To move or to sense? Incorporating somatosensory representation into striatal functions

(1)

HAL Id: hal-01923493

https://hal.archives-ouvertes.fr/hal-01923493

Submitted on 28 Jan 2019

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

To move or to sense? Incorporating somatosensory

representation into striatal functions

David Robbe

To cite this version:

David Robbe. To move or to sense? Incorporating somatosensory representation into striatal functions.

Current Opinion in Neurobiology, Elsevier, 2018, 52, pp.123 - 130. �10.1016/j.conb.2018.04.009�.

�hal-01923493�

(2)

To

move

or

to

sense?

Incorporating

somatosensory

representation

into

striatal

functions

David

Robbe

1,2,3

Along-standinghypothesispostulatesthatthestriatumis essentialfortheconcurrentselectionofadaptiveactionsand repressionofinappropriatealternatives.Here,classicaland recentanatomicalandphysiologicalstudiesarereviewedto showthat,inmammals,thestriatumcandetectdiscrete task-relevantsensorystimuliandcontinuouslytracksomatosensory informationassociatedwiththegenerationofsimple

movementsandmorecomplexactions.Ratherthan

contributingtotheimmediateselectionofactions,thestriatum maymonitorthesensorimotorstateofanimalsbyintegrating somatosensoryinformationandmotor-relatedsignalsona moment-by-momentbasis.Suchfunctioncouldbecriticalfor theprogressiveacquisitionorupdatingofadaptiveactionsand theemergenceofanembodiedsenseoftime.

Addresses

1_{Aix-Marseille}_University,_{De´partement}_de_Biologie,_Parc_Scientifique_de

Luminy,13273Marseille,France

2_{INSERM-Institut}_National_de_la_{Sante´ et}_de_la_Recherche_Me´dicale,

UMR1249,Marseille,ParcScientifiquedeLuminy,13273Marseille, France

3_{INMED-Institut}_de_{Neurobiologie}_de_la_{Me´diterrane´e,}_Parc_Scientifique

deLuminy,13273Marseille,France

Correspondingauthor:Robbe,David(davidrobbe@gmail.com)

CurrentOpinioninNeurobiology2018,52:123–130

ThisreviewcomesfromathemedissueonSystemsneuroscience

EditedbyMichaelLongandRosaCossart

https://doi.org/10.1016/j.conb.2018.04.009

0959-4388/_ã2018ElsevierLtd.Allrightsreserved.

Introduction

Thestriatumisthemainentrypointofthebasalganglia (BG)anditisgenerallyassumedthatthedorsalregionof striatum(DS)contributestomotorcontrol[1].Theexact natureofthiscontributionisdebated[2,3].Along stand-inghypothesisisthattheDSiscriticalforactionselection [4,5]. This hypothesis is mainly based on two striking featuresoftheanatomyandphysiologyofthemammalian BG. First,theoutputnucleiof theBGare mostly com-posed of GABAergic projection neurons that provide a tonic inhibition on the thalamocortical network and on brainstem motor regions [6–8]. Second, theGABAergic

striatalprojectionneurons(SPNs),whichrepresentmore than 95% ofthestriatalneurons, canbedivided intwo distinct classes, on the basis of the BG nuclei they innervate and of the type of dopamine receptor they express[9].Briefly,activationofso-calleddirectpathway SPN(dSPN)willdisinhibitthalamocorticalneuronsand subcortical motor regionstoultimately favormovement generationwhileactivationofso-calledindirectpathway SPN(iSPN)willreinforcetonicinhibitiononBGtargets and repress movements generation [10]. The opposite modulatory power of the direct and indirect-pathways SPN (d/iSPN)oncorticaland subcorticalmotor regions makes of the BGa potential system to select, through disinhibition[11],agivenactionfromasetofcompeting possibilities [4].Despite its popularity, there is still no direct satisfying evidence in support of this theory. In addition, it isnotclear how theorganization of theBG inputs,whichconstantlyprovidesensoryandmotor infor-mationfromtheentirebody,couldbeintegratedinthis theory.Theobjectiveofthisshortreviewistosuggestan alternativeframeworkinwhichanimportantfunctionof theDSisthecontinuousmonitoringofthesensorimotor stateof theanimal.

Inthisreview,thesensorimotorstateofananimalrefers to bothsomatosensoryinformationandefferencecopies of descending motor commands [12]. Somatosensory informationisderivedfromexternalsensorystimulation ofdifferentpartsofthebody(e.g.anairpuffdirectedon thewhiskersorthebackofananimal)andactive move-ments(e.g.locomotiontriggersrhythmicalsensory stim-ulation of the paws contacting the ground; propriocep-tion).Andbyactions,werefertoanorderedsequenceof movements (e.g. the act of turning left iscomposed of severalmovements[2]).

Somatotopic

organization

of

sensory

and

motor

corticostriatal

projections

DSSPN receiveexcitatoryprojectionsmainly fromthe entireneocortexandasetofthalamicnuclei.Earlytracing experimentsrevealed thatspecific regionsofthe cortex projecttospecific,oftennon-overlapping,regionsofthe striatum [13,14].Thetopography of cortico-striatal con-nections roughly respects the rostro-caudal and dorso-medial positions of the cortical regionsproviding these inputs. Forinstance,in rats,theprelimbicregionof the prefrontalcortexprojectstotheventro-frontalregionsof the striatum, the cingulate cortex projects toward the anterior dorsomedial regions of the striatum, and the barrel cortex projects to dorsolateral striatal regions

(3)

[15].Altogether,alargebodyofanatomicalexperiments innon-humanprimates,catsandrats,atdifferentlevelsof theBG,leadtotheimportantconceptthatsensorimotor, associative and limbic information are treated in the striatumthroughsegregatedparallelcircuitsthat consti-tutecortico-basalganglia-thalamo-corticalloops[16–18]. In the BG, one of the main channels of information processing isdedicated to thetreatmentofsensory and motorinformationcomingfromcortex.Itislocatedinthe dorsolateralregionofthestriatum(DLS)inrodentsand theputameninprimates.Thischannelissomatotopically organized[4,17].Forexample,theregionsofthecortices thattriggerarm movementsorrepresentsensory stimu-lation of the arm, project on a specific region of the striatum while cortical regions related to sensory and motorrepresentationofthefaceprojectonamoreventral region [19–23]. Functional evidence of a somatotopic organization of theDLS isalso suggestedfromaseries ofstudiesinwhichextracellularrecordingsofthespiking activityofDLSneuronswereperformedinawakefreely movingratswhileathoroughsomatosensoryexamination (cutaneoustouch,passivemanipulation)ofallaccessible bodyparts(head, vibrissae,paw,chest,chin, snout,ear, shoulder, cheek pad, and trunk) was performed. About halfoftheDLSneuronsincreasedsharplytheirfiringrate inresponsetotheselectivestimulationofagivenpartof thebodyand neuronsrespondingto agivenpartofthe body tended to be located close in space in relative agreementwith anatomicalpredictions[24–27].Finally, thesomatotopicorganizationoftheDLSwasalso appar-entinastudyontheabnormalprocessingofinformation bythe striatum,in whichlocal injections of aselective inhibitor of fast-spiking interneurons in specific striatal subregionsledtotheappearanceofepisodicresttremors ofspecific bodyparts[28].

Whenlooking atrodent stereotaxicatlases, itisstriking thattheDSisoneofthebiggestundividedregionsofthe brain.Itispossible thatthecortico-striatalconnectivity, despiteits topographicalorganization,doesnotallow to delineate clear boundaries between different somatic regions. Indeeditwas shown thatcorticostriatal projec-tions originating from the whisker-related motor and sensory cortices do converge in the DLS, but these projectionsalsodisplaysignificantdivergencethroughout theDS[29–31].Butmaybe thelackof DSsubdivisions findsitsorigininthedifficultytoquantitativelycompare, acrossanimalsandstudies,thepatternofcortico-striatal projections using heterogeneous retrograde or antero-gradetracersinjections.Tworecentstudiesinmicehave usedcomputationalneuroanatomicapproachesto quan-tifydatafromseveralhundreds of well-localizedsingle, doubleandtripleinjectionsofanterogradetracersinthe cortexofmice[32,33].Severalimportantconclusions couldbedrawnregardingtheorganizationof corticostria-talconnectionscarryingsomatosensoryinformation.First,

based on the provenance of the cortical input, it was possible to subdivide theDLS in five subregions func-tionallyrelatedtospecificbodyparts,namelythetrunk, lowerlimbs,upper limb,inner mouth and outer mouth (Figure1a).Second,projectionsfromthewhisker-related barrelcortexweredetectedin thesefivesubregionsand beyond. Third, when two cortical regions displayed strong reciprocal connections (e.g. forelimb motor and sensory cortices) theirstriatal projectionsstrongly over-lapped. Fourth, the delineation of somatotopic bound-aries in the DLS does not reflect step-like pattern of projections: a functionally homogenous cortical region canalsoprovideadiffusearborizationextendingoutside its main site of projection. Fifth, sensory and motor corticostriatalprojectionsarenotlimitedtotheclassically defined DLS but cover most of the DS, including so-called dorsomedial regions[32,33]. Altogether these studiesprovidedquantitativeevidenceinsupportofthe somatotopical organization of the sensory and motor corticalinputtothestriatum(withthenotableexception of whisker-related information) and revealed an unex-pectedprevalenceoftheseinputbeyondthetraditional dorsolateralregionstheywereclassicallyconfinedto.

Processing

of

somatosensory

stimuli

in

the

dorsolateral

striatum

Inregardoftheaforementionedwealthofanatomicaldata showingthatthestriatumisequippedtoprocess somato-sensory information, it is not well known what said processingconsistsof.Thisispartlyduetothechallenge of separatingsensory and motor components associated withmovementgenerationwhenworkingwithbehaving animals. To overcome this issue, several studies used anesthetized rodents and reported that striatal neurons respond to passive deflections of the whisker [34,35,36,37]. Noticeably, Reig and Silberberg [36] performed patch clamp recording of DLS projection neuronsinanesthetizedmiceandexaminedtheresponse oftheseneuronstowhiskerair-puffs,whichconsistedina depolarizationoftheirmembranepotential.Theauthors reported stronger and faster responses for contralateral stimulationcomparedtoipsilateralstimulation.Response amplitudeswerestrongerwhenwhiskerswerestimulated bilaterally,showingthatSPNintegratesensory informa-tioncomingfrombothsidesofthebody.Suchintegration was not present in the barrel cortex. While this study provided originalknowledge onthe integrative sensory capacityofthestriatum,theanesthetizedapproachmakes itdifficulttoconcludeonthepossiblebehavioralfunction ofsuchsensoryprocessing.Still,theintegrationof bilat-eralwhisker-relatedinformationbydSPNwasimpaired indopamine-depletedmice[38],aparticularly interest-ingresultinthecontextofthewell-knownsomatosensory abnormalities observed in Parkinsonian patients [39]. More recently, Sippy et al. [40] directly investigated theroleoftheDLSinhead-restrainedmicetrainedtolick arewardspoutinresponsetosinglewhiskerdeflections.

(4)

Theauthorsperformedpatchclamprecordingandfound thatwhisker-evokeddepolarizationsofSPNwere stron-ger during ‘hit’ trials (in which the mouse successfully lickedinresponsetothewhiskerdeflection)comparedto ‘miss’ trials. Whisker-evoked depolarization were com-posedof fastandslowcomponentsandthefast onewas onlyexpressedbydSPN.Briefoptogeneticstimulationof thedSPNevokedlicking.Altogether,thisworkshowsfor thefirsttimethatdSPNcontributetotheexpressionofa conditioned motor response by signaling a predictive sensory stimuli.

Somatosensory

responses

and

action-related

neuronal

representations

in

the

dorsal

striatum

If neuronal activity in the DLS can be modulated by somatosensory stimulation, either applied externally or resultingfrommovements,suchmodulationsmaybealso apparent during theperformanceof motor tasks.While

earlyelectrophysiologicalstudiesinrodentsandprimates hadreportedmodulationsoffiringoccurringduring move-ments(seealsoPanigrahietal.[41]),aseriesoflandmark articleshaveproposedthat,afterlearning,DLSneurons mainly encode the initiation and ending of prolonged actions,insupportofaroleofthestriatuminthegatingof action [42–45].Usingcell-typespecific calciumimaging infreelymovingmiceitwasalsoshownthatbothdSPN andiSPNincreasedtheiractivityjustpriortoleverpress [46]. Thiswork isnow takenas astrongevidence sup-porting a role of dSPN in action selection while iSPN repress unwantedactions. Still,it remains possible that actioninitiation-relatedandtermination-relatedactivities are primarily driven by the somatosensory dynamics occurring around action initiation and termination (e.g. whisker stimulation, postural/limb adjustments). This possibilityhasnotbeendirectlytestedbutseveralrecent works suggest that this interpretation should not be discarded. Coffey et al. [47] usedoptrodes to identify

Figure1

Somatotopicorganizationofthedorsolateralstriatumandelectrophysiologicalsingatureofsomatosensoryresponses.(a)Thedorsolateralstriatum canbesubdividedinfiveregionsaccordingtothefunctionaloriginofcorticalsensoryandmotorinputs(left,adaptedfrom[33]).Right,the differentregionsofthemousebodyarecolorscodedaccordingtothestriatalbodymap.(b)Schematicofthetreadmill(top).Spikingrasterplot ofaDLSneuronwithstrongrhythmicalmodulationoftheneuronalactivitycoordinatedwiththeoscillatorydynamicsofforelimbmovements duringtreadmilllocomotion(top).Greenlinesindicatetreadmillonset,bluelinegoaltime,orangearrowanddotsindicatebeginningandendof running(modifiedfrom[50]).(c)Spikingrasterplot(top)andperi-eventhistogram(bottom)ofaDLSneuron,alignedtoforepawfootfall,from[26].

(5)

dSPNandiSPNandfoundthatabouthalfoftherecorded neurons in both populations increased their firing rate sharplyinresponsetothepassivestimulationofagiven body part.In addition, itwas shownin micetrained to perform a head-movements task that the firing rate of putative SPN is perfectly correlated with head-move-mentsvelocity[48].Thisresultconfirmedprevious stud-iesin rats inwhich DLSneurons responding topassive head-movementswerealso correlated with head move-ments kinematics [49]. Similarly, in rats performing a learned running sequence on a treadmill, a fraction of theneuronsfiredsynchronouslywithmovementsofthe forelimb (Figure 1b) [50],in agreement with previous reportsofstriatalneuronssensitivetopassiveandactive forelimbmovements(Figure1c)[26,51].Klausetal.used calcium imaging to record the activity of ensemble of identified SPN in the DLSof mice exploring an open field.Theyfoundthatsimilarexploratoryactions(e.g.left turns) were associated with the activation of similar ensemblesof SPNsthatwerenotrandomlylocatedbut spatially close [52], which could be accounted for by somatosensoryresponsestothemovementsthatcompose each action. Using a similar experimental setup, the activity of ensemble of dSPN and iSPN was shown to correlate with animals velocity, on a very slow speed range (from 0 to 10cm, a logarithmic scale was use to revealthecorrelation)[53].Thisencodingof locomotion-related information (see also [54,55]) might be largely explained by an increased rate of somatosensory responses of striatal neurons whenanimals transitioned from resting to a variety of foraging and exploratory behaviors.

Somatosensory

responses

in

the

striatum

bring

experimental

challenges

The fact that a large fraction of neurons in the DS (including identified d/iSPN) responds to a variety of somatosensorystimulationin behavinganimalspresents seriouschallengeswheninterpretingaction-related neu-ronalpatterns ofactivity. While itis temptingto assign causalinterpretationtopatternsofactivityrecordedinthe striatumintermsofactiongeneration/selection/initiation/ chunking/learning,thepossibilitythatthesepatternsare ‘merely’causedbymovementsandtheirassociated sen-soryconsequencesisoftenoverlooked[42–46,52,53,56]. Somatosensoryresponsesinthestriatumarerelatedtothe entirebody,includingthetrunk,neck,paw,innermouth and whiskers, which are difficult to track in behaving animals. Thus,an additional potential confoundis that patternsof neuronalactivityin thestriatum maynotbe primarilyrelatedtoanactionofintereststudiedinatask (alever-press,nose-poke, about of locomotionor even licking)buttosecondarycovertmovementsandsensory stimulation(whiskers,bodypostures,headorjaw move-ments)thatprecede,co-occur,orfollowthisaction.These issues are especially important when investigating the neuralbasesofmotor learning,which isassociatedwith

reorganizationofsomatosensorydynamics.Toovercome thesedifficulties,apossibilityistousebehavioraldesigns allowing a fair comparison between actions performed before and after learning. This approach revealed that striatal population activity recorded in untrained and trained animals performing a similar action (run back andforthonthebeltofamotorizedtreadmill)displayed asimilaroverrepresentationofthebeginningandending ofthataction(Figure2)[50].Thus,therepresentationof these specific action phases is likely to be primarily accounted for by the somatosensory sensibility of the DLS,ratherthanhigherorderprocessessuchaslearning oractionselection/gating.Asecondexperimentalstrategy istocomplementimagingorelectrophysiological record-ingsduringspontaneousorconditionedbehaviorswithan exhaustiveand quantitativesomatosensory examination [24,47,27]. This is difficult to do in freely behaving animals. However, recording striatal activity while, for instance, an experimenter gently rotates the head of a micetowardtheleft,wouldallowtoinvestigatetowhat extenttherepresentationofaleftturnactionisaccounted forbypassivesomatosensory responses.

A

role

for

striatal

somatosensory

responses

in

motor

learning

Thehigh prevalenceof somatosensoryresponses inthe DSshouldnotjustbeseenasachallengebutalsoasakey totheunderstandingofthebehavioralfunction(s)ofthis brainregion.Theintegrationofsomatosensory represen-tation and efference copies of motor programs with rewardpredictionerrorsignalscouldcontributetomotor learningbylinkingagivensensorimotorstatetoanaction that favored the consumption of rewards or prevented unpleasant/dangeroussituations. This associative learn-ingprocesscouldoccurthroughbidirectional dopamine-mediated synaptic plasticity of the cortico-striatal con-nections[9]thatconveysomatosensory informationand efferencecopyofmotorprograms.Theresultingchanges insynaptictransmissionstrengthcouldcontributetothe formationofneuronalensembledistributedacross corti-calandsubcorticalregionssuchasprimaryandsecondary motorcortices,thestriatumandtheBGtargets(thalamus and midbrain/brainstem motor nuclei). Such general hypothesisisinagreement witharecentstudyshowing thatlearningacue-guidedmotorsequenceisassociated with the strengthening of the excitatory connection betweenmotor cortex and dorsolateral dSPN[57]. Itis alsosupportedbythefact thatduring performanceofa simplevoluntaryaction,thespikingactivityof dorsolat-eral SPN is comodulated by movement-related and reward-related information [58]. The functional rele-vanceof asensorimotor tracking function of the dorsal striatum for learning purpose can be illustrated in the context of time estimation tasks. Rodents naturally develop embodied strategies when challenged in such tasks, independently if the time interval to estimate is short [59,60] or long [50]:they progressively refine,by

(6)

trial-and-error,themovementcontentandkinematicsofa motor sequence, until it matches the time interval to estimate(whichistypicallyassociatedwithreward deliv-ery).Inoneoftheaforementionedtime-estimation stud-ies, a lesion of the primary motor cortex in naive rats preventedthelearningoftheembodiedstrategybutthe same lesion was without effect when performed after learning.Itwassuggestedthat,duringlearning,themotor cortex provided a ‘tutor’ signal to subcorticals motor regions[60],mostlikelytotheDLS[61].Inourgeneral hypothesis, efference copy of motor programs and somatosensory responses would tutor DS SPN. During learning,whencorrectorincorrectbehavioralsequences are performed,theoppositemodulatory powerof dopa-mine on dSPN and iSPN activity would facilitate and depressthecellularlinkbetweensensorimotorstatesand motorprograms,respectively.Insupportofthis hypothe-sis,whenmiceweregiventhepossibilitytoself-stimulate their own dSPN by touching a capacitive switch, they quickly repeatedsuchaction. Ontheoppositeasimilar self-stimulationofiSPNinducedavoidanceoftheswitch [62]. Moreover, it was recently shown that thespecific stimulation of dSPN or iSPN while mice performed a simple actionatcertain speedwas sufficienttoproduce

specificandsustainedincreasesordecreasesinthe selec-tionofthisspeed[63].Altogether,thesedifferentstudies support theideathat theDS,bycombining continuous sensorimotor state estimation and bidirectional dopa-mine-basedneuromodulation/plasticity,couldcontribute tolearning,atthelevelofactioncontentandkinematics. Such general function might apply to a wide range of behavioralcontext,thatisbeyondtrial-and-error embod-iedstrategyfor timeestimation.

Conclusion

and

outlook

Compelling evidence in support of prominent somato-sensory responses in the dorsal striatum have been reviewed,alongwiththeexperimentalchallenges associ-ated with such responses and their possible functional implicationintrial-and-errormotorlearning. Itis never-thelessimportanttohighlightthatDSneuronsarenotjust sensitivetospecificsomatosensorystimuli.Theycanalso integrate multimodal information. This has been ele-gantly demonstratedin single-cellrecordingsof SPNin anesthetized mice, in response to visual and whisker stimuli [36]. Early electrophysiological experiments in behaving non-human primates had also shown that DSneuronsrespondtomovementsorsensorystimuliina

Figure2

Over-representationofbeginningandendingofarunningsequenceinnaiveandwell-trainedrats.(a)Ratstrainedinatimeestimationtaskona motorizedtreadmilldevelopedastereotypedbackandforthrunningsequence(left).Hand-guidednaiveanimalscanperformasimilarsequence (right)[50].Blue(gray)tracesaretrajectoriesforcorrect(incorrect)trials.(b)Normalizedaveragefiringrates(sortedaccordingtothetaskphase ofthemaximumfiringrate)ofallthepositivelymodulatedneuronsrecordedinself-trained(left)andnaivehand-guidedanimals(right,modified from[70]).

(7)

context-sensitivemanner[4].Morerecently,duringtime estimation tasks, the activity of individual DS neurons wasreportedtobecomodulatedbytimeand movements-relatedaspect of taskperformance[50,64]. Ithasalso beenproposedthatthestriatumoutperformsthe prefron-talcortexinpredictingelapsedtimebecauseitintegrates incoming information from multiple cortical areas[65]. Two important questions need to be addressed in the futureregardingtheintegrativecapacityofthestriatum. First,howmuchof thisintegration isperformedbyDS neurons versus conveyed by external input? Indeed, whiletheDSreceivesinputsfromanumberof function-allydistinctbrainregions[32,33],theseinputsmight already convey integrated information, as illustrated in thevisualcortexwiththerobustmodulation of visually evoked responsesby locomotion[66].Second, the rela-tivecontributionofcorticalandthalamicinputtostriatal patterns of activityduring behaviorneeds to befurther specified.Whilethepredominanceofthecortical contri-bution to whisker-evoked striatal responses has been recentlydemonstrated[67],thereisalsoconverging evi-dencethatthalamicinputcontributetotheprocessingof somatosensory and motor information in the striatum [32,68]. It has been proposed that, to support motor learning, striatal neurons may integrate complementary information provided by the cortex and thalamus [69]. Progress in our understanding of the striatum function willrequiretotestthepredictionsofsuchprecisemodels [69] and to address theaforementioned gaps in knowl-edge,allthiswhilekeepinginmindthat,evenintheBG, modulations of patterns ofneuronalactivitythat follow movementsareasinterestingthanthoseprecedingthem.

Conflict

of

interest

statement

Nothingdeclared.

Acknowledgements

DRworkissupportedbyaEuropeanResearchCouncil (ERC-2013-CoG—615699_NeuroKinematics,D.R.).Theauthorwouldliketothank DrIngridBureauandMostafaSafaieforcriticalreadingofthemanuscript andpastandcurrentteammembersforstimulantdiscussions.

References

and

reading

Papersofparticularinterest,publishedwithintheperiodofreview, havebeenhighlightedas:

ofspecialinterest ofoutstandinginterest

1. RedgravePetal.:Goal-directedandhabitualcontrolinthe basalganglia:implicationsforParkinson’sdisease.NatRev Neurosci2010,11:760-772.

2. DudmanJT,KrakauerJW:Thebasalganglia:frommotor commandstothecontrolofvigor.CurrOpinNeurobiol2016, 37:158-166.

3. TurnerRS,DesmurgetM:Basalgangliacontributionstomotor control:avigoroustutor.CurrOpinNeurobiol2010,20:1-13.

4. MinkJW:Thebasalganglia:focusedselectionandinhibitionof competingmotorprograms.ProgNeurobiol1996,50:381-425.

5. RedgraveP,PrescottTJ,GurneyK:Thebasalganglia:a vertebratesolutiontotheselectionproblem? Neuroscience 1999,89:1009-1023.

6. TurnerRS,PasquereauB:In TheCognitiveNeuroscience.Edited byGazzanigaM,MangunG.TheMITPress;2014:452-467.

7. DeniauJ-M,MaillyP,MauriceN,CharpierS:Theparsreticulata ofthesubstantianigra:awindowtobasalgangliaoutput.Prog BrainRes2007,160:151-172.

8. HikosakaO,TakikawaY,KawagoeR:Roleofthebasalgangliain thecontrolofpurposivesaccadiceyemovements.PhysiolRev 2000,80:953-978.

9. KreitzerAC,MalenkaRC:Striatalplasticityandbasalganglia circuitfunction.Neuron2008,60:543-554.

10. KravitzAVetal.:Regulationofparkinsonianmotorbehaviours byoptogeneticcontrolofbasalgangliacircuitry.Nature2010, 466:622-626.

11. ChevalierG,DeniauJM:Disinhibitionasabasicprocessinthe expressionofstriatalfunctions.TrendsNeurosci1990, 13:277-280.

12. WolpertDM,GhahramaniZ,FlanaganJR:Perspectivesand problemsinmotorlearning.TrendsCognSci2001,5:487-494.

13. Ku¨nzleH:Anautoradiographicanalysisoftheefferent connectionsfrompremotorandadjacentprefrontalregions (Areas6and9)in_Macaca_{fascicularis.}BrainBehavEvol1978, 15:201-234.

14. Ku¨nzleH:Bilateralprojectionsfromprecentralmotorcortexto theputamenandotherpartsofthebasalganglia.An autoradiographicstudyinMacacafascicularis.BrainRes1975, 88:195-209.

15. McGeorgeAJ,FaullRL:Theorganizationoftheprojectionfrom thecerebralcortextothestriatumintherat.Neuroscience 1989,29:503-537.

16. ParentA,HazratiL-N:Functionalanatomyofthebasalganglia. I.Thecortico-basalganglia-thalamo-corticalloop.BrainRes Rev1995,20:91-127.

17. AlexanderGE,DeLongMR,StrickPL:Parallelorganizationof functionallysegregatedcircuitslinkingbasalgangliaand cortex.AnnuRevNeurosci1986,9:357-381.

18. HaberSN:Theprimatebasalganglia:parallelandintegrative networks.JChemNeuroanat2003,26:317-330.

19. FlahertyAW,GraybielAM:Motorandsomatosensory corticostriatalprojectionmagnificationsinthesquirrel monkey.JNeurophysiol1995,74:2638-2648.

20. FlahertyAW,GraybielAM:Input-outputorganizationofthe sensorimotorstriatuminthesquirrelmonkey.JNeurosci1994, 14:599-610.

21. FlahertyAW,GraybielAM:Corticostriataltransformationsinthe primatesomatosensorysystem.Projectionsfrom

physiologicallymappedbody-partrepresentations.J Neurophysiol1991,66:1249-1263.

22. VertesRP,HooverWB,Szigeti-buckK,LeranthC:Nucleus reuniensofthemidlinethalamus:linkbetweenthemedial prefrontalcortexandthehippocampus.BrainResBull2007, 71:601-609.

23. HooverJE,HofferZS,AllowayKD:Projectionsfromprimary somatosensorycortextotheneostriatum:theroleof somatotopiccontinuityincorticostriatalconvergence.J Neurophysiol2003,89:1576-1587.

24. CarelliRM,WestMO:Representationofthebodybysingle neuronsinthedorsolateralstriatumoftheawake, unrestrainedrat.JCompNeurol1991,309:231-249.

25. MittlerT,ChoJ,PeoplesLL,WestMO:Representationof thebodyinthelateralstriatumofthefreelymovingrat: singleneuronsrelatedtolicking.ExpBrainRes1994,98: 163-167.

(8)

26. WestMOetal.:Aregioninthedorsolateralstriatumoftherat exhibitingsingle-unitcorrelationswithspecificlocomotor limbmovements.JNeurophysiol1990,64:1233-1246.

27. ChoJ,WestMO:Distributionsofsingleneuronsrelatedto bodypartsinthelateralstriatumoftherat.BrainRes1997, 756:241-246.

28. OranY,Bar-GadI:Lossofbalancebetweenstriatal feedforwardinhibitionandcorticostriatalexcitationleadsto tremor.JNeurosci2018,38:2821-2917.

29. HofferZS,AllowayKD:Organizationofcorticostriatal projectionsfromthevibrissalrepresentationsintheprimary motorandsomatosensorycorticalareasofrodents.JComp Neurol2001,439:87-103.

30. AllowayKD,MuticJJ,HofferZS,HooverJE:Overlapping corticostriatalprojectionsfromtherodentvibrissal representationsinprimaryandsecondarysomatosensory cortex.JCompNeurol2000,426:51-67.

31. PanWX,MaoT,DudmanJT:Inputstothedorsalstriatumofthe mousereflecttheparallelcircuitarchitectureoftheforebrain. FrontNeuroanat2010,4:1-14.

32.

HunnicuttthestriatumBJetrevealsal.:Anovelcomprehensivefunctionalexcitatoryorganization.inputElifemap2016,of 5:e19103.

Seeannotationto[33].

33.

HintiryanNeurosciH2016,etal.:19:1100-1114.Themousecortico-striatalprojectome.Nat

Thesestudiesusedcomputationalneuroanatomicapproachesto quan-tify,inthestriatum,theanatomicallabelingresultingfromseveral hun-dredsofwell-localizedsingle,doubleandtripleinjectionsofanterograde tracersindifferentregionsofthecortexandthalamusofmice.Adetailed quantitativeinformationontheorganizationofexcitatoryinputsonthe striatumisprovided.Inregardofthepresentreview,theimportanceof sensorimotorinputtothestriatum,beyondtheclassicaldorsolaterallimit, is remarkable.These studies alsoprovided quantitative evidencefor convergence betweenfunctionallyrelatedsensoryandmotorcortical areasanddelineatedfunctionalterritoriesrelatedtodistinctbodyparts.

34. MoweryTM,HarroldJB,AllowayKD:Repeatedwhisker stimulationevokesinvariantneuronalresponsesinthe dorsolateralstriatumofanesthetizedrats:apotential correlateofsensorimotorhabits.JNeurophysiol2011, 105:2225-2238.

35. PidouxM,MahonS,DeniauJ-M,CharpierS:Integrationand propagationofsomatosensoryresponsesinthe

corticostriatalpathway:anintracellularstudyinvivo.JPhysiol 2011,589:263-281.

36.

Reigstriatum.R,SilberbergNeuron2014,G:Multisensory83:1200-1212.integrationinthemouse

The authors performed whole-cell recordings of identified SPN in responsestounilateralandbilateralwhiskerstimulationandfoundthat DLSSPN,butnotcorticalneurons,integratedbilateralwhisker stimula-tion.TheyalsoreportedthatDMSneuronscouldintegratemultimodal (visualandtactile)information.

37. NagyA,Eo¨rdeghG,Paro´czyZ,Ma´rkusZ,BenedekG: Multisensoryintegrationinthebasalganglia.EurJNeurosci 2006,24:917-924.

38.

KetzefprocessingMetal.:intheDopaminestriatumdepletioninapathway-dependentimpairsbilateralmanner.sensory Neuron2017,94855–865.e5.

Inthisfollow-upstudyofReigandSilberberg[36],theauthorsfoundthat bilateralintegrationofwhiskerstimuliisimpairedindopaminedepleted animalsandisrestoredfollowingL-Dopatreatment.

39. ConteA,KhanN,DefazioG,RothwellJC,BerardelliA: Pathophysiologyofsomatosensoryabnormalitiesin Parkinsondisease.NatRevNeurol2013,9:687-697.

40.

SippysensorimotorT,LaprayprocessingD,CrochetinS,striatalPetersenprojectionCCH:Cell-type-specificneuronsduring goal-directedbehavior.Neuron2015,88:298-305.

Theauthorstrainedmicetolickarewardspoutinresponsetosingle whisker deflectionsandperformedwhole-cellrecordingsand optoge-neticstimulationofidentifiedSPN.Theresultsobtainedsuggestthatthe rapidresponseofdirectpathwaySPNtowhiskerdeflectioniscriticalfor theexpressionofconditionallickingresponse.

41. PanigrahiBetal.:Dopamineisrequiredfortheneural representationandcontrolofmovementvigor.Cell2015, 162:1418-1430.

42. BarnesTD,KubotaY,HuD,JinDZ,GraybielAM:Activityof striatalneuronsreflectsdynamicencodingandrecodingof proceduralmemories.Nature2005,437:1158-1161.

43. JogMS,KubotaY,ConnollyCI,HillegaartV,GraybielAM: Buildingneuralrepresentationsofhabits.Science1999, 286:1745-1749.

44. JinX,CostaRM:Start/stopsignalsemergeinnigrostriatal circuitsduringsequencelearning.Nature2010,466:457-462.

45. JinX,TecuapetlaF,CostaRM:Basalgangliasubcircuits distinctivelyencodetheparsingandconcatenationofaction sequences.NatNeurosci2014,17:423-430.

46. CuiGetal.:Concurrentactivationofstriataldirectandindirect pathwaysduringactioninitiation.Nature2013,494:238-242.

47.

CoffeyprocessingKR,NaderintheM,striatalBawadirectJ,WestandMO:indirectHomogeneouspathways:single bodypartsensitivetypeIIbneuronsmayexpresseither dopaminereceptorD1orD2.EurJNeurosci2017,46: 2380-2391.

The authors used optrodes to d/iSPN neurons and recorded their responsestovarioustypesofstimulation(e.g.cutaneoustouch,passive manipulation, andactivemovements)ofmostoftheaccessiblebody parts(head,vibrissae,paw,chest,chin,snout,ear,shoulder,cheekpad, andtrunk).Theyfoundthatalargefractionofbothd/iSPNrespondedto passivestimulationofspecificpartofthebodyandthatneuronssensitive toagivenbodypartwereclosetoeachother.

48. KimN,BarterJW,SukharnikovaT,YinHH:Striatalfiringrate reflectsheadmovementvelocity.EurJNeurosci2014,40: 3481-3490.

49. TangC,PawlakAP,ProkopenkoV,WestMO:Changesinactivity ofthestriatumduringformationofamotorhabit.EurJ Neurosci2007,25:1212-1227.

50.

RuedacontextualOrozcoandPE,kinematicRobbeD:informationThestriatumtoconstrainmultiplexesmotor habitsexecution.NatNeurosci2015,18:453-460.

Theauthorsusedbehavioralparadigminwhichbothnaiveandtrained ratsperformedarunningsequencewithsimilarkinematics.IntheDLSof trainedanimals,tetroderecordedrevealedintegrativerepresentationsof runningspeed,positionandtime.Stillnaiveandtrainedratsdisplayed similarsequentialpopulationactivityduringrunssuggestingasignificant contributionofsomatosensoryresponsestothesesequentialpatterns.

51. ShiLH,LuoF,WoodwardDJ,ChangJY:Neuralresponsesin multiplebasalgangliaregionsduringspontaneousand treadmilllocomotiontasksinrats.ExpBrainRes2004,157: 303-314.

52. KlausAetal.:Thespatiotemporalorganizationofthestriatum articlethespatiotemporalorganizationofthestriatum encodesactionspace.Neuron2017,951171–1180.e7.

53. BarberaGetal.:Spatiallycompactneuralclustersinthedorsal striatumencodelocomotionrelevantinformation.Neuron 2016,92:202-213.

54. YaminHG,SternEA,CohenD:Parallelprocessingof environmentalrecognitionandlocomotioninthemouse striatum.JNeurosci2013,33:473-484.

55. CostaRM,CohenD,NicolelisMAL:Differentialcorticostriatal plasticityduringfastandslowmotorskilllearninginmice.Curr Biol2004,14:1124-1134.

56. MartirosN,BurgessAA,GraybielAM:Inverselyactivestriatal projectionneuronsandinterneuronsselectivelydelimituseful behavioralsequences.CurrBiol2018:560-573.

57. RothwellPEetal.:Input-andoutput-specificregulationof serialorderperformancebycorticostriatalcircuits.Neuron 2015,88:345-356.

58. IsomuraYetal.:Reward-modulatedmotorinformationin identifiedstriatumneurons.JNeurosci2013,33:10209-10220.

59. GouveˆaTSetal.:Ongoingbehaviorpredictsperceptualreport ofintervalduration.FrontNeurorobot2014,8:10.

(9)

60. KawaiRetal.:MotorCortexIsRequiredforLearningButNotfor ExecutingaMotorSkillArticleMotorCortexIsRequiredfor LearningButNotforExecutingaMotorSkill.2015:1-13http://dx. doi.org/10.1016/j.neuron.2015.03.024.

61. WolffSBE,DhawaleAK,KoR,O¨ lveczkyBP:Distinctrolesfor cortico-andthalamostriatalprojectionsinmotorskilllearning andexecution.SocietyforNeuroscienceAnnualMeeting690.10/ JJ122017.

62. KravitzAV,TyeLD,KreitzerAC:Distinctrolesfordirectand indirectpathwaystriatalneuronsinreinforcement.Nat Neurosci2012,15:816-818.

63.

YttrimovementEA,DudmanvelocityJT:inOpponentthebasalganglia.andbidirectionalNature2016:1-16controlhttp://of dx.doi.org/10.1038/nature17639.

The authors stimulated dSPN and iSPN while head-restrained mice pressedonabar.Thestimulationwasonlypreformedontrialsinitiated inagivenrangeofspeeds.StimulationofthedSPNbiasedthe subse-quentanimalchoicestowardthestimulatedspeedrange.Stimulationof theiSPNbiasedtheanimalspeedchoicesawayfromthestimulated speedrange.

64.

forMellotimeGBMM,intheSoaresstriatum.S,PatonCurrBiolJJ:A2015,scalable25:1113-1122.populationcode

TheauthorsrecordedspikingactivityintheDSwhileratsperformeda lever-pressintervaltimingtask.Theyfoundthatstriatalneuronsactivity integratedbothtask-relevanttimingandsensorimotorinformation.

65. BakhurinKIetal.:Differentialencodingoftimebyprefrontal andstriatalnetworkdynamics.JNeurosci2016,37:854-870.

66. NiellCM,StrykerMP:Modulationofvisualresponsesby behavioralstateinmousevisualcortex.Neuron2010,65: 472-479.

67. ReigR,SilberbergG:Distinctcorticostriatalandintracortical pathwaysmediatebilateralsensoryresponsesinthestriatum. CerebCortex2016,26:4405-4415.

68. AllowayKD,SmithJB,MoweryTM,WatsonGDR:Sensory processinginthedorsolateralstriatum:thecontributionof thalamostriatalpathways.FrontSystNeurosci2017,11:1-19.

69. FeeMS:Oculomotorlearningrevisited:amodelof

reinforcementlearninginthebasalgangliaincorporatingan efferencecopyofmotoractions.FrontNeuralCircuits2012, 6:38.

70. TaoualiW,Rueda-OrozcoPE,RobbeD:Aminority-ruled populationcodingofkinematicsinthestriatum.bioRxiv 2017:130237http://dx.doi.org/10.1101/130237.