Increase of GABAA receptor-mediated tonic inhibition in DG c(3)

466Z.Mtchedlishvilietal./NeurobiologyofDisease38(2010)464–475

isotonicchlorideconditions(ECl=0mV)withaMulticlamp700Bampli er(MolecularDevices,PaloAlto,CA).Wholecellcapacitanceandseriesresistancewerecompensatedby80%.Recordingswereperformedwhenseriesresistanceaftercompensationwas20MΩorless.Accessresistancewasmonitoredwith10ms, 4mVtestpulses,deliveredonceevery60s.Ifatanytimeduringtheexperimenttheseriesresistanceincreasedby25%,therecordingwasterminatedandthedatawerenotusedforstatisticalanalysis.Currenttraceswere lteredat5kHz,digitizedat10kHzusingaDigidata1442digitizer,andacquiredonacomputerharddriveusingpClamp10.1software(MolecularDevices).

Measurementofsynapticcurrents

sIPSCsandmIPSCswereanalyzedwithMiniAnalysissoftware(Synaptosoft,Leonia,NJ).Thedetectionthresholdwassetat vetimesofrootmeansquare.Afterdetection,peakamplitude,frequency,anddecaytimeconstantswereanalyzed.Followingthepointofpeakamplitude,the0to90%regionofcurrentdecaywasassessedby100iterationsofeacheventbygroupanalysisforbestcurve t.DecaytimeconstantsofmIPSCswere tasamonoexponentialfunction,withMiniAnalysissoftware,usingasimplex-basedalgorithm.Approxi-mately500individualeventswereusedfromeachrecordingforanalysisofpeakamplitudesandfrequencyofsIPSCsandmIPSCs,anddecaytimeconstantsofmIPSCs.Measurementoftoniccurrents

Tomeasuretonicinhibition,the30-sepochimmediatelybeforeapplicationofthedrug(baseline)anda30-sepochafter3minofcontinuousperfusionwithadrugwerecompared.Thebuilt-inGaussianfunctionofClampFit(AxonInstruments)wasusedtodetermineGaussian tsintherawtraces(entiresegmentswithsIPSCspresent).All-pointfrequencydistributionhistogramsusinga2-pAbinsizewerederivedfromthese30-sepochs.

ThecomparisonsofobtainedGaussianmeansofall-pointhistogramdistributionswereperformedwithGraphPadPrism4(MountainView,CA).Gaussianmeans(meanbaselinecurrent)wereusedasmeasuresoftoniccurrent.IncontrolDGCs,Gaussian tsofall-pointhistogramswerecomparedtohistogramsderivedfromthesameepochswithsynapticcurrentsremovedbyacustom-designedalgorithminMATLAB(datanotshown).Nosigni cantdifferencewasfoundbetweentheGaussianmeansofall-pointhistogramsandthosederivedfromtheepochswithsynapticeventsremoved.Therefore,wedeemeditsuf cienttouseall-pointhistogramstostudytoniccurrents.EachDGCwastreatedasanestedunitforstatisticalconsiderations.OneDGCwasrecordedperslicetoavoidpotentialcontaminationofadrugusedinapreviousrecordinginthesameslice.Statisticalconsiderations

Tovalidatethesigni canceoftheeffectsofdiazepamandfurosemideonpeakamplitudesofmIPSCs,cumulativeprobability(fraction)distributionsofpeakamplitudesatbaseline(priortodrugapplication)andduringcontinuousapplicationofthedrugwerecomparedwiththeKolmogorov–Smirnov(KS)testforeachDGCusingthebuilt-inKSanalysisoptioninMinianalysissoftware(Synaptosoft,Leonia,NJ).AllDGCstestedwiththeKStestdemon-stratednon-GaussiandistributionsofmIPSCpeakamplitudes.CumulativefractiondistributionsofmIPSCpeakamplitudesalsowerecomparedbetweencontrolandCCIDGCs.Meansofmedianamplitudes,meanfrequencies,andmeandecaytimeconstantswerecomparedwiththetwo-tailedttest±standarderrorofthemean(SEM).WholecellcapacitancecomparisonswereperformedwiththeMann–Whitneytest±SEM.Statisticalsigni cancewassetatpb0.05.

Perfusionandpreparationoftissuesections

At90dayspost-CCI,someanimalsweredeeplyanesthetizedandtranscardiallyperfusedwithphosphatebufferedsaline(PBS;pH7.4),followedby4%paraformaldehydesolution.Thebrainswereremoved,post- xedinthesamesolution,andcryoprotectedin30%sucrose.Grossexaminationofthebrainswasperformedtocon rmthepresenceofTBI.Forty-micron-thickcoronalandhorizontalsectionswerecutonacryostat,collected,andstoredincryoprotectantsolution(phosphatebuffer/ethyleneglycol/glycerol)at 20°C.Standardhistological(Nissl)techniqueswereusedforsubsequentidenti cationofbraincytoarchitectureandmorphologicalchangesfollowingCCI.Nisslstaining

SectionswerewashedinPBS,mountedongelatin-coatedslides,andair-driedovernightatroomtemperature.Slidesweredehydrated,stainedwith0.5%cresylvioletsolution,rinsedindistilledwater,dehydratedagainthroughgradedalcoholsandxylene,andcoveredwithPermountandglasscoverslips.Results

Generalconsiderations

Twenty-threeCCI-injuredand34non-injured(control)Iperformedinourlaboratoryproducesaseverebraininjuryassociatedwithan～1%mortalityrateduringthe rst24h(unpublisheddata);therewasnoadditionalmortalitybeforeanimalsacri ceat90daysafterCCI.Ofthe23injuredanimalsusedinthestudy,11animalsweremonitoredcontinuouslyforaone-weekperiodpriortosacri ceandsubsequentvoltage-clamprecordings;controlanimalswerenotmonitored.Atotalof1918hofvideorecordingswasobtainedduringtheserecordingsessions,ofwhich1078h(924hlightson/154hlightsoff)wereofadequatetechnicalqualityforreview.Noseizurewasdetectedinanyanimal,similartohistoricalvideo-EEGrecordingsofcontrolSprague–Dawleyratsofsimilarage(Kharlamovetal.,2003).Fig.1showsarepresentativeratbrain(Fig.1A)andstructuralandmorphologicalchangesinthehippocampuswithNissl(Figs.1B,C)90daysafterCCI.PropertiesofsynapticreceptorsinDGCsfollowingCCI

DGCsreceivepowerfulinhibitoryinputfromGABAergicinter-neurons,whichformpredominantlyperisomaticsynapsesonDGCs(TraubandMiles,1991;HalasyandSomogyi,1993;Buhletal.,1994;BuckmasterandSchwartzkroin,1995;SolteszandMody,1995;Milesetal.,1996).Decreasedfeed-forwardinhibitioninthedentategyrushasbeenreportedwithindaysafterFPIandhasbeenattributedtolossofhilarGABAergicinterneurons(Tothetal.,1997).Itisimportanttonotethatpreviousstudieshavefocusedonrelativelyshort-termeffectsofTBIonGABAergictransmissioninthehippo-campus,typicallyassessingtissueonetoafewweekspost-TBI.However,neurologicalconsequencesofheadtrauma,suchascognitiveimpairments,posttraumaticstressdisorder,andPTEoftenmanifestthemselvesaftermonthsorevenyearsfollowingtheinitialinsult.Toinvestigatelong-termeffectsofTBIonGABAergictransmission,wholecellpatch-clamprecordingsofsIPSCsandmIPSCswereperformedinratcontrolandCCIDGCs90daysafterlesioning.SlicesweresuperfusedwithoxygenatedACSFthatcontained50μMDL-APVand20μparisonoffrequencyandpeakamplitudeofsIPSCsincontrolandCCIDGCsdidnotrevealsigni cantchanges.ThefrequencyofsIPSCswas4.95±1.27Hzincontrol(n=27DGCs,6animals),and5.58±1.96HzinCCIDGCs(n=16DGCs,4animals,p=0.93,two-tailedttest).Withinthesamegroupofcells,themeans