1

ImpulsivechoicesinmicelackingimprintedNesp55Dent CLa, Humby Tb, Lewis Ka, Plagge Ac, Fischer-Colbrie Rd, Wilkins JFe,

WilkinsonLSa,b&IslesARa*

aBehavioural Genetics Group, MRC Centre for Neuropsychiatric Genetics and

Genomics,NeuroscienceandMentalHealthResearchInstitute,CardiffUniversity,

HadynEllisBuilding,MaindyRoad,Cardiff,UnitedKingdombBehavioural Genetics Group, School of Psychology, Cardiff University, Tower

Building,Cardiff,UnitedKingdomcDepartment of Cellular and Molecular Physiology, Institute of Translational

Medicine,UniversityofLiverpool,Liverpool,UnitedKingdomdDepartmentofPharmacology,InnsbruckMedicalUniversity,Innsbruck,AustriaeRoninInstitute,Montclair,NJ07043,USA

*Authorforcorrespondence:AnthonyR.Isles,E-mailIslesAR1@cardiff.ac.uk

2

ABSTRACT

Genomic imprinting is theprocesswherebygermlineepigenetic events lead to

parent-of-origin specificmonallelic expression of a number of keymammalian

genes.The imprintedgeneNespisexpressed fromthematernalalleleonlyand

encodes for Nesp55 protein. In the brain Nesp55 is found predominately in

discrete areas of the hypothalamus andmidbrain. Previously, we have shown

thatlossofNesp55givesrisetoalterationsinnovelty-relatedbehavior.Herewe

extendthesefindingsanddemonstrate,usingtheNespm/+mousemodel,thatloss

ofNesp55 leads to impulsive choices asmeasuredby a delayed-reinforcement

task,wherebyNespm/+micewerelesswillingtowaitforadelayed,largerreward,

preferring instead to choosean immediate, smaller reward.Theseeffectswere

highlyspecificasperformanceinanothercomponentofimpulsivebehavior,the

abilitytostoparesponseoncestartedasassayedinthestop-signalreactiontime

task, was equivalent to controls. We also showed changes in the serotonin

system,akeyneurotransmitterpathwaymediatingimpulsivebehavior.First,we

demonstrated that Nesp55 is co-localisedwith serotonin and thenwent on to

showthatinmidbrainregionstherewerereductionsinmRNAexpressionofthe

serotoninspecificgenesTph2andSlc6a4,butnotthedopaminespecificgeneTh

inNespm/+mice;suggestinganalteredserotonergicsystemcouldcontribute, in

part,tothechangesinimpulsivebehavior.Thesedataprovideanovelmodeof

action for genomic imprinting in the brain and may have implications for

pathologicalconditionscharacterizedbymaladaptiveresponsecontrol.

3

INTRODUCTION

Genomicimprintingistheprocessbywhichsomegenesaremarkedinaparent

oforiginspecificmannerasaconsequenceofepigeneticeventsthattakeplacein

the mammalian germ line (Ferguson-Smith, 2011). For canonical imprinted

genes, in the somatic cell lineages this epigeneticmarking leads tomonoallelic

expressionfromoneparentalalleleonly.Forsomeimprintedgenes,expression

is solely from thematernal allele, whilst others are solely expressed from the

paternalallele.

There are approximately 150 canonical imprinted genes andnon-codingRNAs

known in the mouse, with similar numbers in humans, although recent next-

generationsequencingexperimentshaveexpanded thisnumber (Wilkinsetal.,

2016). Though relatively small in number, imprinted genes are critical for

normaldevelopmenttotakeplace(Mcgrath&Solter,1984,Suranietal.,1984).

Furthermore, functionally, imprinted genes converge on specific biological

processesthathaveprominentimportanceinmammals,suchasinuterogrowth,

metabolismandbehaviour(Peters,2014,Wilkinsetal.,2016).

The imprinted gene Nesp, encoding Nesp55, is part of the complex GNAS

imprintingclusterandisexpressedexclusivelyfromthematernalallele(Peters

etal., 1999).Nesp is expressed indiscrete areasof themidbrain, including the

Edinger-Westphalnucleus,dorsalRaphénucleus(DRN),andthelocuscoeruleus

(LC), and also in the hypothalamus (Plagge et al., 2005). Our previous work

demonstrated thatmice null formaternally expressedNesp55 show increased

activitywhenplacedinnovelenvironment,butareducedpropensitytoexplorea

novelenvironmentwhengiven thechoice (Plaggeetal.,2005).Alterednovelty

exploration may involve changes in the balance between self-control and

impulsive responding (Flagel et al., 2010, Stoffel & Cunningham, 2008).

Furthermore,Nespisexpressedinkeybrainstructures,namelytheDRNandLC,

thatareimplicatedinmediatingresponsecontrol(Bari&Robbins,2013)andthe

ability towait for a reward (Fonseca etal., 2015), respectively.Apriori, these

datasuggestpossiblelinksbetweenNesp55andimpulsivebehavior.

The term ‘impulsivity’, defined as actionwith limited forethought, is a natural

part of behavior. However, in certain individuals, impulsive behavior becomes

4

pathological and maladaptive and can manifest in psychiatric disease such as

schizophrenia and attention deficit/hyperactivity disorder (ADHD), as well as

addictivedisorderssuchaspathologicalgamblinganddrugaddiction.Agrowing

bodyofevidencesuggeststhatimpulsivityisnotunitary,andinsteadisamulti-

facetedconstructdissociableintermsofbehaviorandunderlyingneurobiology

(Bari & Robbins, 2013). There have been attempts to segregate impulsive

behaviorintotwoseparatecategories,impulsivechoiceandimpulsiveaction

(Broos et al., 2012, Dent & Isles, 2014b). Impulsive choice often manifests as

impulsive decisions resulting from a distorted evaluation of the delayed

consequences of behavior and an increased preference for smaller immediate

rewards over larger delayed rewards; on the other hand, impulsive action

reflectsthemotoricaspectofimpulsivity,andtypicallymanifestsinpoorability

to override a pre-potent response (Humby & Wilkinson, 2011). A number of

behavioral tasks exist to tease apart these discrete aspects of impulsivity

(Humby&Wilkinson,2011),manyofwhicharereadilyavailableinmice(Dent&

Isles,2014b).

Hereweusedadelayed-reinforcementtask(Islesetal.,2003)andastop-signal

reaction time (SSRT) task (Humby etal., 2013), to assay impulsive choice and

impulsive action, respectively, in mice null for maternally expressed Nesp55

(Nespm/+mice).WefoundthatNespm/+miceshowedamarkedincreaseinchoice

impulsivity,manifestindecisionstochooseanimmediate,smallerrewardovera

largerbutdelayedreward,butexhibitednosignificantdifferencesfromcontrol

mice (Nesp+/p) in impulsive action,where subjectshad to suppress anongoing

pre-potentmotorresponse.WealsodemonstratedthatNesp55co-localizeswith

5HT inmidbrain regions, and furtherdemonstrated that thebehavioral effects

observed in the Nespm/+ mice were accompanied by selective effects on the

expression of key serotonergic genes. The data are consistent with a role for

Nesp55 in mediating specific aspects of impulsive behavior, possibly via

interactionswithmidbrainserotonergicpathways.

5

MATERIALSANDMETHODS

Animals

AllprocedureswereconductedinaccordancewiththerequirementsoftheUK

Animals(ScientificProcedures)Act1986,undertheremitofHomeOfficelicence

number30/2673withadditionalethicalapprovalatCardiffUniversity.

As described previously (Plagge et al., 2005), expression of Nesp55 was

eliminatedbyadeletionof26bpatthestartcodonoftheORF(bp2362to2387

in AJ251761). Mice carrying a maternally derived null allele (Nespm/+) lacked

Nesp55 expression, whereas those carrying a paternally derived null allele

Nesp+/p have the same level of expression as wild-type mice (Figure S1). As

before (Plagge et al., 2005), in the present study Nespm/+ were compared to

Nesp+/p mice as themost appropriate control (see Supporting Information for

moredetails)andthesearereferredtoascontrolmicethroughout.

TheNesp55nulllinewasmaintainedonaC57Bl/6J(CharlesRiverLaboratories,

U.K.) background with separate cohorts of Nesp55 and control mice being

utilizedinthetwobehavioraltasks.Subjectsweremalemiceagedbetween4-10

monthsduringtesting;thelargeagerangebeingduetothetimetakentolearn

and perform the operant tasks. Standard laboratory chow was available ad

libitum,but justpriortoandduringtheexperiment,waterwasrestrictedto2h

access per day. This regime maintained the subjects at ≈90% of free-feeding

bodyweight.Allbehaviouraltestingwasperformedinthelightphase(lightson

07:00,for12-hours).

Behavioraltasks

Apparatus

Thedelayed-reinforcementtaskandstop-signalreactiontime(SSRT)taskwere

performedin9-holeoperantchambers(CambridgeCognitionLtd,UK)modified

foruse inmice,aspreviouslydescribed(Humbyetal.,2013, Islesetal.,2003).

For the delayed-reinforcement task holes 3, 5 and 7were open,whereas only

holes4and6wereopenfortheSSRTtask.Themicewerepresentedwithvisual

stimuli (lights) recessed into the holes and were trained to respond to this

6

stimulus with a nose-poke recorded by infra-red beams spanning the hole.

Rewardwaspresentedinarecessedcompartmentconcealedbyapanel,onthe

wall opposite to thenose-poke/stimulus array; animalswere required topush

the panel open in order to consume the reward. The control of stimuli and

recording of the responses weremanaged by an Acorn Archimedes computer

withadditionalinterfacingbyARACHNID(CambridgeCognitionLtd,UK).

Delayedreinforcementtask

Detailsoftheshapingproceduresandbasicaspectsofthedelayed-reinforcement

taskitselfcanbefoundelsewhere(Islesetal.,2003).Briefly,thetaskcomprised

ofthreesequentialblocksof12trials,witheachtrialconsistingofaninitialnose-

poketothecentrallylocatedstimulus,followedbyasecondnose-poketoeither

theleftorrightapertures.Trials1-4inanyblockwere‘forced’informationtrials,

wheretheinitialnose-pokeresultedinpresentationofonlyoneofthetwochoice

options.Thismeasurewasdesignedtoprovidethesubjectswithpriornoticeof

the extent of any delay associated with choosing the large reward. In the

remaining 8 trials of each block, designated as ‘choice’ trials, the initial centre

nose-pokeledtotheoptionofasecondnose-pokeresponsetoeithertheleftor

rightapertures. Oneresponseresulted inthedeliveryofa largereward(50μL

10%solutionofcondensedmilk;NestleLtd,UK),andtheotherinthedeliveryof

small reward (25μL 10% solution of condensed milk). The response

contingencies were kept constant for each mouse, but were counterbalanced

betweensubjects.Inblock1bothresponsesledtothedeliveryofrewardaftera

1s delay. In blocks 2 and 3 increasing delays were introduced between the

responseandthedeliveryofthelargereward(8s,16s,respectively)whereasthe

delaybetweenresponseanddeliveryofthesmallrewardwasfixedat1s. Asa

probe to test the effect of the delays on behaviour, sessions were conducted

wherethedelayassociatedwiththelargerewardwasfixedat1s,equivalentto

thatassociatedwiththesmallreward,throughoutallthreeblocksofthesession.

Thebiasinchoiceofthelargerrewardateachblock(wherebyalwayschoosing

thelargereward=1;neverchoosingthelargereward=0)wasthemainmeasure

usedtodetermineimpulsiveresponding.Additionalmeasurementsthatrelated

to general motoric competence and motivation within the task were also

7

monitored,includingthe‘Start’and‘Choice’latencies,thetimetakentoinitiatea

trialandthetimetakentomakeachoiceonceatrialwasinitiated,respectively.

Alsomeasuredwerethenumberof ‘Non-started’(no initial,centralnose-poke)

and ‘Omitted’ (no secondary, choice nose-poke, following central nose-poke

initiatingtrial)trials.

For the delayed reinforcement task experimental groups were Nespm/+=11,

controls=7.Animalsweretested6-daysperweek.

Stop-signalreactiontime(SSRT)task

Detailsof the shapingproceduresandbasic aspectsof themainSSRT task can

alsobefoundelsewhere(Humbyetal.,2013).TheSSRTtask itselfconsistedof

sessions of 100 trials, which involved both ‘go’ and ‘stop’ trials. Go trials

consisted of rapid double nose-pokes (a ‘go’ response) between two separate

stimulilocations,whichwererewardedwithreinforcement(22ul,10%solution

of condensed milk, Nestle Ltd, UK). 20% of trials were stop trials, pseudo-

randomlydistributedthroughouteachsession,whereastop-signal(65dbwhite

noisefor0.3s)waspresentedbetweenthefirstandsecondnose-pokeresponses.

The aim of the stop-signal was to inhibit (‘Stop’) themouse frommaking the

second (‘Go’) nose-poke, and thenwait for the reward. Failure to refrain from

makingthispre-potentresponsewaspunishedbytheabsenceofrewardand5

secondtimeout(chamberlighton).Atbaseline,thestop-signalswerepresented

concurrently with the initial nose-poke response. To maintain high levels of

performanceofbothgoandstopresponding,thegostimulusdurationandwait

periodtorewarddeliveryinastop-signaltrialweredeterminedindividuallyfor

each subject. To assess the ability to stop once an action had been initiated,

sessionswereimplementedinwhichtheonsetofthestop-signalwaspresented

atdifferentpositionswithintheindividualisedgoresponseofeachmouse.Thus,

thestop-signalwaspseudo-randomlypresented10,40,50,60and90%fromthe

onset of the go response of each subject, with the assumption that stopping

would be more difficult the closer the stop-signal presentation was to the

terminationofthegoresponse.

Theamountofcorrectstoppinginstop-signaltrialsandtheSSRTwerethemain

measures of impulsive responding in this task. The SSRT was calculated by

8

determiningthe50%stoppingabilityforeachsubjectfromtherangeofsessions

in which the stop-signal onset was varied from baseline (full details of this

calculationcanbefoundintheSupportingInformationand(Humbyetal.,2013).

The proportion of correct go-responses, and latency to respond, were also

assessed.Additionalmeasurementsthatrelatedtogeneralmotoriccompetence

andmotivationwithin the taskwerealsomonitored, including the “intitiation”

and“magazine”latencies,thetimetakentoinitiateatrialandthetimetakento

collect the reward.Alsomeasuredwas thenumberof trials completed for any

givensession.

For theSSRTtaskexperimentalgroupswereNespm/+=17,controls=12.Animals

wererun6-daysperweek.

Immunohistochemistry

Mice were terminally anaesthetised using an intraperitoneal injection of

pentobarbitone and then trans-cardially perfused using 4%PFA in PBS. Brains

weredissectedwholeandplacedin4%PFAovernight,thentransferredtoa30%

sucrose solution (in PBS) at 4˚c for 24 hours in order to dehydrate and

cryoprotectthem.Brainswerethenmountedontoamicrotomeplatformusing

anembeddingmatrixandallowedto freeze fully.Serial sectionsof40µmwere

slicedandputinto25-wellcontainerscontainingcryoprotectant(6sectionsper

well). The free-floating sections in cryoprotectant were stored at -20°C until

required.

Dual-labellingimmunofluorescenceofNespwith5HTwascarriedoutinNesp+/+

(i.e.wild-type) free-floating brain sections in order to localize the endogenous

protein. The Nesp55 primary antibody is a well-characterised anti-body

(generatedandobtainedfromtheReinerFischer-ColbrieLab).Specifically,itisa

rabbit anti-NESP55 polycolonal anti-body, recognizing the free terminal end

(GAIPIRRH) ofNESP55 (Ischia etal., 1997). Sectionswerewashed three times

for10mineachin0.1%PBSbeforebeingincubatedfor15minin0.3Mglycine

in 0.1% PBS at room temperature, to neutralise endogenous aldehyde groups.

Sectionswerewashedin0.1%PBSandthenincubatedatroomtemperaturefor

1hourin10%blockingsolution;0.5%BSA(BBInternational,Cardiff,UK),0.5%

9

TritonX-100(v/v,SigmaAldrich)in0.1%PBS.Sectionswerethentransferredto

a solution containing 1:1000 anti-NESP55 and 1:500 anti-serotonin (Abcam)

dilutedina1%blockingsolution;thiswasallowedtoincubatedovernightat4°C

whilst gently shaking. The next day sections were washed three times for 10

minutes in 0.1% PBS. The relevant fluorescent secondary antibodies (Alexa

Fluor;Lifetechnologies)werediluted1/1000in1%blockingsolution,sections

were incubated in this solution in the dark at room temperature for 2 hours,

whilstgentlyshaking.Sectionswerethenwashedin0.1%PBSasbefore(inthe

dark)andtransferredtopolysinecoatedslidesandallowedtodryover-nightin

a dark dust-free environment. The mounted slides were then dehydrated

throughaprocessofincubationinarisingconcentrationofalcohol,followedby

xylene, then cover-slipped and sealed using DPX (Raymond Lamb DPX), and

allowed todryover-night.Tocontrol fornon-specificbindingof thesecondary

antibodies, secondary-only negative controls were carried out alongside all

experiments.

Immunofluorescenceslideswereviewedand imagescapturedusinganupright

fluorescencemicroscope (LeicaDM5000B).Dual-labelled immunofluorescence

images were acquired through separate channels then subsequently merged

usingImageJ(Image>colour>mergechannels).

QuantitativePCR

RNA from macro-dissected brain regions was isolated using standard Trizol

methods. Equal amounts ofRNAwere reverse transcribedusingRNA to cDNA

EcoDry (double primed) premix strips (Takara Bio Europe, France). Gene

expressionwasassessedusingaRotorgene6000withaCAS1200automatedset

up (Corbett Research, U.K.). PCR reactions were carried out using custom

designed primers and Quantace SensiMix NoRef (Bioline). The genes assayed

were Th (For 5ʹ-AGGAGAGGGATGGAAATGCT-3ʹ; Rev

5ʹ-GCGCACAAAGTACTCCAGGT-3ʹ), Tph2 (For 5ʹ-CTGCTGTGCCAGAAGATCATCA-

3ʹ; Rev 5ʹ-TGCTGCTCTCTGTGGTGTCG-3ʹ) and Slc6a4 (For 5ʹ-

TTGTGCTCATCGTGGTCATC-3ʹ; Rev 5ʹ-GTGGCGTACTCCTCCAGCAG-3ʹ).

Additionally, three housekeeping genes were assessed in order to provide a

10

robust measurement of general gene activity. These Gapdh (For 5ʹ-

GAACATCATCCCTGCATCCA-3ʹ; Rev 5ʹ-CCAGTGAGCTTCCCGTTCA-3ʹ), Hprt (For

5ʹ-TTGCTCGAGATGTCATGAAGGA-3ʹ; Rev 5ʹ-AATGTAATCCAGCAGGTCAGCAA-3ʹ)

and β-actin (For 5ʹ-TCTGTGTGGATTGGTGGCTCTA-3ʹ; Rev

5ʹ-CTGCTTGCTGATCCACATCTG-3ʹ).ΔCtvaluesweregeneratedbynormalisingto

thegeometricmeanofthesethreehousekeepinggenes.Allindividualreactions

werecarriedoutintriplicate.RealtimeqPCRdatawasvisualisedusingtheΔΔCt

method(Livak&Schmittgen,2001).

Dataanalysisandstatistics

AllbehavioraldatawereanalysedusingSPSS20(SPSS,USA).Datawereassessed

fornormalityandthenanalysedbyStudent’st-testormixedANOVA,with

between-subjectsfactorsofGENOTYPE(Nespm/+vs.Control),andwithin-subject

factorsDELAY(1s,8s,16s,or1s,1s,1s),CHOICE(choiceoflargeorsmallreward

duringforcedtrialsofthedelayedreinforcementtask),andSTOP-SIGNAL

POSITION(positionofstop-signalrelativetoindividualizedGo-response).For

repeated-measuresanalyses,Mauchly’stestofsphericityofthecovariance

matrixwasapplied;significantviolationsfromtheassumptionofsphericitywere

subjecttotheHuynh–Feldtcorrectiontoallowmoreconservativecomparisons

throughadjustdegreesoffreedom.Duetothenon-parametricnatureofthe

qPCRdata,ΔCtvalueswereanalysedbyMann-WhitneyUtest.Allsignificance

testswereperformedatalphalevelof0.05.

RESULTS

Nespm/+micearemoreimpulsiveinthedelayedreinforcementtask

In thedelayed-reinforcement task, subjectsmadea choicebetweenreceivinga

smallfoodrewardafterashort(1s)delay,oralargerewardafteralongerdelay

(1,8or16s). Inthisway,theextenttowhichsubjectsmadeimpulsivechoices,

indexedbythepreference forchoosingthe immediatebutsmallrewardovera

delayedlargereward,wasdetermined.Increasingthedelaytothelargereward

within a session decreased the likelihood of choosing that reward for both

Nespm/+andcontrolmice(Figure1a,maineffectofDELAY,F2,32=3.84,P=0.032).

11

However,mutantandcontrolmiceshowedsignificantdifferences in theextent

to which they switched their preference to the small, less delayed reward in

responsetoincreaseddelaytothelargereward.Specifically,Nespm/+micechose

the small reward more readily than control mice (Figure 1a, main effect of

GENOTYPE, F1,16=6.74,P=0.019), indicative of increased impulsive responding.

To confirm the specificity of the contingency between reward value and the

delay in reward delivery, performance was also assessed where the cost

associatedwith the large rewardwas removed, thus thedelay associatedwith

the largeand small rewardswasequal (1s) for all trials. In these sessions, all

subjectsdemonstratedanequalhighpreferenceforthelargerewardthroughout

thesession(Figure1b,nomaineffectofDELAY,F2,32=0.478,P=0.624)andthere

were no differences between Nespm/+ and control mice (no main effect of

GENOTYPE,F1,16=0.010,P=0.927).

WealsoconfirmedthatthecontrastingpatternofchoicebehaviorintheNespm/+

mice were not due to any differences between the groups in terms of basic

learning andmotivation to carry out the task, asNespm/+animals acquired the

taskatthesamerateascontrolmice(numberofsessionstolastdayofbaseline,

grandmean=40.3±1.3SEM,t16=1.00,P=0.32).Additionally,variabilitybetween

Nespm/+ and control mice was not related to differences in experiencing the

information trial contingencies. Thus, in the ‘forced’ trials (inwhich no choice

wasavailable)bothNespm/+andcontrolmicemadeequalresponsestothelarge

andsmallreward-relatedstimuliatalldelays(interactionbetweenGENOTYPEx

DELAY x CHOICE, F2,32=1.85, P=0.43; data not shown). Finally, there was no

difference between Nespm/+ and control mice on general measures of task

performance (see Figure S2). As expected, Nespm/+ and control mice

demonstratedageneralmaineffectofDELAYonmanyofthesemeasures(Figure

S2a-d)as follows: increasedstart (F1.7,27.5=22.06,P

12

Nespm/+mice show no differences in impulsive responding in the stop-

signalreactiontimetask

TheSSRTtaskmeasurestheabilitytostopanactiononceinitiatedbypresenting

a stop-signal (in this case an auditory tone)during a rapid transitionbetween

two stimuli locations (the ‘go’ response). Correctly inhibiting the go response

earnedrewardina‘stop’trial,andthedifficultyofstoppingwasmanipulatedby

presenting the stop-signal at different times within the go response. Thus

stoppingwas easiestwhen the stop-signal onsetwas near the start andmore

difficultwhenat theendof thego response.Throughout the training stagesof

the SSRT task, all subjects showed equivalent behavior in learning the task.

Nespm/+ mice acquired the task at the same rate as control animals, as

demonstrated by the similar number of sessions taken to complete the task

(Nespm/+:44.3±6.6,control:46.6±6.8).Theeffectsofalteringthepositionofthe

auditory stop-signal during stop trials on stopping efficiency are illustrated in

Figure2a.InlinewithpreviousSSRTtaskexperimentsinhumansandavariety

of other mammalian species, including mice (Bari et al., 2009, Humby et al.,

2013),stoppingbecameincreasinglydifficultasthestop-signalpresentationwas

movedprogressivelyclosertotheexecutionoftheresponse(10,40,50,60and

90%intotheindividualisedgoresponse)leadingtosystematicreductionsinthe

abilitytostopforbothNespm/+andcontrolmice(Figure2a,maineffectofSTOP-

SIGNALPOSITION,F4,108=43.28,P

13

stop-signal position was moved from baseline (main effect of STOP-SIGNAL

POSITION,F4,108=0.49,P=0.74andF4,108=1.67,P=0.16,fortheamountandspeed

ofcorrectrespondingingotrials,respectively).

Again, Nespm/+ animals acquired the task at the same rate as control mice

(numberofsessions,grandmean=45.3±1.2SEM,t27=0.90,P=0.38).Additionally,

therewasnodifferencebetweenNespm/+andcontrolmiceongeneralmeasures

of task performance during individualized SSRT session). Thus, the number of

trialscompleted(t27=0.10,P=0.92), latencyto initiatea trial(t27=-0.36,P=0.73)

and magazine latency (t27=-1.12, P=0.27) measures were not significantly

different between Nespm/+ and control mice, but indicated a high degree of

stimuluscontrolforbothgroupsinthetask(seeFigureS3).

Nesp55co-localiseswith5-HT;keyserotoningenesshowreducedexpression

inmid-brainofNespm/+mice

Previouswork(Baueretal.,1999,Plaggeetal.,2005),which isreplicatedhere

(Figure S1 d-f), shows thatNesp55 is strongly expressed in the hypothalamus

andareasofthemidbrain,includingtheEdinger-Westphalnucleus,DRN,andthe

LC. Using immunohistochemistry, we examined whether Nesp55 co-localises

with 5-hydroxytryptamine (5HT) in these regions. Qualitative observation

suggeststhatNesp55isindeedco-localisedwith5HTinthemidbrain(Figure3

a-c).Thereappearstobeverylittle,ornoco-localisationofNesp55and5HTin

the hypothalamus (Figure 3 d-f). Additionally, it is clear that Nesp55 is also

expressednon-5HTexpressingcell-types.

To investigatetherelationshipbetweenNesp55and5HT, theexpression levels

ofsome5HT-relatedgeneswereassessedbyqPCRinsamplesofthemidbrainof

Nespm/+andcontrolmice(Figure4).Thisbrainregionwaschosenasitwasthe

areashowingthegreatestco-localisationbetweenNesp55and5HT.Therewere

10-foldreductionsintheexpressionofgenesencodingtryptophanhydroxylase,

Tph2(U=0.01,P=0.021)andtheserotonintransporter,Slc6a4(U=0.01,P=0.021)

inNespm/+micerelativetocontrolsubjects,butnodifferenceintheexpressionof

Th(U=7.0,P=0.886),thegeneencodingtyrosinehydroxylase.

14

DISCUSSION

The imprinted gene Nesp, expressed from the maternal allele only, encodes

Nesp55,andislocatedindiscreteareasofthemidbrainandhypothalamus.Here

wedemonstratethatlossofNesp55expressionfromthematernalallele,butnot

paternalallele, leadstohigher levelsof impulsivechoicebehaviorasmeasured

byadelayedreinforcementtask.TheNespm/+micewerelesswillingtowaitfora

larger but delayed reward, choosing instead the smaller but immediately

availablereward.TheseeffectswerehighlyspecificinthatbehaviorintheSSRT

task, assaying a different form of behavioral control, the ability to stop the

execution of a pre-potent action, was equivalent betweenNespm/+ and control

mice. Inaddition,weshowthatNesp55isco-localisedwith5HTandthatthere

are concomitant reductions in midbrain Tph2 and Slc6a4, but not ThmRNA

expression in Nespm/+ mice, suggesting an altered serotonergic system could

contribute,inpart,tothechangesinimpulsivechoicebehavior.

We have previously shown that Nespm/+ mice have an increased reactivity to

novelty,butareducedpropensitytoexploreanovelenvironmentwhengivena

choice(Plaggeetal.,2005).Alterednoveltyexplorationmayinvolvechangesin

thebalancebetween self-control and impulsive responding (Flageletal., 2010,

Stoffel&Cunningham,2008).Wedecidedtoexploretheimpulsivityphenotype

ofNespm/+mice further by using operant tasks to examine discrete aspects of

impulse control, namely choice impulsivity, as measured on a delayed

reinforcementtask,andimpulsiveaction,asmeasuredontheSSRTtask(Dent&

Isles, 2014b). In the delayed reinforcement task,Nespm/+ mice showed higher

levels of impulsivity, switching their choice of the larger, but increasingly

delayed, reward, more quickly than control mice. Critically though, when the

delayassociatedwith the largerrewardwasequivalent to thatassociatedwith

the smaller reward throughout the session, there was no difference in

performance between Nespm/+ and control mice. Taken together with other

control measures within the delayed-reinforcement task, which indicated that

therewerenodifferencesbetweenNespm/+ and controlmice in their ability to

learnthetaskortheirgeneralmotivationtoperformthetask,thismanipulation

demonstrates that the prime factor controlling the difference in behavior

15

betweenNespm/+ and controlmice on the delayed-reinforcement task,was the

contingency between delay and reward. Conversely, behavior of Nespm/+ and

controlmice on the SSRT task, was equivalent. All mice showed the expected

reductionstoppingasthestop-signalpresentationmovedprogressivelycloserto

theexecutionofthegoresponse,buttherewerenogenotype-relateddifferences

acrossthetaskorinthederivedSSRTmeasure.

Overall, thesedatasuggestthat lossofmaternalNesp55expressionproducesa

specific increase in impulsive choicebehavior,whilst impulsiveaction remains

unaffected.Asaconcept,thedissociationofchoiceandactionimpulsivityisnow

wellestablished.Forinstance,inbredstrainsofmicehavebeenshowntoexhibit

dissociations in terms of impulsive behaviors in a delayed reinforcement task

(Isles et al., 2004), and a measure of action impulsivity in the 5-choice serial

reaction time task (Patel et al., 2006). Furthermore, a cross-species study

examiningbehavior inanumberofdifferent tasks found that, inboth ratsand

humans, measures of impulsive choice and impulsive action did not correlate

(Broos etal., 2012). Our data add to the growing body of evidence suggesting

thatimpulsivityisnotunitary,andinsteadisamulti-facetedconstruct.

Impulsivechoiceandimpulsiveactionarenotonlydissociablebehaviorally,but

alsointermsoftheunderlyingneurobiology(Bari&Robbins,2013).Oneneural

pathwaythathas importantdissociableeffectsondiscreteaspectsof impulsive

behavioristheserotoninsystem(Barietal.,2009,Humbyetal.,2013,Talposet

al., 2006, Winstanley et al., 2004). Although excluded from the key areas of

cognitive control important for mediating impulsive behavior, such as the

frontostriatalcircuitry,Nesp55isstronglyexpressedinanumberofserotonergic

areasofthemidbrain(Baueretal.,1999,Plaggeetal.,2004),andhereweshow

that in some cells Nesp55 and 5HT are co-localised. This is particularly

interesting given the recently established role of serotonergic dorsal raphé

neurons in mediating waiting for delayed reward (Miyazaki et al., 2012a,

Miyazakietal.,2012b), independentfromanyeffectsonreward(Fonsecaetal.,

2015). One mechanism underlying the specificity of impulsive behavior

phenotypeseeninNespm/+micecouldbeviaanimbalanceinserotoninsignaling

intheDRN.

16

QuantitativePCRanalysisrevealedthatinmicelackingNesp55therewasa10-

fold reduction in midbrain expression of Tph2, which encodes tryptophan

hydroxylasetheratelimitingenzymeforbrain5HTsynthesis(Fitzpatrick,1999),

and the serotonin transporter gene, Slc6a4. Expression of Th (tyrosine

hydroxylase) was unaltered, consistent with a specific effect on the serotonin

system. Nesp55 is associated with fast anterograde axonal transport in the

peripheral nervous system and is considered a marker for the constitutive

secretory pathway (Fischer-Colbrie et al., 2002, Li et al., 2002). It is unknown

whetherNesp55influencesthetransportorreleaseofneurotransmittervesicles.

We have previously found no alteration in monoamine levels in brain tissue

taken from the prefrontal cortex and midbrain ofNespm/+ mice (Plagge et al.,

2005). However, thesewhole tissue analysesmaymask the impact of Nesp55

deficiencyonneurobiological indicesmoreclosely related to synaptic function,

suchasextracellularlevelsoftransmittersandchangesinpre-andpostsynaptic

receptor moieties, an idea maintained by the correlative changes in Slc6a4

expression. Moreover, given the known interplay between the serotonin and

dopaminesystemsininfluencingimpulsivebehaviour(Dalley&Roiser,2012),it

is not possible to rule out the action of othermonoamines inNespm/+mice as

beingcausalof thedelayed-reinforcementphenotype.Therefore,atpresentthe

exact neural mechanism(s) by which Nesp55 influences behavior remains

unclear.

Our data provide a novelmode of action for genomic imprinting in the brain,

addingtothelistofbehaviorssensitivetoimprintedgenefunction(Daviesetal.,

2015). Imprinted genes are thought to have evolved as a consequence of

conflictingphenotypic “interests”betweenmaternal andpaternal genes,which

causesanescalatingarmsraceinrelationtoallelicexpression,eventuallyleading

tosilencingofoneorotherparentalallele(Moore&Haig,1991,Wilkins&Haig,

2003). Inadditiontotheallelic level,parentalconflictmayalsobemanifestat

thegeneandtissuelevels(Daviesetal.,2005,Wilkinsetal.,2016).Evidenceof

thisparentalconflictisseenintheoppositeactionofimprintedgenesduringin

utero growth and early post-natal life (Haig, 2004) but as yet there are no

behaviouralexamples.Ourfindingsmayaddsomesupporttothetheorythat,as

aconsequenceofintra-genomicconflict,imprintedgenesexpressedintheadult

17

brain could lead to a “parliamentof themind”with regard todecision-making

(Haig, 1997) in that there are opposing parental interests pulling impulsive

choice in different directions. However, whilst here we show that loss of a

maternallyexpressedgene,Nesp,causesanincreaseinimpulsivechoices,there

is, currently at least, no paternally expressed imprinted gene that has been

shown to influence impulsive responding. A potential candidate here isGrb10,

which shows overlapping expression with Nesp (Dent & Isles, 2014a), and

influencessocialdominance(Garfieldetal.,2011),abehavioraldomainthathas

beenlinkedtoimpulsivity(Davisetal.,2009).Finally,anextensionofthisidea

andourexperimental findings is that, an imbalance inexpressionof imprinted

genes,suchasNesp,maynotonlycontributetofacetsof impulsivitywithinthe

normal range, but also to pathological conditions, such as gambling and drug

addiction, where response control becomes maladaptive. It is therefore, of

particularinterestthataNESPdeletionpolymorphismhasbeenassociatedwith

theoccurrenceofneuropsychiatricillness,includingaddictivebehaviour(Kimet

al.,2000).

REFERENCESBari,A.,Eagle,D.M.,Mar,A.C.,Robinson,E.S.&Robbins,T.W.(2009)Dissociable

effectsofnoradrenaline,dopamine,andserotoninuptakeblockadeonstoptaskperformanceinrats.Psychopharmacology,205,273-283.

Bari,A.&Robbins,T.W.(2013)Inhibitionandimpulsivity:behavioralandneuralbasisofresponsecontrol.Progressinneurobiology,108,44-79.

Bauer,R.,Ischia,R.,Marksteiner,J.,Kapeller,I.&Fischer-Colbrie,R.(1999)Localizationofneuroendocrinesecretoryprotein55messengerRNAintheratbrain.Neuroscience,91,685-694.

Broos,N.,Schmaal,L.,Wiskerke,J.,Kostelijk,L.,Lam,T.,Stoop,N.,Weierink,L.,Ham,J.,deGeus,E.J.,Schoffelmeer,A.N.,vandenBrink,W.,Veltman,D.J.,deVries,T.J.,Pattij,T.&Goudriaan,A.E.(2012)Therelationshipbetweenimpulsivechoiceandimpulsiveaction:across-speciestranslationalstudy.PloSone,7,e36781.

Dalley,J.W.&Roiser,J.P.(2012)Dopamine,serotoninandimpulsivity.Neuroscience,215,42-58.

Davies,J.R.,Dent,C.L.,McNamara,G.I.&Isles,A.R.(2015)Behaviouraleffectsofimprintedgenes.CurrentOpinioninBehavioralSciences,2,28-33.

Davies,W.,Isles,A.R.&Wilkinson,L.S.(2005)Imprintedgeneexpressioninthebrain.Neuroscienceandbiobehavioralreviews,29,421-430.

18

Davis,J.F.,Krause,E.G.,Melhorn,S.J.,Sakai,R.R.&Benoit,S.C.(2009)Dominantratsarenaturalrisktakersanddisplayincreasedmotivationforfoodreward.Neuroscience,162,23-30.

Dent,C.L.&Isles,A.R.(2014a)Brain-expressedimprintedgenesandadultbehaviour:theexampleofNespandGrb10.Mammaliangenome:officialjournaloftheInternationalMammalianGenomeSociety,25,87-93.

Dent,C.L.&Isles,A.R.(2014b)AnOverviewofMeasuringImpulsiveBehaviourinMice.CurrentProtocolsinMouseBiology,4,35-45.

Ferguson-Smith,A.C.(2011)Genomicimprinting:theemergenceofanepigeneticparadigm.Naturereviews.Genetics,12,565-575.

Fischer-Colbrie,R.,Eder,S.,Lovisetti-Scamihorn,P.,Becker,A.&Laslop,A.(2002)Neuroendocrinesecretoryprotein55:anovelmarkerfortheconstitutivesecretorypathway.AnnNYAcadSci,971,317-322.

Fitzpatrick,P.F.(1999)Tetrahydropterin-dependentaminoacidhydroxylases.AnnuRevBiochem,68,355-381.

Flagel,S.B.,Robinson,T.E.,Clark,J.J.,Clinton,S.M.,Watson,S.J.,Seeman,P.,Phillips,P.E.&Akil,H.(2010)Ananimalmodelofgeneticvulnerabilitytobehavioraldisinhibitionandresponsivenesstoreward-relatedcues:implicationsforaddiction.Neuropsychopharmacology:officialpublicationoftheAmericanCollegeofNeuropsychopharmacology,35,388-400.

Fonseca,M.S.,Murakami,M.&Mainen,Z.F.(2015)Activationofdorsalrapheserotonergicneuronspromoteswaitingbutisnotreinforcing.Currentbiology:CB,25,306-315.

Garfield,A.S.,Cowley,M.,Smith,F.M.,Moorwood,K.,Stewart-Cox,J.E.,Gilroy,K.,Baker,S.,Xia,J.,Dalley,J.W.,Hurst,L.D.,Wilkinson,L.S.,Isles,A.R.&Ward,A.(2011)DistinctphysiologicalandbehaviouralfunctionsforparentalallelesofimprintedGrb10.Nature,469,534-538.

Haig,D.(1997)Parentalantagonism,relatednessasymmetries,andgenomicimprinting.ProcRSocLondBBiolSci,264,1657-1662.

Haig,D.(2004)Genomicimprintingandkinship:howgoodistheevidence?AnnuRevGenet,38,553-585.

Humby,T.,Eddy,J.B.,Good,M.A.,Reichelt,A.C.&Wilkinson,L.S.(2013)Anoveltranslationalassayofresponseinhibitionandimpulsivity:effectsofprefrontalcortexlesions,drugsusedinADHD,andserotonin2Creceptorantagonism.Neuropsychopharmacology:officialpublicationoftheAmericanCollegeofNeuropsychopharmacology,38,2150-2159.

Humby,T.&Wilkinson,L.S.(2011)Assayingdissociableelementsofbehaviouralinhibitionandimpulsivity:translationalutilityofanimalmodels.Currentopinioninpharmacology,11,534-539.

Ischia,R.,Lovisetti-Scamihorn,P.,Hogue-Angeletti,R.,Wolkersdorfer,M.,Winkler,H.&Fischer-Colbrie,R.(1997)MolecularcloningandcharacterizationofNESP55,anovelchromogranin-likeprecursorofapeptidewith5-HT1Breceptorantagonistactivity.TheJournalofbiologicalchemistry,272,11657-11662.

Isles,A.R.,Humby,T.,Walters,E.&Wilkinson,L.S.(2004)Commongeneticeffectsonvariationinimpulsivityandactivityinmice.TheJournalofneuroscience:theofficialjournaloftheSocietyforNeuroscience,24,6733-6740.

19

Isles,A.R.,Humby,T.&Wilkinson,L.S.(2003)Measuringimpulsivityinmiceusinganoveloperantdelayedreinforcementtask:effectsofbehaviouralmanipulationsandd-amphetamine.Psychopharmacology,170,376-382.

Kim,S.J.,Gonen,D.,Hanna,G.L.,Leventhal,B.L.&Cook,E.H.,Jr.(2000)DeletionpolymorphisminthecodingregionofthehumanNESP55alternativetranscriptofGNAS1.MolCellProbes,14,191-194.

Li,J.Y.,Lovisetti-Scamihorn,P.,Fischer-Colbrie,R.,Winkler,H.,Dahlstrom,A.,Departmentof,A.&CellBiology,U.o.G.S.j.l.a.g.s.(2002)Distributionandintraneuronaltraffickingofanovelmemberofthechromograninfamily,NESP55,intheratperipheralnervoussystem.Neuroscience,110,731-745.

Livak,K.J.&Schmittgen,T.D.(2001)Analysisofrelativegeneexpressiondatausingreal-timequantitativePCRandthe2(-DeltaDeltaC(T))Method.Methods,25,402-408.

McGrath,J.&Solter,D.(1984)Completionofmouseembryogenesisrequiresboththematernalandpaternalgenomes.Cell,37,179-183.

Miyazaki,K.,Miyazaki,K.W.&Doya,K.(2012a)Theroleofserotoninintheregulationofpatienceandimpulsivity.MolNeurobiol,45,213-224.

Miyazaki,K.W.,Miyazaki,K.&Doya,K.(2012b)Activationofdorsalrapheserotoninneuronsisnecessaryforwaitingfordelayedrewards.TheJournalofneuroscience:theofficialjournaloftheSocietyforNeuroscience,32,10451-10457.

Moore,T.&Haig,D.(1991)Genomicimprintinginmammaliandevelopment-aparentaltug-of-war.TrendsinGenetics,7,45-49.

Patel,S.,Stolerman,I.P.,Asherson,P.&Sluyter,F.(2006)AttentionalperformanceofC57BL/6andDBA/2miceinthe5-choiceserialreactiontimetask.Behaviouralbrainresearch,170,197-203.

Peters,J.(2014)Theroleofgenomicimprintinginbiologyanddisease:anexpandingview.Naturereviews.Genetics,15,517-530.

Peters,J.,Wroe,S.F.,Wells,C.A.,Miller,H.J.,Bodle,D.,Beechey,C.V.,Williamson,C.M.&Kelsey,G.(1999)AclusterofoppositelyimprintedtranscriptsattheGnaslocusinthedistalimprintingregionofmousechromosome2.ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica,96,3830-3835.

Plagge,A.,Gordon,E.,Dean,W.,Boiani,R.,Cinti,S.,Peters,J.&Kelsey,G.(2004)TheimprintedsignalingproteinXLalphasisrequiredforpostnataladaptationtofeeding.Naturegenetics,36,818-826.

Plagge,A.,Isles,A.R.,Gordon,E.,Humby,T.,Dean,W.,Gritsch,S.,Fischer-Colbrie,R.,Wilkinson,L.S.&Kelsey,G.(2005)Imprintednesp55influencesbehavioralreactivitytonovelenvironments.MolCellBiol,25,3019-3026.

Stoffel,E.C.&Cunningham,K.A.(2008)Therelationshipbetweenthelocomotorresponsetoanovelenvironmentandbehavioraldisinhibitioninrats.Drugandalcoholdependence,92,69-78.

Surani,M.A.,Barton,S.C.&Norris,M.L.(1984)Developmentofreconstitutedmouseeggssuggestsimprintingofthegenomeduringgametogenesis.Nature,308,548-550.

Talpos,J.C.,Wilkinson,L.S.&Robbins,T.W.(2006)Acomparisonofmultiple5-HTreceptorsintwotasksmeasuringimpulsivity.JPsychopharmacol,20,47-58.

20

Wilkins,J.F.&Haig,D.(2003)Whatgoodisgenomicimprinting:thefunctionofparent-specificgeneexpression.Naturereviews.Genetics,4,359-368.

Wilkins,J.F.,Ubeda,F.&VanCleve,J.(2016)Theevolvinglandscapeofimprintedgenesinhumansandmice:Conflictamongalleles,genes,tissues,andkin.Bioessays,38,482-489.

Winstanley,C.A.,Dalley,J.W.,Theobald,D.E.&Robbins,T.W.(2004)Fractionatingimpulsivity:contrastingeffectsofcentral5-HTdepletionondifferentmeasuresofimpulsivebehavior.Neuropsychopharmacology:officialpublicationoftheAmericanCollegeofNeuropsychopharmacology,29,1331-1343.

ACKNOWLEDGEMENTS

CLD,KLandARIperformedtheexperiments;TH,JFW,LSWandARIdesignedthe

study;APandRF-Csuppliedreagents;CLD,TH,LSWandARIwrote thepaper.

Thisworkwas funded by a Leverhulme Trust project grant (F/00 407/BF) to

ARI and JFW, which supported CLD. The work was also funded by the MRC

Centre forNeuropsychiatricGenetics andGenomics (G0801418),ofwhichARI,

THandLSWaremembers,andwhichalsosupportedKLviaaPhDstudentship.

Theauthorshavenoconflictsofinteresttodeclare.

21

Figure1.Nespm-/p+miceshowincreasedimpulsivechoicebehaviorinthe

delayed-reinforcementtask.BehaviorofbothNespm/+andcontrolmice

changedacrosssessionblocks(blk)withincreasingdelay,suchthatchoicebias

movedawayfromtheresponseleadingtothelargerewardtowardsthesmall

rewardwithincreasingdelay(a).However,thereweresystematicdifferences

betweenthegroupsintheirbehavior,suchthatNespm/+animals(N=11)

switchedtheirchoicetothesmall,lessdelayedrewardmorequicklythancontrol

mice(N=7).Whenthedelayassociatedwiththelargeandsmallrewardswas

equal(1s)throughoutthesession(b),choicebiaswasconsistentlyhigh(large

rewardchosenapproximately90%ofthetime).Moreover,underthese

conditionstherewerenodifferencesinchoicebiasbetweenNespm/+andcontrol

mice.Datashowsmean±SEMofthreeconsecutivestablesessions;*represents

P

22

Figure2.NodifferencebetweenNespm/+andcontrolmiceperformancein

thestop-signal(SSRT)task.BothcontrolandNespm/+miceshowedan

equivalentabilitytoperformtheSSRTtaskshowingtheexpectedchangein

percentagecorrectrespondingduringa‘stop’trial(a)asthepositionofthestop-

signalwasaltered,buttherewerenodifferencesbetweenNespm/+(N=17)and

controlmice(N=12).Nespm/+andcontrolmicealsoshowedequivalentSSRTsat

50%correctstopping(b).Therewerenogenotypedifferencesforthe‘Go’

responseforbothcohortsofmice,intermsofpercentagecorrectresponding(c)

orresponsespeed(d).Datashowsmean±SEM;**representsP

23

Figure3.Co-localisationofNesp55with5HTinthebrain.Sectionswere

dual-labelledwithanti-bodiesagainstNesp55and5HT,imageswerethen

mergedtogaugecellularco-localisation.Whitearrowsdepictareasofco-

localisation.Therewasevidenceofco-localisationintheregionsofthemidbrain

(a-c),forexamplethelocuscoeruleus.However,therewasverylittleco-

localisationobservedinthehypothalamus(d-e).Allimagesareatx40

magnification

24

Figure4.Altered5HT-relatedgeneexpressioninNespm/+mice.Gene

expressionanalysisofmidbrainmacro-dissectionsfromNespm/+andcontrol

mice(N=4forbothgroups)showedsignificant10-foldreductionsinTph2and

Slc6a4expression,butnodifferenceinThexpression.Datashowsmean±SEM;*

representsP

25

ImpulsivechoicesinmicelackingimprintedNesp55-

supplementarymaterial

Dent CLa, Humby Tb, Lewis Ka, Plagge Ac, Fischer-Colbrie Rd, Wilkinson LSa,b,

WilkinsJFf&IslesARa

aBehavioural Genetics Group, MRC Centre for Neuropsychiatric Genetics and

Genomics,NeuroscienceandMentalHealthResearchInstitute,CardiffUniversity,

HadynEllisBuilding,MaindyRoad,Cardiff,UnitedKingdom

bBehavioural Genetics Group, School of Psychology, Cardiff University, Tower

Building,Cardiff,UnitedKingdom

cDepartment of Cellular and Molecular Physiology, Institute of Translational

Medicine,UniversityofLiverpool,Liverpool,UnitedKingdom

dDepartmentofPharmacology,InnsbruckMedicalUniversity,Innsbruck,Austria

eRoninInstitute,Montclair,NJ07043,USA

Authorforcorrespondence:AnthonyR.Isles,E-mailIslesAR1@cardiff.ac.uk

SUPPLEMENTARYMATERIALSANDMETHODSAnimals

Loci closely linked to a targeted mutation may cause concern with regard to

phenotypicdifferencesbetweenknock-outsandcontrols,sincetheymostlikely

retaintheoriginal129inbredstrainEScellalleles,evenafterseveralgenerations

ofbackcrossingtoadifferentstrain.Foranalysisofmicecarryingahomozygous

mutation of a biallelically expressed gene, this may imply comparing an

associated homozygous 129/Sv genetic background with wild-type C57BL/6J

counterparts.However,theknockoutanalysisofmonoallelicallyexpressedgenes

providestheopportunityforamoredirectcomparisonofheterozygousgroups.

In the case of maternally expressed Nesp, heterozygous deficient mice

(Nespm/+)can be compared with heterozygous Nesp-expressing mice (Nesp+/p)

obtained fromreciprocalcrosses toC57BL/6J. Inbothof thesegroups, the loci

26

surrounding the targeted mutation will comprise compound C57BL/6J and

129/Svalleles,andthus,anypotentialstrainvariationsshouldmanifestequally

in both groups. In contrast, thewild-type littermates of the two heterozygous

groupsarehomozygousC57BL/6Jforthesegenomicregions,couldconfoundthe

interpretation of behavioral assays (Gerlai, 1996), particularly given known

differences in impulsive behavior between inbred strains (Isles et al., 2004).

Thus,weconsidered itmostappropriate tocompareNespm/+withNesp+/pmice

ascontrols.

Immunohistochemistry

For the ImmunohistochemistryanalysisaVectastain®EliteABCKit (PK-6100)

was used on brain sections fromNespm/+ (lacking maternal Nesp55), Nesp+/P

(control, carrying paternal knockout), and wild-type (Nesp+/+) mice. Sections

werefirstwashedinTBS(4x10mins).Theywerethenincubatedinaperoxidase

block on a stirrer for 30minutes (0.6%hydrogenPeroxidase inTBS) to block

endogenousperoxidase.SectionswerewashedinTBSagain(3x10mininTBS),

andthenincubatedinTBSTwith3%normalgoatserum(NGS)(S-1000,vector

labs)onastirrerfor30minutesatroomtemperature.FollowingNGSblocking,

sectionswereincubatedinprimaryantibody(anti-Nesp55)dilutedat1:1000in

TBSTwith3%NGS.Thiswasstirredfor10minutesandthencoveredandstored

overnightat4˚c.The followingdaythesectionswerewashed3x10minutes in

TBSTwith3%NGS. Sectionswere then incubated for1hour in the secondary

antibody diluted 1:200 in TBSTwith 3%NGS at room temperature. Following

thesecondaryantibodyincubationsectionswerewashed3x10minutesinTBST,

andthenallowedtoincubateintheABCcomplex(madeasperkitspecifications)

at room temperature, on a stirrer for 1 hour. Sections were then washed as

before(3x10minutesinTBST),thenwashed2x10minutesin0.05MTrisbuffer.

Sections were then incubated in DAB solution for 35 seconds at room

temperature; theywere then immediatelywashed in coldPBS inorder to stop

thereaction.FinallysectionswerethenwashedinTBSTfor2minutesandleftin

a new change of TBST solution overnight. The following day sections were

mountedon topolysinecoatedslides,andallowedtodryover-night.Themounted

slides were then dehydrated through a process of incubation in a rising

27

concentration of alcohol followed by xylene, then cover-slipped and sealed using

DPX (Raymond Lamb DPX), andallowed todryover-night.All experimentswere

carriedoutalongsidenegativecontrols.

CalculatingtheStopSignalReactionTime

Correct go reaction timeswere determined directly, however the stop-signal reaction

time(SSRT)hadtobederivedfromthedistributionofcorrectgoreactiontimesandthe

proportionofcorrectlystoppedtrials.SSRTsinthetaskwereestimatedemployingthe

standard procedure described in Logan et al. (1984), using data from where the

proportion of correct stop responses was ~50%. For each subject, data from the

sessionsinwhichthestop-signalpositionswerevariedrelativetotheindividualisedgo

reactiontime,wererankedbytheproportionofcorrectstopresponses,anddatafrom

sessionsinwhichthisvaluewasbetween40%and60%(i.e.50%±10%)wereaveraged.

The latency of stopping as defined by the SSRTwas derived from the distribution of

correct go reaction times and the proportion of correctly stopped trials as previously

described (Eagle & Robbins, 2003, Logan, 1994). Hence, for each of the sessions

determined above, the correct go reaction timeswere rank ordered from smallest to

largest and the nth value found, where n is the rank order position based on the

proportionoffailingtostopcorrectlyinstoptrialswascorrectedfortheoccurrenceof

omittedgotrials.Althoughomittedgotrialswerearareoccurrence,thiscorrectionwas

implemented based on the rationale that they could alter the observed inhibition

functionandaffectdeterminationofthenthcorrectgoreactiontimevalueandhencethe

final SSRT (Eagle & Robbins, 2003, Solanto et al., 2001, Tannock et al., 1989). To

determine the SSRT, the time the stop-signal was presented (i.e. ‘mean correct go

reaction time’ x ‘%meanstop-signalposition’)was subtracted from thenth correct go

reactiontimevalue.

SUPPLEMENTARYRESULTS

Initial validation of the Nesp55 anti-body was carried out by completing

immunohistochemistrystainingonbrainsectionsfromWT,Nespm/+andNesp+/p

adultmice(FigureS1).ThestainingrevealedthattheNespanti-bodysuccessfully

targetstheNesp55protein,showingcellspecificstaininginbothWTandNesp+/p

(wherebymaternalNespispresentinthebrain);inthehypothalamus,ponsand

mid-brain regions, consistent with the discrete regions previously described

28

(Plaggeetal.,2005);andshowingnostaining intheNespm/+ sections(whereby

maternalNesp55hasbeendeleted).

Additionalmeasuresinthedelayedreinforcementtask

Inadditiontochoice,anumberofothermeasuresofbasicbehaviorwithin the

delayed reinforcement task were obtained. These relate to general motoric

competence andmotivation within the task. As expected,Nespm/+ and control

micedemonstratedageneralmaineffectofDELAYonmanyof thesemeasures

(Figure S2) as follows: start (F1.7,27.5=22.06, P

29

Gerlai,R.(1996)Gene-targetingstudiesofmammalianbehavior:isitthemutationorthebackgroundgenotype?Trendsinneurosciences,19,177-181.

Isles,A.R.,Humby,T.,Walters,E.&Wilkinson,L.S.(2004)Commongeneticeffectsonvariationinimpulsivityandactivityinmice.TheJournalofneuroscience:theofficialjournaloftheSocietyforNeuroscience,24,6733-6740.

Logan,G.(1994)Ontheabilitytoinhibitthoughtandaction:Ausers'guidetothestopsignalparadigm.InDagenbach,D.&Carr,T.(eds),Inhibitoryprocessesinattention,memory,andlanguage..AcademicPress,SanDiego,pp.189-239.

Plagge,A.,Isles,A.R.,Gordon,E.,Humby,T.,Dean,W.,Gritsch,S.,Fischer-Colbrie,R.,Wilkinson,L.S.&Kelsey,G.(2005)ImprintedNesp55influencesbehavioralreactivitytonovelenvironments.Molecularandcellularbiology,25,3019-3026.

Solanto,M.V.,Abikoff,H.,Sonuga-Barke,E.,Schachar,R.,Logan,G.D.,Wigal,T.,Hechtman,L.,Hinshaw,S.&Turkel,E.(2001)TheecologicalvalidityofdelayaversionandresponseinhibitionasmeasuresofimpulsivityinAD/HD:asupplementtotheNIMHmultimodaltreatmentstudyofAD/HD.JAbnormChildPsychol,29,215-228.

Tannock,R.,Schachar,R.J.,Carr,R.P.,Chajczyk,D.&Logan,G.D.(1989)Effectsofmethylphenidateoninhibitorycontrolinhyperactivechildren.JAbnormChildPsychol,17,473-491.

30

FigureS1

FigureS1ImmunohistochemistrystainingforNespincoronalsectionsoftheadultmouse

brain.(a-c)ShowsstaininginWT,Nesp+/p(usedascontrols)andNespm/+adultmice;showingthe

presenceofdiscreteexpressionofNesp55 in thehypothalamus inbothWT(a)andNesp+/p(b)

mice,butnotNespm/+mice(c).(d-f)WTmousebrainprobedusinganti-Nesp55antibodyshows

distinct staining in the hypothalamus, specifically the DM, dorsomedial hypothalamic nucleus,

Arc, arcuatehypothalamicnucleus, and theLH, lateral hypothalamic area (d); in themidbrain,

specifically the EWn, Edinger-Westphal nucleus (e); and in the pons, specifically the LC, Locus

Coeruleus(f).

31

FigureS2

FigureS2AdditionalbehavioralmeasuresofperformanceofNespm/+andcontrol

mice in the delayed reinforcement task. a Start latency; b Choice latency; c Non-

startedtrials;dPanelpushes.Alldatarepresentmeans±s.e.m.

32

FigureS3

FigureS3AdditionalbehavioralmeasuresofperformanceofNespm/+andcontrol

miceintheSSRTtask.aNumberoftrialsinasessionwithindividualisedSSRT(when

thestopcueispresented50%intotheindividual’sGo-reactiontime); b Initiationand

magazinelatencyinasessionwithindividualisedSSRT.Alldatarepresentmeans±s.e.m.