ContentslistsavailableatScienceDirect
Behavioural Brain Research
j o ur na l h o me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / b b r
Research report
Possible contribution of epigenetic changes in the development of schizophrenia-like behavior in vasopressin-deficient Brattleboro rats
Kornél Demeter
a,1, Bibiána Török
a,1, Anna Fodor
a, János Varga
a, Szilamér Ferenczi
b, Krisztina J. Kovács
b, Ildikó Eszik
c, Viktor Szegedi
c, Dóra Zelena
a,∗aDepartmentofBehavioralNeurobiology,InstituteofExperimentalMedicine,Budapest,Hungary
bLaboratoryofMolecularNeuroendocrinology,InstituteofExperimentalMedicine,Budapest,Hungary
cBiologicalResearchCenter,HungarianAcademyofSciences,Szeged,Hungary
h i g h l i g h t s
•AVPdeficientBrattlebororatstrainisavaluablemodeltostudyschizophrenia.
•LackofAVPresultedinchangesofhistoneacetylationinthefrontalbrainandhippocampus.
•Influencinghistoneacetylationmayprovidetherapeuticalternativesforschizophrenia.
a r t i c l e i n f o
Articlehistory:
Received14September2015
Receivedinrevisedform4December2015 Accepted9December2015
Availableonline15December2015
Keywords:
Schizophrenia Epigenetics Vasopressin Brattleboro
a b s t r a c t
Schizophrenia-likesymptomsweredetectedinvasopressin-deficient(di/di)Brattlebororats,anditwas alsosuggestedthatschizophreniamighthaveanepigeneticcomponent.Weaimedtoclarifyifepigenetic changescontributetoschizophrenia-likebehaviorofthisstrain.
Behavioral(locomotionbytelemetry,cognitionbynovelobjectrecognition,socialrecognitionand socialavoidance test,attentionbypre-pulse inhibition)andepigenetic differenceswerecompared betweenwildtypeanddi/dianimals.DNAmethyltransferase1(DNMT1),DNMT3a,aswellasCOMT, GAD,VGLUT1,5HT2A,BDNFmRNAlevelsinprefrontalbrainregionandhippocampuswerestudiedby qRT-PCR.Histone3(H3)andH4acetylation(Ac)werestudiedbywestern-blotfollowedbyregionspecific examinationofH3lysine9(K9)acetylationbyimmunohistochemistry.
Impairedcognitive,socialandattentionbehaviorofdi/diratsconfirmedschizophrenia-likesymptoms inourlocalcolony.Thepan-AcH3immunoreactivitywaslowerinprefrontalregionandelevatedinthe hippocampusofdi/dianimals.Wefoundlowerimmunopositivecellnumberinthedorsalpeduncular prefrontalcortexandtheventrallateralseptumandincreasedAcH3K9immunoreactivityinCA1region ofdi/dianimals.TherewerenomajorsignificantalterationsinthestudiedmRNAlevels.
WeconfirmedthatBrattlebororatisagoodpreclinicalmodelofschizophrenia.Itsschizophrenia- likebehavioralalterationwasaccompaniedbychangesinH3acetylationintheprefrontalregionand hippocampus.Thismaycontributetodisturbancesofmanyschizophrenia-relatedsubstancesleadingto developmentofschizophrenia-likesymptoms.Ourstudiesconfirmedthatnotasinglegene,ratherfine changesinanarrayofmoleculesareresponsibleforthemajorityofschizophreniacases.
©2015ElsevierB.V.Allrightsreserved.
1. Introduction
Schizophrenia(SCZ),oneofthemajorpsychoses,affectsmore than1%ofthewholehumanpopulation[1,2].SCZpatientsarechar-
∗ Correspondingauthorat:HungarianAcademyofSciences,InstituteofExper- imental Medicine,Department ofBehavioral Neurobiology, H-1083Budapest, Szigonyutca43,Hungary.Fax:+3612109954.
E-mailaddress:zelena.dora@koki.mta.hu(D.Zelena).
1 Theseauthorscontributedequally.
acterizedby delusions, hallucinations,disorganized speechand behavioretc.[3].Thesesymptomsalsoincludesocialdysfunction, whichadditionallyaffectsfamilymembersandburdenthesociety.
Despitetheimportance,thereisnocureforSCZ.Manypartially effectivetherapiesexist,butnonecanguaranteeacuretoaspe- cificcaseofthedisease.Themechanismofallrecentlyavailable antipsychoticdrugswasdiscovered60yearsago[4,5],theirmain therapeuticeffectistheblockadeofdopaminergicreceptors[6].
A fundamental barrier tonewdrug developmentfor thetreat- mentofmajorpsychiatricdisorderswasandstillisanincomplete
http://dx.doi.org/10.1016/j.bbr.2015.12.007 0166-4328/©2015ElsevierB.V.Allrightsreserved.
understandingoftheethiopathogeneticmechanismsunderlying thesymptomatologyofthesediseases.
Theepigenetichypothesis providesanintegrative theoryfor SCZ [7–9] unifying the multifactorial origin under a common framework[10,11].Geneticstudieshavelinkedmanygenesand chromosomalregionsdistributedthroughoutthegenometoSCZ, butnosingleorsmallnumberofgenesaccountsforthemajor- ity of cases [6,12]. According to the biopsychosocial model of Engel perinatal development is a determining factor for later occurrence of psychiatric disorders [13]. Especially during this period(butalsolaterinlife)environmentallyinducedepigenetic changesmayactivate(e.g.,histoneacetylationbyacetyltransferase (HAT)) or inhibit (e.g., DNA methylation by DNA methyltrans- ferases(DNMT)orhistonedeacetylationbydeacetylases(HDAC)) theexpressionofdifferentgenes[14,15].Thereforeonthesame geneticbackground different susceptibilityto SCZ-inducing life eventsmaydevelop[16].Asdevelopmentaldisturbancesarecore features of SCZ,and epigenetic changes occurglobally in early developmentacting asinterface betweentheenvironment and thegenome[17],we mightassume thatepigenetic rather than geneticchanges have generalethiopathogenic role. Indeed, the first suggestion on the epigenetic aspect of SCZ goes back to theearly60s,whenenhancingDNAmethylationbymethionine administration was shown to exacerbate psychotic symptoms [18].Sincethenmanystudiesfoundepigeneticmodifications in SCZpatients,aswellasinanimal models(forareviewseee.g., ourbook chapter[14]).The importance ofepigenetic theory is that epigenetic processes are highly dynamic unlike the DNA sequenceofacell,whichisstableandstronglyconserved.There- foreepigenetic disruptions may develop relativelyfast and are potentially reversible, thus a realistic target for pharmacologi- calintervention.Indeed,thefactthatantipsychoticsareclinically effectivejustafterprolongedtreatment,and—similarly—moretime isneededforthedevelopmentofepigeneticchanges,suggeststhat oneof themaintarget ofthesedrugs mightbetheepigenome [19–21].
Developmentof newdrugs requires animal models.Genetic modelshave the advantage that the SCZ-inducing intervention doesnotinterferewiththetreatmentdrug.However,onemight assumethattheperinataldevelopmentoftheseanimalsissimi- larlyimportantleadingtoepigeneticchangesonadisturbedgenetic background.Herewewerefocusingontheargininevasopressin (AVP)deficient(di/di)Brattlebororatstrain.Reducedlevelofthis socialhormone[22–26],oritscarrier,neurophysin[24]wasfound inSCZpatients.LaterstudiesrevealedanassociationbetweenAVP gene[27] orAVP1areceptor [28,29]and somebehavioral dys- function(e.g.,pre-pulseinhibition(PPI)).Moreover,someofthe SCZ-symptomswereamelioratedbytreatmentwithAVPanalogs [30–35].However,otherstudiesfoundelevatedlevels[36,37]or nobeneficial effect of theAVP treatment [38], therefore direct pathogenicroleofAVP-deficiencyinthedevelopmentofsymptoms isnotevident.
Nevertheless, the AVP-deficient Brattleboro rat shows SCZ- relatedsymptoms,which developspontaneously. Thisstrain,in additiontoexhibiting adiabetesinsipidus-likephenotypechar- acterizedbypolydipsiaandpolyuria,havebeenshowntohavea numberofcognitiveandbehavioral abnormalitiesthatareanal- ogoustothose seenin SCZpatients, includingabnormalities in memory,emotionalreactivity,socialrecognition,motivationand attention[39–44].AsAVPisnotclearlyconnectedtodiseaseeti- ology,epigeneticcontributionismorelikelyleadingtosecondary changesindifferentneurotransmittersystems[45,46].
Wehypothesized,thatepigeneticchangesassomeparticular DNAandhistonemodifications,maycontributetothedevelop- ment of SCZ-like symptoms in this animal model. We focused onthefrontal partof thebrain (FB) as themain areathought
tobeinvolvedinSCZ[47]andassociatedmostlywithepigenetic changesinthisdisorder[14,48,49].However,thecurrentlitera- turedoessuggestthatalterationsarenotisolatedtoafewbrain regions,butarecharacterizedbyabnormalitieswithinbrainnet- works[50],likethehippocampo-prefrontalcortex (PFC)system [51].Thus,we studiedthehippocampus(HC)aswell,asoneof themain centerof memory[52], becausecognitivedeficitsare mostpredictiveoflong-termoutcomesofSCZ.Accordingtorecent findingscorticalGABAergicdeficitdominatesSCZpathophysiology [53],thereforeweexamineditsdetectionmarker,thesynthetiz- ingenzymeglutamatedecarboxylase(GAD1),theisoformknown tobedownregulatedinPFC[54]andHC[55]ofSCZpatients.In relationwiththedopaminergic, glutamatergicand serotonergic neurotransmittersystems,cathecol-o-methy-transferase(COMT), vesicular glutamate transporter1 (VGLUT1) and 5HT2A recep- tormRNAlevelswerestudiedrespectively.Neurodevelopmental changeswereexaminedbychangesinbrainderivedneurotrophic factor(BDNF)mRNA.Methylationchangesofthesespecialgenes wereobservedmostlyinrelationtoSCZ[14].
2. Materialsandmethods
2.1. Animals
Adult,maleBrattlebororats(∼330g,10–12weeksold)were maintainedinourinstitute[56]inacolonystartedfrombreeder ratsfromHarlan,Indianapolis,IN,USA.Ratswerekeptinconven- tionalplastic cages (30×40cm)among controlledenvironment (23±1◦C, 50–70%humidity, 12h light starting at 07:00h) and givencommercialratchow(CharlesRiver,Budapest,Hungary)and tapwateradlibitum.We comparedtheAVP deficienthomozy- gous(di/di)ratswithdiabetesinsipidustohomozygousnormal (+/+)controlrats.Allexperimentswereconductedinthemorn- ingbetween09:00and12:00h.Theexperimentswereperformed inaccordancewithregulationssetbytheEuropeanCommunities CouncilDirective(2010/63/EU)andwereapprovedbyourInstitu- tionalAnimalCareandUseCommittee.
2.2. Biotelemetry
Locomotoractivitywasmonitoredviaimplantedbiotelemetry emitters(MinimitterCo.,Bend,OR,USA)asdescribedearlier[57].
Inbrief,theemitterwasplacedintotheabdominalcavityofrats throughamidlineabdominalincisionunderKetamin(50mg/kg;
ProdulabPharmab.v.,Netherlands),Xylasine(10mg/kg;Produlab Pharma b.v., Netherlands), and Pipolphen (5mg/kg; Egis Phar- maceuticals PLC, Hungary) anesthesia. Biotelemetric recordings started2weekslaterandweremadebymeansofa12-channel VitalViewsystem(MinimitterCo.).Dailyactivitywasmonitoredfor 24h,startedinthefirsthourofdarkperiod,visualizedinZeitgeber time.
2.3. Behavioraltests
2.3.1. Novelobjectrecognitiontest(NOR)
The rats were singly habituated to the experimental cages (41.3×26×29.8cm,GeoMaxi,Ferplast,Italy)withfreshbedding 1hbeforestartingthetest. Fortheobject relatedequivalentof socialdiscriminationwedevelopedanexperimentaldesigncom- parabletothesocialparadigminbothtimecourseandtestsettings [58].Twodifferentobjectswereused:a62gtinboxandan80g bottleof tomatosauce. Oneof this objects(Object1)waspre- sentedfor4min(samplingphase),thanremoved,and30minlater thesame(Object1)andadifferentobject(Object2)wereintro- ducedtotheratfor4min.Theobjectswerethoroughlycleaned withalcoholbeforeeach animal.Thetestswerevideotapedand
analyzedlaterbyanexperimenterblindtothetreatmentsbymeans ofacomputer-basedeventrecorder(H77,Budapest,Hungary).In behaviordirectedtowardtheobjectwedistinguishedsniffingand gnawing asanimportantcomponent. Toexcludeobjectprefer- encethetwoobjectswererandomlyusedasObject1orObject 2.Discriminationindexwascalculatedasfollows:(timepercent- ageObject2−timepercentageObject1)/(timepercentageObject 1+timepercentageObject2)×100.Theresultoftheindexchanges between−100%and+100%,where0=nodiscrimination.Normally the animals spend more time with the new stimulus (novelty effect),thustheindexaroundorbelow0isasignofmemorydeficit.
2.3.2. Socialrecognitiontest(SR)
LikeintheNORtest,afterthehabituationperiod,insteadofthe objects,ajuvenilerat(average20daysold)wasplacedintothecage for4min(samplingphase)[59].Aftera30-mininter-trialinterval thejuvenilefromtrial1(old)plusanoveljuvenilewereplacedthere for4min(juvenilesweremarkedwithdifferentcolors).Both4min werevideo-recordedandlatertheamountofinteractionwasmea- sured.Thefirstsessionshowssocialinteractiontendencyinanovel environmentandthesecondoneshowsthediscriminationcapabil- itybetweenthetwojuveniles.Discriminationindexwascalculated asaboveinNOR.
2.3.3. Socialavoidance(SA)
The test was performed as described earlier [57]. Briefly, rats were studied in two plastic cages connected by a slid- ing door.The subject was placed in the smallercage (surface:
15cm×50cm×40cm)fora3minhabituationperiod.Thelarger cage (40cm×40cm×40cm) was divided into two equal com- partmentsby atransparent,perforated plasticwall. Thedistant compartmentcontainedalargeunfamiliarmale.Afterthehabit- uationperiod,theslidingdoorwasremoved,andthesubjectwas allowedtoexplorethecagefor5min.Thetestapparatusdidnot permitphysicalcontactbetweentheexperimentaland stimulus animals.Behaviorwasvideo-recordedfromaboveandanalyzed later.Threevariableswererecorded:thefrequencyofvisitsmade tothecompartmentcontainingtheopponentaswellasduration andtheinteractionwithunfamiliarmale(opponententriesand time%spentwithopponent,respectively).
2.3.4. Prepulseinhibition(PPI)
TheColbourneInstruments(USA)AcousticStartlesetupwas usedinourexperiments.Afteraweightcalibrationsubjectswere placed in a test cage on an instrument measuring the startle response(byrecordingsmallchangesinweight) insideasound attenuatedchamber.Following5minofhabituationsubjectswere presented40mslong,120dBacousticstimuli(noise)for5timesin every20sectostandardizestartle.Fivetrialtypeswerethenpre- sentedduringtesting:(i)40mslong120dBstartlepulsewithout prepulse,(ii–iv)40ms120dBpulsepreceded80msbya20mspre- pulse(tone)ofthreevaryingintensity(73,77or81dB)and(v)atrial withnoprepulseand0dBpulse.Trialswererepeateduntilevery typeoftrials waspresented5 times.Different animalsreceived different trial types in a different, randomized order. Immedi- ately afterwards other four trial types were randomlyapplied with75dB prepulseand increasing intervals(30, 100, 300and 500ms)betweenpre-pulseandpulse(ISI).Theprogramautomat- icallyrecorded thestartleresponse.Responsetothe0dBpulse wastheweightofthesubjectandwassubtractedfromsubsequent startleresponsedata.Meanofthestartleresponsetothe120dB pulsewithoutprepulsewascalculatedforeverysubjectsandcon- sidered100%,fromwhichprepulseinhibition(PPI)wascalculated bythefollowingformula:PPI=100−((startleafterprepulse/startle
withoutprepulse)×100).MeanPPIvaluesweregivenfordifferent prepulseintensities.
2.4. Polymerasechainreaction(PCR)
Afterdecapitationbrainswerequicklyremoved,wholeFB(ros- tral partof thebrain from Bregma0mm)and dorsal HC were hand-dissectedandflash-frozenindryice,putinasterilenuclease freeplastictubeandkepton−80◦Cuntilprocessing.Frozentis- suesampleswerehomogenizedinTRIReagentSolution(Ambion, USA)and totalRNA wasisolatedwithQIAGENRNeasyMiniKit (Qiagen,Valencia,CA,USA)accordingthemanufacturer’sinstruc- tion.ToeliminategenomicDNAcontaminationDNaseItreatment wereusedand100lRNase-freeDNaseI(1unitDNase)(Thermo Scientific, USA)solutionwasadded.Samplequality controland thequantitativeanalysiswerecarriedoutbyNanoDrop(Thermo Scientific,USA).AmplificationwasnotdetectedintheRT-minus controls.ThecDNAsynthesiswasperformedwiththeHighCapac- ity cDNA Reverse Transcription Kit (Applied Biosystems, USA).
Primers for the comparativeCt experiments were designed by PrimerExpress3.0Program.Glyceraldehyde 3-phosphatedehy- drogenase(GAPDH)wasusedashousekeepinggene.TheDNMT1, GAD1,catechol-O-methyltransferase(COMT),vesicularglutamate transporter-1(VGLUT1,SLC17A7gene),serotoninreceptorsubtype 2A(5HT2A),brain-derivedneurotrophicfactor(BDNF)andGAPDH primers(Microsynth,Switzerland)wereusedintheReal-TimePCR reactionwithFastEvaGreen® qPCR MasterMix (Biotium,USA), while DNMT3a(Rn01027162g1,LifeTechnologies) and GAPDH (Rn01775763g1,Life Technologies) by LightCycler 480Probes Master (Roche) on ABI StepOnePlus instrument. The following primerpairsweredesigned:
DNMT1forward:AGCATTCCCGTACAGAGCAG reverse:CGGGTTGAGCTTTGGGATTG GAD1forward:TGAATCGAGCCCGTTCCTG reverse:GGCTACGCCACACCAAGTAT COMTforward:CGTGTTAAAACCCGTGTCTGC reverse:AGCCAACGGCATCTCCTCAA
VGLUT1forward:CCACGACCAATGTGCGAAAG reverse:GAGTATCCGACCACCAGCAG
5HT2Aforward:AGCTGATATGCTGCTGGGTTT reverse:CACCGGTACCCATACAGGA
BDNFforward:AAACGTCCACGGACAAGGCA reverse:TTCTGGTCCTCATCCAGCAGC GAPDHforward:ACAGCCGCATCTTCTTGTGC reverse:GCCTCACCCCATTTGATGTT
2.5. Westernblot
Brainwere handledasin case ofPCRand thefrozentissues were homogenized in 200lof ice-cold extractionbuffer con- taining250mMsucrose,50mMTris,pH7.5,25mMKCl,0.5mM PMSF,0.9mMNaB,aswellasproteaseinhibitors(proteaseinhibitor cocktail, Roche,Basel, Switzerland) and phosphatase inhibitors (phosphataseinhibitorcocktail1,Sigma).Thenuclearfraction(pel- let)wasseparatedbycentrifugationat7700×gfor1min(4◦C), andre-suspendedin1ml0.4NH2SO4 and incubatedfor30min (4◦C).Sampleswerecentrifugedat 14,000×gfor 30min(4◦C).
250ltrichloroaceticacid(with4mg/mldeoxycholate)wasadded tothesupernatant,andincubatedfor30min(4◦C)toprecipitate protein.Sampleswerethenspunat14,000×gfor30min(4◦C)to pelletprotein.Pelletswerewashedfor5minwith1mlacidified acetone(0.1%HCl),thenfor5minwithacetone.Betweenwashes, proteinwascollectedbycentrifuging5minat 14,000×g(4◦C), andaspiratingsupernatant.Afterthelastwash,thepelletwasre- suspendedin200l10mMTris,pH8.0,andincubatedfor15minat roomtemperature.Proteinconcentrationsweredeterminedusing
aNanodrop.Equivalentamountsofprotein(10g)foreachsam- plewereresolvedwithSDS-PAGEusing4–15%gradientTris–HCl gels.Afterelectrophoresis,proteinsweretransferredtonitrocel- lulosemembranesfor2hat50V.Membraneswereincubatedin blockingbuffer(LI-COR,Lincoln,Nebraska,USA)1hatroomtem- peraturetoblocknon-specificbinding.Theblotswerereactedwith primaryantibodies(H3,ab1791,Abcam;AcH4K8,ab15823,Abcam, AcH4K12,07-595,Millipore,1:5000;H4,ab10158,Abcam,1:10,000 and AcH3, 06-599, Millipore, 1:20,000) overnight at 4◦C, then immunolabeledbychemiluminescenceandfollowedbydetection withChemiDocTMXRS+SystemwithImageLabTM Software.The AcH3/H3bandsweredetectedat17kDaandtheAcH4bandwas detectedat10kDa.Astherewasnoappropriatepan-AcH4anti- bodyavailable,wemeasuredtwospecificH4acetylationsite,the lysineacetylationat8(AcH4K8)or12(AcH4K12)position.Ratios ofAcH3orAcH4tototalH3orH4werecalculatedforeachsample andanalyzedacrossconditions.
2.6. Immunohistochemistry(IHC)
Rats,whounderwentPPI,weredeeplyanesthetizedbyamix- turedescribedbybiotelemetryandtrans-cardiallyperfusedwith 100mlsalinefollowedby300mlicecold4%(w/v)paraformalde- hyde(PFA;MolarChemicalsLtd.,Hungary)inphosphatebuffered saline(PBS).Aortadescendenswaspinchedbypeanforceps.The brainswereremovedfromtheskull,post-fixedfor1dayinPFA at4◦Candcryoprotectedin30%glucose(w/vinPBS)containing 0.1%(w/v)sodium-azide(Sigma–Aldrich,Inc.,Hungary).Sixseries of30mfrozensectionswerecutinthefrontalplaneonaslid- ingmicrotome.FloatingsectionswereincubatedinPBScontaining 0.5%TritonX-100and0.5%H2O2for30min.Non-specificantigens wereblockedby2%bovineserumalbumin(BSA;Sigma–Aldrich)in PBSfor30minatroomtemperature.Sectionswereincubatedfor 72hat4◦Cwithanti-histoneH3(Acetyl-Lys9)antibodymadein rabbit(1:5000,SAB4500347Sigma–Aldrich,Inc.,Hungary),diluted inblockingsolution.AfterthoroughPBSwashingsectionswere incubatedfor1hin biotinylatedanti-rabbitIgGsecondary anti- body(1:500)(VectorLaboratories).Next,sectionswereincubated inavidin–biotincomplex(1:1000)(ABCVectastainElitekit,Vec- torLaboratories)dilutedin 0.05MTrisbufferedsaline(TBS,pH 7.6)for1hatroomtemperature.H3K9acimmuno-positivecells werevisualizedbynickelenhanced3,3-diaminobenzidine(DAB).
Sections were incubated for equal time in Tris-buffered solu- tioncontaining0.2mg/mlDAB,0.1%nickel-ammonium-sulphate and0.003% H2O2.Enzymaticreactionwasstoppedbythorough TBSwashing.Sectionsweremountedonglassslidesinchrome- gelatin solution [0.5% (w/v) gelatin (Sigma–Aldrich) and 1mM Chromium(III)potassiumsulfatedodecahydrate(Sigma–Aldrich)], dehydrated by mixtures of xylol isomers and covered by DPX mounting medium (Sigma–Aldrich, Inc., Hungary). Microscopic imagesweredigitizedbyOLYMPUSCCDcamera,andstainedpar- ticleswere countedby meansof theScionImagesoftware. The methods of quantification of immune-positive cells seein Ref.
[60].
2.7. Statisticalanalysis
Data were analyzed by analysis of variance (ANOVA) using theSTATISTICA 12softwarepackage(Tulsa,OK,USA).One(fac- tor genotype) or repeated measure (factor genotype and time or new-old, or latency) ANOVA was conducted. For post-hoc analysisNewman–Keulstestwasused.Wehaveanalyzedthecor- relationbetweenimmunohistochemicalandPPIdatabyPearson correlation.Datawereexpressedasmean±SEMandthelevelof significancewassetatp<0.05.
Fig.1. Dailylocomotoractivity(counts/minutesineachhour)wasmonitoredin vasopressin-deficient(di/di)Brattlebororatsfor24hbybiotelemetricequipment.
Measurementwasstartedinthe1sthourofdarkperiod.Nosignificantdifferences wereobservedbetween+/+anddi/dianimals.Bothgenotypesfollowthenormal circadianrhythm.Greybackgrounddenotesthedarkperiod.##p<0.01vs.light period.
3. Results
3.1. Locomotoractivity
Thetelemetricdata didnot showanysignificantdifferences betweenthegenotypesatanystudiedtimepointoftheday(Fig.1).
Thelocomotoractivityfollowedthenormallifecycles,thus,both genotypesweresignificantlymore activeduringthedarkphase thaninlight(Ftime(23,115)=10.00;p<0.01).Thegenotypedidnot influencethiscircadianvariation(nogenotype×timeinteraction).
3.2. Novelobjectrecognitiontest
In the sampling phase of NOR +/+ and di/di animals spent sametimewiththeObject1(old)(Fig.2A).Duringthetestphase (Fig. 2B) repeatedmeasureANOVA showeda significantdiffer- encebetweenthetimespentsniffingthenewandtheoldobject (Fnew-old(1,14)=4.73;p<0.05)withatendencyofitsinteractionwith thegenotype(Fnew-old×genotype(1,14)=4.24;p=0.059).Post-hocanal- ysisshowedthatduringthetestphasedi/dianimalsspentsame timewitholdandnewobject,while+/+ratssniffedmorethenew one(p<0.01).Furthermore,+/+anddi/dianimalsspentalmostsim- ilartimewiththenewobject,butdi/diratsexploredsignificantly moretimetheoldonethanthe+/+(p<0.05).Thecalculateddis- criminationindexshowedthatdi/diratshadsignificantlyweaker abilitytodistinguishnewfromoldobject (Fgenotype(1,14)=11.52;
p<0.01)(Fig.2C).
3.3. Socialrecognitiontest
Similarly toNOR,in thesamplingphase of SR+/+ and di/di animals interacted the same time with the Animal 1 (old) (Fig.2D).RepeatedmeasureANOVAshowedasignificantdiffer- encebetweenthetimespentsniffingthenewandtheoldjuvenile (Fnew-old(1,11)=21.61; p<0.01),which wasdetectableonlyin+/+
animals (Fnew-old×genotype(1,11)=8.01; p<0.05). Post-hoc analysis showedthatduringthetestphasedi/dianimalsspentthesame timewitholdthanwithnewjuvenile,while+/+ratsinteractmore withthenewone(p<0.01)(Fig.2E).Discriminationindexrevealed thatdi/diratshadsignificantlyweakerabilitytodistinguishold animalfromthenew(Fgenotype(1,11)=14.17;p<0.01)(Fig.2F).
0 5 10 15 20 25 30 35 40 45
+/+ di/di
Exploration of object (time%)
Genotype
Test phase
new
old
*
##
0 5 10 15 20 25 30 35 40 45
+/+ di/di
Discrimination index
Genotype
**
A
B
C
0 5 10 15 20 25 30 35 40 45
+/+ di/di
Exploration of object (time%)
Genotype
Sampl ing phase
0 5 10 15 20 25 30 35 40 45 50
+/+ di/di
Discrimination index
Genotype
**
0 5 10 15 20 25 30 35 40 45 50
+/+ di/di
Time spentwithsniffing(%)
Genotype
Sampl ing phase D
E
F
0 5 10 15 20 25 30 35 40 45 50
+/+ di/di
Time spent with sniffing (%)
Genotype
Test pha se
##
newold
Fig.2.Novelobject(A–C)andsocialrecognitiontests(D–F).(A)Duringthesamplingphasevasopressin-deficient(di/di)andcontrol(+/+)Brattlebororatsobservedsame time(intime%ofwholetest)theunknownobject.(B)Inthetestphasedi/diratsspentsignificantlylesstimewiththenewobject,theydidnotdistinguishnovelobjectfrom thefamiliarone.Moreover,di/dianimalsspentsignificantlymoretimewitholdobjectthan+/+.(C)Discriminationindexconfirmedthatdi/diratswereunabletodistinguish newfromoldobject.(D)Insamplingphase+/+anddi/dianimalsspentsametimewithsniffingofconspecific.(E)Duringtestphase+/+ratsinteractedsignificantlymore timewithnewjuvenile,whiledi/diratscouldnotdistinguishnewandoldanimals.Thisbehaviorwasconfirmedbydiscriminationindex(F),wherethevalueofdi/dirats wascloseto0.##p<0.01time%sniffingnewvs.oldobject(B)oroldjuvenile(E);*p<0.05,**p<0.01genotypedifference.
3.4. Socialavoidancetest
The AVP-deficient di/di rats enter less frequently (Fgenotype(1,13)=5.55; p<0.05) (Fig. 3A), spent less time (Fgenotype(1,13)=5.10; p<0.05) in the large compartment (Fig. 3B) and sniffed significantly less the unfamiliar male (Fgenotype(1,13)=4.69;p<0.05)thantheir+/+counterpairs.
3.5. Prepulseinhibitiontest
Prepulseinhibitionwaslesseffectiveindi/diratscomparedwith +/+animals, especiallyat lowerprepulseintensity level(73dB:
Fgenotype(1,18)=7.10; p<0.05, 77dB: Fgenotype(1,18)=2.44; p=0.14, 81dB:Fgenotype(1,18)=4.18;p=0.06)(Fig.3C).Weinvestigatedthe effectofincreasinginter-stimulusinterval(ISI)at75dBprepulse
Fig.3. Socialavoidance(A,B)andprepulseinhibition(C,D)testsinvasopressin-deficient(di/di)Brattlebororats.(A)Thedi/diratsenteredsignificantlylesstimeintothe largecage,containingtheseparatedunfamiliarmale.(B)Thedi/diratsspentandinteractedlesstime(intime%ofwholetest)withtheunfamiliarconspecificthandi/dirats.
(C)Increasedprepulseintensityreducedthedifferencebetweengenotypes.(D)Increasinginterstimulusinterval(ISIlatency)followedby75dBprepulsetonedecreasesthe inhibitionin+/+rats,whiledidnotchangethePPIindi/dirats.*p<0.05,**p<0.01genotypedifference.##p<0.01vs.30msISIlatency.
and 120dBpulse on PPI (Fig. 3D). The response of +/+ rats at increasinglatencyinverselycorrelatedwithPPI.Thereweresignif- icantdifferencesbetweengenotypes(Fgenotype(1,17)=9.76;p<0.01) as thePPI of di/dirats did not react tochanges in ITI. It was reflectedalsobyasignificantinteractionbetweenISIandgeno- type (FISI×genotype(3,51)=4.25;p<0.01).Post-hoc analysisshowed asignificantdecreaseofPPIin+/+rats,whenwecomparedthe 30msISI with300ms and 500ms latency(p<0.01 bothcases).
Moreover,thegenotypedifferencedisappearedat300and500ms latencies.
3.6. qRT-PCR
Foldchangesofexpressionofdnmt1,dnmt3a,gad1,comt,vglut1, 5ht2aandbdnfgeneswereinvestigatedinFBandinHC,correlated to+/+animals(Fig.4AandB).DNMT1inthefrontalareashowed significantlylowerlevelindi/dirats(Fgenotype(1,10)=8.71;p<0.05), whileintheHCandtheother6genetranscriptinbothinvestigated area,showednosignificantdifferences.
3.7. Westernblot
IntheFBofdi/diratsthepan-AcH3/H4ratiowassignificantly lowerthanin+/+ones(Fgenotype(1,7)=7.32;p<0.05)(Fig.4C).Onthe otherhand,intheHCthispan-AcH3/H4ratiowasmuchhigherin di/diratsthanin+/+ones(Fgenotype(1,12)=23.87;p<0.01).AcH4K12 andAcH4K8ratiocomparedtoH3didnotshowsignificantdiffer- encesneitherintheFBnorintheHC.
3.8. Immunohistochemistry
Thefollowingbrainregionswereinvestigated(Fig.5E):nucleus accumbenscore(AcbC)and shell(AcbS)regions, prelimbiccor- tex (PrL), infralimbic cortex (IL) and dorsal peduncular cortex (DP),dorsal(LSD),intermediate(LSI)andventral(LSV)partofthe lateralseptum, aswell asCA1, CA2and CA3 fieldsof the dor- salHC.
Inthedi/dianimalsAcH3K9immunostainingshowedsignifi- cantlylesslabeledcellsinthedorsalpeduncularcortex(DP)ofthe PFCthaninthe+/+rats(Fgenotype(1,16)=5.24;p<0.05)(Fig.5A).In theotherinvestigatedPFCareas(PrLandIL)nodifferenceswere detectedbetweenthegenotypes.AcH3K9immunohistochemistry didnotshowsignificantdifferencesbetweendi/diand+/+animals neitherinthewholeAcbnorintheseparatelyinvestigatedcore andshell(Fig.5B).InthelateralseptumonlyinLSVcompartment weresignificantlylessAcH3K9immunopositivecellsindi/dithan in+/+rats(Fgenotype(1,16)=5.37;p<0.05)(Fig.5C).IntheCA1there wassignificantlymoreAcH3K9labeledcellindi/diratsthan+/+
(Fgenotype(1,15)=5.81;p<0.05),whileinCA2andCA3therewasno genotypeeffect(Fig.5D).
TherewasanegativecorrelationbetweentheISIparameterof PPItestandthenumberofAcH3K9immunopositivecellsintheCA1 (vs.75dB100msISIr=−0.63,p<0.01,vs.75dB300msISIr=−0.50, p<0.05).MarginalnegativecorrelationwasmeasuredbetweenPFC andthe75dB300msISI(r=−0.44,p=0.078).
Wehad correlatedthevaluesoftheinvestigatedbrainareas (seeTable1).Therewasasignificantpositivecorrelationbetween DPareaandAcbC(r=0.47,p<0.05),LSI(r=0.48,p<0.05)andLSV
RT-PCR
0 1 2 3
FB HC FB HC FB HC
AcH3/H4 H4K12/H3 H4K8/H3
Normalized protein ratio
Histone ratio in brain area
**
*
A
B
C
West ern blot
0 0.2 0.4 0.6 0.8 1 1.2 1.4
DNMT1 DNMT3a GAD1 COMT VGLUT1 5TH2A BDNF
Fold change
Genes
Hippoc ampus
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
DNMT1 DNMT3a GAD1 COMT VGLUT1 5HT2A BDNF
Fold change
Genes
Frontal part
+/+di/di
*
Fig.4.mRNAlevelidentifiedbyqRT-PCRinthefrontal(A)andhippocampal(B)partofthebrainofvasopressin-deficient(di/di)Brattlebororatsandproteinratioofacetylated histonesmeasuredbywesternblot(C).(A)AtfrontalpartDMNT1leveldecreasedsignificantlyindi/dianimals,whileDNMT3a,GAD1,COMT,VGLUT1,5HT2AandBDNF mRNAlevelwereidenticalinthetwogenotypes.(B)IntheHCnogenotypeeffectwasobserved.(C)H3comparedtototalH4significantlydecreasedinthefrontalbrainarea (FB)andincreasedinthehippocampus(HC)ofdi/dirats.HistonemodificationonH4(acetylationonlysine8or12)comparedwithwholeH3proteinleveldidnotshow significantdifferences.*p<0.05,**p<0.01genotypedifference.
(r=0.55,p<0.05).Thecoreregionoftheaccumbenssignificantly correlatedwithCA3(r=0.61,p<0.05),whiletheshellregionpos- itivelycorrelatedwithLSD(r=0.50,p<0.05),LSI(r=0.70,p<0.01) andLSV(r=0.50,p<0.05).
4. Discussion
Combiningbehavioralstudieswithmolecularbiologicaltech- niques we demonstrated that in Brattleboro rat the lifelong
Fig.5.Acetylatedlysine9ofhistone3(AcH3K9)immunohistochemistryinvasopressin-deficient(di/di)Brattlebororatbrain.(A)Amongtheprefrontalcortex(PFC)regions inthedorsalpeduncularpart(DP)di/diratsrevealedlowerlevels,whileprelimbic(PrL)andinfralimbic(IL)regionshowedequalnumberofAcH3K9immunopositivecell numberinbothgenotypes.(B)Nucleusaccumbens(Acb)shell(AcbS)andcore(AcbC)didnotshowsignificantdifferencebetweenthegenotypes.(C)Inthelateralseptumonly theventral(LSV)butnotthedorsal(LSD)orintermediate(LSI)partsshowedalterations.(D)IntheCA1regionoftheHCtheAcH3K9levelwasincreased.(E)3Dreconstruction andpositionoftheinvestigatednucleiintheratbrain.Forbetteroverviewonlyuni-lateralnucleiwasshown,fromcaudo-lateralview.Redcolorindicatesthesignificant increase,whilebluethesignificantdecreaseofthenumberofAcH3K9immunopositivecellsintheappropriatenuclei.*p<0.05genotypedifference.(Forinterpretationof thereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)
frameshiftmutationinthegenecodingfortheneuropeptideAVP resultedinepigenetic alterationinthefrontalpartofthebrain (DP,LSV)andinthehippocampus(CA1),whichcanunderlinethe appearanceofSCZ-likebehaviorinBrattlebororats.
Earlierobservations(forreferencesseeintroduction)andour recentstudyconfirmedlotsofsimilaritiesbetweenSCZpatients andtheSCZ-likebehaviorofdi/dirats.However,previouswork concentratedonasingletest(mostlyonPPI),butSCZisaspectrum disorder,thus,wehavetotakeintoconsiderationawiderangeof behaviors[61].Inthissense,ourworkisthefirstsummarypre- sentinginasinglecolonyawiderangeofSCZ-likebehaviorinthe
Brattlebororatfromalldomains(positive,negativeandcognitive).
Cognitiveimpairmenthasdrawnattentionasparticularlyimpor- tantbeingoneofthemoretreatmentresistantcomponents.Inthe presentstudycognitivesymptomswerestudiedbyNORandSR, showingthatourcolonyhasSCZ-likedeficienciessimilarlytopre- viousobservations[42,58,62–64].Anothernegativesymptom,the lackofinteresttoconspecificwasconfirmedbySAtest,whichwas neverusedinthisstrainbefore,althoughsocialimpairmentwas alreadydescribedinBrattlebororat[58,65,66].Inourcolonythe attentiondeficitwasalsoconfirmedbyPPItestasthemainbehav-
Table1
MatrixofthecorrelationamongtheAcH3K9immunohistochemicallystainedbrainareas.
FB PFC PrL IL DP Acb AcbC AcbS LS LSD LSI LSV HC CA1 CA2 CA3 FB
PFC PrL
IL +
DP + +
Acb
AcbC +
AcbS +
LS + +
LSD +
LSI + + + +
LSV + + + +
HC CA1 CA2
CA3 + + +
Lightgreycellsindicatethenon-significant,whiledarkgraycellsthesignificantcorrelations.“+”indicatesthatthecorrelationispositive.Blackfilledcellssignthemeaningless comparison.FBmeansfrontalpartofthebrain(PFC+Acb+LS).PFCmeansthesummarizeddataofPrL+IL+DP.AcbisthesummaryofAcbCandAcbS,LSisLSD+LSI+LSV, whileHCisthesummaryofCA1+CA2+CA3.
Abbreviations:Acb—nucleusaccumbens,AcbC—Acbcore,AcbS—Acbshell;PFC—prefrontalcortex,PrL—prelimbiccortex,IL—infralimbiccortex,DP—dorsalpeduncularcortex;
LS—lateralseptum,LSD—LSdorsalpart,LSI—LSintermediatepart,LSV—LSventralpart;HC—hippocampus,CA1-3—cornuammonisofhippocampus.
ioralmodelofSCZusednotonlyinpreclinical,butalsoinclinical research[67,68].
The negative correlation between PPI and PFC AcH3K9 immunoreactivityinourcolonywasingoodagreementwiththe roleofepigenome,especiallyhistoneacetylation,incognitivepro- cesses[69,70].Althoughinhumansthemethylation“maintaining”
enzyme,DNMT1[14,71],aswellasthedenovomethyltransferase DNMT3[72]wereincreasedinthePFCofSCZpatients,inBrattle- bororatsanoppositechange,namelyadecreasewasdetectable inDNMT1ofthefrontalregion.Thereforewe werefocusing on acetylation,whilemethylation processesseemedtohavesubtle contributiontothedevelopmentof SCZ-likesymptomsinBrat- tlebororats.Moreover,baseduponourWesternblotanalysis,H3 ratherthanH4acetylationseemedtobeimportant.Anelevated HDAC1levelinthePFC,observedin postmortembrainsamples ofSCZpatient[73],mightcontributetoreducedH3 acetylation intheFBofdi/dianimals,however,itwasnotexaminedinour study.Wehavechosentostudytheacetylationofasinglelysine residue(AcH3K9),becauseAoyomaetal.[19]usingapharmaco- logicalmodel ofSCZ (piperidinehydrochloride(PCP)inmouse) founddecreasedlevelofAcH3K9inthePFC.Moreover,previous observationfoundadecreaseofAcH3K9inhumanSCZpatients [74].Ourresultsindi/diratswereingoodagreementbothwiththe pharmacologicalmodelandhumandata.
However,despitetheGABAergicoriginhypothesis[75]wewere unabletofindalterationintheGABAsynthetizingenzyme,GAD1 mRNAlevel.AlthoughdeficitinGABAergicsystemisoneofthe mostcommonlydescribedabnormalityinSCZpatients[76],and GAD1thoughttobedownregulatedbecauseoftheincreasedlevel ofDNMT1[71],but inBrattlebororatsanenhancedratherthan reducedGABAlevelwasreported[77]ingoodagreementwiththe foundlowerDNMT1levels.Thereforewemightassumethatthe detectedloweracetylationintheFBcoulddecrease othergene expressionleadingtoreducedadaptivecapacitytothesuddenly changingenvironment.Assummarizedin[14]alterationofDNA methylationmaybeinvolvedinepigeneticalterationsoftheglu- tamateandserotoninneurotransmittersystemaswell.Wehave
investigatedthemRNAlevelofkeycomponentofthesesystems theVGLUT1and5HT2Areceptor,respectively,butnosignificant changeswereobserved.Furthermore,weexaminedmRNAlevel ofBDNFgenethatknowntoplayanimportantroleincognition.
Previousstudiesshowedthehypermethylationofitsgenesinthe etiologyandpathogenesisofSCZ[9],butwedidnotfindsignificant alterationindi/dianimals,either.
BothinhumanpatientsandanimalmodelsofSCZdysfunction ofthedopaminergicsystemwasobserved[78]andthecurrently usedantipsychoticsantagonizedopaminereceptors[79].In the PCP model of SCZ the low AcH3K9 caused dysfunction of the dopaminergic systems and the atypical antipsychotic clozapine ameliorated the AcH3K9 level mainly on D1 receptor positive cellsinthePFC[19].Itisalsoknown,thatD1signalingregulates histonemodification[80]andwealsohavefoundsignificantpos- itivecorrelation(seeTable1)inthenumberofAcH3K9positive cellsbetweenelementsofthedopaminergicsystem(DP,Acb,LS).
AlthoughwedidnotmeasuredsignificantalterationinthemRNA levelofdopaminedegradingenzymeCOMTneitherintheFBnor intheHC,wecanhypothesizethatinBrattlebororatsthelackof AVPinducesepigeneticmodificationleadingtodisturbancesofthe dopaminergicsystem,whichcontributestotheSCZ-likebehavioral alterations.Indeed,higherdopaminecontent[81,82]andupregu- lateddopaminereceptors[45]werereportedindi/dianimals.
Inthefollowingwehaveexaminedindetailsthosebrainareas, whicharethoughttobeinvolvedinthedevelopmentofSCZ.We havefoundsignificantdecreaseinthenumberof AcH3K9posi- tivecellsintheDPofdi/dirats.TheroleoftheDPislessknown and investigated than the neighboring IL and PrL despite their structural similarities [83]. Ventral part of the PFC specifically responsibleforaflexibleshiftingtonewstrategiesrelatedtospatial cues,furthermore—onthebasisofitsconnectionswithautonomic centers—fortheintegrationofinternalphysiologicalstates with salientenvironmentalcuesfortheguidanceofbehavior[84].DP sharesprojectionsrelevanttorewardcircuitrysuchasglutamater- gicafferentstotheventraltegmentalarea[85].Furthermore,theIL andDPbothprojectheavilytotheAcbS[86].
ThereislotsofevidencethattheAcb(becauseitsdopaminecon- tent)isinvolvedinthepathologyofSCZ[87,88].Althoughalteration ofdopaminereceptorsintheAcbwasfoundindi/dirats[45],but wedidnotfindsignificantgenotypedifferenceinthenumberof AcH3K9labeledcells.ItseemsthatinBrattlebororat,AcH3K9alter- ationisnotinvolvedinthedevelopmentofsymptoms,butchanges atotherlysineresiduesmighthavesomeimportance.
LSVplaysanimportantroleinvariousbehavioralprocesses[89], integratessensorystimuliconveyingthisinformationtorespon- sible brain areas to direct motivating behaviors. LSV, showing significantlylower AcH3K9reactivity in di/dithan controlrats, wasactivated(measuredbyc-Fosimmunochemistry)duringPPI [90].MedialPFCsendsglutamatergicprojectionstoLS[89],thus, thelower cell activityof DP induced by decreased acetylation mightleadtoreducedacetylation/activationofLSVneurons.LSis stronglyinterconnectedwiththeHCpredominantlybyinhibitory GABAergicneurons[91],whichconnectionhasanestablishedrole inlearningandmemory[92].Moreover,thesepto-hippocampal (CA1)GABAergicneuronswereactivatedduringlocomotionand salientsensoryeventinbehavingmice[93].
Thus,itisnotsurprising,thatreducedinhibitoryGABAergictone intheHCleadtoenhancedpan-AcH3andespeciallyAcH3K9level intheCA1ofdi/diratscomparedtocontrolanimals.Accordingto theliteratureduringHCdependentmemoryformation(e.g.,con- textualfearconditioningorNOR)AcH3levelincreasedintheCA1, whileAcH4wasunaltered[69].ThissuggeststhatH3acetylation occursduringchronicevents[9],likethechronicstressofdiabetes insipidusindi/dirats.Thesedata,alongwithpostmortemanaly- ses[55,73],suggestedthathistonemodificationsintheHCmight alsocontributetothebehavioralalterationinSCZthroughaberrant regulationofoneormoregenes.Indeed,epigeneticmechanisms wereinitiallydescribed for theirabilitytopromote differentia- tionincludingneurogenesisandHCisoneofthetworegionsin whichgenerationofnewfunctionalneuronsfromneuralstemcells occursthroughouttheadultlife.Adulthippocampalneurogenesis contributestolearningandmemory,corefeaturesofSCZ[94,95].
Positivesymptomssuchasauditoryhallucinationshavebeen correlated with abnormal activation patterns in PFC [96], fur- thermoredelusionalstatesandnegativesymptoms(e.g.,avolition and apathy) also appear to be involved in prefrontal dysfunc- tion[97]. Thevolume ofthe graymatter [98]and thecerebral bloodflow[99]isreducedinthePFCofSCZpatients, knownas hypofrontality.OurSCZmodel,thedi/diBrattlebororathasalso smallerbrainvolume,compared to+/+animals[100].Although hyperlocomotion is thought to reflect positive symptom, but, despitesomepreviousobservations(inadults[101]andinpups [102,103]),wewereunabletodetectchangesinlocomotionnot onlyindifferenttestsituations[104,105],butalsoduringalong- termobservation by biotelemetry for 24h. Nevertheless, when weusedahyperlocomotion-inducingNMDAantagonist(ketamine ip10mg/kg/2mlsalinerightbeforethe15minopenfieldtest;a modelofpositivesymptoms[106])thedi/diratsdidnotshowan enhancementoflocomotion(numberoflinecrossingincontrol:
530.7±74.6;indi/di:335.7±63.2),whichcanbesupposedinan alreadySCZ-likesubject[61].
6. Conclusion
WecanproposethattheAVPdeficientBrattlebororatstrainis avaluablemodel(i)tostudythedevelopmentofSCZ-likebehav- ior,toestablishnewtestbatteryforSCZ,(ii)followthechangesin epigeneticstateofaffectedgenes,and(iii)testnewlydiscovered antipsychotics.ThisishotspotasinfluencingHDACsmayprovide therapeuticalternativesfortreatingmanyofthesymptomsasso- ciatedwithSCZ,particularlycognitivedeficits[107].Ourstudies
confirmedthatnosinglegene,moreprobablyfinechangesinan arrayofmoleculesareresponsibleforthemajorityofSCZcases.
Acknowledgements
ThisstudywassupportedbyHungarianResearchFoundOTKA NN71629toDZ;K109622toKJK;109744toSzFandbyBolyaigrant toDZandVSz.
References
[1]D.Bhugra,Severementalillnessacrosscultures,ActaPsychiatr.Scand.
Suppl.429(2006)17–23.
[2]T.Nagai,D.Ibi,K.Yamada,Animalmodelforschizophreniathatreflects gene-environmentinteractions,Biol.Pharm.Bull.34(2011)1364–1368.
[3]American,Psychiatric,Association,DiagnosticandStatisticalManualof MentalDisorders(DSM-5®)(2013).
[4]A.Carlsson,M.Lindqvist,Effectofchlorpromazineorhaloperidolon formationof3methoxytyramineandnormetanephrineinmousebrain,Acta Pharmacol.Toxicol.20(1963)140–144.
[5]J.M.vanRossum,Thesignificanceofdopamine-receptorblockadeforthe mechanismofactionofneurolepticdrugs,Arch.Int.Pharmacodyn.Ther.
160(1966)492–494.
[6]SchizophreniaWorkingGroupofthePsychiatricGenomicsC,Biological insightsfrom108schizophrenia-associatedgeneticloci,Nature511(2014) 421–427.
[7]D.M.Svrakic,C.F.Zorumski,N.M.Svrakic,I.Zwir,C.R.Cloninger,Risk architectureofschizophrenia:theroleofepigenetics,Curr.Opin.Psychiatry 26(2013)188–195.
[8]D.P.Gavin,R.P.Sharma,Histonemodifications,DNAmethylation,and schizophrenia,Neurosci.Biobehav.Rev.34(2010)882–888.
[9]N.Tsankova,W.Renthal,A.Kumar,E.J.Nestler,Epigeneticregulationin psychiatricdisorders,Nat.Rev.Neurosci.8(2007)355–367.
[10]A.Petronis,A.D.Paterson,J.L.Kennedy,Schizophrenia:anepigenetic puzzle?Schizophr.Bull.25(1999)639–655.
[11]I.I.Gottesman,J.Shields,D.R.Hanson,Schizophrenia:TheEpigeneticPuzzle, CambridgeUniversityPress,Cambridge,England,1982.
[12]C.L.Smith,A.Bolton,G.Nguyen,Genomicandepigenomicinstability,fragile sites,schizophreniaandautism,Curr.Genom.11(2010)447–469.
[13]G.L.Engel,Thebiopsychosocialmodelandtheeducationofhealth professionals,Ann.N.Y.Acad.Sci.310(1978)169–187.
[14]D.Zelena,Co-regulationandepigeneticdysregulationinschizophreniaand bipolardisorder,in:J.Minarovits,H.H.Niller(Eds.),Patho-Epigeneticsof Disease,SpringerScience+BusinessMedia,2012,pp.281–347.
[15]S.Akbarian,Epigeneticsofschizophrenia,Curr.Top.Behav.Neurosci.4 (2010)611–628.
[16]A.Must,Z.Janka,S.Horvath,Schizophrenia,environmentandepigenetics, Neuropsychopharmacol.Hung.13(2011)211–217.
[17]E.Li,Chromatinmodificationandepigeneticreprogramminginmammalian development,Nat.Rev.Genet.3(2002)662–673.
[18]G.G.Brune,H.E.Himwich,Effectsofmethionineloadingonthebehaviorof schizophrenicpatients,J.Nerv.Ment.Dis.134(1962)447–450.
[19]Y.Aoyama,A.Mouri,K.Toriumi,T.Koseki,S.Narusawa,N.Ikawa,etal., Clozapineamelioratesepigeneticandbehavioralabnormalitiesinducedby phencyclidinethroughactivationofdopamineD1receptor,Int.J.
Neuropsychopharmacol.17(2014)723–737.
[20]M.G.Melka,C.A.Castellani,B.I.Laufer,R.N.Rajakumar,R.O’Reilly,S.M.
Singh,OlanzapineinducedDNAmethylationchangessupportthedopamine hypothesisofpsychosis,J.Mol.Psychiatry1(2013)19.
[21]E.Dong,D.R.Grayson,A.Guidotti,E.Costa,Antipsychoticsubtypescanbe characterizedbydifferencesintheirabilitytomodifyGABAergicpromoter methylation,Epigenomics1(2009)201–211.
[22]S.O.Frederiksen,R.Ekman,C.G.Gottfries,E.Widerlov,S.Jonsson,Reduced concentrationsofgalanin,argininevasopressin,neuropeptideYandpeptide YYinthetemporalcortexbutnotinthehypothalamusofbrainsfrom schizophrenics,ActaPsychiatr.Scand.83(1991)273–277.
[23]A.Jobst,S.Dehning,S.Ruf,T.Notz,A.Buchheim,K.Henning-Fast,etal., Oxytocinandvasopressinlevelsaredecreasedintheplasmaofmale schizophreniapatients,ActaNeuropsychiatr.6(2014)347–355.
[24]J.K.Mai,K.Berger,M.V.Sofroniew,Morphometricevaluationof neurophysin-immunoreactivityinthehumanbrain:pronounced inter-individualvariabilityandevidenceforalteredstainingpatternsin schizophrenia,J.furHirnforschung34(1993)133–154.
[25]I.Elman,S.Lukas,S.E.Shoaf,D.Rott,C.Adler,A.Breier,Effectsofacute metabolicstressontheperipheralvasopressinergicsystemin schizophrenia,J.Psychopharmacol.17(2003)317–323.
[26]L.H.Rubin,C.S.Carter,J.R.Bishop,H.Pournajafi-Nazarloo,L.L.Drogos,S.K.
Hill,etal.,Reducedlevelsofvasopressinandreducedbehavioralmodulation ofoxytocininpsychoticdisorders,Schizophr.Bull.40(2014)1374–1384.
[27]O.Teltsh,K.Kanyas-Sarner,A.Rigbi,L.Greenbaum,B.Lerer,Y.Kohn, Oxytocinandvasopressingenesaresignificantlyassociatedwith