Edinburgh Research Explorer

Everyday memory: towards a translationally effective method ofmodeling the encoding, forgetting and enhancement of memory

Citation for published version:Nonaka, M, Fitzpatrick, R, Lapira, J, Wheeler, D, Spooner, PA, Corcoles-Parada, M, Munoz-Lopez, M, Tully,T, Peters, M & Morris, RGM 2017, 'Everyday memory: towards a translationally effective method ofmodeling the encoding, forgetting and enhancement of memory', European Journal of Neuroscience.https://doi.org/10.1111/ejn.13637

Digital Object Identifier (DOI):10.1111/ejn.13637

Link:Link to publication record in Edinburgh Research Explorer

Document Version:Peer reviewed version

Published In:European Journal of Neuroscience

General rightsCopyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s)and / or other copyright owners and it is a condition of accessing these publications that users recognise andabide by the legal requirements associated with these rights.

Take down policyThe University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorercontent complies with UK legislation. If you believe that the public display of this file breaches copyright pleasecontact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately andinvestigate your claim.

Download date: 04. Apr. 2020

1

Dr 47 – RGMM, MN - FINAL

ResearchReport/TechnicalSpotlight/NewMethod

Everydaymemory:towardsatranslationallyeffectivemethodof

modelingtheencoding,forgettingandenhancementofmemory.

MioNonaka1,RichardFitzpatrick1,JenniferLapira3,DamianWheeler3,PatrickA.

Spooner1,MartaCorcoles-Parada1,MónicaMuñoz-López2,TimTully3,

MarcoPeters3andRichardG.M.Morris1,4

1CentreforCognitiveandNeuralSystems,

EdinburghNeuroscience,TheUniversityofEdinburgh,1GeorgeSquare,Edinburgh,EH89JZ,Scotland(UK)

2HumanAnatomyLaboratory,

FacultyofMedicineandRegionalCentreforBiomedicalResearch,UniversityofCastilla-La-Mancha,

14AvenidaAlmansa,02071Albacete,Spain

3DartNeuroScienceLLC12278ScrippsSummitDrive,SanDiego,CA92131,USA

4InstituteforNeuroscience,

CSIC-ULM,Alicante,03550Spain

CorrespondingAuthor:r.g.m.morris@ed.ac.uk

KeyWords:Learning,event-arena,spacedtraining,

RNAseq,gene-expression

2

ABSTRACT

Thetestingofcognitiveenhancerscouldbenefitfromthedevelopmentof

novelbehaviouraltasksthatdisplaybettertranslationalrelevancefordaily

memory,andpermittheexaminationofpotentialtargetsinawithin-subjects

mannerwithlessvariability.Wehereoutlineanoptimizedspatial‘everyday

memory’task.Wecalibrateitsystematicallybyinterrogatingcertainwell-

establisheddeterminantsofmemory,andconsideritspotentialforrevealing

novelfeaturesofencoding-relatedgeneactivation.Ratsweretrainedinanevent

arenainwhichfoodwashiddeninsandwellsinadifferentlocationeveryday.

Theyfoundthefoodduringaninitialmemory-encodingtrialandwerethen

requiredtorememberthelocationin6-alternativechoice-orprobetrialsat

varioustime-pointslater.Trainingcontinueddailyoveraperiodof4months,

realisingastablehighlevelofperformanceandcharacterisedbydelay-

dependentforgettingover24h.Spacedbutnotmassedaccesstomultiple

rewardsenhancedthepersistenceofmemory,asdidpost-encoding

administrationofthePDE4inhibitorRolipram.QuantitativePCRandthen

genomewide-analysisofgene-expressionledtoanewobservation-stronger

gene-activationinhippocampusandretrosplenialcortexfollowingspacedthan

massedtraining.Inasubsidiarystudy,aseparategroupofanimalsreplicated

aspectsofthistrainingprofile,goingontoshowenhancedmemorywhentraining

wassubjecttopost-encodingenvironmentalnovelty.Distinctivefeaturesofthis

protocolincludeitspotentialvalidityasamodelofmemoryencodingused

routinelybyhumansubjectseveryday,andthepossibilityofmultiplewithin-

subjectcomparisonstospeedupassaysofnovelcompounds.

3

INTRODUCTION

Transientmemoriesareformedaswegoaboutourdailylifeathomeorwork.

Theygenerallyoccuragainstthebackgroundofrelevantpriorknowledge.Manyof

these‘everyday’memories,beingoflittleconsequence,areforgottenrapidlyas

Ebbinghausfirstreportedwithexperimentsonmemorytrainingfornonsensesyllables

(Ebbinghaus1885).However,asubsetofeverydaymemoriesmayberetained

overnightorforlonger(Wixted2004).Whatisresponsiblefortheselectivityofdaily

memory?Mightbetterunderstandingoftherelevantneuralmechanismsbehelpfulfor

thedevelopmentofeffectivecognitiveenhancers?Thismethodsreport/technical

spotlightarticlehighlightsanoptimizedprotocolformonitoringeverydaymemoryin

animals.Itsdevelopmentwas,inpart,inspiredbythe“RivermeadEverydayMemory

Task”usedinclinicalpracticeandresearch(Wilson,Cockburnetal.1991).

Memoryofsuccessiveitemsovertimereflectsvariationsintheeffectivenessof

encodingand/orretentionofmemory.Synergisticorcompetitiveeffectshavelongbeen

studied,suchastheclassicalphenomenaofretro-orproactiveinterference(Mullerand

Pilzecker1900,McGeoch1932,Underwood1969).Neurobiologicalstudieshave

identifiedchangesinmemoryretentioncausedbyongoingeventsatthetimeof

encoding(suchaschangesinattentionormetaplasticity:(AbrahamandBear1996,

Kentros,Hargreavesetal.1998),byeventsshortlybeforeorafterencodingthattrigger

plasticityprotein-synthesis(synaptictaggingandcapture:(FreyandMorris1997,

RedondoandMorris2011),andbyneocorticalmechanismsengagedforlongtime

periodsafterencoding(suchassystemsconsolidation:(Squire1992,McGaugh2000,

DudaiandMorris2001).Alterationsintheeffectivenessofmemoryretrievalcanalso

occurinassociationwithdifferentialaccesstoappropriateretrievalcues(Tulving1983,

SpearandRiccio1994).Tounderstandtheneuraldynamicsofeverydaymemory

processinginmoredetail,weneedbetter,theoreticallyguidedanimalmodels.Wenow

reportanewandpotentiallyusefulmodel.

Twokeyprincipleshaveguidedourapproach.Thefirstisthat,justlikehumans

facingdailymemorychallenges,theanimalsshouldnotbe‘experimentallynaïve’.A

strikinglyhighproportionofbehaviouralneurosciencestudiesusenaïveanimals,many

forgoodreasonbutmanyoutofhabit.Memoryinexperimentallynaïveanimalsis

frequentlytestedintaskssuchnovelobjectrecognition,conditionedtasteaversion,or

variousparadigmsoffearconditioning.Inthisstudy,wedeliberatelyallowmemory

4

processingtooccuragainstabackgroundofexperimentallycontrolledpriorexperience

thatpermitsrepeatednovelencodingandretrievalacrossdays.Thisbackgroundlikely

affectstheefficiencyofencodingorconsolidation(orboth).Asecondprinciplewasto

developaprotocolinwhichwhatisrememberedfromonemomenttothenextis

usuallybutnotalwaysforgotten.Thisisnotonlymore‘realistic’ofeverydaylife,where

weforgetagreatdeal,butalsomotivatedbyaspecifictheoreticalmodelofmemory

formationbasedonthetime-courseofactivity-dependentsynapticplasticity(Morris

2006).Whilevariouswidelyused‘everyday’taskscouldbeincludedinsuchaprotocol,

includingrecognitionmemory,ourstartingpointwasspatialmemory–analogousto

the“wheredidIleavemyglasses?”kindofmemory(asexaminedintheRivermead

Test).Trainingtookplaceeveryday,overweeksandmonthsinarepetitivemanner,

withonlyoccasionaltestsofthememorabilityofaspecificitem.

The‘delayedmatching-to-place’protocolinthewatermaze(SteeleandMorris

1999,daSilva,Bastetal.2014)isanexampleofaneverydaymemorytaskasdefined,

butanappetitivelymotivatedalternativeismorerealisticthanataskinvolvingescape

fromwater.Weusedan‘eventarena’inwhichratsaretrainedtodigforfoodin

sandwellswhoselocationvariedfromdaytoday.Eachday,formemoryencoding,they

explorethearenatofindoneortwosamplesandwells(ofwhichonlyonesandwellwas

rewarded),bothlocatedinavirtualarrayofupto43locationsacrosssessions.For

memoryretrieval,atanintervalof60-90minlater,arewardedchoicetrialwastypically

givenwith6sandwellsofwhichonlythecorrectonefromthesampletrialwas

rewarded.Win-staytrainingcontinuedlikethisovermanydaysandweeksasthe

animalsfirstlearnedthetaskandthenperformeditwell–learninganeweachdayand

rapidlyforgettingwherethefoodhadbeenovertheprecedingdays.Apartfromdaily

choice-trials,memorywasalsotestedoccasionallyusingananalogoustechniquetothe

watermaze-non-rewardedprobetestsatvariousdelaysupto24h.Inthese,thetime

spentdiggingateachof6availablesandwells,noneofwhichcontainedaccessiblefood

duringthetest,wascarefullymeasured.

Suchaprotocolinvolvesnewspatialmemoryencodingeveryday–wherethe

‘event’islearningwheretodigeachdayandthenrapidlyforgettingwherethefoodhad

beenovertheprecedingdays(Bast,daSilvaetal.2005).Withmodestrewardonthe

sampletrial,delay-dependentforgettingresultsinthememoryofthedailylocationof

foodbeingatchancelevelswithin24h.Thepresentprotocolhastheimportant

5

refinementthatthesampleprotocolhastwosandwellsratherthanjustone.Pilotwork

revealedthatthisreducedvariabilitybecausetheanimalslearnnottokeepdiggingat

oneSandwell,buttoswitchbetweensandwellsasanoptimumstrategytosecure

rewardduringbothmemoryencodingandretrieval.Aspectsoftheencodingprotocol

werethensystematicallyvariedinwaysthatservedtocalibratethenewtask(singlevs.

multiplerewards,thetemporalspacingofmultiplerewards)toidentifywhetherthese

wouldaffectmemoryretention.Toexplorethefuturepotentialofthenewtechnique,

RT-qPCRandRNA-Seqanalyseswereconductedtoexaminetraininginducedgene-

induction(ShengandGreenberg1990,LanahanandWorley1998).Assessmentofthe

effectofaphosphodiesterase4inhibitor,theputativecognitiveenhancerRolipram

(Rutten,Prickaertsetal.2006)furtherservedtovalidatethebehaviouralprotocol.A

separatecohortofanimalswastrainedinasubsidiarystudytoestablishareplication

andtoexploretheimpactuponretentionofpost-encodingnoveltythatisknownto

driveimmediate-earlyandplasticity-proteinsynthesis(Guzowski,McNaughtonetal.

1999,Guzowski,McNaughtonetal.2001,Wang,Redondoetal.2010,Takeuchi,

Duszkiewiczetal.2016).

6

METHODS

Subjects

Themainstudyused24adultmaleListerHoodedratsweighing200g-250gon

arrival.Theywereapproximately2monthsoldandtrainingtookplaceoveraperiodof

4months;allbutoneratcompletedthetesting(dataisreportedonn=23).Theywere

grouphoused(3-4/cage;12-hlight-darkcycle;alltraininginthelightphase).After

habituationtotheanimalfacilitywithfreefeedingandregularhandling,theywerefood

restrictedtomaintainbodyweightat90%ofthefreefeedingweightthroughoutthe

trainingandtestingperiod.Ifananimal’sweightdroppedbelow85%,itwas

individuallyfedtogain>85%bodyweight.Thesubsidiarystudyused11adultmale

ListerHoodedrats(250gonarrival)maintainedidentically.Allanimalexperiments

wereperformedinaccordancewiththeregulationsandguidelinesoftheUKHome

Officeregulations(PPLnumber,60/4566).

Apparatus–theeventarena

Allexperimentswereconductedusingan‘eventarena’(1.6mx1.6m)whose

floorsurfaceconsistedof49squarePlexiglaspanels(20x20cm).Someofthesepanels

hadholesof6cmdiameterintowhichsandwellscouldbeplaced(Fig.1A,Fig.S1A).The

floorpanelswerecleanedregularlyandreplacedindifferentlocationstorandomizeany

residualolfactorybias.Therewereprominent2Dand3Dextra-arenacues(Fig.S1B),

andtwointramazelandmarks(square-towerinrow4,column2;golf-balltowerinrow

4,column6).Thearenahadtransparentside-walls(30cmhigh),andfourstart-boxesat

thecenteroftheNorth(N),S,EandWwallsmadeofblackPlexiglas(transparentlid,

carpetedwithsawdust,availablewatercup,andautomateddoors,andthusdarker

insidethanthemainarena).Aceilingmountedvideocamerawasusedtomonitor

animalmovementsandchoices,andthisconnectedtoaDVDdriveandaLabView

event-recordingsystem.

InsertFigure1abouthere

Plexiglassandwells(6cmdiameter,4cmdepth),thatcontainedhiddenfood

pellets,wereplacedinthefloorpanelswithholes(Fig.1B,Fig.S1C).Tomaskthesmell

7

ofthereward,thesandwellswerefilledwithbirdsandmixedwithGaramMasala(P&B

(Foods)Ltd.,150g/5kgsandinitially,with3g/5kgaddedeverysession(toreplenish

themaskingeffect,andthesandlostduringcleaning).Eachsandwellhadaspherical

plasticbowlatthebottominwhichrewardpellets(5x0.5g)wereplacedtobe

accessible(‘rewardedsandwell’).Thisplasticbowlalsomadeitpossibleforanequal

numberofrewardpelletstobeplacedunderneath,andtherebyinaccessible(‘non-

rewardedsandwell’).Theplasticbowlshadholesandsowereporoustoodors,

ensuringthattherewarededandnon-rewarededsandwells,notonlycontainedthe

samenumberofrewardpelletsatapproximatelythesamedepthinthesandbutwould

alsosmellthesame(Fig.1B,Fig.S1C).Thesandwellwasdesignedwithanextrapanelat

thebottomtoenableeasychangingofthebottompellets,withfreshpelletsusedin

everysession.Extensiverandomizingandcounterbalancingwasconductedtominimize

anyolfactoryartifacts;thisincludedensuringthatthesandwellsusedinthesampletrial

werenotusedforthechoicetrialofthesamesession;wipingthearenafloorwithan

alcoholimpregnatedtowelbetweeneachanimaltrial,andbeforechoiceandprobe

trials.

Behaviourtraining,sampletrialsandchoicetrials

Therewasaseriesofdistinctphasesoftraining:(1)habituation,(2)trainingto

dig,(3)maintrainingsessionswithsampleandchoicetrials,andinterspersedtests

involving(4)within-subjectmanipulationofdiverseparametersthatmightaffectthe

strengthorpersistenceofmemory(e.g.spacingofrewardavailabilityinasampletrial)

withassociatedmemoryprobetests.

Habituation(session-8):Theanimals,at90%offree-feedingbodyweight,were

broughttothearenaandplacedinthestart-boxfor30s,andthenallowedtoexplorein

thearenafor10min,beforebeingplacedbackinthestart-boxfor30s,andthengivena

rewardpelletinthehomecage(Fig.1C).Trainingtodig(sessions-7to-1):Theanimals

weregivenonepelletinthestart-box,thenallowedtoexploretolookforthereward

pellets(BioServ,SupremeMini-Treat0.5gm)atthesinglesandwellplacedinarandom

locationofthearena,andtaughttobringthepellet,oneatatime,backtothestart-box

ofthedaytoconsume3-5pellets(Fig.1C).Pelletswereinitiallyvisibleonthesand,and

thenburiedlowerdaybyday,encouragingtheanimalstodigdeeptofindthepelletsat

thebottomofthebowl.Bytheendofhabituationsessions,alltheratshadlearnedtodig

8

withconjointexperienceofthesandwellsinthehomecage.Theyalsostartedtolearn

thewin-staystrategyofreturningtoarewardedsandwellthatwasexplicitlytrained

duringthesubsequentsessions.

Trainingsessionswithsampleandchoicetrials(sessions1-81):Trainingthen

commenced,andrepeatedacrossdays,5-7daysperweek,forseveralmonths.A

trainingsessionconsistedofasampletrial(ST=memoryencodingtrial)and,60-90min

later,achoicetrial(CT=memoryretrievalchoicetrial)(Fig.1D).OnST,theratswere

presentedinthearenawithonerewardedsandwell(sessions1-18)ortwosandwellsof

whichonlyonecontainedaccessiblereward(session18onward).Theratswerefirst

givenafoodpellet(0.5g)ofaspecificflavorinthestart-box.Thedoorwasthenopened

permittingaccesstothearenawheretheratcouldfindanddiguptherewardpelletthat

itthenimmediately,orwithgentleencouragementforthefirstfewtrials,tookbackto

thestart-boxtoeat(Whishaw,Kolbetal.1983).Thereafter,between1and5pellets

wereavailableduringanST,asdescribedinlaterprotocols.OnCTs,theratswere

presentedwith6sandwellsinthearena,withthecorrectsandwellbeinginthesame

locationastherewardedsandwelloftheearlierST(thus,thecorrectsandwelliscalled

the‘cued’sandwellinCTandPTalthoughthisisamemorycue;thereisnoexplicit

sensorycuemarkingit)(MovieS1).

Eachfloorpanelwitha6-cmholeforsandwellplacementcouldbeplacedinany

slotinthearena,theonlyconstraintsbeingthatthelocationsoftheintramazecues

(Row4,Columns2and5)andthosedirectlyinfrontofeachofthe4possiblestart-box

locationswereneverused(i.e.43possibleoutof49).Thesandwelllocations(map)

changedacrosssessionstoexcludeproceduralsearchingstrategiessuchasright/left,

close/far,specificquadrants,centre/corner,ordistancefromtheintramazecues.The

rewardedandnon-rewardedlocationswerecounterbalancedacrossanimalswithina

day.Twelvedifferentflavorsofrewardpelletwerealsoused,tohelpdiscriminability

andsopotentiallytoenhancememory(theseflavorswerecurry,blueberry,bacon,

nutmeg,anise,maple,garlic,ginger,cinnamon,mustard,coffee,andpeach).Flavors

wererepeatedonlyafter12sessions.

Behaviourwascarefullymonitoredthroughoutforerrorsandcorrect-digsfrom

whichaperformanceindex(PI)andLatencywerecalculated.OnCTs,theexperimenter

notedtheidentityandnumberofsandwellsatwhichananimaldug(errors)before

diggingatthecuedsandwell(correctdig),andthetimebeforediggingatthecued

9

sandwell.Thechancelevelforerrorsis2.5,butweaimedinthis6-alternativeforced-

trialprocedureforastablelevelofperformanceataround1error/trial.ThePIwas

calculatedas100*(5-errors/5)%,forwhichchanceis50%.Thelatencyto‘first-dig’at

thecorrectsandwellintheCTdroppedgraduallytoastablelevelof<20s(Fig.S1D),

exceptingonthesinglecontrolsession(S60).(Seeresultssectionfordetails.)Thiswasa

sessiontocheckforolfactoryartifactsbydeletingthesampletrialandrunningonlythe

choice-trial.Itwasexpectedthatperformancewouldbeslowandatchance.

Probetrialsessions

Oncestableperformancewasreached,occasionalProbeTrial(PT)sessionswere

interleavedwithfurtherS+CTtraining–thesebeingassociatedwithvariationsin

protocolfortheSTonawithin-subjectbasis(seebelow).Thesesessionsconsistedofan

STandthen,afteramemoryinterval,thePTitself.All6sandwellshadinaccessiblefood

pellets(NonrewardedsandwellsinFig.1D).Theanimalscoulddigatoneormoreof

thesandwells,butrewardwasinaccessibleateverylocation.Effectivememorywas

preferentialdiggingatthepreviouslyrewardedsamplelocation(cuedlocation).The

ratsweregivenonepellet(0.5g)inthestart-box,andthenallowed60secinthearena

(MovieS2).Diggingtimewasdefinedasdirectcontactwiththesandwiththeforepaws;

incontrast,sniffing,anosepoke,oragentlefoot-printassociatedwithonlyavisitwere

notconsideredas‘digging.’Thediggingtimescoring(0.1sresolution)wasperformed

inblindmannerbythesingleexperimenter(MN,mainstudy;MC,subsidiarystudy),and

checkedwithfurtheroff-lineblindscoringbyanindependentobserver(RF,seeFig.1F

foroneexampleofexperimenter/observercross-correlation).After60sec,thedoorof

thestart-boxwasclosedandtheratsreturnedtoitbyhand.Toreduce‘extinction’,the

experimenterthenplaced3accessiblepelletsatthebottomofthesandwellatthe

correctlocationinthearena,openedthestart-boxdoor,andtheratwasagainallowed

tofindandretrieveall3pellets.Allprobetestswereinterleavedwithtwoormore

regulartrainingsessions(STsfollowedbyCTs)(Fig.1D).

Memoryparametersexaminedbehaviourallyandbiochemically

OncetheanimalshadreachedstableasymptoticperformanceondailySTandCT

sessions,aseriesofPTswereconductedprecededbydistinctmanipulationsonorafter

thedailysampletrials(STs).

10

First,timedelays:weexaminedmemoryduration–schedulingthePTina

counterbalancedmanner24min,2.4h,and24hafteranSTinwhichtheanimalswere

rewardedwithasingle0.5gfoodpellet.Thisandfuturetestsalsoservedtocheckon

theaccuracyofmemoryscoringduringthePTs.

Second,massedvs.spaced:wecomparedtheclassicparameterofmassedvs.

spacedaccesstorewardonST,with3pelletsavailableoneachST.Afterretrievingthe

firstrewardpellet,andreturningtothestart-boxtoeatit,thestart-boxdoorwasclosed.

Thedoorwasthenre-openedforaccesstothetwofurtherpelletsaftereither30s

(massed)or10min(spacedaccess)–theseaccessdelaysbeingscheduledwithin-

subjectsinacounterbalancedmanner.Aftercollectingandeatingthethirdpelletinthe

start-box,theanimalswerereturnedtothehomecage.Betweenrewards,the

experimenterquicklycleanedanysanddugupbytheanimal.Thiscleaningprocesswas

introducedtolimittheemergenceofanystrategyinvolvingpreferentialdiggingata

sandwellsurroundedbysand,ratherthanguidancebyspatialmemory(MovieS3).

Third,biochemicalassayfortask-associatedgeneexpression:welookedat

theimpactofmemoryencodingongeneexpression.Asampletrialwasgivenusing

eitherthemassedorspacedtrainingprotocolwithtotalof3rewardpelletsretrieved.

Weconsideredcarefullytheappropriateprotocolsandtime-delaysbeforesacrificefor

conductingthisassay,decidingtokeeptheintervalafterthe2ndrewardpellet

consistentacrossconditions(seeFig.4A).Thismeansthat,asaconfoundingofthe

intervalbetweenrewardpellets,thereislongerinthestart-boxforthemassed

condition,andthismayhave,insomewaycontributedtothedifferentialpatterntobe

reported.Thereareseveralpotentialwaystoaddressthisproblem,butallinvolve

addeddifficulties(e.g.equatingtimeinthestart-boxbytakingtheanimalsinandout;

leavingtheanimalsinthestart-boxforalongperiodafterthesampletrialbefore

returningtothehome-cage,whichwouldhaveaddednovelty.Wechosethesimplest

protocol.Thus,forty-five(45)minafterthe3rdrewardpellet,theanimalswerebriefly

anaesthetizedinanisofluoranechamberanddecapitated.Thebrainswererapidly

takenoutandplacedinice-coldPBSfor30s.Acoronalblock(A/Pfrom-2mmto-7mm

fromBregma)wascutusingachilledbrainmatrix,dorsalhippocampusand

retrosplenialcortexweredissected,andsnapfrozenintheliquidnitrogen.Thesamples

wereshippedtoDartNeuroScience,andthenanalyzedusingRT-qPCRandRNA-Seq

(seebelow).

11

Fourth,cognitiveenhancement:priorto(3)above,welookedattheimpactof

i.p.injectionofaphosphodiesterase4inhibitor(PDE4inhibitor)Rolipramattwotime

pointsaftermemoryencoding.Thedrugsolutionwasfreshlypreparedonthedayof

injectioninthesamewayasdescribedbyRuttenetal.(Rutten,Prickaertsetal.2006).

Rolipram(Tocris,#0905)wasfirstdissolvedinethanol,withinjectablegradewater

addedtoduringrigorousvortexing.Thissolutionwasthenmixedwith2%Tylose

(Sigma,MH300,#93800)inwaterthathadbeenpreparedbyovernightgentlemixing

onarotatorat4oC.Afinalconcentrationof68.1mM(0.019mg/mL)in5%Ethanol/1%

Tylosewascreated,with0.03mg/kginjectedintraperitoneallyinthecontrolroomof

thelaboratoryat45minor3haftertheST(2-sandwellprocedure,onepelletretrieval).

Fifth,noveltyexposure:asbraintissuewastakeninstep3abovetomeasure

differentialgeneexpression,theimpactofnoveltyonretentionwastestedinthe

subsidiarystudy.Environmentalnoveltyinvolvedplacingtheanimalsfor5minintoa

largesquarePerspexbox30minafteradailyST(post-encodingnovelty).Thebox

measured1.4msquare,andcontaineddifferentfloorsurfacesonsuccessivevisits.Rats

areverysensitivetothesurfaceonwhichtheywalkandtheuseofverydifferent

surfacestothatofthemaineventarenaevokedactivityandexplorationoneach

occasion,providedthesearenotscheduledtoofrequently.Twoconditionswere

compared–onepairoftestsinwhicheithernoveltyexposureorremainingthehome-

cagefollowed30minafteranSTwith1pelletreward(Lowreward);andanotherwith

3pelletsreward(Highreward).Thetwo-sandwellprotocolwasusedthroughoutfor

which1pelletrewardisinsufficienttoproducemuchmorethanatransientmemory.

RT-qPCRandRNA-Seq

RNAExtraction:FrozentissueswerehomogenizedinQiazol(Qiagen)using

LysingMatrixDceramicspheresintandemwithFastPrep24TM5Ghomogenizer(MP

Biomedicals).RNAextractionwashstepsandelutionwereperformedroboticallywith

Qiacube(Qiagen)withanRNeasyMinikit(Qiagen)andDNasetreatment(Qiagen)as

permanufacturer’sinstructions.

QuantitativePCR(qPCR):TheDNasetreatedRNAswerereversetranscribedwith

randomhexamers(TaqmanReverseTranscriptionKit,ThermoFisherScientific).qPCR

reactionswererunintriplicateusingtheTaqmanFastAdvancedMasterMix

(ThermoFisherScientific)withTaqmanprobesfortheimmediateearlygenes,Arc,c-Fos,

12

Npas4andthereferencehousekeepinggenesTbp,Gapdh,andActb(ThermoFisher

Scientific)onaStepOnePlusRealTimePCRSystem(ThermoFisherScientific).We

performed40cyclesofPCR:1X(50°Cfor2min,95°Cfor20s)and40X(95°Cfor1s,

60°Cfor20s),anddeterminedrelativemRNAlevelsbynormalizingcyclethreshold(Ct)

valuesofthetargetandreferencehousekeepinggenesusingthe∆∆Ctmethod.Briefly,

wecalculated∆Ct(thedifferenceinCtbetweentargetgenesandreferencegenes)for

experimentalandcontrolconditions.Thedifferencebetweenthese∆Ctvaluesfor

controlandexperimentalconditionsisthe∆∆CtandchangesinmRNAlevelsaredirectly

proportionalto2-∆∆Ct.

RNA-Seq:RNAconcentrationandqualityweremeasuredwithaQubit

Fluorometer(LifeTechnologies)andaBioanalyzer(Agilent),respectively.RNA-Seq

librarieswerepreparedwithnon-strandedNEBNext®Ultra™RNALibraryPrepKitfor

Illumina®(NEB).Briefly,mRNAwaspurifiedfrom200ngtotalRNAandthen

fragmentedat94°C.cDNAwaspreparedwithrandomhexamers.Illumina-compatible

adaptorsanduniqueindexeswereaddedandDNAfragmentswereamplifiedwith12

roundsofPCR.Ampurebeads(BeckmanCoulter)wereusedforsizeexclusion.Libraries

wererunontheBioanalyzerandhadmeanfragmentsizesbetween384and457bp.3-

20pMofeachlibraryunderwentpairedend(2x50)sequencingonaHighSeq2000

(Illumina)with8librariesmultiplexedperlane.FastQfileswerealignedtotheEnseml

transcriptomeviaBowtieanddifferentialexpressionwasdeterminedusingDE-Seq.To

avoidcomplicationsfromfalsepositives,wesetalowp-valuecut-offofp<0.0001for

significanceandfocusedouranalysisonthosehitsthatwereconcordantbetween

hippocampusandretrosplenialcortex.

Randomising,CounterbalancingandStatisticalAnalysis

Behaviouralexperimentsrequirecarefulattentiontodetailwithrespecttoboth

theirconductandanalysis.Pointstonoteincludethatthecomparisonsofconditions

areallwithin-subjectsuchthatthereisnoissueofrandomisingtheallocationof

animalsacrossgroups–thereisonlyonegroupinboththemainandsubsidiarystudies.

Counterbalancingisalsosimpleinprinciplebutcomplicatedinpractice,theaimbeing

toensurethat,foranyoneofthefourstart-boxpositionsthatareusedacrossdays,half

theanimalsaretrainedoneachdaywiththecuedsandwellinanearlocationandthe

otheranimalsinafarlocation.Similarly,theplacementofthenon-rewardedsample

13

maybenearorfar,andtotheleftorrightoftherewarded(correct)samplelocation.

Withrespecttostatisticalanalysis,thedataalwayscompliedwithassumptionsof

homogeneityofvarianceallowingparametricANOVAstobeusedthroughout.Some

analysesfocusedonlyonchoiceperformanceusingthePerformanceIndex(PI),while

othersfocusedon(1)timespentdiggingattherewardedlocation(hereafter‘cued

location’),or(2)therelativediggingtimeatthecuedlocation,thenon-rewardedsample

location(non-cuedtrained)andtheremainingincorrectsandwells(non-cueduntrained

locations).Notetheterm‘cued’doesnotrefertoanystimulusmarkingtherewarded

location;tothecontrary,allcueingiswithinmemory.Asweusedawithin-subjects

design,Greenhouse-Geissercorrectionsofdegreesoffreedomwereroutinelyused

withinSPSStoensureaconservativeapproachtotheanalysis,buttheiruseleadsto

somewhat‘odd’valuesfordegreesoffreedom,asapparentbelow.

14

RESULTS

Theratsrapidlylearnedtorunfromthestart-boxesintothearenaandtodigfor

foodthattheywouldalwayscarrybacktothe(darker)start-boxtoeat.Following

habituationduringwhich,asnotedintheMethods,somelearningwouldhaveoccurred,

webegantrainingwithasinglesamplesandwellastheencodingtrial(ST)followedbya

choicetrial(CT),butswitchedafter18sessionstousingtwosamplesandwellsofwhich

onlyonewasrewarded(Fig.1E).

Performanceonthe6-sandwellCTs,circa1hafterthesampletrial,was

measuredusingtheperformanceindex(PI;seeMethods).Thiswasalreadyjustabove

chanceinsomeanimalsafterhabituation,butthenassumedbyallanimalsand

sustainedatanaveragelevelof77.9%fromsession10onwardswithameanz-score

relativetochanceof4.8(Fig.1E).ThisremarkablestabilityofthePIacross3monthsof

testingisimportantasitconstitutesthebackgroundagainstwhichthesuccessivedaily

1-trialencodingcanbeevaluated.Onsession60,thecontrolconditionwasscheduledin

whichtheSTwasnotgivenbutaCTstillrun;theanimalsshouldthennotknowwhich

sandwellwascorrectandshouldfalltochance.The‘virtual’correctsandwelllocation

wasdeterminedaccordingtothesamemappingruleasinregulartraining–namelythat

anynewrewardedsamplesandwelllocationcouldnotbeinthesamequadrantinthe

arenaasthatoftheprevioussession.Alltheconditionbiases-left/rightfromthe

startbox,corner/center,nearfarfromthestartbox-werecounterbalancedamong4

sub-batcheswithinthesession.Insuchdesign,theperformanceindexofsession60was

calculatedusingthe‘virtual’error.Virtualerroristhenumberof‘virtual’errorsmade

beforediggingatthe‘virtual’correctone(whichwasinpracticerewarded).Ananalysis

ofthePIsfromsession58to62showsthat,aspredicted,performancedroppedto

chanceonsession60(repeatedmeasures(RM)-ANOVA:F=9.52,df2.9/61.2,

p=0.000036;degreesoffreedomadjustedbyuseofGreenhouse-Geissercorrection;

one-samplet-testvs.chance:t22=1.16,p=0.26forsession60andt22>4,ps<0.0001for

allothersessions;Fig.1E).InterspersedPTs(grayshading)werethereafterusedasthe

primarymeasureofmemory.ThescoringduringPTswasdouble-checkedforasub-set

ofanimalsusingindependentblind-scoring.Averyhighcross-correlation(R2=0.992)

wasobservedbetweentheinitialsemi-blindscorederivedfromscoringduringtesting

(byMN)andthatderivedfromoff-lineblindanalysisofDVDrecordeddatabyan

independentobserver(RF;Fig.1F).Thus,theratssuccessfullyencodedthelocationof

15

therewardedsandwellduringanST(withoneortwosamplesandwells);thisencoding

wasnecessarytomakeacorrectchoiceduringthesucceedingCT;andPTscoringwas

objective.

Theuseofasinglesandwell(containing1-5rewardpellets)asthesampletrial

(ST)hadtheunintendedeffectofencouraging,duringCTsandPTs,persistentdiggingat

whicheversandwellwaschosenfirst.Asthisincludedtrialswhenerrorsweremade,

persistentdiggingataninitiallyselectedlocationresultedinhighvariabilityoftest

scores.Theswitchtousing2sandwellsonSTsresultedinastrategychangetofrequent

shiftingbetweensandwellstosecurereward.Thisgreatlyreducederrorvariability

duringlaterPTsatthecostofaslightdecreaseinaveragetimespentdiggingatthe

correct(cued)sandwell.

Delay-dependentforgettingwasobservedaftersampleencodingusingboththe

1-and2-sandwellprocedures,eachinvolvingasinglesmallreward(1pellet).Two

counterbalancedseriesof3within-subjectPTs(nofoodavailable)wereconductedat

delaysof24min,2.4hand24hafterSTs,with2interveningregulartrainingsessions

(consistingofSandCTs,Fig.2A).Memorywasmeasuredintermsofthepercenttime

spentdiggingatthecorrect(cued)vs.theincorrect(non-cued)sandwellsduringthe

PTs(Fig.2B-C).TheoverallpooledrepeatedmeasuresANOVAofthetwosetsofdata

showedsignificantdelay-dependentforgettingover24h(F=3.87,df1.6/35.7,

p=0.038);Greenhouse-Geissercorrectionforrepeatedmeasures),withastrong

downwardlineartrendovertime(F=10.12,df1/22,p=0.0043)).Subsequentt-tests

comparingcorrect(cued)sandwellperformanceateachmemoryintervalindicated

abovechanceperformanceinthePTsafterboth1-sandwelland2-sandwellsample

trialsat24minand2.4h(1sandwell:t22=5.12and4.95,ps=0.00004and0.00006,

respectively)forboth;2sandwell:t22=4.34and3.24,ps=0.00027and0.0039but

chanceperformanceat24h(1SWt22=1.96,p=0.063;2SWst22=1.73,p=N.S.).

InsertFigure2abouthere

Theimpactofmassedvs.spacedaccesstorewardwasthenexamined,using

onlythe2-sandwellprocedureforSTs(Fig.3A).Memoryretentionimprovedwhen

rewardavailabilitywasincreasedfrom1to3pellets,witheithermassedaccessto

reward(inter-reward-intervalof30sec)orspacedaccess(IRI=10min)(MovieS3).

16

However,whilemassedrewardproducedgoodmemoryat2.4h,therewasstill

forgettingover24hr.Incontrast,spacedrewardhadadramaticeffecton24hmemory

whichwasbothabovechanceat24handnotmeasurablydifferentfromthatat2.4h

(Fig.3B).Importantly,theANOVAcomparingmassedvs.spacedrewardandtime-

intervalsrevealedasignificantinteraction(F=7.39,df1/22,p=0.013).Separatet-

testsshowedthatmemorywasabovechanceat24honlyfor3spacedrewards(Spaced:

t22=4.52,p=0.00017;Massed:t22=1.73,p=0.099),whereaslocationmemoryat2.4h

didnotdifferaftermassedvs.spacedreward(Massed=35.3± 3.3%;Spaced=32.1±

2.7%;N.S.p=0.45).Thus,theimpactofspacedrewardseemstobeonthepersistenceof

everydaymemoryat24hratherthanimmediatememorystrengthat2.4h.

Overthe3rewardsoftheST,wealsoexaminedtherateofdeclineoferrorsin

choosingbetweenthe2sandwells(Fig.3C).Tobeexpected,theanimalswereatchance

forthe1strewardpellet,butimprovedforpellets2and3.Ifanything,thetrendwasfor

thisimprovementtobeslightlygreaterformassedthanspacedtrials,butthisdidnot

reachsignificance(F=1.15,df1/22,p=0.295).However,theimprovementacross

successivepellets(1st,2ndand3rd)wassignificant(F=48.51,df2/44,p=7.4E-12).

InsertFigure3abouthere

Becausememoryconsolidationrequiresproteinsynthesisandtranscription

(DavisandSquire1984,AlberiniandKandel2015),wethenexaminedwhetherdaily

encodingofspatialmemoryresultedinchangesingeneexpression,andmore

interestingly,ifthereweredifferencesintheresponsetomassedandspacedtraining.

Attheendofstudy,ratsweregiveneithermassedorspacedaccesstothereward

presentationsinthe2-sandwellprotocolandeuthanized45minafterthesecondof

threetrials(Fig.4A;anintervalchosentominimizeanypossibleimpactofslight

differencesintimingbetweenparadigms).Thethreegroups(Control(noSTgivenand

Naive),Massed,andSpaced)wereassignedsoasnottodifferwithrespecttopriortask

performance(PIsforCtrl,MassedandSpacedwere77.4,77.8,77.5%,respectively

(one-wayANOVA,N.S.);andPTaveragescoreswere30.1,29.4,28.6%,respectively

(alsoN.S.).Numberofsubjects=11,8,8(Note:Thenaïverats(n=4)didnotdifferfrom

thetrainedanimals(n=7)givenno-STpriortoeuthanasiaandtheRT-qPCRanalyses,

sopooledtogetherasacontrolgroup(Ctrln=11)).Ouranalysisfocusedonthedorsal

17

hippocampusandtheretrosplenialcortex-tworegionsthatarecriticalforspatial

navigationandmemory(VannandAggleton2002,FranklandandBontempi2005,Vann,

Aggletonetal.2009)(Fig.4A).

InitiallyweperformedRT-qPCRonthreeimmediate-earlygenes(IEGs),Arc,c-

FosandNpas4aswellasthreehousekeepinggenes(HKP),Gapdh,ActbandTbp.All

threeIEGsweremarkedlyupregulatedbymassedandspacedtraininginbothtissues,

whereasthelevelsofthehousekeepinggeneswereunchanged(Fig.4B-C).TheDDCt

wascalculatedwithnormalizationtoGapdh,butsimilardatawasobservedwheneither

oftheothertwohousekeepinggenes(ActbandTbp)wasusedasareference.Thedata

fromallthreeIEGswaspooled,andthatforthethreeHKPgenes,andanRM-ANOVA

conductedacrossthethreeconditions(Ctrl,Massed,Spaced).Thisrevealedahighly

significantIEGvs.HKPxConditionsinteractioninbothbrainregions(HPC:F=24.08,df

2/24,p=000002;RSC:F=14.13,df2/24,p=0.000088).Thisjustifiedlookingatthe

IEGsalone,revealingsignificanteffectsoftrainingcondition,i.e.betweenControl,

MassedandSpacedconditions(HPC:F=19.64,df2/24,p=0.000009;RSC:F=13.17,df

2/24,p=0.000138).Subsequentorthogonalcomparisonsshowedthat,forHPC,the

Controlconditionwassignificantlylowerthanthecombinedmassedandspaced

conditions(F=34.1,df1/24,p<0.001)andthatactivationwithspacedrewardwas

higherthanwithmassed(F=5.12,df1/24,p<0.05).Basedonthesedata,wedecidedto

performRNA-Seqexperimentsonthesamesamples.

ToensurethatourRNA-SeqdataprovidedaccuratemeasurementofmRNA,we

plottedthenormalizedFPKMvaluescalculatedwithRSEMforthegenesalreadytested

withqPCR(Fig.4D-E).WefoundthattheRNA-SeqandRT-qPCRmethodologiesgave

nearlyidenticalresults,andproceededtomeasuretheeffectsoftrainingonallprotein

codinggenesusingDE-Seq2.Inthemassedcondition,theexpressionof45geneswas

increasedinHPCand82inRSC.Ofthesegenes,29weresignificantlyinducedinboth

structures.Inthespacedcondition,wefound113upregulatedgenesinHPCand137in

RSC-with54genesinducedinboth.Becauseanyrandomfalse-positivesinagiven

tissuewouldbeexceedinglyunlikelytoreplicateinindependentRNAsamplesfrom

anothertissue,weexaminedfurtheronlythe29‘Massed’and54‘Spaced’genesthat

wereup-regulatedinboththeHPCandRSC,andfoundthatofthegenesmeetingour

strictcriteriainmassedtraining,27werealsoup-regulatedinspaced(Fig.4F).Because

moregeneswereupregulatedinthespacedconditionthaninthemassed,wewantedto

18

determineiftherewereanygenesonlyregulatedbyspacedfoodreward.Wefocusedon

the56genesthatweresignificantinbothtissueswitheithertrainingparadigmand

correlatedthefold-changeinducedbymassedandspacedtraininginHPC(Fig.4G)and

RSC(Fig.4H).Thisanalysisshowedthatthegenesthatwereup-regulatedsignificantly

inratstrainedwithspacedtrialswerealsoup-regulatedbymassedtraining,butdidnot

alwaysreachourstrictstatisticalsignificancecut-off.Thisindicatesthatthe

constellationofgenesinducedbymassedandspacedtrainingissimilar.However,a

linearfitofthedata(R2=0.94and0.96forHPCandRSC,respectively)revealedthatthe

relationshipbetweentheeffectsofmassedandspacedtrialsfollowedapowerlawof

1.33inbothtissues,indicatingalargerincreaseingeneexpressionfollowingspaced

foodreward(TableS1).Only2ofthe56geneswereincreasedtoagreaterextentinthe

massedconditionand,interestingly,thesetwogenesweresignificantinbothHPCand

RSC.Incontrast,ahighernumberofgeneswereincreasedtoagreaterextentafter

spacedtraining.Orthogonalcomparisonsacrossall56genesrevealedthat7geneswere

significantlymoreinducedinspacedthanmassedinbothHPCandRSC,allofwhichare

CREB-targetgenes;14genesweresignificantlymoreinducedinspacedthanmassedin

eitherHPCorinRSC;someofthesearealsoCREB-targets.Amongstthose,weidentified

severalmembersoftheEgrfamilyofgenes,Dusp1(aregulatoroftheMAPKpathway),

andtheheatshockproteinsHspa1a(Hsp70)andHspb1(Hsp27),chaperonescriticalfor

thefoldingofnewlysynthesizedproteinsandcellulardifferentiation(Fig.4G-H).These

dataindicatethatthemassedandspacedtrainingparadigmsinitiallyrecruitvery

similargeneexpressionresponses,butthatthemagnitudeoftheresponseisgreater

withspacedtraining.

InsertFigure4abouthere–RT-qPCRandRNA-Seq

TheimpactoftheputativecognitiveenhancerRolipram(0.03mg/kg)wasthen

investigatedusingtwopost-encodinginjectiontimes,45minand3h,usingonlysingle

rewardencodingintheSTtomaximizeforgettingover24h(Fig.5A;butstillusing2

sandwellswithonlyonerewardedintheencodingtrial).Noenhancementwas

observedwiththe3hpost-encodinginjections,butastrikingimprovementwas

apparentat45minrelativetovehicleinjections(28.5±2.8%,comparedto19.8±2.4%).

AnoverallANOVAofdiggingintheprobetrialsatjustthecued-locationofthesample

19

trialshowedatrendintheinteractionbetweenDrugandTime-Delay(F=3.87,df1/22,

p=0.061842).However,anRM-ANOVAcomparingdiggingattheCued-trained,Non-

cuedtrained,andtheNon-cueduntrainedsandwellswassignificantforthe45minpost-

encodinginjection(F=10.92,df1.415/31.12,p=0.001),togetherwithaninteraction

reflectingtheimpactofRolipramcomparedtovehicle(F=3.67,df1.615/35.54,p=

0.044)(Fig.5B).Theanalysisofperformancewiththe3hpost-encodinginjections

showednodrugeffect(F<1),althoughtheusualdifferenceacrosssandwells(F=9.36,df

1.42/31.24,p=0.0019).Thus,theretentionof‘everyday’spatialmemorycanbe

enhancedbyinhibitionofPDE4andsuchaugmentationofmemoryispossiblein

behaviourallyexperiencedanimals.

InsertFigure5abouthere

Thesubsidiarystudyinvolvedtrainingaseparatecohortofanimalsinsameway

toinvestigatetheimpactof5minofpost-encodingenvironmentalnoveltyonmemory

retentionafterweak(1pellet)orstrongreward(3-pellet,massed)inthe2-sandwell

sampletrialprocedure(Fig.6A).TheexcellentstabilityofperformanceonTraining

DaysisshownintheFig.S2.Afteraninitiallearningperiod,dailyperformancewasat

oraround75-85%throughoutalongseriesofmemoryprobetests(Chance

performanceis50%).Thisstabilitygivesconfidencethatprobetestsconductedearly

orlateintrainingwereagainstastableperformancebaseline.

AsshowninFig.6B-C,forthestrongprotocol,therewasexcellentmemoryfor

therewardedsamplelocationat24min(58.2± 7.1%;chance=16.7%).Memory

declinedover24htochance(25.6± 6.3%)asexpectedformassedreward(an‘internal

replication’).However,if5minnoveltyexposurewasscheduled30minpost-encoding,

memorywassustainedabovechance(39.2± 4.0%).Thedataplottedisanaverageof

thefirst60secsearching,during2separateprobetestsforthe24htests,withonlyone

testbeingsufficientforthe24mintest.TheRM-ANOVAofdiggingacrossthesandwell

locationsshowedawithin-subjectsconditionseffect(F=6.72,df1.97/19.70,p=

0.0061)thatdeclinedinamonotonicfashionacrossconditions(24min,24h+Novelty,

24h-only:LinearF=11.89,df1/10,p=0.062).Thus,retentionwasgreaterwithpost-

encodingnovelty.Separatetestsestablishedthatperformancewasabovechanceinthe

24minand24h+Noveltyconditions(p=0.00017and0.00022,respectively),butat

20

chanceforthe24hcondition(p=0.19,NS).Noveltyscheduled30minafterencoding

appearedtoimprovetheabilityoftheanimalstodiscriminatetherewardedandnon-

rewardedsamplelocations(Fig.6C;plotting%timespentattheNon-cuedtrained

locationfromthatattheCuedlocation,whileignoringtheotherNon-cueduntrained

locations.Specifically,thisdiscriminationisclear24minafterencodingbutatchance

after24h.However,post-encodingnoveltyliftsthecapacitytodistinguishNon-cued

trainedfromCued-trainedatthe24hrtestinterval(ANOVArevealedadeclineacross

memorydelays(F=4.16,df1.99/19.95,p=0.031),decliningmonotonically(LinearF=

7.98,df1/10,p=0.018).Subsequentplannedorthogonalcomparisonsrevealedthatthe

24minand24h+NoveltyconditionsdidnotdifferfortheCuedlocation(F=2.01,df

1/20,N.S.),butthemeandiscriminationbetweenthetwosamplelocationsinthesetwo

conditionswassignificantlygreaterthanthatinthe24honlycondition(F=6.29,df

1/20,p<0.01).

Separateprobetestswereconductedusingtheweak1-pelletrewardprotocol

(Fig.6D,E).Theserevealedmodestmemoryat24min,nothingat24h,andnoimpact

ofnoveltyoneitherabsolutelevelsforeachsandwellorwithrespecttothedifference

score(ANOVAs:F=1.66,df2/20,p=0.216(N.S.);andF=2.69,df2/20,p=0.092).

Takentogether,thesefindingsimplythatpost-encodingnoveltycanaugmentretention

buttherehastobesomeminimallevelofmemoryforthistobeeffective.

InsertFigure6abouthere

DISCUSSION

Thisreport/technicalspotlightreportsaneverydaymemoryprotocolthat

enables1-trialmemorytobeinvestigatedrepeatedlyovermonthsinexperienced

animalstherebyenablingwithin-subjectcomparisons.Inthisstudy,wehaveconducted

both‘calibration’experimentsinrelationtowell-establishedprinciplesofmemory,and

a‘prospective’studyillustratingitspotentialforfuturework.Bywayofcalibration,we

observe:(1)delay-dependentforgettingofone-trialappetitivelymotivatedspatial

memoryover24h;(2)morepersistentmemoryovertime(i.e.betterretention)with

multiplespacedaccesstorewardbutnotmassedaccesstoreward;(3)enhanced

retentionwhenthePDE4inhibitorRolipramwasadministeredshortly(45min)after

21

encoding.Withrespecttoits‘prospective’useforfuturediscoveryresearch,wealso

report(4)upregulationofspecificIEGsandawiderpatternofgene-inductionin

associationwitheverydaymemorytrainingthatisquantitativelygreaterforspaced

comparedtomassedexperience.Toourknowledge,thisisthefirstreportofsuch

differentialgeneexpressionasafunctionoftrialspacinginmammals.Thesedatawere

securedagainstabackgroundofdemonstrablystablelevelsofchoiceperformance(our

PerformanceIndex(PI)measure),withtheobjectivescoringoftheMemoryProbeTests

(PTs)securedbymeansoftwoindependentbutwell-correlatedobservers.The

reliabilityofthetaskwasreplicatedinthesecondcohortofanimalswhowentonto

display,usingthe2-sandwellprotocol,that(5)post-encodingnoveltyenhances

retention.Weconsidertheseresultsinrelationtothevalueofour‘everydaymemory’

protocol,itslikelygenerality,andthusitspotentialformakingdiscoverieswithrespect

tocognitiveenhancement.

Everydaymemoryischaracterizedbysubstantialdailyforgetting

DavidMarrsuggested,asanassumptionwithinhismathematicalmodelof

‘archicortex’(Marr1971),thathumansmayencodeamaximumof1memorypersec

throughouttheday,withovernightconsolidationintoneocortex,therebyrequiringa

maximaldailystoragecapacityinhippocampusof105-see(Willshaw,Dayanetal.

2015).Thisassumption,whilelikelyinaccurateindetail,servesasareminderthatthe

mammalianepisodic-likememorysystemshouldbecapableofprocessingnumerous

eventsthroughthedaywithminimalinterference.Thisideaiscapturedinseveral

behaviouraltasks,includingspatialmemoryastestedhere((Olton,Beckeretal.1979,

SteeleandMorris1999,Bast,daSilvaetal.2005),andrecognitionmemorytasksas

developedfornon-humanprimates(MishkinandDelacour1975)anditsrodent

equivalent(EnnaceurandDelacour1988).Ourapproachto‘everydaymemory’

incorporatestheideathatdistincteventscanbeencoded,rememberedandforgotten

eachdayincludingsuchinformationas“what,where,when,whatcontext”(Griffiths,

Dickinsonetal.1999,EacottandEaston2007).

Againstthebackgroundofstabilityacrosstesting,ourfirstfinding–delay-

dependentforgettingof1-trialmemoryover24h-capturesadefiningfeatureofthis

modelofeverydaymemory.Ouruseofthe2-samplesandwellprocedureextendsthe

earlierobservationofdelay-dependentforgettingwithasinglesandwell(Bast,daSilva

22

etal.2005),butnowusingawithin-subjectsprotocolruneverydayfor4+months.We

canthinkoftheprotocolasonethatincludesdailyeventsthatmightbeautomatically

remembered–beingplacedintheapparatus,waitingforthestart-boxdoortoopen,

leavingthestart-box,searching,digging,findingfood,returningtothestart-box,and

eatingthefood.Wecannottestforeachofthese‘everyday’memories,butspatial

memoryinthearenaservesasanindex.Duringtrainingsessions,theanimalsaregiven

theopportunitytochoosebetween6sandwellsapproximately1hrafterencoding.

Behaviourally,theydisplaya‘win-stay’choicecorrectlyfromveryearlyintraining

(therewere7daysof‘habituation’precedingthathelpedteachthisstrategy),eventually

reachingperformanceofaround1errorpertrialthroughouttheremainderoftraining

(circa80%).Thestabilityofperformanceisreflectedinaremarkablez-score,relativeto

chance,of>4.5,thatwasshown,usingasuitablecontrolcondition,wasnotduetoany

‘olfactory’artefact.Whenthechoice-trialwasoccasionallysubstitutedbyaprobetestin

whichreinforcementwasnotavailablefor60sec,theyshowstrongpreferentialdigging

atthecorrectsandwellat24minandat2.4h,butmemorydeclinedtochanceover24h.

The2-sandwellsampleprocedureresultsinnotablylowvariabilityinprobetrial

performanceacrossanimalsastheyacquiredthesandwell-shiftingstrategyinthedaily

STsasdiscussedabove.Thestabilityofperformance(Fig.1E)andthesustained

motivationoftheanimalsshouldenablereliablewithin-subjectcomparisonofdrugs

regardlessofthesequenceofdrugtestingovertime.Performancestabilitywas

replicatedinthesubsidiarystudy(Fig.S2).

Asfurthercalibrationofthetask,weobservedthatmemoryimprovedwith

accessto3spacedrewardsonanencodingtrial(i.e.theanimalcollectsthe3reward

pelletsonebyone).Theimprovementof24hmemoryoccurredwithaninter-

reinforcementinterval(IRI)of10minbutnotwhenonly30sec.Thisfindingconfirmsa

largebodyoffindingsindicatingthatspacedtrainingproducesbettermemorythan

massedtraininginhumansandanimals(Ebbinghaus1885,Greene1989,Commins,

Cunninghametal.2003),aneffectalsoobservedinstudiesoflong-termpotentiation

(Kramar,Babayanetal.2012)thatlikelyreflectstheneedtoengagelong-termmemory

aswellasworkingmemoryduringacquisition.Apotentialcomplicationisthatthe

quantitativeparameterofinter-rewardinterval(30secvs.10min)isconfoundedby

totaltimespentinthestart-boxthatmayhaveeffectsonshort-termsatiation,arousal

orfrustration.However,withsuchasmallproportionoftheanimal’sdailydiet

23

consumedduringthetask,itwasnosurprisethattheanimalsatethepelletsreadilyand

within30secandshort-termsatiationseemsunlikely.Insofarasarousalmayhave

beenaffected,arguablytheshorterintervalwouldhavesustainedarousalbetterthan

the10minwaitinthestart-box,unlessthelongerintervalof10mininducedfrustration.

However,wesuspectthatthemaincontributoristhatreturningtothecorrectsandwell

forthesecondandthirdrewardpelletrequiresretrievalfromlong-termmemoryin

onlythespacedcondition.Retrievalduringanencodingsessionisknowntoenhance

latermemory–theso-called‘testingeffect’(KarpickeandRoediger2008)–andalso

causesdifferentialmemory-relatedgeneexpression.

Enhancedgenetranscriptionassociatedwithtrial-spacingandenhancedmemory

Theprospectivevalueofthisnewbehaviouraltechniqueisshowninthe

observationofgradedupregulationofIEGsgenessuchasArc,c-FosandNPAS-4,and

heatshockproteins,Per1,andDusp1,withnochangeincontrol‘housekeeping’genesin

thehippocampusandretrosplenialcortex.Notonlywasgene-expressionhigherin

animalstrainedinthetaskshortlybeforeeuthanasia,butwasalsosignificantlyhigher

withspacedthanmassedtraining-afindingobservedwithbothRT-qPCRandRNA-Seq.

Toourknowledge,thisisthefirststudyinrodentstofindmoregeneexpressionin

spacedovermassedtraining.Apotentialqualificationisthatwetestedgene-expression

45minafterthemiddletime-pointofthesampletrial,whichconstitutesa9.5min

differencebetweenthefirstandthelasttrialsofmassedorspacedtrainingrespectively.

Whilethisminortimingdifferencecouldpossiblyaccountforthedifferentialexpression

ofhighlytransientIEGs,itisunlikelythatthemanygenesfoundinourRNA-Seqresults

shouldallshowsuchtimesensitivity.Interestingly,inthefruitflydrosophila

melanogaster,memoryencodedwithasinglesessionofolfactoryconditioningis

augmentedbyover-expressionofthetranscriptionfactorCREB(cAMP-response

elementbindingprotein)(Yin,DelVecchioetal.1995).Thisfindingwascorroborated

morerecentlybystudiesdemonstratingthatspaced,butnotmassed,trainingleadsto

CREBdependentactivationofthecalcium/calmodulin-dependentproteinkinaseII

(CaMKII)andperiodgenesintwodorsal-anterior-lateral(DAL)neuronsindrosophila,

andthatproteinsynthesisinDALneuronsisrequiredforLTMconsolidation(Chen,Wu

etal.2012).Inorthogonalexperimentsinmice,eitherspacedtrainingormassed

trainingcombinedwithover-expressionofCREByieldsstrongerfearmemorythan

24

massedtrainingalone(Josselyn,Shietal.2001).Together,theseresultsindicatethat

spacedtrainingisassociatedwithtranscriptionalactivationininvertebratesand

vertebratesalike.

Massedandspacedtrainingtrialsbothincreasedgene-expression,butthe

responsewasmorepronouncedafterspacedencoding.Thedifferentialeffectsof

massedandspacedtrainingonsignal-transductioncascadesforgene-inductionare

poorlyunderstood.However,individualtrialsmayresultincalciumsignalsthatare

alreadyabovethresholdfortriggeringCaMKII,PKAorothersignalingcascades.Insuch

scenario,thespacingofrewardwouldenablesuchcascadestobeactivatedthreetimes

whereas,withmassedtrials,thismaynotoccurduetoarefractoryperiod.Asspaced

trainingcansometimescompensateforthereductionofCREB(e.g.inCREBa/d

knockoutmice(Kogan,Franklandetal.1997,Bourtchouladze,Lidgeetal.2003)),there

isthepossibilitythatotherpathwaysmaybeinvolvedinmemorypersistence.Non-

transcriptionalmechanismsshouldalsobeconsideredsuchasactin-polymerizationand

recruitmentofspinesforencodingduringspacedtraining(Kramar,Babayanetal.

2012),aswellasproteinphosphatases.SpacedlearningtrialsinhibitPP1activityvia

PKAphosphorylationofinhibitor1inmiceleadingtostrongeractivationofCREBand

PKAphosphorylationofGluR1(Genoux,Haditschetal.2002)andindrosophilaspaced

traininginducesrepetitivewavesofMAPKactivationinaSHP2dependentmanner

(Pagani,Oishietal.2009).Dopaminemayalsoplayacriticalroletheregulationof

memorypersistenceafterspacedtraining,somethingsuggestedbyworkinfruitflies

(Placais,Trannoyetal.2012),buttoourknowledgenotyetexploredinmammals.

Clearly,additionalworkisrequiredtofosterourunderstandingofthespacingeffecton

memorypersistence–aphenomenonfirstdescribedmorethan130yearsago.

Cognitiveenhancement:impactofRolipram

WealsoobservedthatthePDE4inhibitorRolipramenhancedeverydaymemory

at24hwhengiven45minafterencodingtrials,butnotwhengiven3haftertraining.It

isalsoworthnotingthat,inmuchoftheliteratureexaminingtheimpactofRolipramon

memory,severalpreviousstudieshavereportedpositiveeffectswhenthedrugis

appliedpriortothelearningtrial,forexample(Barad,Bourtchouladzeetal.1998,

ZhangandO'Donnell2000).Thismayeventuallyprovetobea‘valid’wayforcognitive

enhancerstobeusedinhumans,but,fromtheperspectiveofscientificanalysis,it

25

carriestheambiguityofbeingaprocedurethatconfoundsapotentialimpacton

encoding(attention,memoryacquisition)andaneffectonconsolidation(post-

acquisitionmechanisms).Ourprimaryinterestisthelatterand,forthatreason,weonly

testedpost-trialinjectionsofRolipram.Theeffectivenessofthe45mintime-pointis

alsoofinterestasvariouslinesofresearchonlong-termpotentiationinvitro(at32oC)

suggestthatsynaptictags,setatthetimeofmemoryencoding,havealife-timeof

around1.5h(FreyandMorris1998a).Synaptictagssequesterplasticity-related

proteins(PRPs).IfRolipramupregulatestheavailabilityofPRPsfromthecAMP-PKA-

CREBpathway,thenanintervalof45miniswithinthetimewindowforeffective

capturebyasynaptictag.TheeffectivenessofRolipramasacognitiveenhancerhas

beendemonstratedinvarioustestsofmemoryinnaïveratsandmice,including

contextualconditioning(Barad,Bourtchouladzeetal.1998),objectrecognition

(Bourtchouladze,Lidgeetal.2003,Rutten,Prickaertsetal.2006)andobjectlocation

memory(Rutten,VanDonkelaaretal.2009).Ourstudynowextendsthisefficacy

spectrumtoeverydayspatialmemoryinratswithextensivepriormemorization

experience.WhileRolipramisnotsuitableforclinicaldevelopmentbecauseofits

emesis-inducingpotential(ED50(p.o.)=0.03mg/kg(humannausea),0.48mg/kg(rat

pica);PercieduSert,Holmesetal.2012),nextgenerationPDE4inhibitorswithalesser

emeticpotentialhavebeendevelopedforCNSindications(Tully,Bourtchouladzeetal.

2003,Burgin,Magnussonetal.2010,Peters,Bletschetal.2014).

Theoreticalimportanceoftheimpactofpost-encodingenvironmentalnovelty

Noveltyisofteninvestigatedinrelationtoreward,asinnowclassical

experimentsinwhichananimal’sexpectationsofrewardareviolatedand,forexample,

greaterrewardismadeunexpectedlyavailable(SchultzandDickinson2000).Thisis

nowknowntoactivatedopaminergicsignallingintheventraltegmentalarea(VTA)that

promoteseffectivelearning(MontagueandSejnowski1994).Adifferentkindofnovelty

isunrelatedtothetaskinhand,suchasexposuretoanovelstimulusorenvironmentat

somepointbeforeoraftermemoryencodingofaseparatetask(MoncadaandViola

2007,Wang,Redondoetal.2010).Thisnoveltyaugmentstheretentionoftheunrelated

buttemporallyproximatetaskinadopamine-dependentmannerwithdifferentialparts

oftheCaMKinasepathwayinvolvedintaggingandplasticity-proteinsynthesis

(RedondoandMorris2011).Theneuromodulatorymechanismsarestillunclear,butan

26

intriguingnewpossibilityisthatnoveltyismediatedbyactivationofneuronsinthe

locuscoeruleus(LC)(Takeuchi,Duszkiewiczetal.2016)thatcanco-releasedopamine

(SmithandGreene2012,Kempadoo,Mosharovetal.2016).Thisparadoxical

dopaminergicmodulationcouldtriggerthesynthesisofplasticity-relatedproteins

linkedtothestabilisationandthusretentionofmemory(FreyandMorris1997,

RedondoandMorris2011).Todate,severaltestsdemonstratingenhancedmemory

withpost-encodingnoveltyhaveusedmemoryofasinglestimulusorevent(inhibitory

avoidance,digginginasandwell,object-placememory),butwesoughttotestwhether

noveltycouldalsoenhancethediscriminabilityoftworecentevents,andtheselective

retentionoftherewardedeventrelativetoanon-rewardedevent.Thesubsidiarystudy

usedthe2-sandwellencodingprocedurewith3massedrewardswithonlyone

sandwellrewarded.Whilememoryofthecuedsandwelllocationwasatchanceafter24

h,post-encodingnoveltyenhancedretentionofthecuedsandwellwithoutenhancing

memoryofthenon-cuedtrainedlocation.

Evaluatingcognitiveenhancement

Muchhasbeenwrittenaboutmeritsandpossibledrawbacksofcognitive

enhancement,andaboutthestrategiesforachievingitrangingfrombehavioral

approachessuchasthe‘method-of-loci’andretrievalpractice(RoedigerandButler

2011),throughtodrugsthatactatthetimeofencodingsuchasAMPAkines(Lynch,

Palmeretal.2011)orduringconsolidation(Barco,Pittengeretal.2003,Tully,

Bourtchouladzeetal.2003).Afinalpointrelatestothesuitabilityofanimalmodelsfor

testingputativeenhancers.Acriticalfactorisourpresentinabilityto‘back-translate’

fromhumanclinicaldatatoanimalmodels–largelybecausedrugsthatreliablyenhance

memoryinhumanshaveyettobefound.Minordifferencesinbehaviouralprotocolsin

animalmodelsmayseemtobeunimportantcomparedtotheexcitementofnovel

targetsidentifiedusingthelatestgenetictechniques.However,wenowassertthatthey

haveanimportantplaceinsuccessfullyidentifying‘hits’andavoiding‘falsepositives’.

Thepresentprocedurecanbecontrastedwithseveralothertasksofwhichone,context

fearconditioning,iswidelyusedbyvirtueofbeingsucharapidassayforthescreening

ofgenesanddrugcandidates.However,notonlyisitaslightlyoddtaskcomparedto

anythingthatnormallyhappenstoapersonduringtheday,andthusofdoubtful

‘constructvalidity’,itisgenerallytestedinexperimentallynaiveanimals.

27

The‘lost-in-translation’problemthathasledtotoofewdrugsbeingdeveloped

forhumanusedespitepromisingpre-clinicalresearchmaybeexacerbatedbyafailure

torecognizethatadifferencebetweenanimalandhumantrialswithrespectto

cognitionisnotonlythespeciesbutalsothemannerinwhichthetestingisdone.Prior

knowledgeisakeyparameter.Recentworkonschemalearninginanimalsandhumans

(Tse,Langstonetal.2007,vanKesteren,Ruiteretal.2012)hasbroughtintosharpfocus

theissueof‘priorknowledge’ineverydaymemoryfunction.Thepresenttask,whilenot

schematicinnature,nonethelessreliesonlookingatsingleepisodememoryencoding

inanimalsthathavehadextensivepriortraining.Wenowhaveanoptimizedversionof

theeverydaymemorytaskfeaturing:1)sustainedmotivationandperformanceover

months,2)statisticaladvantages(within-animaldesignandlessvariabilityinprobetest

scoresamonganimals),3)objectivescoringwithgoodinter-observeragreement.

Hopefully,thistaskcouldbridgethegapbetweenrodenttestsandhumantrialsbetter

thanmanyincurrentuseinpreclinicaltrialsThegeneexpressionanalysesnotonly

revealedforthefirsttimeinmammalsthatspacedencodingeventsinducedmoregene

expressionbutthegeneticresourcesobtainedfromRNA-seqwouldoffercluesto

potentialtargetsformemoryenhancement.Ourapproach,possiblysupplementedby

addingnon-spatialrecognitionmemorytotheprotocol,offerstheprospectofrevealing

theeffectivenessofnewdrugsthatcouldtranslateeffectivelyintosuccessfulPhaseII

trialsforhumans.

28

ACKNOWLEDGEMENTSANDROLES

ThisworkwasfundedbyDartNeuroScienceLLCandanERCAdvanced

InvestigatorGrantheldbyRGMMandGuillenFernandez(ERC-2010-AdG-268800-

NEUROSCHEMA).JL,DW,TTandMPareemployeesofDartNeuroScienceLLC.MNwas

supportedbyJSPSPostdoctoralFellowshipsforResearchAbroad.MN,MPandRGMM

designedtheexperiments;PSconstructedtheapparatus;MNandRFranthestudy,and

MCthesubsidiarystudy;JLandDWconductedtheanalysesofgene-expression;MN,MP,

DW,MM-LandRGMMwrotethemanuscript.

CONFLICTSOFINTEREST

JL,DW,MPandTTareemployeesofDartNeuroscienceLLCwhichisinvolvedin

researchanddevelopmentassociatedwithcognitiveenhancement

FIGURELEGENDS

FIGURE1.The‘eventarena’apparatusandperformancedata.

(A)Thearenawas1.6mx1.6mandmadeofclearplexiglass.Itwasopentoextramaze

cuesinthelaboratory.Ithad4start-boxes,withentrydoors,anda7x7gridwhere

floorpanelscanbeplaced.Somefloorpanelshavecircularholesintowhichsandwells

couldbeplaced.Aceilingmountedvideocameraenabledobservationoftheanimals.

(B)Thecircularsandwellswereconstructedwithsemi-circularbowlswithholes(left).

Therewardfood(showninbrown)couldeitherbeinthebowl(accessible)orunderthe

bowl(inaccessible)andwerecoveredwithsandmixedwithmaskingflavor(middleand

right).(C)Habituationprotocol.Onthefirstday[session(-8)],thearenalocationslots

werefilledwithclosedpanels,andtheratshabituatedtothestart-box(1min)andto

thearena(10min).Duringthefollowinghabituationsessions[sessions(-7)to(-1)],one

sandwellwasopeninthearena(invariablelocations).Rewardpelletswereplacedon

topofthesandwells,butgraduallyburieddeepinthesandtoencouragedigging.The

ratswereallowedtoshuttlebetweenthestart-boxandsandwellstoretrievemorethan

onepellet.(D)Trainingschedules.TrainingsessionconsistsofST(greenshade)andCT

thatfollowed60-90minlater.Thelocationoftherewardedsamplesandwell(filled

circle)stayedintheCT.The2-sandwellsampleprotocolisalsoshownasexample.

Whenstableperformancewasachieved,PTsessionswereinsertedonceineverythree

29

sessions.Aprobetrial(PT)isoneinwhichthefiverewardpelletsinallsixsandwells

areinaccessible(opencircles).(E)Performanceoftheanimals(n=23)thatcompleted

allphasesoftraining.Noteseparatephases(colors),thesessionsonwhichprobetests

(PTs)werescheduled(gray),andthe5-sessioncontrolphase(S58-S62)inwhichno

sampletrial(ST)wasgivenonthethirdsession(S60,pinkshade).(F)Inter-observer

correlationofprobetestperformanceinasubsetof12ratstrainedinthetask.X-axis

showsscoressecuredbyMNinasemi-blindmannerduringthetestitself,whileY-axis

showslaterblindscoringbyanindependentobserver(RF).R=0.996,p<0.001).Means

±1S.E.M.SeealsoFigureS1andMovieS1.

FIGURE2.Delay-dependentforgettingandinter-observercorrelation.

(A)WiththeSTgivenasasinglesandwellsample(greenshadingovereventarena,left

panels),wemeasuredintheprobetest,aftervariousmemorydelays,thepercenttime

diggingatthecorrectsandwell(blackarrow)andthatattheother5sandwells.Forthe

two-sandwll(2SW)procedure(rightpanels),diggingwasmeasuredatall6sandwells,

butnowdistinguishedbetweenthecorrectsandwell,thenon-rewardedsandwellinthe

ST(grayarrow),andtheuntrained4sandwells.(B)Forthe1SWprocedure,therats

displayeddelay-dependentforgettingover24h.(C)WiththeSTgivenasa2SWsample

(oneofwhichwasrewarded),delay-dependentforgettingover24hwasalsoobserved,

buttheplotnowseparatesthecued(rewarded)andnon-cued(non-rewarded)

sandwellsoftheSTfromtheother4sandwells.Theaveragevariabilityincorrect

digginginBwas6.9%whereaswasonly3.6%inC.***:p<0.001;****:p<0.0001.

Means±1S.E.M.Chancelevel=16.7%.SeealsoMovieS2.

FIGURE3.Impactofmassedversusspacedaccesstorewardduringthesampletrial.

(A)Experimentaldesignonprobetestswith30sor10minintervalsbetweenthe3

rewardpelletsofthesinglesampletrial.(B)Probetestperformanceat2.4hand24h

asafunctionofmassedvs.spacedtraining.Notesustained24hmemoryforspaced

access.(C)PlotofPerformanceIndex(PI)acrossthe3separateentriestothearena

duringthesampletrial(ST).Theanimalswereatchanceforthe1stpellet(indicating

thatthesandwelldesignwassuccessfulinpreventingodorartifacts),butdeclinedfor

pellets2and3.Thedeclineshowedatrendtowardsafasterdeclineinerrorswithina

30

trialformassedaccess,despitelongtermretentionover24hbeingmoreeffectivefor

spacedaccess.***:p<0.001;****:p<0.0001.Means±1S.E.M.SeealsoMovieS3.

FIGURE4.Geneupregulationassociatedwithtaskperformance.

(A)Timelineofmassedorspacedsampletrainingandtissuecollection(left).Coronal

ratbraintoillustratelocationofRSCandHPC(right).Atissueblockof4.0mmAPwas

dissectedout,butonlybregma-3.80mmisshownforsimplicity.(B)(HPC)and(C)

(RSC):RT-qPCRanalysisfromHPCtissueshowedupregulationofIEGs(Arc,c-Fos,and

Npas4)aftereithermassedorspacedsampletrialswhencomparedtountrained

controls,whereastwohousekeeping(HPK)genes(ActbandTbp)remainedunchanged.

Gapdhwasusedastheloadingcontrolandthedatashownnormalizedtothecontrol

valueofeachgene.ForthethreeIEGs,therewassignificantlygreaterupregulationin

thespacedthanmassedrewardcondition.(D)(HPC)and(E)(RSC):NormalizedFPKM

valuesfromRNA-SeqanalysisofthesamesetsofIEGsandHPKs(includingGapdh).

NotethesimilartrendsasinBandC,respectively,forthismethodofgene-quantification.

(F)RNA-Seqanalysisofgeneinduction.Venndiagramshowingthenumberofgenes

significantlyinduced(p<0.0001cut-off)aftermassed(left)andspaced(middle)

training.ForgenesinducedinbothHPCandRSCinthemassed(29‘massed’genes)and

thespacedconditions(54‘spaced’genes),2geneswereinducedpreferentiallyby

massedtrainingonly,27genesbyspacedtrainingonly,and27genescalledforboth

massedandspaced(right).(G)Correlationofgene-inductionlevelsformassedvs

spacedconditions(HPC,56genes).Thetwogeneswithmoreexpressioninmassedthan

spacedarehighlightedingray,andcertaincommonIEGsofnoteamongstthe54‘spaced’

genesarenamedandhighlightedinred.Dottedlineindicatesy=xline.Greenlineisthe

linearregressionofall56datapoints.Thedeviationfromunityto1.33reflectsoverall

strongerinductionofgenesafterspacedtraining.(H)SameasGbutforRSCtissue.

FIGURE5.Thephosphodiesterase4inhibitorRolipramenhancesmemoryretention

(A)Experimentaldesignonprobetests24hlaterwithi.p.injectionofRolipram(blue)

at45minor3hafterthesinglesampletrial(ST)ofsinglepelletretrieval.(B)Probetest

performanceat24hmemoryforRolipram/vehicleinjectedgroups.Rolipraminjected

45minaftertheSTsuccessfullyenhancedthememoryat24h.**:p<0.01;Means±1

S.E.M.(Seethemaintextforstatistics.)

31

FIGURE6.Post-encodingenvironmentalnoveltyenhancesmemoryretention.

(A)Experimentaldesignonprobetests(PT)24minor24haftersampletrial(ST)of

eitherHighrewardorLowrewardwithorwithout5minofnoveltyexploration.(B)

Probetestperformance24minor24hafterSTwithorwithoutpost-encodingnovelty,

using3-pelletreward(Highreward).Blackfilledbarsarecued-location,light-greyare

non-cuedtrained;agreysurroundtothebarsdenotesnovelty.(C)Differencescoresof

relativesearchingattheCuedlocationvs.Non-cuedtrainedlocationateachmemory

delay(Highreward).Blackandgreystripingtodenotesubtraction.(D)Probetest

performance24minor24hafterSTwithorwithoutpost-encodingnovelty,using1-

pelletreward(Lowreward).Noteonlytransientmemoryat24min.(E)Difference

scoresofrelativesearchingattheCuedlocationminusthattotheNon-cuedtrained

locationateachmemorydelay(Lowreward).**:p<0.01;Means±1S.E.M.Seealso

FigureS2.

32

REFERENCES

Abraham, W. C. and M. F. Bear (1996). "Metaplasticity: The plasticity of synaptic plasticity." Trends in Neurosciences 19(4): 126-130.

Alberini, C. M. and E. R. Kandel (2015). "The regulation of transcription in memory consolidation." Cold Spring Harb Perspect Biol 7(1): a021741.

Barad, M., R. Bourtchouladze, D. G. Winder, H. Golan and E. Kandel (1998). "Rolipram, a type IV-specific phosphodiesterase inhibitor, facilitates the establishment of long-lasting long-term potentiation and improves memory." Proc Natl Acad Sci U S A 95(25): 15020-15025.

Barco, A., C. Pittenger and E. R. Kandel (2003). "CREB, memory enhancement and the treatment of memory disorders: promises, pitfalls and prospects." Expert Opin Ther Targets 7(1): 101-114.

Bast, T., B. M. da Silva and R. G. M. Morris (2005). "Distinct contributions of hippocampal NMDA and AMPA receptors to encoding and retrieval of one-trial place memory." J Neurosci 25(25): 5845-5856.

Bourtchouladze, R., R. Lidge, R. Catapano, J. Stanley, S. Gossweiler, D. Romashko, R. Scott and T. Tully (2003). "A mouse model of Rubinstein-Taybi syndrome: defective long-term memory is ameliorated by inhibitors of phosphodiesterase 4." Proc Natl Acad Sci U S A 100(18): 10518-10522.

Burgin, A. B., O. T. Magnusson, J. Singh, P. Witte, B. L. Staker, J. M. Bjornsson, M. Thorsteinsdottir, S. Hrafnsdottir, T. Hagen, A. S. Kiselyov, L. J. Stewart and M. E. Gurney (2010). "Design of phosphodiesterase 4D (PDE4D) allosteric modulators for enhancing cognition with improved safety." Nat Biotechnol 28(1): 63-70.

Chen, C. C., J. K. Wu, H. W. Lin, T. P. Pai, T. F. Fu, C. L. Wu, T. Tully and A. S. Chiang (2012). "Visualizing long-term memory formation in two neurons of the Drosophila brain." Science 335(6069): 678-685.

Commins, S., L. Cunningham, D. Harvey and D. Walsh (2003). "Massed but not spaced training impairs spatial memory." Behav Brain Res 139(1-2): 215-223.

da Silva, B. M., T. Bast and R. G. Morris (2014). "Spatial memory: behavioral determinants of persistence in the watermaze delayed matching-to-place task." Learn Mem 21(1): 28-36.

Davis, H. P. and L. R. Squire (1984). "Protein synthesis and memory: a review." Psychol Bull 96(3): 518-559.

Dudai, Y. and R. G. M. Morris (2001). To consolidate or not to consolidate: what are the questions? Brain, Perception and Memory: Advances in Cognitive Sciences. J. Bolhuis. Oxford, OUP: 147-162.

Eacott, M. and A. Easton (2007). "On familiarity and recall of events by rats. ." Hippocampus 17: 890-897.

Ebbinghaus, H. (1885). Memory: A contribution to experimental psychology. . New York, Dover Press.

Ennaceur, A. and J. Delacour (1988). "A new one-trial test for neurobiological studies of memory in rats. 1: Behavioral data." Behav Brain Res 31(1): 47-59.

Frankland, P. W. and B. Bontempi (2005). "The organization of recent and remote memories." Nat Rev Neurosci 6(2): 119-130.

Frey, U. and R. G. M. Morris (1997). "Synaptic tagging and long-term potentiation." Nature 385(6616): 533-536.

33

Frey, U. and R. G. M. Morris (1998a). "Weak before strong: Dissociating synaptic tagging and plasticity-factor accounts of late-LTP." Neuropharmacology 37(4-5): 545-552.

Genoux, D., U. Haditsch, M. Knobloch, A. Michalon, D. Storm and I. M. Mansuy (2002). "Protein phosphatase 1 is a molecular constraint on learning and memory." Nature 418(6901): 970-975.

Greene, R. L. (1989). "Spacing effects in memory: Evidence for a two-process account." Journal of Experimental Psychology: Learning, Memory and Cognition 15: 371-377.

Griffiths, D., A. Dickinson and N. Clayton (1999). "Episodic memory: what can animals remember about their past?" Trends in Cognitive Sciences 3: 74-80.

Guzowski, J. F., B. L. McNaughton, C. A. Barnes and P. F. Worley (1999). "Environment-specific expression of the immediate-early gene Arc in hippocampal neuronal ensembles." Nat Neurosci 2(12): 1120-1124.

Guzowski, J. F., B. L. McNaughton, C. A. Barnes and P. F. Worley (2001). "Imaging neural activity with temporal and cellular resolution using FISH." Curr Opin Neurobiol 11(5): 579-584.

Josselyn, S. A., C. Shi, W. A. Carlezon, Jr., R. L. Neve, E. J. Nestler and M. Davis (2001). "Long-term memory is facilitated by cAMP response element-binding protein overexpression in the amygdala." J Neurosci 21(7): 2404-2412.

Karpicke, J. D. and H. L. Roediger, 3rd (2008). "The critical importance of retrieval for learning." Science 319(5865): 966-968.

Kempadoo, K. A., E. V. Mosharov, S. J. Choi, D. Sulzer and E. R. Kandel (2016). "Dopamine release from the locus coeruleus to the dorsal hippocampus promotes spatial learning and memory." Proc Natl Acad Sci U S A 113(51): 14835-14840.

Kentros, C., E. Hargreaves, R. D. Hawkins, E. R. Kandel, M. Shapiro and R. V. Muller (1998). "Abolition of long-term stability of new hippocampal place cell maps by NMDA receptor blockade." Science 280(5372): 2121-2126.

Kogan, J. H., P. W. Frankland, J. A. Blendy, J. Coblentz, Z. Marowitz, G. Schutz and A. J. Silva (1997). "Spaced training induces normal long-term memory in CREB mutant mice." Curr Biol 7(1): 1-11.

Kramar, E. A., A. H. Babayan, C. F. Gavin, C. D. Cox, M. Jafari, C. M. Gall, G. Rumbaugh and G. Lynch (2012). "Synaptic evidence for the efficacy of spaced learning." Proc Natl Acad Sci U S A 109(13): 5121-5126.

Lanahan, A. and P. Worley (1998). "Immediate-early genes and synaptic function." Neurobiol Learn Mem 70(1-2): 37-43.

Lynch, G., L. C. Palmer and C. M. Gall (2011). "The likelihood of cognitive enhancement." Pharmacol Biochem Behav 99(2): 116-129.

Marr, D. (1971). "Simple memory: A theory for archicortex." Phil. Trans. Roy. Soc. Lond. B: 262: 23-81.

McGaugh, J. L. (2000). "Memory--a Century of Consolidation." Science 287(5451): 248-251.

McGeoch, J. A. (1932). "Forgetting and the law of disuse." Psychological Review 39: 352-370.

Mishkin, M. and J. Delacour (1975). "An analysis of short-term visual memory in the monkey." Journal of Experimental Psychology: Animal Behavior Processes 1: 326-334.

34

Moncada, D. and H. Viola (2007). "Induction of long-term memory by exposure to novelty requires protein synthesis: evidence for a behavioral tagging." J Neurosci 27(28): 7476-7481.

Montague, P. R. and T. J. Sejnowski (1994). "The predictive brain: temporal coincidence and temporal order in synaptic learning mechanisms." Learning and Memory 1: 1-33.

Morris, R. G. M. (2006). "Elements of a neurobiological theory of hippocampal function: the role of synaptic plasticity, synaptic tagging and schemas." Eur J Neurosci 23(11): 2829-2846.

Muller, G. E. and A. Pilzecker (1900). "Experimentelle beitrage zur lehre und gedachtniss." Zeitschrift fur Psychologie 1: 1-288.

Olton, D. S., J. T. Becker and G. E. Handelmann (1979). "Hippocampus, space, and memory." Brain and Behavioral Science 2: 313-365.

Pagani, M. R., K. Oishi, B. D. Gelb and Y. Zhong (2009). "The phosphatase SHP2 regulates the spacing effect for long-term memory induction." Cell 139(1): 186-198.

Percie du Sert, N., A. M. Holmes, R. Wallis and P. L. Andrews (2012). "Predicting the emetic liability of novel chemical entities: a comparative study." Br J Pharmacol 165(6): 1848-1867.

Peters, M., M. Bletsch, J. Stanley, D. Wheeler, R. Scott and T. Tully (2014). "The PDE4 inhibitor HT-0712 improves hippocampus-dependent memory in aged mice." Neuropsychopharmacology 39(13): 2938-2948.

Placais, P. Y., S. Trannoy, G. Isabel, Y. Aso, I. Siwanowicz, G. Belliart-Guerin, P. Vernier, S. Birman, H. Tanimoto and T. Preat (2012). "Slow oscillations in two pairs of dopaminergic neurons gate long-term memory formation in Drosophila." Nat Neurosci 15(4): 592-599.

Redondo, R. L. and R. G. M. Morris (2011). "Making memories last: the synaptic tagging and capture hypothesis." Nat Rev Neurosci 12(1): 17-30.

Roediger, H. L., 3rd and A. C. Butler (2011). "The critical role of retrieval practice in long-term retention." Trends Cogn Sci 15(1): 20-27.

Rutten, K., J. Prickaerts and A. Blokland (2006). "Rolipram reverses scopolamine-induced and time-dependent memory deficits in object recognition by different mechanisms of action." Neurobiol Learn Mem 85(2): 132-138.

Rutten, K., E. L. Van Donkelaar, L. Ferrington, A. Blokland, E. Bollen, H. W. Steinbusch, P. A. Kelly and J. H. Prickaerts (2009). "Phosphodiesterase inhibitors enhance object memory independent of cerebral blood flow and glucose utilization in rats." Neuropsychopharmacology 34(8): 1914-1925.

Schultz, W. and A. Dickinson (2000). "Neuronal coding of prediction errors." Annu Rev Neurosci 23: 473-500.

Sheng, M. and M. E. Greenberg (1990). "The regulation and function of c-fos and other immediate early genes in the nervous system." Neuron 4(4): 477-485.

Smith, C. C. and R. W. Greene (2012). "CNS dopamine transmission mediated by noradrenergic innervation." J Neurosci 32(18): 6072-6080.

Spear, N. E. and D. C. Riccio (1994). Memory: Phenomena and Principles. Massachusetts, Allyn and Bacon, Needham Heights.

Squire, L. R. (1992). "Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans." Psychol Rev 99(2): 195-231.

35

Steele, R. J. and R. G. M. Morris (1999). "Delay-dependent impairment of a matching-to-place task with chronic and intrahippocampal infusion of the NMDA-antagonist D-AP5." Hippocampus 9(2): 118-136.

Takeuchi, T., A. J. Duszkiewicz, A. Sonneborn, P. A. Spooner, M. Yamasaki, M. Watanabe, C. C. Smith, G. Fernandez, K. Deisseroth, R. W. Greene and R. G. M. Morris (2016). "Locus coeruleus and dopaminergic consolidation of everyday memory." Nature 537: 357-262.

Tse, D., R. F. Langston, M. Kakeyama, I. Bethus, P. A. Spooner, E. R. Wood, M. P. Witter and R. G. M. Morris (2007). "Schemas and memory consolidation." Science 316(5821): 76-82.

Tully, T., R. Bourtchouladze, R. Scott and J. Tallman (2003). "Targeting the CREB pathway for memory enhancers." Nat Rev Drug Discov 2(4): 267-277.

Tulving, E. (1983). Elements of Episodic Memory. New York, Oxford University Press.

Underwood, B. J. (1969). "Attributes of Memory." Psychological Review 76: 559-573.

van Kesteren, M. T., D. J. Ruiter, G. Fernandez and R. N. Henson (2012). "How schema and novelty augment memory formation." Trends Neurosci 35(4): 211-219.

Vann, S. D. and J. P. Aggleton (2002). "Extensive cytotoxic lesions of the rat retrosplenial cortex reveal consistent deficits on tasks that tax allocentric spatial memory." Behav Neurosci 116(1): 85-94.

Vann, S. D., J. P. Aggleton and E. A. Maguire (2009). "What does the retrosplenial cortex do?" Nat Rev Neurosci 10(11): 792-802.

Wang, S. H., R. L. Redondo and R. G. Morris (2010). "Relevance of synaptic tagging and capture to the persistence of long-term potentiation and everyday spatial memory." Proc Natl Acad Sci U S A 107(45): 19537-19542.

Whishaw, I. Q., B. Kolb and R. J. Sutherland (1983). The analysis of behavior in the laboratory rat. Behavioral approaches to brain research. T. E. Robinson. New York, Oxford University Press: 141-211.

Willshaw, D. J., P. Dayan and R. G. M. Morris (2015). "Memory, modelling and Marr: a commentary of Marr (1971) 'Simple memory: a theory of archicortex'." Phi. Trans. R. Soc. B. 370.

Wilson, B., J. Cockburn and A. Baddeley (1991). The Rivermead Behavioural Memory Test. Bury St. Edmunds, Thames Valley Test Company.

Wixted, J. T. (2004). "The psychology and neuroscience of forgetting." Annu Rev Psychol 55: 235-269.

Yin, J. C., M. Del Vecchio, H. Zhou and T. Tully (1995). "CREB as a memory modulator: induced expression of a dCREB2 activator isoform enhances long-term memory in Drosophila." Cell 81(1): 107-115.

Zhang, H. T. and J. M. O'Donnell (2000). "Effects of rolipram on scopolamine-induced impairment of working and reference memory in the radial-arm maze tests in rats." Psychopharmacology (Berl) 150(3): 311-316.

Per

form

ance

inde

x

0

20

40

60

80

100

1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

1SW 2SW 2SW, Massed / Spaced 2SW, Pharmacology

Nonaka et al. - Figure 1

Control

Session

A

E

Event arena

Training data

F Inter-observer correlation

0

5.0

10.0

0 5.0 10.0

MN

RF

(s)

(s)

non-cued control sessionProbe trial session

60 - 90 min Memory

Delay

duration : 60 sec

Session

Sample trial (ST) Choice trial (CT) Sample trial (ST) Probe test (PT)

Training session Probe test session

ST

CT

ST

CT

ST

CT

ST

CT

ST

CT

ST

CT

ST

CT

ST

CT

ST

CT

ST

PT

ST

CT

ST

CT

ST

CT

ST

CT

ST

PT

ST

PT

ST

PT

......

Training schedule

Habituation S(-7) - S(-1)

Stable performance stagepre-training

Habituation S(-8)C

D

20 cmFloor panels

BSandwell design Rewarded sandwellNon-rewarded

sandwell

Sandwells

Locationin event arena:

1 cm

10 cm

A Probe test sessions after 1 or 2SW-sample encoding

Memory Delay

% Cued % Non-cued % Cued % Non-cued % Cued % Non-cued

%di

g tim

e

24 h2.4 h

C 2SW-sample

B 1SW-sample

Nonaka et al. - Figure 2

24 min

**** N.S.

1SW 2SW

Memory Delay

ST PT ST PT

****

chance

0

20

40

60

80

100

%Cued %Non-cuedtrained

%Non-cueduntrained

24 h

%Cued %Non-cuedtrained

%Non-cueduntrained

2.4 h

%Cued %Non-cued trained

%Non-cueduntrained

*** %di

g tim

e

24 min

*** N.S.

0

20

40

60

80

100

2.4 h

Spaced

PT

24 h

C

0

0.5

1st 2nd 3rd

Err

or

Massed

Spaced

Reward

B

2.4 h 24 h 2.4 h 24 h

Massed Spaced

% d

ig ti

me

% Cued% Non-cued (trained)

Nonaka et al. - Figure 3

ST

10min 10min

PTA

30s 30s

2.4 hPT

24 h PT

Massed

PT

% Non-cued (untrained)

N.S.****

*******

0

20

40

60

80

Arc c-Fos Npas40

1

2

3

4ControlMassedSpaced

0

1

2

3

4

5

Retrosplenial cortex, qPCR (DDCt)

Gapdh0

1

2

3

4

Hippocampus, RNAseq (FPKM)

0

1

2

3

4

5

HKPIEG

1.0 2.00.0

0.5

1.0

1.5

2.0

2.5Hippocampus (log2 fold change)

Dusp1

Egr1

Egr2

Egr4

c-Fos

Hexim1

Hspb1

Junb

Nr4a1

Per1

Rbm4

Sik1

0.0 1.0 2.0 3.00.0

0.5

1.0

1.5

2.0

2.5

3.0

Arc

Dusp1

Egr1

Egr2

Egr4

c-Fos

Fosb

Hexim1

Hspa1a

JunbPer1

Rbm4

Sik1

y = 1.33x - 0.028R = 0.95642

B

16 5329

HPC RSC

Massed

59 8354

HPC RSC

Spaced

2 2727

Massed Spaced

HPC & RSC

D

F

G

C

E

H

A

gene induction (Massed) gene induction (Massed)

gene

indu

ctio

n (S

pace

d)

gene

indu

ctio

n (S

pace

d)

ControlMassedSpaced

mR

NA

(nor

m. t

o co

ntro

l)

Retrosplenial cortex (log2 fold change)

Retrosplenial cortex, RNAseq (FPKM)

ControlMassedSpaced

ControlMassedSpaced

Hippocampus, qPCR (DDCt)

y = 1.33x - 0.024R = 0.94912

1.5 2.5 0.5 1.5 2.50.0 0.5

Hspa1a

Actb Tbp Arc c-Fos Npas4 Actb Tbp

HKPIEG

HKPIEG

HKPIEGArc c-Fos Npas4 Actb Tbp GapdhArc c-Fos Npas4 Actb Tbp

RSC

HPC

Bregma -3.80mm

mR

NA

(nor

m. t

o co

ntro

l)

mR

NA

(nor

m. t

o co

ntro

l)

mR

NA

(nor

m. t

o co

ntro

l)

Nr4a1Hspb1

Fosb

Arc

Nonaka et al. - Figure 4

Massed

Spaced

45 min

Tissue collection

10 min 10 min

30s 30sProtocol

B

A

45 min

3 h

Rolipram / Vehicle

Vehicle Rolipram Vehicle Rolipram

i.p. 45 min i.p. 3 h

% Cued

% Non-cued (trained)

Rolipram / Vehicle

% d

ig ti

me

Nonaka et al. - Figure 5

ST PT

24 h

% Non-cued (untrained)

PT

0

10

20

30

40

50

60

70

24 min 24 h24 h + Novelty

24 min 24 h24 h + Novelty

High reward at encoding (3 pellets)

% d

ig ti

me

% d

ig ti

me

-20

-10

0

10

20

30

40

50

Cue

d - N

on-c

ued

train

ed (%

)

-20

-10

0

10

20

30

40

50

Cue

d - N

on-c

ued

train

ed (%

)

B C Difference score (High reward)

% Cued% Non-cued (trained)% Non-cued (untrained)

Low reward at encoding (1 pellet)D E Difference score (Low reward)

24 min

Low reward

PT

24 h

24 min PT

24 h PT

High reward

PT

Novelty

Novelty

1 pellet

24 h PT

24 h PT

3 pellets

24 h 24 h +Novelty

24 min

24 h 24 h+Novelty

24 min

30 min

30 min

A Nonaka et al. - Figure 6

ST

ST

**

0

20

40

60

80

100

0

20

40

60

80

1

Nonaka et al. - Supplemental Figures, Table, and Movie Legends

Figure S1 (related to Figure 1)

Training apparatus. (A) Photo of the event arena with a random 6 locations open onto

which 6 sandwells were placed. (B) Photo of the distal spatial cues (the patterns on the

walls and hanging cues) in the event arena room. (C) Photo of the sandwells. Rewarded

sandwell (left) has accessible pellet hidden under the sand (not shown in the picture),

and the non-rewarded sandwell (right) has inaccessible pellets below the porous plastic

bowls in the sandwell. (D) Latency to �irst dig at the correct sandwell in the choice trials

throughout the training sessions. Means ± 1 S.E.M.

Figure S2 (related to Figure 6)

Plot of performance index (PI) across 52 sessions of training with the second cohort of

animals used in the novelty study. After 11 sessions of triaining with a single sandwell

during the sample trial, they were switched to a 2 sandwell sample trial, and this used

during the later novelty phase. Note stability of above chance performance throughout

the probe test sessions shown with grey shading. See also Figure 1E. Means ± 1 S.E.M.

Table S1

Statistical information for RNAseq data.

Movie S1 (related to Figure 1)

Training sessions from ceiling camera view.

Two representative cases of the normal training session with 2-sandwell sample, 1-

pellet retrieval paradigm for sample trials (ST). In STs, the rats choose the two

sandwells randomly; in half the cases they �irst dig at the correct sandwell and never dig

the other sandwell, while in other cases they visit and dig the correct sandwell as the

2nd choice. Both cases are shown in this video.

Movie S2 (related to Figure 2)

Probe trial sessions of 2-sandwell protocol from ceiling camera view.

2

(same as above) and the PT that followed 24 h after, indicating the memory decay over

24 h. In the PTs, the six sandwells are identical, but the cued location is indicated with

red arrow and text.

Movie S3 (related to Figure 3)

Probe trial sessions of Massed vs Spaced from ceiling camera view.

Two representative cases of Probe trial (PT) sessions for comparing massed and spaced

access to reward. The �irst case shows the massed ST training (2-sandwell, 3 pellet

retrieval, 30 s inter-reward interval (IRI)) and the PT that followed 24 h later, and the

second case shows the spaced ST training (2-sandwell, 3 pellet retrieval, 10 min-IRI)

and the PT that followed 24 h later.

A

B

Nonaka et al. - Figure S1

Rewarded sandwell Non-rewarded sandwell

(accessible pellets) (inaccessible pellets)

C Sandwell

N E S W

Spatial cues

Event arena

Intramaze cues

0

60

120

1 5 10 15 30 35 40 45 50 55 60 65 70 75 80

Late

ncy

(s)

20 25Session

Probe trial sessionnon-cued control session

D

12

3

4

5

6

3

Nonaka et al. - Figure S2

1 5 10 15 30 35 40 45 5020 25

100

0

50

Per

form

ance

inde

x

Probe trial session

1SW 2SW 2SW, Novelty

Chance

Session

4

‘Massed’ vs ‘Spaced’ comparison in HPC: Best-fit values ± SE

Slope 1.333 ± 0.0397 Y-intercept -0.02377 ± 0.02279 X-intercept 0.01783 1/slope 0.7501 95% Confidence Intervals

Slope 1.254 to 1.413 Y-intercept -0.06947 to 0.02192 X-intercept -0.01729 to 0.04973 Goodness of Fit

R square 0.9543 Sy.x 0.09867 Is slope significantly non-zero?

F 1127 DFn, DFd 1, 54 P value <0.0001 Deviation from zero? Significant Equation Y = 1.333*X - 0.02377 ‘Massed’ vs ‘Spaced’ comparison in RSC: Best-fit values ± SE

Slope 1.333 ± 0.03644 Y-intercept -0.02788 ± 0.02664 X-intercept 0.02091 1/slope 0.7501 95% Confidence Intervals

Slope 1.26 to 1.406 Y-intercept -0.0813 to 0.02554 X-intercept -0.02002 to 0.05853 Goodness of Fit

R square 0.9612 Sy.x 0.1243 Is slope significantly non-zero?

F 1338 DFn, DFd 1, 54 P value <0.0001 Deviation from zero? Significant Equation Y = 1.333*X - 0.02788

Nonaka et al. - Table S1

5