ebook img

The Median Preoptic Nucleus Reciprocally Modulates Activity of Arousal-Related and Sleep ... PDF

15 Pages·2007·1.12 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview The Median Preoptic Nucleus Reciprocally Modulates Activity of Arousal-Related and Sleep ...

1616•TheJournalofNeuroscience,February14,2007•27(7):1616–1630 Behavioral/Systems/Cognitive The Median Preoptic Nucleus Reciprocally Modulates Activity of Arousal-Related and Sleep-Related Neurons in the Perifornical Lateral Hypothalamus NataliaSuntsova,1,2,4RubenGuzman-Marin,1,2SunilKumar,1,3Md.NoorAlam,1,2RonaldSzymusiak,1,3and DennisMcGinty1,2 1ResearchService,VeteransAffairsGreaterLosAngelesHealthcareSystem,NorthHills,California91343,Departmentsof2Psychologyand3Medicine, UniversityofCaliforniaLosAngeles,LosAngeles,California90095,and4A.B.KoganResearchInstituteforNeurocybernetics,RostovStateUniversity, Rostov-on-Don344091,Russia Theperifornical–lateralhypothalamicarea(PF/LH)containsneuronalgroupsplayinganimportantroleincontrolofwakingandsleep. Amongthebrainregionsthatregulatebehavioralstates,oneofthestrongestsourcesofprojectionstothePF/LHisthemedianpreoptic nucleus(MnPN)containingasleep-activeneuronalpopulation.ToevaluatetheroleofMnPNafferentsinthecontrolofPF/LHneuronal activity,westudiedtheresponsesofPF/LHcellstoelectricalstimulationorlocalchemicalmanipulationoftheMnPNinfreelymoving rats.Single-pulseelectricalstimulationevokedresponsesin79%ofrecordedPF/LHneurons.Nocellswereactivatedantidromically. Direct and indirect transynaptic effects depended on sleep–wake discharge pattern of PF/LH cells. The majority of arousal-related neurons,thatis,cellsdischargingatmaximalratesduringactivewaking(AW)orduringAWandrapideyemovement(REM)sleep, exhibitedexclusivelyorinitiallyinhibitoryresponsestostimulation.Sleep-relatedneurons,thecellswithelevateddischargeduring non-REMandREMsleeporselectivelyactiveinREMsleep,exhibitedexclusivelyorinitiallyexcitatoryresponses.ActivationoftheMnPN viamicrodialyticapplicationofL-glutamateorbicucullineresultedinreduceddischargeofarousal-relatedandinexcitationofsleep- relatedPF/LHneurons.DeactivationoftheMnPNwithmuscimolcausedoppositeeffects.TheresultsindicatethattheMnPNcontains subset(s)ofneurons,whichexertinhibitorycontroloverarousal-relatedandexcitatorycontroloversleep-relatedPF/LHneurons.We hypothesizethatMnPNsleep-activeneuronalgrouphasbothinhibitoryandexcitatoryoutputsthatparticipateintheinhibitorycontrol ofarousal-promotingPF/LHmechanisms. Keywords:perifornicallateralhypothalamus;medianpreopticnucleus;electricalstimulation;microdialysis;neuronalactivity;sleep– wakingcycle Introduction (QW)andnon-rapideyemovement(NREM)sleep(Estabrooke Theposteriorlateralhypothalamuscontainsseveralinteracting etal.,2001;Torteroloetal.,2001;Alametal.,2002;Koyamaetal., neuronalpopulationscriticallyimplicatedinthecontrolofwak- 2003;Leeetal.,2005;Mileykovskiyetal.,2005).Currentevidence ing,includinghistaminergic,glutamatergic,andpeptidergicneu- suggestsaroleofGABAergicinhibitioninthesleep-relatedinac- rons(Lin,2000;GerashchenkoandShiromani,2004).Peptider- tivation of arousal-related PF/LH neurons and involvement of gic cells producing hypocretins/orexins located within thismechanisminsleepcontrol.Itwasshownthatreversemicro- perifornical–lateralhypothalamic(PF/LH)areaarethoughttobe dialysis of the GABAA receptor antagonist bicuculline into the oneofthekeycomponentsofbrainarousalnetwork(deLecea PF/LHincreasedboththetimespentawakeandthenumberof andSutcliffe,2005;JonesandMuhlethaler,2005;Sakurai,2005; c-Fos-positive hypocretin-immunoreactive and nonhypocretin Saperetal.,2005). PF/LHcells(Alametal.,2005a).ApplicationoftheGABA re- A ThemajorityofPF/LHneurons,includinghypocretincells, ceptoragonistsintotheposterolateralhypothalamus(Linetal., exhibitincreasedneuronaldischargeandc-Fosimmunoreactiv- 1989;Sallanonetal.,1989)orthePF/LH(Thakkaretal.,2003) ity during active waking (AW) compared with quiet waking causedhypersomnia. RecentevidencesuggeststhatGABA-mediatedcontrolofthe PF/LHcouldoriginateinsleep-activeneuronslocatedwithinthe ReceivedMarch27,2006;revisedJan.8,2007;acceptedJan.9,2007. preoptic area/basal forebrain. GABAergic cells showing sleep- ThisworkwassupportedbytheMedicalResearchServiceoftheDepartmentofVeteransAffairs,NationalInsti- tutesofHealthGrantsMH63323,MH47480,HL60296,andNS-50939,andtheJ.ChristianGillinResearchGrantfrom relatedc-Fosimmunoreactivityarefoundwithintheventrolat- theSleepResearchSocietyFoundation(N.S.). eralpreopticarea(VLPO)(Sherinetal.,1996)andthemedian CorrespondenceshouldbeaddressedtoDennisMcGinty,ResearchService(151A3),VeteransAffairsGreaterLos preopticnucleus(MnPN)(Gongetal.,2004).Sleep-activeneu- Angeles,HealthcareSystem,16111PlummerStreet,NorthHills,CA91343.E-mail:[email protected]. ronslocatedwithinthesenucleiexhibitaspecificpatternofen- DOI:10.1523/JNEUROSCI.3498-06.2007 Copyright©2007SocietyforNeuroscience 0270-6474/07/271616-15$15.00/0 hanceddischargeduringbothNREMandrapideyemovement Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity J.Neurosci.,February14,2007•27(7):1616–1630•1617 (REM)sleep(Szymusiaketal.,1998;Suntsovaetal.,2002),which thenacclimatedtoexperimentalmilieufor3d.Experimentswereper- isoppositetothatdemonstratedbywake-activePF/LHneurons. formedonfreelymovinganimals.EEGandEMGsignalswererecorded Warming of the preoptic area suppresses discharge of wake- bipolarly using Polygraph model 78 amplifiers (Grass Instruments, activePF/LHneurons(Methipparaetal.,2003),whereasinhibi- Quincy,MA)withpassbandssetat1–30and100–1000Hz,respectively. Neuronalactivitywasrecordedextracellularlyusingbipolarderivations tionofthisareawithmuscimolinducesFosinhypocretinand frommicrowires(impedanceat1kHz,500–800k(cid:2))andamplifiedbya nonhypocretinPF/LHneurons(Satohetal.,2004).Recentana- differentialACamplifier(model1700;A-MSystems,Carlsborg,WA) tomicalstudiesshowedthattheMnPNinputtothePF/LHisone withlowandhighcutofffiltersof10Hzand10kHz,respectively.During oftheheaviestamongthestructuresinvolvedincontrolofbe- recordingsessions,themicrowireswereadvancedin25–30(cid:1)msteps, havioralstates(Yoshidaetal.,2006).MnPN–PF/LHprojection untilactionpotentialswithsignal-to-noiseratio(cid:2)3wereobserved. neurons include a subset of cells exhibiting sleep-related c-Fos Bioelectricalsignalsweredigitizedandstoredonharddriveforoff-line immunoreactivity(Uschakovetal.,2006).Thesefindingsallow analysisusingMicro1401dataacquisitioninterfaceandSpike2software ustohypothesizethatMnPNplaysanimportantroleinsleep- package(CambridgeElectronicDesign,London,UK).Polygraphicdata relatedinhibitorymodulationofhypocretinandnonhypocretin weredigitizedatasamplingrate256Hzandunitactivitydataat10or25 arousal-relatedPF/LHneurons. kHzforwaveformandwavemarkdatachannels,respectively. Electrical stimulation. Electrical stimulation of the MnPN was per- An alternative or complementary source of inhibition of formedwith200(cid:1)sconstant-currentsquarepulsesusingA-65Timer/ PF/LH arousal-related cells could be local GABAergic neurons stimulatorcoupledwithSC-100constant-currentmonophasicstimulus (Rosinetal.,2003)and/orcellscontaininganinhibitorypeptide isolationunit(WinstonElectronicsCompany,Millbray,CA).Stimula- melaninconcentratinghormone(MCH).Thelatterareintercon- tionparameterswererestrictedbecauseMnPNstimulationatcurrent nectedwithhypocretinneurons(Bayeretal.,2002)andarelikely intensities (cid:4)150–200 (cid:1)A triggered absence-like seizures within a few to be a subset of PF/LH sleep-active cells (Verret et al., 2003; seconds.TotestPF/LHneuronalresponses,thefollowingnonepilepto- Modirroustaetal.,2005). genicstimulationparadigmswerechosen.MnPNwasstimulatedat100 TodeterminethefunctionalinfluenceoftheMnPNonPF/LH (cid:1)A current intensity with single pulses (0.5 pulses/s) or 5 s trains of neuronalpopulationswithdifferentrolesinbehavioralstatecon- stimulifollowedby150msperiod.Trainstimulationswereseparatedby trol,weexaminedtheeffectsofelectricalstimulation,chemical atleast20srecoveryperiods.Stimulationfrequencywithinthetrainswas stimulation,andchemicalinhibitionoftheMnPNonsingle-unit chosenbasedonmedianfrequencyofdischargeofMnPNsleep-related neuronsinquietwakingdeterminedfromthepreviousdata(Suntsovaet activitywithinthePF/LHinrats. al.,2002). Theeffectivecurrentspreadaroundthetipsofbipolarside-by-side MaterialsandMethods electrodeshasnotbeenestimated.However,itislikelytobelessthanor Animals.MaleSpragueDawleyrats(300–350g;8–10weeksofage)were atthemostcomparablewiththespreadfromatipofamonopolarelec- housed individually in Stand-Alone Raturn Animal Handling System trode,whichwasestimatedasthesquarerootofthecurrentdividedby (BioanalyticalSystems,WestLafayette,IN)placedinsideanelectrically thesquarerootoftheexcitabilityconstant(Tehovniketal.,2006).Ap- shielded,sound-attenuatedchamber.Ratswerekeptina12hlight/dark plyingthisequation,monopolarstimulationwithparametersusedinour cyclewithlightsonat8:00A.M.,designatedasZeitgebertime0(ZT0). studywouldbeexpectedtoactivatethemostexcitableneuronswithin0.6 Animalshadadlibitumaccesstofoodandwater.Allexperimentswere mmradius.Theeffectivecurrentspreadfromamonopolarelectrodehas performedinaccordancewiththeNationalResearchCouncilGuidefor beenestimatedas250–500(cid:1)mbasedonsingle-cellrecordingsorusing the Care and Use of Laboratory Animals. Animal use protocols were behavioralmethods(Wise,1972;BagshawandEvans,1976;Tehovniket reviewedandapprovedbytheInternalAnimalCareandUseCommittee al.,2006).Comparedwithsomeofthesestudies,weusedlargerelectrode oftheVeteransAffairsGreaterLosAngelesHealthcareSystem. tips,whichcouldresultinasmallereffectivecurrentspread(Follettand Surgery.Surgicalprocedureswereperformedunderketamine/xylazine Mann,1986).Thisissupportedbyourfindingthat150–200(cid:1)mshiftsin anesthesia(80/10mg/kg,i.p.respectively)andasepticconditions. electrodepositionfromtheMnPNmarginsinmediolateralordorsocau- Forbehavioralstateassessment,stainless-steelscrewEEGelectrodes daldirectionsledtodisappearanceofEEG-synchronizingeffectofhigh- wereplacedintotheskulloverthefrontalandparietalcortexandtwo frequency (100–200 pulses/s) MnPN stimulation and recruiting re- Teflon-coatedstainless-steelEMGwireswereimplantedintothedorsal sponsesevokedbylow-frequencyMnPNstimulation. neckmuscles. Themainadvantageofthemethodofelectricalstimulationisitshigh To record a single-unit activity within the PF/LH, a preassembled temporalresolutionallowingthedistinctionofantidromic,monosynap- constructionwasused.Itconsistedof10Formvar-insulatedstainless- ticandpolysynapticresponses.Alimitationislackofselectivityforneu- steelmicrowires(20(cid:1)m)insertedintothe23-gaugeguidecannulathat rons. Possible excitation of axonal terminals and fibers of passage in wasattachedtoamechanicalmicrodriveanchoredtotheminiatureelec- additiontoneuronalcellbodiesresultsinuncertaintyregardingtheori- tricalconnector.Aholewastrephinedintheskull,centeredatanterio- ginofobtainedresponses.Toovercomethisproblem,wealsoapplied posterior (AP), (cid:1)3.14; mediolateral (ML), 1.2 (Paxinos and Watson, localchemicalactivationandinhibitionofneuronalcellbodiesandden- 1998).Theguidecannulawasloweredinsidethebraintopositionitstip drites within the MnPN. This method, however, has lower temporal 3mmabovethetarget.Afterfixationoftheentireassemblytotheskull, precisionandpermitsassessmentofonlyrelativelylong-termtonicin- the microwires were advanced through the guide cannula to a point fluencesofMnPNafferents. correspondingtothePF/LHdorsalmargin[horizontal(H),8.2]. Reversemicrodialysis.MnPNcellularactivitywasmanipulatedchemi- TostimulatetheMnPNelectrically,animalswereimplantedwithpar- callyusingmicrodialyticapplicationofL-glutamate,theGABAAreceptor allelbipolarFormvar-insulatedtungstenelectrodes(80(cid:1)m;10–20k(cid:2)at antagonist bicuculline methiodide, and the GABA receptor agonist A 1000Hz).Electrodewirestargetedthemostrostroventalanddorsocau- muscimol(allfromSigma,St.Louis,MO)toexcite,disinhibit,andin- dalportionsoftheMnPN(AP,0.0;H,7.0;andAP,(cid:1)0.46;H,5,respec- hibittheMnPNneurons,respectively. tively)(Swanson,1998)toprovidestimulationoftheentirestructure The microdialysis probe (cuprophane semipermeable membrane (Fig.1A). length,1mm;outerdiameter,0.24mm;molecularweightcutoff,6000 To activate or inhibit the MnPN neuronal activity chemically, rats Da;CMA/11;CMAMicrodialysis,NorthChelmsford,MA)wasinserted wereimplantedwithaguidecannulaforsubsequentinsertionofmicro- intotheimplantedguidecannula24hbeforestartinganexperiment.The dialysisprobe.Theguidecannulawasimplantedinthemidlineata15° tipoftheprobeextended3mmbeyondthetipoftheguidecannula.The anglefromtheverticalwiththetiplocated3mmfromthetarget(AP, positioningoftheprobewaschosentominimizedamagetothenucleus (cid:3)0.12;ML,0;H,7). (Fig.1D).Thetipoftheprobe(coveredwithgluefor0.3–0.5mm)was Recording.Theratswereallowedtorecoverfromsurgeryfor7d,and placedwithinthemostrostralpartoftheMnPN.Theopenpartofthe 1618•J.Neurosci.,February14,2007•27(7):1616–1630 Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity microdialysismembranewasincontactwitha dorsalsurfaceofthenucleus. The probe was continuously perfused at a flow rate of 2.0 (cid:1)l/min with filtered artificial CSF(aCSF)containingthefollowing(inmM): 145NaCl,2.7KCl,1.3MgSO ,1.2CaCl ,and2 4 2 Na HPO ,pH7.2.Alldrugsweredissolvedin 2 4 this solution. Gastight Bee Stinger syringes were filled with perfusates and mounted on BabyBeeSyringePumpsconnectedtoBeeHive Pumpcontrollers(BioanalyticalSystems,West Lafayette,IN).Tochangesyringesduringmi- crodialysiswithoutinterruptingflow,aliquid switch(UniSwitch;BioanalyticalSystems)was used.Afterdeliveryofsubstances,theperfusion solution was switched back to aCSF and the recordingcontinuedforanother45–90minor untilthedischargeratewasatbaselinelevels. To examine the effects of MnPN chemical activationanddisinhibitiononPF/LHneuro- nalactivity,1mML-glutamate,50(cid:1)Mbicucul- line,and50(cid:1)Mmuscimolconcentrationshave beenchoseninpreliminaryexperiments.These concentrations,ononehand,werebelowthe thresholdforinductionofEEGandbehavioral manifestationsofabsence-likeseizuresbut,on theotherhand,wereshowntobeeffectiveto modulatebehavior,neuronaldischarge,c-Fos immunoreactivity, and neurotransmitter re- lease while applied by reverse microdialysis into the other structures (Karreman and Moghaddam, 1996; Materi and Semba, 2001; Westetal.,2002;Satohetal.,2004;Alametal., 2005a).ExcessesofeitherL-glutamateorbicu- culline can potentially inactivate instead of 3 Figure 1. Anatomical localization of stimulating elec- trodes,microdialysisprobes,andrecordedneurons.A,D, Schematicdrawingshowingtargetedplacementofstimulat- ingelectrodes(A)andmicrodialysisprobes(D)atsagittalsec- tionoftheratbrain.B,E,Histologicallyverifiedlocationsof thetipsofstimulatingelectrodes(B,asterisks)andmicrodi- alysisprobes(E,verticalbars)plottedondiagramsofcoronal sections.Thelocationsofthetipsofstimulatingelectrodes withinrostralandcaudalMnPNinBareshownonthetopand bottompanels,respectively.TheMnPNishighlightedwith bluecolor.C,F,PhotomicrographsofNissl-stainedsections showingthelocationsofstimulatingelectrodesandmicrodi- alysisprobes.Thearrowsindicatethesitesofelectrolyticle- sionswithintherostral(C,top)andcaudal(C,bottom)MnPN andtrackofthetipofmicrodialysisprobewithintherostral MnPN(F).Scalebar,250(cid:1)m.G,H,Locationsofneuronswith differentsleep–wakedischargepatternsrecordedwithinthe PF/LHduringexperimentswithelectricalstimulation(G)and microdialyticperfusion(H)oftheMnPN.Forbettervisualiza- tion,AW/REMsleep-relatedneurons(circles)areshownon theleftsideandAW-related(triangles),NREM/REMsleep- related (diamonds), REM sleep-related (rectangles), and state-indifferent(asterisks)unitsareshownontherightside ofthedrawings.Theperifornicalareaisoutlinedwithbrown. I,Photomicrographsofstainedforhypocretin-1histological sectionsshowingthelocationsofmicrowiretracks(arrows) within lateral (1) and medial (2) parts hypocretin- immunoreactiveneuronalfield.Scalebar,50(cid:1)m.MS,Medial septalnucleus;ac,anteriorcommissure;och,opticchiasm;f, fornix;VDB,nucleusoftheverticallimbofthediagonalband; 3V,thirdventricle;Br.,bregma. Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity J.Neurosci.,February14,2007•27(7):1616–1630•1619 stimulateMnPNneuronsbecauseofdepolarizationblock.However,sig- thebeginningofdrugapplication.ThepercentagesoftimespentinAW, nificantlyhigherconcentrationsofthesedrugshavebeenusedpreviously QW,NREM,andREMsleepwerecalculatedperhour. tophysiologicallyactivatebrainregions(GeorgesandAston-Jones,2001; ToquantifytheeffectofchemicalmanipulationsoftheMnPNonthe ZhangandFogel,2002;Hanamori,2003;Puigetal.,2003).Theconcen- dischargeofthePF/LHneurons,meanfiringrateswerecalculatedfor trationofbicucullineusedinourstudy(0.025(cid:1)g/(cid:1)lintheperfusion sleep–wakestatesduringbaselineaCSFperfusionanddrugadministra- fluidand(cid:5)10timeslessinthetissue)wasmuchlowerthanreportedto tion. Average firing rates obtained during different sleep–wake states produceadepolarizationblockage(0.75(cid:1)g/(cid:1)l)inmicroinjectionexper- before and during drug application were compared with two-way iments(Perieretal.,2002).Inaddition,inourpreliminarystudies,we repeated-measuresANOVAwithtwowithin-subjectsfactors.Onefactor never observed inhibitory responses of preoptic area neurons to wasdefinedasasleep–wakestatewithfourlevels(AW,QW,NREM,and L-glutamateandbicucullinelocallymicrodialyzedatconcentrationsused REMsleep).Theotherfactorwasdrugtreatmentwithtwolevels(aCSF in this study. These preliminary experiments and our previous data perfusionanddrugadministration).Incasesofviolationofsphericity (Alametal.,2005a)basedonc-Fosanalysisallowedustoestimatethe assumption (Mauchley sphericity test), to avoid erroneous results of effective spread of drugs delivered by reverse microdialysis (0.5–0.75 univariateANOVA,amultivariateapproachwasappliedtotestthesig- mmaroundtheprobe). nificanceofunivariaterepeated-measuresfactorswithmorethantwo Dataanalysis.Todeterminethestate-relatedchangesindischargeof levels(Wilk’smultivariatetest).Inpresenceofasignificantmaineffect, therecordedcells,themeanfiringrateswerecalculatedfor1015–60s posthoccomparisonsweredoneusingTukey’sHSDtest. artifact-freeperiodsofAW,QW,NREM,andREMsleep,whichwere Resultsareexpressedthroughoutasmean(cid:7)SEM.Thelevelofsignif- identifiedonthebasisofEEGandEMGparametersusingstandardcri- icancewassetatp(cid:6)0.05. teria(Timo-Iariaetal.,1970).Recordedcellswereclassifiedintogroups Histology.Attheendofarecordingsession,underdeepanesthesia withdifferentsleep–wakefiringrateprofilesbasedontheresultsofa (pentobarbital;100mg/kg,i.p.),microlesionsweremadeatthetipof K-meansclusteranalysisfollowedbyevaluationofstate-relatednessfor microwires(20(cid:1)Aanodaldirectcurrent;15–20s)atthemostventral eachmemberofthecluster(Suntsovaetal.,2002).One-wayANOVA recordedsite.Afterinjectionwithheparin(500U,i.p.),animalswere followedbyTukey’shonestlysignificantdifference(HSD)posthoctest wasusedtoexaminetheinterstatedifferencesinthemeanfiringratesof perfusedtranscardiallywith30–50mlof0.1Mphosphatebuffer,pH7.2, followedby300mlof4%paraformaldehydeinphosphatebuffercon- individual neurons and groups of cells with different sleep–wake dis- chargeprofiles. taining15%saturatedpicricacidsolution,100mlof10%sucrose,and TheorthodromicresponsesofPF/LHneuronsevokedbyMnPNstim- finally100mlof30%sucroseinphosphatebuffer.Thebrainswerere- ulationwerecharacterizedbymeasuringtheonsetlatencyandduration movedandequilibratedin30%sucrose.Serialcoronalsections(40(cid:1)m) of stimulation-induced effects from peristimulus time histograms werestainedforNissl(cresylviolet)toverifythepositionofthestimu- (PSTHs).TogeneratePSTHs,atleast80trialsofsingle-pulsestimulation latingelectrodes.Toverifythepositionofmicrowiretrackswithinthe performedduringAWwereselected.Thetrialsweredividedinto1ms hypocretinneuronalfield,sectionsthroughthePF/LHwereimmuno- bins,andactionpotentialsoccurringwithineachbinweresummedover stainedforhypocretin-1asdescribedpreviously(Alametal.,2002). all stimulation trials. Significant excitatory or inhibitory poststimulus Reconstructions of tracts of stimulation electrodes, microdialysis neuronalresponseswereidentifiedinthosebinswhosecountswere(cid:4)2 probes,andmicrowiresweremadewiththeaidofNeurolucidaimaging SDsaboveorbelow,respectively,thebaselinemean(averagedovera system(MicroBrightField,Colchester,VT)guidedbytheratbrainatlases prestimulus period of 200 ms). For slowly firing cells, inhibition was ofPaxinosandWatson(1998)andSwanson(1998)forPF/LHareaand definedbya(cid:4)50%decreaseinthefiringratewithrespecttothepre- theMnPN,respectively.Accordingtoreconstructions,stimulatingelec- stimulus value. The neuronal response onset latency and offset were trodes were correctly positioned in five rats (Fig. 1B,C). From these definedasthefirstoffiveconsecutivebinsmeetingtheresponsecriterion animals,87cells,recordedwithinhypocretin-immunoreactiveneuronal andnolongermeetingthecriterion,respectively.Fivemillisecondwin- fieldwithsignal-to-noiseratio(cid:2)3(Fig.1G),wereexaminedfortheir dowsallowedustoidentifybothshort-andlong-durationresponses. sleep–wake discharge patterns and responses to the MnPN electrical GiventhatconductiontimefromtheMnPNtothePF/LHisunknown, stimulation. The vast majority (85%; n (cid:8) 74) of these neurons were neuronal responses with onset latencies less than (cid:5)15 ms potentially recordedlateraltothefornixandwereconsideredtobeintheLH.The couldbeconsideredasantidromicormonosynaptic,takingintoaccount restofthecellswererecordeddorsalormedialtothefornix,withinthe possibilityoflowconductionvelocity(StockerandToney,2005)and perifornical and dorsomedial hypothalamic nucleus. Localizations of slowsynaptictransmission.Short-latencyexcitatoryresponseswereop- bothmicrodialysisprobes(Fig.1E,F)andmicrowires(Fig.1H)within erationallydefinedasmonosynapticiflatencyjitter(SDofthetimefrom thetargetswerehistologicallyconfirmedinthreerats.Responsesof54 thebeginningofthestimulusartifacttothebeginningofthefirstevoked PF/LHneuronswithidentifiedsleep–wakedischargeprofilestochemical spike)was(cid:6)0.5ms.Antidromiccriteriaincludedtheabilitytofollow stimulationand/orinhibitionoftheMnPNwereanalyzed. single pulses with a constant latency ((cid:6)0.1 ms latency jitter) and to followhigh-frequencystimulustrainsat100–200pulses/s. ToestimatetheeffectsofMnPNtrainstimulationonthefiringrateof Results PF/LHneurons,PSTHs(0.1sbinwidth;250bins;summationofatleast EEGresponsestoMnPNelectricalstimulation 30stimulationtrials)andrasterdisplaysweregenerated,whichprovided TheeffectsoftheMnPNsingle-pulseandtrainstimulationon graphicrepresentationofneuronalactivityoccurringfrom5sbeforeto PF/LHneuronalactivitywerestudiedduringactivewaking.Five 15saftertheendofthetrainstimulation.TogeneratePSTHs,pulses secondtrainstimulationoftheMnPN(200(cid:1)s,100(cid:1)Asquare initiatingtrainsofstimuliwereusedastriggers.Themeanfiringrates werecalculatedfor5sepochsbeforeandduringstimulationaswellasfor pulseswith150msperiod)evokedrhythmicactivityresembling threeconsecutiveepochsthatfollowedthestimulation.Thesemeasure- recruitingresponsesorspindle-likeactivityincombinationwith mentsweredoneforeachindividualtrialandforaveragedtrials.The high-amplitude(cid:3)waves.Trainsofhigh-frequencystimuli(100– statisticalsignificanceofchangesinthefiringrateforeachindividual 200pulses/s)alsotriggeredEEGsynchronizationwithinthefre- neuron and for subsets of the cells was determined using one-way quencyrangeofsleepspindles.Singlestimuluspulsesofthesame repeated-measuresANOVAwithtimeasarepeatedmeasurefollowedby durationandintensityatfrequency0.5pulses/sfailedtoinduce Tukey’sHSDposthoctest. EEGsynchronizationinAW.However,inQWandNREMsleep, To quantify the sleep–wake parameters during baseline conditions stimulation with these parameters evoked slow (180–300 ms) andchemicaltreatmentsoftheMnPN,EEGandEMGrecordingswere scored on the basis of the predominant state within each 10 s epoch wavesfollowedbyrhythmicafterdischarges(seriesofwavesthat during1hofpredrugaCSFperfusionandduring1hperiodstartingfrom occurredat10–25Hzandlastedupto1.5s). 1620•J.Neurosci.,February14,2007•27(7):1616–1630 Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity Table1.SummaryofelectrophysiologicalresponsesofPF/LHneuronstoasingle-pulseMnPNstimulation Initialresponse Secondresponse Inhibition Excitation Excitation Inhibition Number Latency Number Latency Duration Number Duration Number Duration Sleep–wakedischargeprofile ofcells (ms) Duration(ms) ofcells (ms) (ms) ofcells (ms) ofcells (ms) AW/REMsleep-relatedwithSPFM(n(cid:8)41) 31 6.5(cid:7)0.9 87.1(cid:7)7.7 10 44.2(cid:7)5.9 AW/REMsleep-relatedwithMFM(n(cid:8)16) 6 24.0(cid:7)1.1 119.3(cid:7)25.5 8 3.6(cid:7)0.4 5.3(cid:7)1.6 8 74.6(cid:7)14.4 AW-related(n(cid:8)5) 4 6.8(cid:7)2.5 82.3(cid:7)22.1 NREM/REMsleep-related(n(cid:8)9) 9 9.0(cid:7)1.8 9.1(cid:7)1.8 7 60.6.(cid:7)11.2 REMsleep-related(n(cid:8)9) 7 7.4(cid:7)1.2 11.4(cid:7)3.3 3 49.7(cid:7)8.1 State-indifferent(n(cid:8)7) 4 21.0(cid:7)1.9 30.5(cid:7)8.7 PF/LHneurons:sleep–wakedischargepatternsandresponses tionfollowedbyinhibition(Table1).Thepercentagesandrep- toMnPNelectricalstimulation resentativeexamplesoftheseresponsesareshowninFigure2,C The electrophysiological responses of 87 PF/LH neurons were andD–F,respectively. analyzed. Exclusively inhibitory reactions or initial inhibitions were InagreementwithpreviousPF/LHunitactivitystudies(Alam foundin37cells(82%ofresponsiveunits).Thevastmajorityof etal.,2002;Koyamaetal.,2003),theexaminedcellswereheter- theseresponderswereSPFMneurons(n(cid:8)31).Theoverallmean ogeneousintermsoftheirsleep–wakedischargeprofiles.Based responselatencyanddurationofexclusivelyandinitiallyinhibi- ontheresultsofclusteranalysisfollowedbyevaluationofstate- toryresponsesareshowninTable1. relatedness for each member of the cluster, PF/LH cells were Twenty-sevenAW/REMsleep-relatedcells(21SPFMneurons subdivided into five groups designated in accordance with the and6MFMcells)exhibitedpureinhibitoryresponses(Figs.2D, state(s) in which unit discharge was the highest, namely: 3B).K-meanclusteringanalysisshowedexistenceofthreeclus- AW/REMsleep-related,AW-related,NREM/REMsleep-related, tersofunits(n (cid:8)10;n (cid:8)11;n (cid:8)6)differinginlatencyof 1 2 3 REMsleep-related,andstate-indifferentneurons. response(2.0(cid:7)0.2,11.5(cid:7)0.9,and24.0(cid:7)1.1ms,respectively). Single-pulsestimulationoftheMnPNperformedduringAW ThefirsttwoclustersincludedSPFMneurons,whereasthelong- eliciteddischargechangesin69ofthe87PF/LHcells(79.3%).No est latencies were found in MFM cells, which constituted the cells were activated antidromically. Although the predominant thirdcluster(Fig.2G).Overallmeanonsetlatencyandduration MnPN-inducedtransynapticeffectwasinhibitory,neuronswith ofpureinhibitoryresponseswere10.4(cid:7)1.7and105.1(cid:7)9.1ms, differentsleep–wakedischargeprofilesexhibiteddifferentreac- respectively. tionstostimulation.Characteristicsoftheseresponsesaresum- Biphasic inhibitory–excitatory responses (Fig. 2E) were marizedinTable1. found only in SPFM cells (n (cid:8) 10; 18%). These neurons were groupedintotwoclusters(n (cid:8)4;n (cid:8)6)withdifferentlatency AW/REMsleep-relatedneurons ofinitialinhibitoryresponse1(1.5(cid:7)20.3and10.0(cid:7)0.7ms,re- Sleep–wake discharge pattern. Fifty-seven PF/LH cells (65.5%) spectively)(Fig.2H).Theperiodofdischargereductionlastedfor firedathigherratesduringAWandREMsleepcomparedwith 57.7(cid:7)8.6msandwasshorter(p(cid:6)0.01,Student’sttest)thanin QW and NREM sleep. The interstate differences in the mean purelyinhibitedSPFMcells(101.1(cid:7)9.3).Themeandurationof firing rate of these cells as a group and sleep–wake discharge postinhibitoryexcitationwas44.2(cid:7)5.9ms. profiles of individual units are shown in Figure 2, A and B, Biphasic excitatory–inhibitory responses were found exclu- respectively. sivelyinMFMneurons(n(cid:8)8;14%)(Figs.2F,3D).Theywere Themajority(n(cid:8)41;71.9%)ofAW/REMsleep-relatedcells excitedwithalatencyrangingfrom2to5ms(Fig.2I).Themean exhibited a single-pulse firing mode (SPFM) across all sleep– latencyanddurationofbothexcitatoryandinhibitorycompo- wake states (Fig. 3A). The rest of AW/REM sleep-related units nentsoftheseresponsesareshowninTable1. (n(cid:8)16;28.1%)generatedsinglepulsesduringwakingandREM During5sMnPNtrainstimulation,thevastmajority(84.2%) sleep and a combination of single pulses and high-frequency ofAW/REMsleep-relatedcellssignificantly(p(cid:6)0.05,Tukey’s (400–900pulses/s)burstdischargesduringNREMsleep.These HSD post hoc test) decreased their firing rates to 7.1–74.8% of cellswereidentifiedwithinthePF/LHforthefirsttimeandwere prestimulation firing. No neurons were facilitated. The overall designatedasmixedfiringmode(MFM)neurons(Fig.3C).An meanfiringrateofrespondingAW/REMsleep-relatedcellswas abilityofPF/LHneuronstogenerateburstsduringNREMsleepis reducedby42.7%duringstimulation(p(cid:6)0.001,Tukey’sHSD compatible with the findings that some LH neurons express a posthoctest).Dischargeremaineddecreasedby31.8%for5safter low-voltage-activatedcalciumcurrent(Fanetal.,2000)andre- theendofstimulustrains(p(cid:6)0.001)andby15.4%duringa ceiveinhibitoryinputsfromthereticularthalamus(Baroneetal., subsequent 5 s epoch (p (cid:6) 0.05) (Fig. 2J). The group of cells 1994).ThepercentageofAW/REMsleep-relatedcellswashigher inhibitedbytrainstimulationincludedboththeneuronsreactive amongtheneuronsrecordedwithinLH(70.3%)thanamongthe tosingle-pulsestimulationandunitsunresponsivetosinglestim- neurons recorded within the medial part of hypocretin- uli(Fig.2K,L).Inthelattercase,itislikelythattemporalsum- immunoreactiveneuronalfield(38.5%).AllcellswithMFMwere mation of postsynaptic potentials was needed to suppress dis- foundwithintheLH. chargeofthetargetneuron. Responses to stimulation. Forty-five AW/REM sleep-related cells(79%)respondedtoasingle-pulseMnPNstimulation.The AW-relatedneurons responsesfellintothreecategories:(1)apureinhibition,(2)an Sleep–wakedischargepattern.FivePF/LHcells(5.8%)exhibited initialinhibitionfollowedbyexcitation,and(3)aninitialexcita- maximumfiringratesduringAW.Theinterstatedifferencesin Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity J.Neurosci.,February14,2007•27(7):1616–1630•1621 the mean firing rate of these cells as a group and sleep–wake discharge profiles ofindividualunitsareshowninFigure4, AandB,respectively.AllAW-relatedneu- ronsexhibitedasingle-pulsefiringmode ineachstageofthesleep–wakecycle. Four AW-related cells were recorded dorsal or medial to the fornix (Fig. 1G). Their percentage within this region was 30.8%,whichiscomparablewiththatpre- viouslyreported(Alametal.,2002).Lat- eraltothefornix,onlyoneof74recorded neurons (1.4%) was AW-related. Hypo- cretin cells exhibit an AW-related dis- charge profile (Lee et al., 2005; Mileyk- ovskiyetal.,2005).Thelowpercentageof AW-relatedneuronswithinthelateraldi- visionofhypocretinfieldimpliesthatlat- eralhypocretincellsareeitherselectively active during specific waking behaviors underrepresented in our experimental conditionsorhaveadifferentsleep–wake dischargeprofile. Responses to stimulation. Four AW- related neurons responded to a single- pulseelectricalstimulationoftheMnPN withpureinhibition(Fig.4C,D,H).Two cells responded with 2 and 3 ms latency, andtheothertwocellswereinhibited10 and12msafterstimulusonset(Fig.4E). The mean onset latency and duration of responsesareshowninTable1.AllAW- relatedunitswereinhibitedduringMnPN train stimulation. The firing rates de- creasedto26.4–64.0%ofbaselinefiring. TheoverallmeanfiringrateofAW-related units decreased by 58.3% (p (cid:6) 0.001, Tukey’sHSDposthoctest)duringstimu- lationandremainedreducedduringtwo consecutive5sepochssubsequenttostim- ulationperiodby51.0%(p(cid:6)0.001)and 30.1% (p (cid:6) 0.05) compared with pre- stimulationfiring(Fig.4F). Figure2. SummaryofdischargecharacteristicsofAW/REMsleep-relatedPF/LHneuronsandtheirresponsestoMnPNelectrical stimulation.A,Groupmean(cid:7)SEMdischargerates(spikes/second)ofAW/REMsleep-relatedcellsacrosssleep–wakestates.The Sleep-relatedneurons meanfiringrateinAWandREMsleepwassignificantlyhigherthaninQWandNREMsleep(one-wayrepeated-measuresANOVA Sleep–wake discharge patterns. Eighteen followedbyTukey’sHSDtest).B,Meandischargeratesofindividualneurons.C,Percentagesofdifferenttypesofresponsesto PF/LH neurons (20.7%) were sleep- MnPNelectricalstimulationexhibitedbyneuronswithSPFMandMFM.Notethelowproportionofnonresponsive(NR)unitsand related. These cells were identified as ei- thehighpercentageofcellsexhibitingpureinhibitoryresponses(Inh)inbothSPFMandMFMsubsetsofneurons.TherestofSPFM ther NREM/REM sleep- or REM sleep- neuronsexhibitedinhibitory/excitatoryresponses(Inh/Exc),whereastheremainderofMFMneuronsrespondedwithinitial related, depending on the sleep state(s) excitationalwaysfollowedbyinhibition(Exc/Inh).D–F,Representativeexamplesofrasterplots(toppanel)andPSTHs(bottom panel)correspondingtoanInh(D),Inh/Exc(E),andExc/Inh(F)responsesaftersingle-pulseelectricalstimulationoftheMnPN.For duringwhichtheyexhibitedelevateddis- allPSTHshereandinsubsequentfigures,electricalpulses(200(cid:1)s;100(cid:1)A)wereappliedattime0(arrowheads),andbinwidth chargecomparedwithwaking. was1ms.Thescaleonthey-axisisinspikes/second.Rasterdisplayshowstheoccurrenceofspikesforeachtrialaccumulatedinthe NREM/REM sleep-related neurons PSTH.They-axisintherasterplotshowsthenumberoftrials.G–I,Histogramsshowingdistributionsofonsetlatenciesforpure (n(cid:8)9;10.3%)wereforthefirsttimeiden- inhibitory(G),initiallyinhibitory(H),andinitiallyexcitatory(I)responses.Noteasubstantialproportionofshort-latencyre- tified within the PF/LH. They increased sponses.J–L,ResponsesofAW/REMsleep-relatedunitstotrainstimulationoftheMnPN.J,Groupmean(cid:7)SEMfiringrateof48 their discharge during both NREM and unitscalculatedforfiveconsecutive5sepochsstartingwiththeepochimmediatelyprecedingstimulationperiod.Duringstimu- REMsleepcomparedwithwakingstates. lation200(cid:1)s,100(cid:1)ApulseswereappliedtotheMnPNwith150msperiodfor5s(horizontalbarnexttox-axis).Notethatthe REM sleep-related neurons (n (cid:8) 9; groupmeanfiringratesignificantlydecreasedduringstimulationandcontinuedtobereducedfor10saftercompletionof 10.3%)increasedfiringratesduringREM stimulation.K,Inhibitionbytrainstimulationofaneuronunresponsivetosingle-pulsestimulation.PSTH(0.1sbinwidth, sleep compared with waking states and summationof30stimulationtrials)andrasterdisplayprovidegraphicrepresentationofactivityofthiscelloccurringfrom5s beforethebeginningto10saftertheendofthetrainstimulation.L,Firingratehistogram,unitactivity,EMG,andEEGduringa NREMsleep.Activationofcellsoccurred singlestimulationtrialmarkedbythearrowinK.*p(cid:6)0.05;**p(cid:6)0.01;***p(cid:6)0.001;ns,nonsignificant. 8–46sbeforeREMsleeponset.Theinter- statedifferencesinthegroupmeanfiring 1622•J.Neurosci.,February14,2007•27(7):1616–1630 Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity ratesofsleep-relatedcellsandsleep–wake discharge profiles of individual units are showninFigure5,AandB,respectively. All sleep-related neurons exhibited a tonicfiringmodeacrosssleep–wakestates. The percentages of sleep-related cells amongneuronsrecordedmedialandlat- eraltothefornixwerecomparable. Responses to stimulation. All NREM/ REMsleep-relatedcellandsevenofnine REMsleep-relatedneuronswererespon- sivetotheMnPNsingle-pulsestimulation exhibitingeitherexclusivelyexcitatoryor excitatory–inhibitoryresponses(Fig.5C). Table 1 shows the overall mean latency anddurationofexcitatoryresponsesand themeandurationofpostexcitatoryinhi- bitions calculated separately for NREM/ REMsleep-andREMsleep-relatedunits. None of these parameters differs signifi- cantly between subsets of sleep-related cells. Pure orthodromic excitations were found in two NREM/REM sleep-related neurons(Figs.5D,6B)andinfourREM sleep-related cells with latencies ranging from2to18ms.SevenNREM/REMsleep- andthreeREMsleep-relatedneurons(Fig. 5E)wereinitiallyexcitedwith2–15msla- tencies.Thedistributionoflatenciesofex- clusivelyandinitiallyexcitatoryresponses of NREM/REM sleep- and REM sleep- relatedunitsisshowninFigure5F. Fourof14sleep-relatedcellsrespond- ingwith(cid:6)15msonsetlatencywereoper- ationallydefinedasmonosynapticallyac- tivated because they exhibited (cid:6)0.5 ms latency jitter and ability to follow high- frequency stimulation at 70–85 pulses/s (see Fig. 6B). The rest of the cells were classifiedaspolysynapticallyactivated. 4 withEEGdesynchronizationonarousalsfromNREMsleep (a).B,Responseofthesameneurontoasingle-pulsestim- 2 ulationoftheMnPN.PSTH(bottompanel),rasterplot(middle panel),andarawtrace(toppanel)ofasinglestimulationtrial markedbyarrowaredisplayed.Theneuronrespondedwitha short-latency(3ms)pureinhibition.C,ThedischargeofMFM cellacrossthesleep–wakingcycle.c,ExpandedtracingfromC markedbyhorizontalbar.1–4,Unitactivitytracingscorre- spondingtoepochsmarkedbyverticalbarsinC:1,NREM sleep;2,transitiontoREMsleep;3,REMsleep;4,AW.Note thatMFMneuronsdramaticallyincreasedtheirfiringrates 18–42sbeforeREMsleeponset(c)primarilybecauseofan increaseindensityofhigh-frequencyburstdischarges(c, Figure3. RepresentativeexamplesoftheresponsesevokedinAW/REMsleep-relatedPF/LHneuronswithSPFMandMFMby compare1,2).D,Responseofthesamecelltoasingle-pulse MnPNelectricalstimulation.A,ThedischargeofSPFMcellacrossthesleep–wakingcycle.Hereandinsubsequentfiguresshowing MnPN stimulation. The neuron responded with a short- polygraphicrecordingsthatillustratethesleep–wakedischargepatternsofPF/LHneurons,thefiringratehistogram(Rate), latency(4ms)excitationfollowedbyinhibition.Thetoppanel electromyogram(EMG),andelectroencephalogram(EEG)aredisplayed.Thescaleonthey-axisinthefiringratehistogramsisin ontheleftsideshowsrowtracesoftwosinglestimulation spikes/second.Thesleep–wakestates(wakefulness,NREMsleep,transitionfromNREMtoREMsleep,REMsleep)areindicated trialsmarkedbyarrowsinrasterdisplay(middlepanel)and undertheEEGtracingsusingthicksolid,dashed,dot-dash,anddottedlines,respectively.Extracellularlyrecordedunitactivity ontheright-handsidefragmentofthePSTHfromthebottom (Unit)isshownonexpandedtracingfromthesectionsmarkedbyhorizontalbars(a,a).Notethatthecellbelongstothesubset panelshowingathigherresolutiontheinitialpartofthe 1 2 ofSPFMcells(n(cid:8)15)thatdramaticallyincreasedtheirdischarge8–23sbeforeREMsleeponset(a)andalmostsimultaneously response. 1 Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity J.Neurosci.,February14,2007•27(7):1616–1630•1623 Theresponsesofsleep-relatedneurons to train stimulation of the MnPN were tested on six NREM/REM sleep- and on fiveREMsleep-relatedunits.Firingrates ofoneNREM/REMsleep-andthreeREM sleep-related neurons were significantly affectedbytrainstimulations(p(cid:6)0.001, one-wayrepeated-measuresANOVA).All reactions were excitatory and strictly timed to the stimulation period. During stimulation, NREM/REM sleep-related unitincreaseditsdischargeby137%and REM sleep-related neurons by 30–46% (mean,38.9(cid:7)4.7%)comparedwiththe prestimulation firing rate (p (cid:6) 0.001, Tukey’sHSDposthoctest).Thefiringrate was at its maximum at the beginning of stimulationanddecreasedtothebaseline level after 2–3 s. Every cell responsive to trainstimulationexhibitedexclusivelyex- citatoryreactionstosingle-pulsestimula- tion, whereas the two cells that did not show significant responses during train stimulationexhibitedbiphasicexcitatory/ inhibitoryresponsestosinglepulses. State-indifferentneurons Sleep–wake discharge pattern. Seven PF/LHcells(8%)withasingle-pulsefiring modedidnotshowsignificantchangesin the mean firing rate across sleep–wake states(F (cid:8)1.9;p(cid:4)0.05).Individual (3,18) cells exhibited (cid:6)25% between-state differences. Responses to stimulation. Four state- indifferentneuronsrespondedtoasingle- pulse stimulation of the MnPN. They showed pure orthodromic excitations, whichstarted17–26msafterMnPNstim- ulation.Themeanlatencyanddurationof theseresponsesareshowninTable1. Train stimulation of the MnPN had significant effects only on discharge of those cells that showed excitatory re- sponse to single-pulse stimulation. The meanfiringratesofindividualneuronsin- creased during stimulation by 61–105% (mean,88.3(cid:7)9.5%)comparedwiththe baseline.Ateachofthethreepoststimula- tionintervals(5seach),thefiringratedid notdiffersignificantlyfromtheprestimu- lationvalue(p(cid:4)0.05,Tukey’sHSDpost Figure4. SummaryofdischargecharacteristicsofAW-relatedPF/LHneuronsandtheirresponsestoMnPNelectricalstimula- tion.A,Groupmean(cid:7)SEMdischargeratesofAW-relatedneuronsacrosssleep–wakestates.ThemeanfiringrateinAWwas hoctest). highercomparedwithQWandbothphasesofsleep(one-wayrepeated-measuresANOVAfollowedbyTukey’sHSDtest).B,Mean ComparisonoftypesofPF/LHneuro- dischargeratesofindividualneurons.C,Percentagesofdifferenttypesofresponsestosingle-pulseelectricalstimulationofthe nalresponsesandtheirratiosinfiverats MnPN.NotethatallresponderswereinhibitedbyMnPNstimulation.Inh,Inhibitoryresponses;NR,nonresponsive.D,Rasterplot with different localization of stimulating andPSTHcorrespondingtoaresponsethatappearedwith10msonsetlatency.E,Histogramofdistributionofonsetlatenciesof electrodetipswithintheMnPN(Fig.1B) responses.Notethepresenceoflatenciescompatiblewithamonosynapticlinkage.F,Groupmean(cid:7)SEMfiringrateofAW- didnotrevealdifferencesinresponsesof relatedunitscalculatedforfiveconsecutive5sepochsbefore,during,andaftertrainstimulationoftheMnPN.Notethatneurons either arousal- or sleep-related units. In wereinhibitedduringstimulationperiod(markedwithhorizontalbarnexttox-axis)andfor10sthereafter.G,Thedischargeof cases when stimulating electrodes were AW-relatedcellacrossthesleep–wakingcycle.g,g,ExpandedtracingsfromGmarkedbyhorizontalbars.Onthetoprightcorner 1 2 misplaced((cid:5)200and400(cid:1)mlaterallyor oftheexpandedtracingssuperimposedareallactionpotentialsthatappearedduringcorrespondingfragmentsofrecordings.Note 150 (cid:1)m dorsally/200 (cid:1)m caudally from thatneuronincreasesitsfiringratebeforearousalfromNREMsleep(g)andawakeningfromREMsleep(g).H,PSTHandrasterplot 1 2 showingshort-latency(2ms)inhibitoryresponseofthiscelltosingle-pulseMnPNstimulation.*p(cid:6)0.05;***p(cid:6)0.001. MnPN margins), we observed different 1624•J.Neurosci.,February14,2007•27(7):1616–1630 Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity EEG- and PF/LH neuronal responses. Low-frequency stimulation of these sites failed to induce EEG-synchronizing re- sponses,whereashigh-frequencystimula- tion caused EEG desynchronization. Among arousal-related units, we did not find cells exhibiting inhibitory responses with(cid:6)15mslatency.NREM/REMsleep- relatedneuronswereeitherunresponsive (twoofthreecells)orinhibitedbystimu- lationofextra-MnPNsites. ResponsesofPF/LHneuronsto chemicalstimulationandinhibitionof theMnPN Activation of MnPN neuronal popula- tionswithoutinfluenceonpassingfibers wasachievedby1hmicrodialyticperfu- sionofexcitatoryaminoacidL-glutamate ortheGABA receptorantagonist,bicu- A culline,atconcentrationsof1mMand50 (cid:1)M, respectively. Experiments were con- ductedduringthedarkphaseofthelight/ darkcycle(ZT13–ZT16),becauseduring applicationofthedrugsinthelightphase, sustained waking episodes were absent Figure5. Summaryofdischargecharacteristicsofsleep-relatedPF/LHneuronsandtheirresponsestoMnPNelectricalstimu- and,therefore,theeffectsofMnPNchem- lation.A,Groupmean(cid:7)SEMdischargeratesofNREM/REMsleep-relatedneurons(blackbars)andREMsleep-relatedneurons ical stimulation on discharge of wake- (hatchedbars).ThemeanfiringrateofNREM/REMsleep-relatedcellssignificantlyincreasedinQWversusAW,inNREMsleep active PF/LH neurons could not be esti- versusQW,andinREMversusNREMsleep.ThemeandischargeofREMsleep-relatedneuronsdidnotdiffersignificantlybetween mated. Table 2 shows the percentage of AW,QW,andNREMsleepandincreasedinREMsleepcomparedwitheveryotherstate.B,Individualmeandischargeratesof time spent in different sleep–wake states NREM/REMsleep-relatedneurons(solidlines)andREMsleep-relatedneurons(dashedlines)acrosssleep–wakestates.C,Per- 1 h before and during perfusion of centagesofdifferenttypesofresponsestosingle-pulseelectricalstimulationoftheMnPN.Notethatallresponderswereexclu- L-glutamateandbicucullineaveragedfor sivelyorinitiallyexcitedbyMnPNstimulation.D,E,RasterplotsandPSTHscorrespondingtoapureexcitatoryresponseof allexperimentsconductedduringthedark NREM/REMsleep-relatedcell(D)andanexcitatory–inhibitoryresponseofREMsleep-relatedcell(E)tosingle-pulseMnPN stimulation.F,HistogramsshowingdistributionofonsetlatenciesofresponsesofNREM/REMsleep-relatedcells(blackbars)and phase. REMsleep-relatedcells(hatchedbars).Notethepresenceofcellsrespondingwithshortlatencieswithinbothsubsetsofsleep- ToassesseffectsofMnPNdeactivation relatedcells.NR,Nonresponsive;Exc,excitatory;Exc/Inh,excitatory/inhibitoryresponses.*p(cid:6)0.05;***p(cid:6)0.001. on PF/LH neuronal activity, the MnPN was microdialytically perfused with the ingthetreatments(n(cid:8)12)werepooledtogetherwiththecells GABA receptoragonist,muscimol,duringthelightphase(ZT3– A recordedthroughouttheentireexperiment.Forallneuronsde- ZT6)ofthelight/darkcycle.Applicationofthisdrugat50(cid:1)M creasingdischargeinresponsetochemicaltreatments(n(cid:8)6), concentrationfor15minresultedinaprolongedAWstate,which theabsenceofdataonrecoveryofthebaselinedischargeisem- appeared 1–3 min after the start of muscimol perfusion and phasizedbecauseofuncertaintyregardingthecauseofchangesin lastedfor35–84min(mean,61.3(cid:7)3.9).Duringthe1hperiod theactivityoftheseneurons. that started from the beginning of drug application compared with1hpredrugaCSFperfusion,animalsspent91%ofthetime AW/REMsleep-relatedcells inAW,NREMsleepwasdramaticallyreduced(p(cid:6)0.001,one- Effects of chemical stimulation and deactivation of the MnPN wayrepeated-measuresANOVA)andREMsleepepisodeswere werestudiedon35AW/REMsleep-relatedPF/LHneurons. almostabsent(Table2). Thechangesinthegroupmeanfiringratesof12and8AW/ Perfusion of the MnPN with 50 (cid:1)M muscimol was used to REM sleep-related units in response to administration of evaluate responses of AW-, AW/REM sleep-related, and state- L-glutamateandbicuculline,respectively,areshowninFigure7, indifferentunits.Toexamineresponsesofsleep-relatedneurons, AandB.Theneuronalresponsestothesetreatmentsweresimilar. 20(cid:1)Mmuscimolwasappliedfor1hduringthelightsonphaseof Two-wayrepeated-measuresANOVAinbothcasesrevealedsig- thelight/darkcycle(ZT3–ZT6).Withthislowerconcentration, nificantmaineffectsofwithin-subjectsfactors(sleep–wakestate sustainedperiodsofNREMandREMsleepwerepresentduring anddrugtreatment)ondischargeofAW/REMsleep-relatedcells drugadministration(Table2). and significant interaction between the factors. During The effects of chemical stimulation and inhibition of the L-glutamateandbicucullineadministration,themeanfiringrate MnPNwerestudiedon36PF/LHcellsrecordedthroughoutall of AW/REM sleep-related cells decreased compared with aCSF thestagesoftheexperiment(baselineaCSFperfusion,drugtreat- perfusion(F (cid:8)45.7,p(cid:6)0.001;andF (cid:8)24.5,p(cid:6)0.01, (1,11) (1,7) ment,andpostdrugaCSFperfusion).Amongthesecells,allre- respectively). The effects of both treatments depended on the sponsive units recovered their baseline discharge during post- state of the sleep–waking cycle (L-glutamate: F(3,9) (cid:8) 5.1, p (cid:6) drugaCSFperfusion.Inaddition,wearereportingresponsesof 0.05,Wilk’smultivariatetest;bicuculline:F (cid:8)4.0,p(cid:6)0.05). (3,21) 18neurons,whichwererecordedduringbaselineanddrugtreat- Discharge suppression during L-glutamate and bicuculline ad- mentsonly.Neuronsincreasingornotchangingdischargedur- ministrationwasstatisticallysignificantonlyinAW(p(cid:6)0.001 Suntsovaetal.•EffectsoftheMnPNonPF/LHNeuronalActivity J.Neurosci.,February14,2007•27(7):1616–1630•1625 L-glutamate on discharge of an AW- relatedPF/LHunit.Duringdrugadminis- tration,comparedwithpredrugaCSFper- fusion,thecellexhibiteda98%reduction initsfiringrateandbecamevirtuallysilent in AW, whereas during QW and NREM sleepthechangesofdischargeratedidnot exceed25%.REMsleepdidnotoccurdur- ing drug perfusion. L-Glutamate treat- mentchangedthefiringrateofoneoftwo cells, which were lost during postdrug aCSF perfusion. This neuron decreased dischargeonlyinAWto57%ofbaseline firing. Inresponsetobicucullineadministra- tion,oneunitalmostceasedtodischarge during the whole period of drug perfu- sion, whereas the other cell showed 44% decreaseinitsfiringrateexclusivelydur- ingAW. During MnPN perfusion with musci- mol, one neuron increased discharge by 43%,andtheotherwasunresponsive. Sleep-relatedcells Atotalof11sleep-relatedPF/LHneurons weretestedforresponsestoMnPNperfu- sionwithL-glutamate(n(cid:8)2),bicuculline Figure6. RepresentativeexampleoftheresponseofanNREM/REMsleep-relatedPF/LHneurontoMnPNelectricalstimulation. (n(cid:8)3),andmuscimol(n(cid:8)6). A,Thedischargeofthecellacrossthesleep–wakingcycle.a,ExpandedtracingsfromAmarkedbyhorizontalbar.Notethat Microdialytic application of L- NREM/REMsleep-relatedneuronsceasedtodischargeordramaticallyreducedtheirfiringratesafewsecondsbeforearousal.B, glutamateresultedinadramaticincrease ResponseofthiscelltoMnPNelectricalstimulation.Rasterplot,PSTH,superpositionofrawtracesoffiveconsecutivestimulation inthefiringrateofaNREM/REMsleep- trials(1),andneuronalresponsesto80(2)and100pulses/s(3)repetitivestimuliaredisplayed.Noteshortlatency(2ms),lowjitter ofresponse(1),andabilityofthecelltofollowrepetitivestimuliat80pulses/s(2).At100pulses/s,theneuronrespondsonlyto relatedneuronto250,500,260,and190% everyotherstimuli(3).Thesefindingssuggestmonosynapticactivation. ofbaselinefiringinAW,QW,NREM,and REM sleep, respectively (Fig. 9A). An REMsleep-relatedcellincreaseddischarge and p (cid:6) 0.01, respectively; Tukey’s HSD test). Individually, in by28and42%inNREMandREMsleep,respectively. AW in response to L-glutamate and bicuculline treatments, 92 Duringbicucullineperfusion,twoNREM/REMsleep-related and85%ofneurons,respectively,exhibitedchangesintheirac- unitsincreasedtheirfiringratesby44and54%inAWandby44 tivitymeetingtheresponsecriterion((cid:4)25%changeinthefiring and120%inQW.DuringNREMsleep,thechangesintheactivity rate).Nocellswereactivated.DuringL-glutamateandbicucul- ofonecelldidnotmeettheresponsecriterion;theotherneuron lineperfusion,responsivecellsreducedtheirdischargeto34.7(cid:7) increasedfiringby63%.DuringREMsleep,spikeactivityofthese 4.0%(range,11–59%)andto58.5(cid:7)4.3%(range,44–73%)of cellsincreasedby25and29%.AnREMsleep-relatedneuronwas baseline firing, respectively. The effect of L-glutamate on dis- unresponsivetobicucullinetreatment. chargeofanindividualcellacrosssleep–wakestatesisshownin In response to muscimol perfusion, the discharge of one of Figure8A.Allcellsrecordedduringbaselineanddrugtreatment two NREM/REM sleep-related cells decreased by 43 and 25% only(n(cid:8)3)reducedtheirdischargeinresponsetoL-glutamate duringNREMandREMsleep,respectively,andby(cid:6)25%during (n(cid:8)2)andbicuculline(n(cid:8)1)treatmentsto29–45%ofbaseline waking(Fig.9B).Aunitthatwasnotrecordedduringpostdrug firingandwerenotincludedintostatisticalanalysis. aCSFperfusionalsoexhibitedareductionoffiringto27,34,and TheeffectsofMnPNperfusionwithmuscimolwerestudied 45%ofbaselinevaluesduringQW,NREM,andREMsleep,re- on 12 AW/REM sleep-related cells. During the muscimol- spectively.TwooffourREMsleep-relatedcells(onewithverified induced AW state, the mean firing rate of these neurons as a recoveryofthebaselinedischarge,theothernotrecordedduring group increased (F (cid:8) 23.9, p (cid:6) 0.001, one-way repeated- postdrug aCSF perfusion) exhibited 37 and 28% decreases in (1,11) measuresANOVA)(Fig.7C).Individually,75%ofcellsincreased firingratesexclusivelyduringREMsleep.TheremainingREM theirdischargeonaverageto220.8(cid:7)48.5%(range,134–600%) sleep-related neurons did not exhibit criterion responses to of baseline firing; the rest of the cells were unresponsive. The muscimol. effect of muscimol perfusion on discharge of AW/REM sleep- State-indifferentneurons relatedneuronisshowninFigure8B. Intotal,responsesofeightstate-indifferentPF/LHunitstochem- AW-relatedcells icaltreatmentsoftheMnPNwereanalyzed. ResponsesofsevenAW-relatedPF/LHneuronstomicrodialytic ThefiringrateoffivePF/LHstate-indifferentneuronswasnot application of L-glutamate (n (cid:8) 3), bicuculline (n (cid:8) 2), and affectedbyMnPNperfusionwithL-glutamate(n(cid:8)2)andmus- muscimol(n(cid:8)2)intotheMnPNwerestudied. cimol(n(cid:8)3)by(cid:4)10%.Onecellincreaseddischargeby28%in Figure 8C shows the effect of MnPN perfusion with responsetomicrodialyticapplicationofL-glutamate.MnPNper-

Description:
and Sutcliffe, 2005; Jones and Muhlethaler, 2005; Sakurai, 2005;. Saper et . chosen based on median frequency of discharge of MnPN sleep-related neurons . scored on the basis of the predominant state within each 10 s epoch during 1 .. trials marked by arrows in raster display (middle panel) and.
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.