicDomainDedicationwaiver(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthisarticle,unlessotherwisestated.Huetal.BMCMedicalGenetics (2018) 19:75 https://doi.org/10.1186/s12881-018-