ABBREVIATIONS:GABA, @-aminobutyncacid;SNpr,substantianigraparsreticulata;FZP,flurazepam;EPSP,excitatorypostsynapticpotential;DZP diazepamISO isoguvacineMUS muscimolTHIP 4,5,6,7-tetrahydroisoxazolo[5,4-cJ-pyridin-3-olBAC baclofen.

0022-3565/92/2621-0204$03.00/OTHE JOURNALOFPHARMACOLOGYANDExpxa@srr@ THERAPEUTIC5Copyright C 1992 by The American Society for Pharmacology and Experimental Therapeutics

ReductioninPotencyof Selective‘¿�y-AminobutyricAcidAAgonistsandDiazepaminCAl Regionof in VitroHippocampalSlicesfromChronicFlurazepam-TreatedRats1

XIANG-HUI XIE and ELIZABETH I. TIETZDepartmentof Pharmacology,MedicalCollegeof Ohio,Toledo,OhioAcceptedfor pu@ation March27, 1992

VoL262, No.1Printed in U.S-A.

ABSTRACTThe potency and efficacy of selective y-aminobutyric a@ldA(GABAA)agonists (GABA, muscimol, isoguvacine and 4,5,6,7-tetrahydr@soxazoIo-[5,4-cJ-pyridin-3-ol),the GABAÃagonist, baclofen,and the benzodiazepineagonist, diazepam, were examined usingextracellular recording techniques in in vitro hippocampal slicesfromrats sacrificed 2 days after 1 week of flurazepam treatment.Population spikes elicited by stimulation of Schaffer collateralswere recorded in the CAl pyramidal cell region with NaCIcontaining glass micropipettes. GABA agonists were superfusedin increasing concentrations for 5 mm. Drug responses, averagedover the last 2 mm for each concentration,were comparedtothepredrugbaseline.GABAAagonists,butnotbaclofen,showed

a significant,2-fold, decrease in potency, but not efficacy, toreduce CAl-evoked responses in treated vs. control slices. Thebenzodiazeplne effect was evaluated by the shift in the isoguvaane dose-response curve in the absence, then presence, ofdiazepam. A reduction in diazepam potency was demonstratedin vitro by a significantlyreducedshift in the isoguvacinecurveby 300 nM, but not 1 aiM,diazepamafter chronicbut not acutein vWopretreatment.The results indicateda s@ectiveGABA,agonist subsensitivfty and diazepam tolerance in hippocampusafter 1 week of flurazepam treatment and establish the hippocampal slice preparation as a valuablesubstrate for investigatingsynaptic mechanisms of benzodiazepine tolerance.

Prolonged administration of benzodiazepineS to animals results in the development of functional tolerance, a reduction inthe sensitivityof the central nervoussystemto benzodiazepineactions (ci. Rosenberg and Chiu, 1985). The pharmacologicalactionsof benzodiazepines,which bind to an allostericsite onthe central GABAAreceptor,are mediatedby the enhancementofa GABA-activated C1 conductance (Haefely, 1985). Althoughthe precisecellularand molecularmechanismsunderlyingbenzodiazepine tolerance are not well understood, many changesat the GABAA receptor have been detected after chronic benzodiazepine treatment. These include downregulation of benzodiazepine binding sites (Tietz et al., 1986, 1989;Miller et at.,1988),upregulation of low-affinity GABAA receptors (Gallageret at., 19Mb) and a reducedcoupling betweenGABA andbenzodiazepinebinding sites (Gallager et at.; 1984a;Tietz et al.,1989).Decreasedexpressionofthe mRNAs for the a1(Heningeret at., 1990; Kang and Miller, 1991) and @2(Heninger et al.,

Receivedfor publication February 4, 1992.1 This work was supported by Grant RO1-DA-04075 from the National Insti

tute on Drug Abuseand a BiomedicalResearch Support Grant S07-RR-050700from the Department of Health and Human Services. Portions of this manuscripthave been published in abstract form: Fed. Am. Soc. Exp. Biol. J. 5: A499, 1991and Soc. Neurosci. Abstr. 17: 264, 1991.

1990)subunits of the GABAA receptor have also been reported.Changes in GABA function associated with these receptorrelated modifications include a decreasein GABA agonistmediated Cl influx (Miller et at., 1988; Marley and Gallager,1989) and a reducedability of benzodiazepinesto enhanceGABA-stimulated Cl influx in brain microsacs (Yu et at., 1988).Benzodiazepine tolerance and GABA agonist subsensitivityhave alsobeen observedin brains of chronicbenzodiazepinetreated rats using in vivo electrophysiological (Gailager et at.,1985;Wilson and Gallager, 1988;Tyma et at., 1988)and behavioral methods (Tietz and Rosenberg, 1988; Ramsey et at., 1991).Alterations in the GABAA receptor with chronic benzodiazepinetreatment and the accompanyingchangesin GABA functionshow a regional heterogeneity and have been localized to,amongother brain areas,cerebralcortex, dorsal raphé,SNprand hippocampus (Gallager et at., 1985;Tietz et at., 1986;Milleret at., 1988; Tyma et at., 1988; Wilson and Gallager, 1988).

The hippocampusis an important site of benzodiazepineactions (Haefely, 1985), in particular anticonvulsant actions(Rock and Taylor, 1986).Recently, by using an vitro hippocampal slice preparation, proven to be a useful substrate for thestudy of GABA/benzodiazepinepharmacology(Kemp et at.,1986;Kemp et at., 1987),we demonstrated a significant reduc

204

at Univ O

f Toledo on July 30, 2009

jpet.aspetjournals.orgD

ownloaded from

HippocempalBenzodiazeplneTolerance 2051992

tion in paired-pulse inhibition in CAl region of hippocampusof rats sacrificed 2 days, but not 7 days, after 1 of week FZPtreatment. Moreover, there was a significant prolongationofthe field EPSP half-decay time 2 days after FZP treatmentwithout a change in its initial slope or maximal amplitude (Xieand Tietz, 1991).Thesefindingsare consistentwith an impairment in endogenousGABA functionin hippocampusof chronicbenzodiazepine-treated rats which coincides with benzodiazepine anticonvulsant tolerance in the whole animal (Rosenberget at.,1985,1991).

One purposeof this studywasto directly test the hypothesisthat GABAA 1@Ceptor function is impaired in in vitro hippocam

pal slicespreparedfrom chronicbenzodiazepine-treatedratsbyevaluating the ability of exogenous GABAA and GABAB agonists to inhibit CAl pyramidal cell-evoked responses.Becausebehavioral studies from our laboratory had shown that theregulation of benzodiazepine and GABAA agonist effects weredissimilar in SNpr after chronic benzodiazepine treatment(Tietz and Rosenberg,1988;Ramseyet at., 1991), experimentswerealsodesignedto investigatethepossibilityof differentialregulation in hippocampus by measuring the ability of severalGABA agonists to inhibit CAl-evoked responsesafter chronicbenzodiazepine treatment. The electrophysiological actions ofbenzodiazepines can be measured in the hippocampal slicepreparation by their ability to potentiate the effect of GABAagonists (Kemp et at., 1987).To establish whether benzodiazepine tolerance could be measured locally in hippocampus afterchronic benzodiazepine treatment, the effect of DZP was testedin hippocampal slices prepared from chronic benzodiazepinetreated rats by its capacity to shift the dose-responsecurve ofthe GABAAagonist,ISO.

Materials and Methods

Chronic and acute benzodiazepine treatment. The 1-weekFZPtreatmentusedin thisstudyhasbeenshownto reSultin tolerancetothe antipentylenetetrazol effects of the benzodiazepines (Rosenberg etal., 1985, 1991), subsensitivity to the behavioral actions of benzodiazepine and GABA agomstsmicroinjectedinto SNpr (Tietz and Rosenberg,1988),suppressionof spontaneousfiring of SNpr neurons(Tymaet at., 1988)and downregulationof benzodiazepinereceptorsin severalbrain regionsincluding SNpr and hippocampus(Tietz et al., 1986).This chronictreatmentdoesnot produceanyovertbehavioraleffects,e.g.,ataxia, or does it result in spontaneous or precipitated withdrawaleffects(Tietz and Rosenberg,1988).Detailsof the chronic treatmentregimen have been provided previously (Rosenberg et at., 1985, 1990;Tietz et at.,1986;Tietz andRosenberg,1988).

After a 2-day adaptation period during which rats received a 0.02%saccharin water vehicle, male Sprague-Dawley rats (initial weight, 175—200 g) were offered FZP for 7 days (100 mg/kg x 3 days and 150 mgIkg x 4 days) in their drinking water. Only rats that consumed anaverage of 100 mg/kg/day for the 1-week treatment period were included in the study (Tietz and Rosenberg, 1988; Ramsey et a!., 1991;Rosenberg et at., 1991). After the FZP treatment, rats received saccharm vehicle for 2 days until sacrifice. Paired-handled controls received saccharin water for the entire treatment period.

In order to assure that the effects observed in hippocampal sliceswere specific to chronic benzodiazepine treatment, an acute dose of theFZP active metabolite, desalkyl-FZP, was given to another group ofrats. The acute desalkyl-FZP dose to be administered was determinedin preliminary radioreceptorassaystudiescarried out as describedpreviously (Xie and Tietz, 1991). Briefly, ethanol extracts of whole

brain from rats given several different doses ofdesalkyl-FZP (1-10 mgIkg p.o.) or 1 week of chronic FZP treatment were compared for theirability to displace2 nM [3HJFNPbinding.The levelof benzodiazepine

activity in the brain immediatelyafter the end of treatment wasestimatedfroma standardDZP displacementcurveandexpressedinDZP equivalents,nanogramsof DZP per/gramof brain. A singledoseof desalkyl-FZP(2.5mg/kgp.o.) resultedin a levelof benzodiazepineactivity in brain (162.4 ng/g of DZP; 0.57 zM, n 2) equivalent to thatfound in the brains of rats after 1 week of chronic FZP treatment(161.1 ±35.9 ng/g DZP; 0.57 @M,n = 6). Twelve hours before intubation, the food was removed from the cage. Two days before sacrifice,2.5 mg/kg of desalkyl-FZP was administered by gastric intubation inan emulsion ofpeanut oil, water and acacia (4:2:1). Control rats receivedthe same volume of the emulsion vehicle by gastric intubation.

Hippocampal slice preparation. Hippocampal slices (400 tim)from treated or control rats were cut on a tissue chopper (Stoelting,Wood Dale, IL) from the left dorsal hippOCampUs. Slices were placedin ice-cold pregassed (95% 02-5% C02) buffer [(mM): NaCl, 120; KC1,5.0; MgSO4,1.3; NaH2PO4,1.2; CaCl2,2.4; NaHCO3,26.0;and Dglucose, 10.0; 288 mOsm, pH 7.6], then transferred to a recordingchambermaintainedat 33-34°C.After sliceswereallowedto equilibratefor at least 1 hr at an interfacewith buffer and humidified gas,theslicesweresuperfusedwith warm,gassedbuffer for 30mm.

Extracellular CAl-evoked responses. CAl population spikes(fig. 1)elicitedby stimulationof the Schaffercollateralpathwaywererecorded in the stratum pyramidale of CAl with 2 M NaC1-filled glassmicroelectrodes (2—5megohms). A concentric, bipolar, tungsten-stunulating electrode was placed into the stratum radiatum near the CAllCA2 border and a monophasic, O.1-msec square-wave pulse (3-15 V)was delivered at 30-sec intervals with an S44 stimulator (Grass Instrument Co., Quincy, MA) connected to a stimulus isolation unit. Theamplified (1000-fold), filtered (10 Hz, 3 kHz) population spikes, monitored on an audio monitor, were digitized using a MacLAB analog-todigital converterunit (World PrecisionInatruments,Inc., Sarasota,FL) interfaced with a MAC II computer, and were displayed usingMacLAB software which emulated a digital oscilloscope. Evoked response data was stored on computer disk for subsequent analysis.

A stimulus-response curve [population spike (millivolts) us. stimulusintensity (volts)] was constructed for each electrode placement byincreasingstimulusintensity manuallyin 0.5-to 1-V incrementsfromthreshold (the stimulus intensity required to elicit a 1-mV potential)until the maximum population spike amplitude was achieved. Halfmaximum (1/2 Max), the stimulus intensity required to elicit a population spike 50% of maximum,was usedas the standardstimulusintensity. Healthy responses were selected for analysis based on themagnitudeof themaximumevokedresponse(>2.7mV)andtheabsenceof multiple population spikes before and after drug application. Aftera stable response was obtained, the last four responses were averagedto serve as the predrug base line. Only one slice was used from each

treated or control rat.GABA agonist effects. GABA agonistsweredissolvedin distilled

water, at a concentration 100 to 1000 times the desired final concentration and weredeliveredto the perfusatewith a calibratedsyringepump (Raze!, Stamford, CT) at a rate of 10 to 100 @tl/min.Eachconcentration was perfused for 5 mm to ensure that a maximal responsewas achieved and the last four responses at each concentration averagedto minimize inter-response variability. Drug doses were added cumu

latively and the dose-response curve was constructed by plotting drugconcentration against the percentage reduction of the baseline population spike amplitude. Drugs were washed out with buffer for 20 to 25mm until the postdrug response returned to 95 to 99% of the predrugresponse.

In preliminary studies in slices from control rats, the effect ofsuperfusion time on the magnitude of the evoked response was evaluated in the presence of buffer alone or 2.5 gzM ISO. There was a smallincreasein the sizeof the populationspikeover 60 mm during bothbuffer and ISO perfusion (7.5% vs. 6.6%). The doses of each GABAagonistto be testedwerechosenin other preliminary studies(MIJS,0.5—20zM; ISO,2.5—50tiM; THIP, 25-500tiM; GABA,60 @iM-1mM;and BAC, 1-12 tiM). The effects of GABA in treated and control sliceswere determined in the presence of the GABA uptake inhibitor, nipe

at Univ O

f Toledo on July 30, 2009

jpet.aspetjournals.orgD

ownloaded from

206 XIeandTletz Vol. 262

cotic acid (200 tiM). The IC@for nipecotic acid blockade of GABAuptake in rat brain slices was reported to be 9 ±1 @M(KrogsgaardLarsen and Johnston, 1975). Nipecotic acid alone (200 tiM) had noeffect on the population spike whereas 500 @tMdepressed baselineevoked responses. MUS, ISO, THIP (Brehm et a!., 1979) and BAC(Boweryet al., 1983)arenot substratesfor GABAuptakesites.

Benzodiazepineagonisteffects.Becausethebenzodiazepinesactby potentiationof GABA function,benzodiazepineswouldbeexpectedto potentiate the inhibitory effects of GABA on the CAl populationspike in a dose-dependent manner as shown by Kemp et at. (1986). Theaction of DZP was tested in slices from saccharin-treated control ratsby its ability to shift the GABAA agonist, ISO, dose-response curve (fig.4). After superfusion with ISO (2.5-20 NM), DZP (100 nM-i MM,dissolved in 1% ethanol) was added to the superfusate for 20 mm (20Mi/mm) before a 2nd ISO dose-effect curve was determined in thepresence of DZP. A single slice was used for each DZP concentrationtested. The experiment was repeated 3 to 8 times per DZP concentration. The results of these studies served as the basis for the DZPconcentrationschosenfor studyin FZP-treatedandcontrolslices.

Data analysis. The population spike was measured from the midpoint ofthe tangent to the field EPSP onset and offset to the maximumnegative point of the population spike (fig. 1) (Adamec et at., 1981;Tuff et al., 1983; Xie and Tietz, 1991). GABA agonist potency wasmeasured by the concentration of agonist which produced 50% inhibition of the population spike (EC@).Because the concentrations of ISOand DZP chosen for studies in hippocampal slices derived from treatedand control rats did not always result in 50% inhibition, the shift ofISO response by DZP was measured as the dose of ISO to produce 40%inhibition of the population spike (EC@o)in the absence, then presence,of DZP. The ability of the benzodiazepineto potentiate the ISOresponse was plotted as the —¿�logdose ratio [ECho (ISO + benzodiaze

pine)/EC@@,(ISOalone)Jvs.benzodiazepineconcentration(nanomolar)(fig.4) (Kemp et a!., 1987). The differences in the concentrationresponse curves among groups were analyzed by a two-way repeatedmeasures analysis of variance. Post-hoc comparisons were made byBonferronit testbetweencontrolandcorrespondingtreatmentgroups.Changes were deemed significant if P < .05.

Drugs. Drugswereobtainedfromthe followingsources:DZP anddesalkyl-FZP (Hoffmann-La Roche Inc., Nutley, NJ); GABA, nipecoticacid and BAC (Sigma Chemical Co., St. Louis, MO); MUS (ResearchOrganics, Cleveland); ISO hydrochloride and THIP hydrochloride (Research Biochemicals, Inc., Wayland, MA).

Results

Effect of GABA and benzodiazepine agonists on theCAl-evoked response. Electrical stimulation of Schaffercollaterals elicited a synchronous firing of the CAl pyramidalcells that could be measuredextracellularly. An example ofextracellular recordingsand the effect of the selectiveGABAAagonist, ISO, to inhibit the population spike in slices fromcontrol and FZP-treated rats is shown in figure 1. GABAagonists inhibited the CAl population spike in in vitro rathippocampus in a dose-related manner with relative potencies(MUS > BAC > ISO > THIP > GABA) similar to that reportedby Kemp et al. (1986) (figs. 2 and 3). The specific GABA uptakeblocker, nipecotic acid (200 MM),increased GABApotency 2.5-fold (fig. 2). Consistent with previous findings (Kemp et at.,1986), BAC, a selective GABAB agonist, produced 100% inhibition of the CAl population spike.

As shown in figure 4, DZP dose-dependently potentiated theinhibitory effect of ISO on the CAl population spike. Higherconcentrations of DZP (>1 MM) had direct, presumably non

specific, inhibitory effects on the amplitude of the CAl population spike (data not shown) (Kemp et at., 1987). The 1%ethanol vehicle alone had no effect on the base-line response.

During perfusion of slices with 1% ethanol over 60 min, thesize of the population spike increased by 6.9%, similar to thechange in the evoked responseover time during perfusion withbuffer alone. Based on the results of DZP dose-response studies(fig. 4), DZP at concentrations of 300 nM (ECro) and 1 @iM(EC1®)were chosen for studies in slices from FZP-treated andcontrol rats. Choosing concentrations on the linear (ECro) andmaximal (EC1@) portions of the DZP concentration-responsecurveallowedfor detectionof changesin DZP potencyand/orefficacy.

Effects of chronic FZP treatment on GABA agonistactions. GABA (fig. 2, P = .047) and the selective GABAAagonists,MUS, ISO and THIP (fig. 3), showedsignificant (P< .001) decreases in potency in hippocampal slices derived fromrats sacrificed 2 days after ending 1 week of FZP treatment.This was indicatedby a rightward shift in the dose-responsecurves for their ability to inhibit the CAl population spike intreated compared to control slices (ECse in micromolar: GABA,186.4 vs. 112.8;MUS, 6.1 vs. 2.4; ISO, 23.6 vs. 5.2; and THIP,238.3vs.106.7;n = 6/group).However,therewasnosignificantdifference(P = .56) in the CAl responseto the selectiveGABABagonist, BAC, between treated (ECro, 7.2 MM) and control slices(EC@,6.5 MM) (fig. 2).

Effect of chronic FZP treatment on benzodiazepineagonist actions. DZP was used to evaluate the developmentof tolerance in vitro inasmuch as it is a full agonist at thebenzodiazepine receptor in this preparation (Kemp et at., 1987)and because anticonvulsant tolerance could be measured toDZP in vivo after chronic FZP treatment (Rosenberg et at.,1991). Two days after 1 week of FZP treatment, there was asignificant decreasein the ability of DZP to shift the ISO doseresponse curve (2.5—20 MM) in treated (1.3-fold shift) vs. control

(1.8-fold shift) slices (fig. 5). The decreasedability of DZP toshift the ISO responsewas observed at 300 nM DZP (n = 8/group, P = .02), but not 1 @MDZP (n = 6/group, P = .49;treated. 2.5-fold shift; control: 2.4-fold shift), indicating a decrease in DZP potency, but not efficacy in CAl region afterchronic FZP treatment.

Effect of acute desalkyl-FZP treatment on GABA andbenzod.iazepine agonist actions. Two days after a singledoseof desalkyl-FZP (2.5 mg/kg p.o.), slicesfrom both acutelytreated and control rats were tested. There was no significantdifference(P = .73) in the ability of ISO to decreasethe CAlevoked responsein slices from acute desalkyl-FZP vs.emulsionintubated rats (fig. 6). There was also no significant difference(P = .40) betweengroupsin the ability of 300 nM DZP to shiftthe ISO dose-responsecurve(fig. 6; 2.0-foldvs.1.7 fold-shift;n= 4/group). There was no change in the effect of ISO or ofDZP in hippocampal slices from rats sacrificed 2 days afterpretreatment with a single dose of DZP (10 mg/kg p.o. in 0.5%TWEEN 80, n = 3/group) vs. control slices (data not shown).This acutetreatment alsoresultsin benzodiazepinebrain levelsequivalent to that found after 1 week of FZP treatment (Tietzet at.,1986).

Discussion

The principal findingsof this studywerea reductionin bothGABAAand benzodiazepineagonisteffectsin CAl regionof invitro hippocampal slices from rats 2 days after cessation of 1weekof FZPtreatment. There wasa significantdecreasein theactions of superfused GABAA agonists (GABA, ISO, MUS and

at Univ O

f Toledo on July 30, 2009

jpet.aspetjournals.orgD

ownloaded from

HippocampalBenzodiazepineTolerance 2071992

CONTROL

.@@

Fig.1. Representative extracellularevoked potentials(populationspikes) re

RECOVERY corded in CAl pyramidal celllayer in controlandtreatedhippocampalslicesduringsuperfusionof ISO (2.5—50MM).Horizontaldistancebetweendots = 5 msec;verticaldistancebetweendots = 2 mV.

Fig. 3. MUS (A, is), ISO (U, 0) and THIP (S. 0) concentration-responsecurves in hippocampalslices from control rats (A, R, •¿�)and ratssacrificed48 hr after 1 week of FZP treatment (& 0, 0). Drugs (10—i00MI)were added to the superfusate (2 mI/mm), incrementing the concentrationevery5 mm.Drugresponses,averagedoverthe last2 mm,werecalculatedas a fractionof the predrugbaseline.Eachpoint representsthe mean±S.E.M.for eachconcentration.Eachcurvewas generatedfromsixslicestakenfromsix rats/groupper drugtested. Therewas asignificant decrease in potency between treated and control sl@esforeachdrug(ECsein micromolar:MUS,6.1vs.2.4; ISO,23.6vs.5.2;andTHIP,238.3vs. 106.7).

BASELINE__ @M@MIL*MI@aM1 WEEK FZP TREATED

BASELINE 2.@.5.sM i_sM 1.@M LtaM LIaM

C

a

@0

0LU

RECOV@Y

C00.

Fig.2. CumulativeBAC(•,0) andGABA(A, @;plus200MMnipecoticacid)concentration-responsecurvesin hippocampalslicesfromcontrolrats(A,•¿�)andratssacrificed48 hrafter1 weekof FZPtreatment(&0). Drugs (10—100MI)were added to the superfusate (2 mI/mm), incrementingthe concentrationevery5 mm.Drugresponses,averagedoverthe last 2 mm,were calculatedas a fractionof the predrugbaseline.GABAshoweda significantdecreaseinpotencyintreated(ECse,186.4MM)vs. controlslices(ECsc,112.8MM).Therewas no significantdifferenceintheresponseto BACbetweentreated(ECse,7.2MM)&id COfltI'OIslices(ECse,6.5MM).Eachpointrepresentsthemean±S.E.M.for eachconcentration.Eachcurvewasgeneratedfromsix slicestakenfromsixrats/group.Theresponseto GABAalone(ECse,294.7MM;intheabsenceof nipecoticacid)is alsoshown(——¿�@,threeslices/threerats).

[@tM] (pM]

and was only observedafter superfusionwith a DZP concentration of 300 nM (ECsc), not 1 MM (ECioo), indicating a

decreasein DZP potency, not efficacy (fig. 5).This is the first report of a reduction in both GABAA and

benzodiazepine agonist potency, rather than efficacy, after a 1-weekchronicbenzodiazepinetreatment. The presenceof residual FZP or its active metabolites in the slice, which would resultin an apparent increase in GABA and benzodiazepine agonistpotency, could not have affected these observationsbecauseresidual benzodiazepines were not detectable in hippocampus

2 days after withdrawal from 1-week FZP treatment (Xie andTietz, 1991). These data suggestthat there is an alteration inthe affinity of both GABA and benzodiazepinebinding sitesduring chronic benzodiazepine treatment. This conclusion issupported by the finding that the rank order of potency of

THIP) to inhibit population spikes recordedextracellularlyfrom CAl pyramidal cells (figs. 2 and 3). GABAAagonistconcentration-responsecurveswere shifted 2-fold to the rightwithout a changein maximal effect, indicating a decreaseinGABAA agonist potency, but not efficacy. The reduction in thepostsynaptic actions of GABA in hippocampal slices was selective for the GABAA receptor because there was no change inthe action of the GABABagonist,BAC, to inhibit CAl-evokedpotentials after chronic FZP treatment (fig. 2). In addition, theability of the benzodiazepineagonist,DZP, to shift the ISOdose-response curve was reduced in chronically treated as compared to control slices. The tolerance to DZP measured inhippocampal slices was produced by chronic FZP adininistration, but not by a single desalkyl-FZP pretreatment (fig. 6),

at Univ O

f Toledo on July 30, 2009

jpet.aspetjournals.orgD

ownloaded from

1VUA

80

60

40

20

0 ,1

80

208 XleandTietZ Vol.262

the finding that the allosteric interaction between GABA andbenzodiazepinereceptorsiteswas not altered in hippocampalmembranes after chronic FZP administration (Tietz et at.,1989).

The current results extend our previous findings of a reduction in paired-pulseinhibition and a prolongationof the EPSPhalf-decay time in CAl region of hippocampus 2 days after theend of 1-week chronic FZP treatment (Xie and Tietz, 1991).The subsensitivityto exogenousGABAA and benzodiazepineagonists in the CAl region supports the hypothesis that thereduction in paired-pulse inhibition in the hippocampus wasdue to a decreasein endogenousGABA-mediated inhibitionand suggeststhat the decreasedinhibition is due, at least inpart, to a decreasedpostsynaptic responseto GABA, althougha change in GABAA presynaptic receptors cannot be ruled out(Waldmeir and Baumann, 1990). The significant correlationbetween GABAA agonist potency and the average duration ofCl channel opening in spinal cord cultures (Barker and Mathers, 1981) would suggestthat the reduction in GABAA agonistactionsin the CAl pyramidal cell regionof the hippocampusafter chronic FZP treatment may be the result of an alterationin Cl channel kinetics rather than a decrease in conductance.With respect to benzodiazepine actions, lowering benzodiazepine binding affinity by photoinactivation of the benzodiazepinereceptorreducedthe ability ofchlordiazepoxideto increaseGABA-induced C1 conductance (Gibbs et at., 1985). Chronictreatment of spinal cord and cortical cell cultures with clona

zepam resulted in a decreasein benzodiazepine affinity for thereceptor (Sher et at., 1983; Sher, 1986) and was associated with

a reduced ability of DZP to facilitate GABA-induced conductance (Sher et at., 1983). The reducedpotencyof DZP in hippocampalslicesfrom FZP-treated rats might resultin a similarchange.Intracellular current and voltage clamp recordings fromCAl pyramidal cellsin slicesfrom benzodiazepine-treatedratsshould further elucidate the synaptic mechanisms underlyingthe impairmentin GABA-mediatedinhibition in the hippocampus of benzodiazepine-tolerant rats.

0

0

0,0

0.6

0.5

0.4

0.3

0.2

0.1

0.0

-0.1

Dlazepam[nM]Fig.4. Concentration-responsecurvefortheabilityof DZPto increasethepotencyof ISOinhippocampalslicesfromcontrolrats.DZPconcentrationwasplottedagainstthe —¿�log[DoseRatiol.[DoseRatio]= EC@(ISO+ DZP)/EC@(ISO).Eachpointis themean±S.E.M.of threetoeightdeterminations.

GABA (Kemp et al., 1986) and benzodiazepine(Kemp et al.,1987) actions in hippocampal slices was highly correlated withtheir ability to displace [3H]GABA and [3H]R015—l788binding, respectively, to hippocampal membranes. The lack ofchange in DZP efficacy is consistent with the reversal ofbenzodiazepine receptor downregulation in hippocampal pyramidal cell regions within 2 days after the end of chronic FZPtreatment (Tietz et at., 1986) and suggest that benzodiazepinereceptor downregulation (Rosenberg and Chiu, 1981; Tietz et

at., 1986; Miller et al., 1988) may not be necessaryfor themaintenance ofbenzodiazepine tolerance in hippocampus. Furthermore,the absenceof a decreasein benzodiazepineefficacy,which hasbeenshownto be related to the ability of GABA toalter the affinity ofbenzodiazepines for their receptor (“GABAshift―)in hippocampus (Kemp et at., 1987) is consistent with

100 1000

B

Ca&:@LUI-

C.)0C0

.0

.CC

60

40

20

Fig.5. DZPwas tested in slices2 daysafter1weekofFZPtreatmentbyitsabilityto shifttheISOdose-responsecurve.Aftersuperfusionof ISO (2.5-20 MM),DZP (300nM or 1 MM)wassuperfusedfor20 mmbeforea secondISOconcentration-effectcurvewas determinedin the presenceofDZP.Incontrolslices(A andC),300 nMDZP(A;n = 8 slices/8rats)and1MMDZP(C;n = 6 slices/6rats)significantlyShiftedthe ISOcurve(ISOalone,•¿�;ISO+ DZP,0). In treated slices (B and D), the abilityof DZPto shift the ISOconcentration-response curve (ISOalone,A; ISO + DZP =L@)was significantlydecreasedwith 300nMDZP(B;n=8shces/8rats)butnotlMMDZP(D;n= 6 slices/6rats).Eachpointrepresentsthe mean±S.E.M.for eachconcentration.

10 100

80

60

40

20

1'0

Isoguvacine [piM] Isoguvaclne[@tM]

at Univ O

f Toledo on July 30, 2009

jpet.aspetjournals.orgD

ownloaded from

HippocampalBenzodiazeplneTolerance 2091992

ment. These studies had demonstrated the differential regulation of the circling responseto several GABAA agonistsandFZP microinjectedinto SNpr (Tietz and Rosenberg,1988;Ramsey et at., 1991).The time course for the development andthe reversal of tolerance to FZP and subsensitivityto MUSweredissimilar(TietzandRosenberg,1988)andsubsensitivitywas demonstrated to GABA, but not to THIP (Ramsey et al.,1991). A possible basis for this hypothesis may be found inother studies in the hippocampus (Alger and Nicoll, 1982;Woodson et at., 1989).A recent quantitative immunocytochemical study showed a heterogeneous population of GABAergic

cells in the rat hippocampal formation. Two different types ofGABA irnmunoreactive neurons were identified in the CA1/CA2 interface region (Woodsonet at., 1989). The existenceoftwo distinct, i.e., synapticand extrasynaptic,GABAAreceptorshave been proposed to mediate GABA-induced hyperpolarization and depolarization,respectively(Alger and Nicoll, 1982).THIP preferentially inducedhyperpolarization(Alger and Nicoll, 1982) and has been found to label a subset of GABAreceptors (Falch and Krogsgaard-Larsen, 1982). In hippocampal slicesfrom chronicFZP-treated rats, the potencyof eachGABAA agonist was reduced to a similar extent (2-fold) 2 daysafter chronic benzodiazepine treatment. Although the shift inISO potencywasinitially 4-fold (fig. 3), this wasreducedto 2-fold when the number of test dosesused was decreasedfrom 8to 4 (figs. 5 and 6). This was due to the increase in potency of

ISO in control sliceswhen the greater number of doseswasused,perhaps due to a cumulative drug effect. The shift in theISO curvein treated sliceswasmaximal (EC@@ 20 MM) usingboth paradigms. Both subsensitivity to GABAA agonists andtolerance to DZP were detected concurrently in the hippocampus, suggesting that regulation of their actions may not followdifferent time coursesasthey did in SNpr. Studiesat differenttime pointsafter chronictreatment wouldbenecessaryto rejectthe hypothesisthat responsesto GABAA agonistsand benzodiazepines are regulated differentially in the hippocampus.Intracellular recordingmethodsand localizedpressureejectiontechniques would be a logical approach to further investigatethe possibility of differential regulation in the hippocampusafter chronic benzodiazepine treatment.

Numerous experiments have demonstrated regional variations in the function of the GABAA receptor complex afterchronic benzodiazepine treatment. For example, subsensitivity

to iontophoretically applied GABA was demonstrated in dorsalraphé(Gallager et at., 1984;Wilson and Gallager, 1988)but notin SNpr neurons (Wilson and Gallager, 1987). Furthermore, nochanges were found in the GABA iontophoretic currents required to inhibit CA3 pyramidal cell firing in the hippocampusof chronic DZP-treated rats (Lista et at., 1990). The exact

mechanisms mediating benzodiazepine tolerance still remainunknown. Regional variation in benzodiazepine and GABAreceptor binding and changes in their allosteric interactionhave also been associated with chronic benzodiazepine treatment (Gallager et at., l984a, 1985;Tietz et al., 1986;Miller etat., 1988;Tyma et at., 1988;Wilson and Gallager, 1988;Tietzet at., 1989). The regionalvariations in changesin the GABAcomplex during chronic benzodiazepine treatment may reflectthe differential distribution of GABAA receptorsubunit (aie,$1—3, Y2 and 6) mRNAs in the brain, specifically in hippocampal

laminae (Möhleret at., 1990). Variable expression of the genesencodingfor GABAA receptorsubunitscould lead to the synthesis of a GABAA receptor with an altered affmity for GABA

@100@

a@80@

@60

0

C0

@ 0@ .@ . U

C 1 10 100

@ Isoguvaclne[pM]Fig. 6. The ability of DZP (300 nM) to shift the ISO concentrationresponsecurve(2.5—20MM)was testedin vitro 2 daysafter a single,acutepretreatmentwith deSaIkyI-FZP(2.5 mg/kg p.0.).Therewas nosignificantdifferencebetweenthe ISOconcentration-responsecurvegeneratedin treated(A; EC@= 13.7MM)vs. controlslices(•;EC@=10.2 MM)or inthe actionof 300 nMDZPto shiftthe ISOcurve(ISO+DZP:treated, @,EC@= 7.0 MMvs. control;0, EC@= 6.1 MM).Eachpointrepresentsthemean±S.E.M.for eachconcentration.Eachcurvewas generatedfrom four sl@estaken from four rats/groupper drugtested.

The lack of changein the actionsof BAC in treated slicessuggests that the impairment in GABA-mediated inhibition inthe hippocampusis a selectivedysfunctionmediatedby GABAA, rather than GABAB receptors.GABAB receptorswhichhavebeenlocalizedtotheCAl areaoftheratbrain(Chuetat.,1990)mediate both pre- and postsynaptic GABA actions (Blaxter and Carmen,1985).

Activation of postsynaptic GABAB receptors results in thehyperpolarization of hippocampal CAl pyramidal cells (Newberry and Nicoll, 1984; Reynolds et at., 1990) through anincrease in K' conductance (Newberry and Nicoll, 1984; Andrade et at., 1986).Thus, the primary action of superfusedBAC,as found in this study (fig. 2), was inhibition of CAl-evokedresponses (Kemp et at., 1986). GABAB receptors are not, likeGABAA receptors, coupled to the benzodiazepine receptorlinked Cl channel (Haefely, 1985). Therefore, chronicbenzodiazepine treatment would not be predicted to modulate postsynaptic GABAB receptor function. To our knowledge, this isthefirstdemonstrationofthepresumedselectivityoftheeffectof chronic benzodiazepine treatment on postsynaptic GABAAreceptor function. Nonetheless, changes in the presynapticactions of BAC as the result of chronic benzodiazepine treatment cannotbe ruled out. BAC has been reportedto decreaseendogenousGABA release(Waldineier et al., 1988),possibly asa result of presynaptic calcium channel blockade (Blaxter andCarlen, 1985). Changes in the presynaptic actions ofBAC mightbe predicted from the report that chronic DZP treatment canmodify [3H]GABA release (Hitchcott et at., 1990). However,recentdata fromour laboratoryhasshownthat KC1-evoked[3H]GABA release is unchanged in in vitro hippocampal slices2 days after withdrawal from chronic FZP treatment (Tietz andXie, 1991).

Behavioral studies from our laboratory in chronic FZPtreated rats led us to test the hypothesis that differentialsubsensitivityto the actions of GABAA and benzodiazepineagonists may be exhibited after chronic benzodiazepine treat

at Univ O

f Toledo on July 30, 2009

jpet.aspetjournals.orgD

ownloaded from

210 Xieandlietz Vol. 262

affinity GABA recognition site followingchronic benzodiazepinetreatment.Eur. J. Pharmacol.98: 159-190,1984b.

GIBBS, T. T., CHAN, C. Y., CwxowsKl, C. M. AND FARB, D. H.: Benzodiazepinereceptor photoaffinity labeling Correlation of function with binding. Eur. J.PharmacoL110: 171—180,1985.

HAEFELY,W.: The biological basis ofbenzodiazepine action. In The Benzodiazepines: Current Standards for MediCal Practice, ed. by D. E. Smith and D. ItWesson,pp. 7-41, MTP Press, Lancaster, 1985.

HENINGER, C., SArro, N., TALLMAN, J. F., GARRETT, K. M., Vrrnic, M. P.,DUMAN, R. S. AND GALLAGER,D. W.: Effects of continuous diazepain adininistration on GABAAsubunit mRNA in rat brain. J. MoL Neurosci. 2: 101—107,1990.

HITCHCOTF,P. K., PHi, S. E., EKWURU,M. ANDNEAL,M. J.: Chronic diazepamtreatment in rats causes long-lasting changes in central (3HJ-5-hydrozytryptamine and [“CJ--y-aminobutyric acid release. Br. J. PharmacoL 99: 11-12,1990.

KANG, I. AND MILLER, L. G.: Decreased GABAA receptor subunit mRNA concentrations following chronic lorazepam administration. Br. J. Pharmacol.103: 1283—1287,1991.

KEMP, J. A., MARSHALL, G. R., WONG, E. H. F. AND WOODRUFF, G. N.: Theaffinities, potencies and efficaciesof some benzodiazepine-receptoragonists,antagonists and inverse-agonists at rat hippocampal GABAA-TeCeptOrs.Br. J.Pharmacol. 91: 607—608,1987.

KEMP, J. A., MARSHALL, G. R. AND WOODRUFF, G. N.: Quantitative evaluationof the potencies of GABA-receptoragonists and antagonists using the rathippocainpal slice preparation. Br. J. PharmacoL 87: 677-684, 1986.

KROOSGAARD-LARSEN,P. ANDJOHNSTON,G. A. R.: Inhibition of GABA uptakein rat brain slicesby nipecoticacid,variousisoxazolesand related compounds.J. Neurochem. 25: 792-802, 1975.

LISTA, A., BLIER, P. AND M0NTIGNy, C. D.: Benzodiazepine receptor modulatingserotonergic neurotransmission in rat hippocampus do not desensitize afterlong-term diazepam treatment. Brain Bee. 526: 161-164, 1990.

MARLEY,R. J. AND G@u@cast, D. W.: Chronic diazepam treatment producesregionally specific changes in GABA-stimulated chloride influx. Eur. J. Pharmacol. 159: 217—223,1989.

MILLER,L. G., GREENBLAVF,D. J., BARNHILL,J. G. ANDSH@gR, R. I.: Chronicbenzodiazepine administration. I. Tolerance is associated with benzodiazepinereceptor downregulation and decreased -y-aminobutyric acid@receptor function.J. PharmacoL Ezp. Ther. 246: 170-176, 1988.

MOHLER, H., MALHERBE, P., DRAGUHN, A. AND RICHARDS, J. G.: GABAAreceptors: Structural requirements and sites of gene expression in mammalianbrain. Neurochem.Rae. 15: 199-207,1990.

NEWBERRY, N. R. AND NICOLL, R. A.: Direct hyperpolarizing action of baclofenon hippocampalpyramidalcelia.Nature (Londj 308: 450-451,1984.

RAMSEY,V. A., Tinrz, E. I. AND ROSENBERG,H. C.: Chronic flurazepaindifferentially regulates a behavioral effect of GABA agonists Pharmacol.Biochem. Behav. 38:659-663,1991.

REYNOLDS,J. N., WU, P. H., KHANNA,J. M. AND CARLEN,P. L.: Ethanoltolerance in hippocampal neurons: Adaptive changes in cellular responses toethanol measuredin vitro.J. Pharmacol.Exp. Ther. 252: 265—271,1990.

ROCK, D. M. AND TAYLOR, C. P.: Effects of diazepam, pentobarbital, phenytoinandpentylenetetrazolon hippocampalpaired-pulseinhibitionin vivo.Neurosci.Lett. 65: 265—270,1986.

ROSENBERG, H. C. AND CHIU, T. H.: Regional specificity of benzodiazepinereceptor down-regulation during chronic treatment of rats with flurazepani.Neurosci.Lett. 24: 49-52, 1981.

ROSENBERG,H. C. ANDCHIU,T. H.: Time course for development of benzodiazepine tolerance and physical dependence. Neurosci. Biobehav. Rev. 9: 123-131, 1985.

ROSENBERG, H. C., T@n'rz,E. I. AND CHIU, T. H.: Tolerance to the anticonvulsantaction of benzodiazepines: Relationship to decreased receptor density. Neuropharmacology 24: 639-644, 1985.

ROSENBERG, H. C., Tirrz, E. I. AND CHIU, T. H.: Differential tolerance to theanti-pentylenetetrazol activity of benzodiazepinesin flurazepam-treatedrats.PharmacoLBiochem.Behav.39: 711-716,1991.

ROSENBERG, H. C., TIETz, E. I., ZHANG, H. AND CHIU, T. H.: Tolerance topositive and negative modulators of GABA function measured in substantianigra of benzodiazepine tolerant rats. Life Sci. 46: 519-525, 1990.

SHER, P. K.: Long-term exposure of cortical cell cultures to clonazepain reducesbenzodiazepine receptor binding. Exp. NeuroL 92: 369-368, 1986.

SHER, P. K., STUDY, R. E., MAZZETFA,J., BARKER, J. L AND NELSON, G. P.:Depression of benzodiazepine binding and diazepam potentiation of GABAmediated inhibition after chronic exposure ofspinal cord cultures to diazepam.Brain Res. 268: 171-176,1983.

TIETz, E. I., CHIu, T. H. AND ROSENBERG, H. C.: Regional GABA/benzodiazepine receptor/chloride channel coupling after acute and chronic benzodiazepinetreatment.Eur.J. PharmacoL167: 57—65,1989.

TIETz, E. I. AND ROSENBERG, H. C.: Behavioral measurement of benzodiazepinetolerance and GABAergic subsensitivity in the substantia nigra pars reticulata.Brain Baa. 438: 41-51, 1988.

TIETz, E. I., ROSENBERG, H. C. AND CHIU, T. H.: Autoradiographic localizationof benzodiazepine receptor downregulation. J. PharmacoL Exp. Ther. 236:284-292, 1986.

TizTz, E. I. ANDXIE, X-H.: KC1-evoked GABA release unchanged in in vitro

and benzodiazepineagonists. Chronic benzodiazepinetreatment has been reported to decreasethe expression a1 (Heningeret at., 1990;Kang and Miller, 1991)and -y2mRNAs (KangandMiller, 1991) in cerebralcortexalthoughnot in cerebellum,or interestingly, in the hippocampus. The latter findings maysuggestthe lack of involvement of an alteration in geneexpression in the induction and maintenance of benzodiazepine tolerance in the hippocampus or that localized changes in thehippocampus, a very heterogeneousbrain region, may be difficult to detect using molecular techniques in homogenizedtissue.Because most GABAA subunit mRNAs are expressed in thehippocampus, including the CAl region (Möhleret at., 1990),it is also possible that the various subunit mRNAs which havebeenanalyzedthus far in the hippocampusare not thosewhosevariable expressionare associatedwith benzodiazepinetolerance phenomenon.

The current findings,i.e.,the reducedpotencyof GABAAandbenzodiazepineagonists in the CAl region of in vitro hippocampus from chronic FZP-treated rats, establish the hippocampalslice preparation as a valuable substrate for investigating benzodiazepine tolerance mechanisms. The functional tolerancemeasured in the hippocampus correlates with the time courseof the tolerancemeasuredto the antipentylenetetrazolactionsof the benzodiazepinesin vivo (Rosenberget at., 1985, 1991).By usingintracellular electrophysiologicaltechniquesin combination with biochemical methods, i.e.,autoradiographic binding and in situ hybridization techniques, this preparation willallow us to further investigate the changes in the GABAAreceptorand its functions which are associatedwith chronicbenzodiazepinetreatment to gain a better understanding of thesynaptic mechanisms underlying benzodiazepine tolerance.

Acknowledgments

The authors wouldlike to acknowledgethe technical assi8tanceof WilliamJ.Ferencak. We would also like to thank Dr. Peter F. Sorter (Hoffmann-La RocheInc.) for the gifts of DZP and desalkyl-FZP.

References

ADAMEC, R. E., MCNAUGHTON, B., RACINE, R. AND LIVINGSTON, K. E.: Effectsof diazepamon hippocampalexcitabilityin the rat: Actionin the dentate area.Epilepsia(Basal)22: 205—215,1981.

ALGER,B. E. AND NICOLL,R. A.: Pharmacological evidence for two kinds ofGABAreceptorson rat hippocampalpyramidalcellsstudiedin vitro.J. PhysioL(Lond.)328: 125-141,1982.

ANDRADE,R., MALENKA,R. C. ANDNICOLL,R. A.: A G protein couples serotoninand GABA5 receptors to the same channels in hippocampus. Science (Wash.DC) 234: 1261-1265, 1986.

BARKER, J. L. AND MATHERS, D. A.: GABA analogues activate channels ofdifferent duration on cultured mouse spinal neurons. Science (Wash. DC) 212:358—360,1981.

BLAXTER,T. J. ANDCARLEN,P. L.: Pre- and postsynaptic effects of baclofen inthe rathippocampalslice.BrainRes.341: 195-199,1985.

BOWERY, N. G., HILL, D. R. AND HUDSON, A. L.: Characteristics of GABABreceptor binding sites on rat whole brain synaptic membranes. Br. J. PharmacoL78: 191—206,1983.

BREHM, L., KROGSGAARD-LARSEN,P. AND JACOBSEN,P.: GABA uptake inhibitors and structurally related “¿�pro-drugs.―In GABA Neurotransmitters: Pharmacological, Biochemical and Pharmacological Aspects, ad. by P. KrogsgaardLarsen, J. Scheel-Kruger and H. Kofod, pp. 247-262, Academic Press, NewYork,1979.

CHU, D. C. M., ALBIN,R. L., YOUNG,A. B. ANDPENNEY,J. B.: Distributionand kinetics of GABABbinding sites in rat central nervous system:A quantitative autoradiographic study. Neuroscience 34: 341-357, 1990.

FALCH, E. AND KROGSGAARD-LARSEN, P.: The binding of the GABA agonist[‘HITHIPto rat brainsynapticmembranes.J. Neurochem.38: 1123-1129,1982.

GALLAGER,D. W., LAKOSKI,J. M., GONSALVES,S. F. ANDRAUCH,S. L.: Chronicbenzodiazepine treatment decreases postsynaptic GABA sensitivity. Nature(Lond.) 308: 74-76, 1984a.

GALLAGER,D. W., MALCOLM, A. B., ANDERSON, S. A. AND GONSALVES, S. F.:Continuous release of diazepam: Electrophysiological, biochemical and behavioral consequences. Brain Res. 342: 26-36, 1985.

GALLAGER, D. W., RAUCH, S. L. AND MALCOLM, A. B.: Alterations in a low

at Univ O

f Toledo on July 30, 2009

jpet.aspetjournals.orgD

ownloaded from

Hippocampal Benzodiazeplne Tolerance 211WILSoN, M. A. ANDGALLAGER,D. W.: GABAergic subsensitivity ofdorsal raphe

neurons in vitro after chronic benzodiazepine treatment in vivo. Brain Res.473: 198—202,1988.

W00DS0N, W., NrFECKA,L. ANDBEN-Am, Y.: Organization of the GABAergicsystem in the rat hippocampal formation: A quantitative immunocytochemicalstudy. J. Comp. Neurol. 280: 254—271,1989.

XIE, X.-H. ANDTIETZ,E. I.: Chronic benzodiazepine treatment of rats inducesreduction of paired-pulse inhibition in CAl region of in vitro hippocampus.Brain Baa.561: 69-76, 1991.

YU, 0., CHIU, T. H. AND ROSENBERG, H. C.: Modulation ofGABA-gated chlorideion flux in rat brain by acute and chronic benzodiazepine administration. J.Pharmacol. Exp. Ther. 246: 107—113,1988.

Send reprint requests to: Elizabeth I. Tietz, Ph.D., Associate Professor,Department of Pharmacology, Medical College of Ohio, P.O. Box 10008, Toledo,OH 43699.

1992

hippocampus after chronic flurazepam treatment. Soc. Neurosci. Abstr. 17:1601, 1991.

TUFF, L. P., RACINE, R. J. AND ADAMEC, R.: The effects of kindling on GABAmediated inhibition in the dentate gyrusof the rat. I. Paired-pulsedepression.Brain Bee. 277: 79-90, 1983.

TYMA,J. L, ROSENBERG,H. C., TIETz, E. I. ANDCHIU, T. H.: Effects of chronicflurazepam treatment on firing rate of rat substantia nigra pars reticulataneurons. Brain Res. 453: 344-348, 1988.

WALDMEIER, P. C. AND BAUMANN, P. A.: Presynaptic GABA receptors. Ann.N.Y. Acad. Sci. 604: 136-151, 1990.

WALDMEIER, P. C., Wiclu, P., FELDTRAUER,J.-J. AND BAUMANN, P. A.: Potential involvement of a baclofen-sensitive autoreceptor in the modulation of therelease of endogenous GABA from rat brain slices in vitro. Naunyn-Schmiedeberg's Arch. Pharmacol. 337: 289-28@, 1988.

WILSoN, M. A. ANDGALLAGER,D. W.: Effects of chronic diazepam exposure onGABA sensitivity and on benzodiazepine potentiation of GABA-mediatedresponses of substantia nigra pars reticulate neurons of rats. Eur. J. Pharmacol.136: 333—343,1987.

at Univ O

f Toledo on July 30, 2009

jpet.aspetjournals.orgD

ownloaded from