468Z.Mtchedlishvilietal./NeurobiologyofDisease38(2010)464–475

andBetz,2000;KittlerandMoss,2003;Alldredetal.,2005)andinphasicGABAergicinhibition(NusserandMody,2002).However,someγ2subunitexpressionhasbeendocumentedinextrasynapticlocations(Somogyietal.,1996;Nusseretal.,1998;Sassoè-Pognettoetal.,2000).TheinhibitorynetworkofthedentategyrusundergoessubstantialperturbationsafterTBI,suchaslossofGABAergicinterneuronsandsproutingofexcitatory bers(Santhakumaretal.,2000,2001;Golaraietal.,2001).Anetresultoftheseperturbationsisincreasedneuronalactivityinthetissuesurroundingtheinjurysite.Becauseincreasedneuronalactivityinhippocampalneuronscausesincreasedinternali-zationofγ2subunit-containing,butnotδsubunit-containingGABAARs(Goodkinetal.,2008;Terunumaetal.,2008;JoshiandKapur,2009),wetestedthepossibilitythattonicandphasiccurrentsmightbealtereddifferentiallyafterCCI.γ2subunit-containingreceptorsaresensitivetothebenzodiazepinediazepam,whereasα4/δsubunitsrenderthereceptorsdiazepam-insensitive.

RecordingsofmIPSCsfromcontrolandCCIDGCswereperformedwith300nMdiazepam.TheeffectsofdiazepamonmIPSCfrequency,peakamplitude,anddecaytimeconstantsat23°Cand34°CaresummarizedinTable1.Therewasnosigni canteffectonmIPSCfrequencybydiazepamat23°Cincontrol(1.03±0.18Hzvs.0.96±0.19Hz,n=9,4animals,p=0.43,two-tailedttest)andCCIDGCs(0.67±0.09Hzvs.1.21±0.21Hz,n=10,4animals,p=0.11,two-tailedttest),andat34°Cincontrol(3.10±0.84Hzvs.3.59±1.22Hz,n=16,3animals,p=0.57,two-tailedttest)andCCIDGCs(2.44±0.67Hzvs.1.42±0.27Hz,n=7,2animals,p=0.17,two-tailedttest).PeakamplitudeanddecaytimeconstantsofmIPSCswerealtereddifferentiallybydiazepamincontrolandCCIDGCsat23°CandinCCIDGCsat34°C(Figs.2A–H).At23°C,diazepamincreasedthemeanofmedianpeakamplitudeofmIPSCsinbothcontrolandCCIgroups,buttheincreasewas20%higherincontrolDGCsthaninCCIDGCs.Thesigni canceofthediazepameffectonpeakamplitudewasvalidatedinindividualDGCsbycomparingcumulativefractiondistributionsbytheKStest,beforeandafterdiazepam.TheKStestsuggestedastatisticallysigni canteffectofdiazepam(pb0.05)inallcontrolandCCIDGCs,thusrejectingthenullhypothesisthattheamplitudepopulationswereequalbeforeandafterdiazepam.BecausemIPSCamplitudesareskewed(demonstrateanon-Gaussiandistribution),meansofmedianpeakamplitudesweresubsequentlycompared.IncontrolDGCs,diazepamincreasedtheamplitude(26.81±2.20pAand42.60±1.22pA,n=9,3animals,p=0.031,two-tailedttest)andinCCIDGCs(33.46±2.98pAand46.13±1.09pA,n=10,5animals,p=0.047,two-tailedttest).At34°C,diazepamdidnotincreasepeakamplitudeincontrolDGCs(62.10±0.73pAvs.59.85±0.67pA,n=16,3animals,p=0.11two-tailedttest).However,diazepamdecreasedpeakamplitudeinCCIDGCs(53.10±0.59pAvs.36.06±7.09pA,n=7,2animals,p=0.03,two-tailedttest).

IncontrolDGCs,diazepamprolongeddecaytimeconstantsby21.5%at23°C(5.12±0.29msand6.59±0.08ms,n=9,5animals,p=0.002,two-tailedttest)andby21%at34°C(5.97±0.07msand7.22±0.19ms,n=12,6animals,p=0.001,two-tailedttest).InCCIDGCs,diazepam'sprolongationofdecaytimeconstantswasnotsigni cant:8.9%at23°C(5.48±0.03msand5.61±0.08ms,n=10,5animals,p=0.093,two-tailedttest)andby0.4%at34°C(6.59±0.12msand6.62±0.98ms,n=9,4animals,p=0.12,two-tailedttest).

PotentiationofinhibitionofsynapticcurrentsbyfurosemideafterCCIThediureticfurosemideisanoncompetitiveantagonistofGABAARsthatdemonstratesanenhancedaf nitytoα4subunit-containingrecombinantreceptors(Waffordetal.,1996)anddoesnotdiscriminatebetweenγandδsubunit-containingreceptors(KorpiandLuddens,1997).Recordingswereperformedwithfurosemideat23°Cand34°Candwerequalitativelysimilar;datafromrecordingsat34°Carepresented.TheeffectsoffurosemideontypicalrecordingsofmIPSCsare

presentedinFigs.3A–D.Thesigni canceofthefurosemideeffectonpeakamplitudewasvalidatedinindividualDGCsbycomparingcumulativefractiondistributionsbytheKStest,beforeandafterfurosemide.TheKStestsuggestedastatisticallysigni canteffect(pb0.05)offurosemideinallCCI,butnotincontrol,DGCs.Furosemide(100μM)wasbath-appliedtocontrolDGCsanddidnotalterthemeanofmedianpeakamplitudes(49.80±4.20pAatbaseline,and45.20±4.40pAinthepresenceoffurosemide,n=6,2animals,p=0.32,two-tailedttest),frequency(0.39±0.05Hzatbaseline,and0.36±0.06Hzinthepresenceoffurosemide,n=6,p=0.85,two-tailedttest),ordecaytimeconstants(6.42±0.30msatbaseline,and7.13±0.44msinthepresenceoffurosemide,n=6,p=0.13,two-tailedttest).InCCIDGCs,100μMfurosemidedecreasedthemeanofmedianpeakamplitudes(51.30±0.80pAatbaseline,and43.50±5.30pAinthepresenceoffurosemide,n=5,2animals,pb0.001,two-tailedttest).Furosemidedidnotalterthefrequency(0.42±0.09Hzatbaseline,and0.38±0.11Hzinthepresenceoffurosemide,n=6,p=0.34,two-tailedttest)ordecaytimeconstants(5.26±0.30msatbaselineand5.40±0.20msinthepresenceoffurosemiden=5,p=0.73,two-tailedttest).IncreaseoftonicinhibitioninDGCsafterCCI

InordertoassesspotentialalterationsintoniccurrentsinCCIDGCs,measurementsofwholecellcapacitancewereobtainedtoensurethatpotentialdifferencesintoniccurrentwerenotattribut-abletosigni cantdifferencesincellmembraneareas.MeasurementsofwholecellcapacitancerevealedsimilarvaluesincontrolandCCIDGCs(46.40±6.50pFincontrol,n=36,9animals,and36.30±8.50pF,n=27,8animals,inCCIDGCs,p=0.11,Mann–Whitneytest),suggestingrelativehomogeneityincellmembraneareasincontrolandCCIDGCgroups.

4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol(THIP)isalowaf nityagonistofGABAARswithpreferentialaf nityforδsubunits(Adkinsetal.,2001;Brownetal.,2002).1and3μMTHIPwereusedforpotentiationand20μMbicucullinewasusedforblockadeoftoniccurrents.Therecordingswereperformedat34°C.Quantita-tivemeasurementoftoniccurrentwasperformedbycomparingGaussiandistributionsofall-pointhistogramsof30-sepochsimmediatelybeforetheopeningoftheTHIP-containingreservoir(baseline)and3minafteropeningthereservoir.Bothconcentra-tionsofTHIPincreasedbaselinenoiseandshiftedholdingcurrentinbothcontrolandCCIDGCs.ThedataoftoniccurrentalterationarepresentedinTable2.1μMTHIPshiftedholdingcurrentincontrolDGCsfrom98.7±30.4pAto114.7±20.4pA(13%increase,n=13,6animalsp=0.0005,two-tailedttest)whereasinCCIDGCs,THIPshiftedholdingcurrentfrom119.8±34.6pAto176.3±42.2pA(47%increase,n=9,4animals,p=0.0003,two-tailedttest).Tocon rmthat1μMTHIPcausedagreaterpotentiationoftoniccurrentinCCIDGCs,werepeatedthisexperimentbyincreasingtheconcentrationofTHIPbyahalf-log.3μMTHIPshiftedholdingcurrentincontrolDGCsfrom88.9±20.6pAto233.7±50.4pA(162%increase),n=11,3animals,p=0.0005,two-tailedttest,Figs.4A,B),whereasinCCIDGCs,holdingcurrentwasshiftedfrom111.6±30.8pAto334.0±69.6pA)(198%increase,n=7,2ani-mals,p=0.0001,two-tailedttest,Figs.4C,D).AsteeperrisetimeofTHIP-inducedcurrentinCCIDGCssuggeststhat,comparedtocontrolDGCs,activationofalargernumberofreceptorsbyTHIPresultsinahigherrateofcurrentamplitudeincrease.

InordertocorroboratethepresenceoflargertoniccurrentinCCIDGCs,20μMbicucullinewasbath-appliedtothesameDGCswhereholdingcurrentwaspotentiatedby1and3μMTHIP.WefoundthatsubstantialamountsoftimewererequiredforTHIP-inducedcurrenttoequilibratetoasteadylevel,likelyduetoaveryslowdesensitizationrate.Inmanycases,thequalityoftherecordingduringprotractedapplicationofTHIPwouldoftendeterioratetosuchanextentthatapplicationofbicucullinewasnotfeasible.Ouraimwastodetectthe