The Journal of Neuroscience January 1966, 6(l): 120-126

Endogenous Dopamine Can Modulate Inhibition of Substantia Nigra Pars Reticulata Neurons Elicited by GABA lontophoresis or Striatal Stimulation

B. L. Waszczakl and J. R. Walters Experimental Therapeutics Branch, National Institute of Neurological and Communicative Disorders and Stroke, Bethesda, Maryland 20205

Previous reports from this laboratory have described an ability The existence of dopamine in dendrites of substantianigra pars of iontophoretically applied dopamine to attenuate the inhibi- compacta dopamine neurons was first described in 1975 by tory effects of iontophoresed GABA on neurons of the substan- Bjorklund and Lindvall. Since their report, considerableeffort tia nigra pars reticulata. This finding raised the question of has been expended in establishingthat the transmitter is phys- whether endogenous dopamine, released from dendrites of iologically and pharmacologically releasable from this pool neighboring pars compacta dopamine neurons, might act as a (Cheramy et al., 1981; Geffen et al., 1976; Korfet al., 1976; neuromodulator which diminishes the inhibition of pars retic- Nieoullon et al., 1977) and, more recently, in elucidating the ulata neurons evoked by either GABA iontophoresis or electri- physiological consequencesof this release.It has long been rec- cal stimulation of the striatonigral GABAergic pathway. ognized that dendritically releaseddopamine might exert au- Extracellular, single-unit activity of pars reticulata neurons toregulatory influences on the dopamine neurons themselves. was recorded in male rats anesthetized with chloral hydrate. In However, it is also conceivable that dopamine releasedwithin one set of studies, d-amphetamine, a drug reported to release the nigra might influence the functions of other neuronal ele- dopamine from nigral dendrites, was administered intravenous- ments within this region. Severalrecent extracellular, singleunit ly (1.6 mg/kg) during regular, intermittent iontophoretic pulses recording studies (Ruffieux and Schultz, 1980; Waszczak and of GABA. As had been previously observed with iontophoresed Walters, 1983, 1984) have focusedon this latter concern- how dopamine, i.v. amphetamine significantly lessened the inhibi- dopamine affects the function of the output neuronsof the sub- tion of reticulata neurons produced by GABA application. This stantia nigra pars reticulata, a strategic region in basalganglia change was reflected by a decrease in GABA’s inhibitory po- information-processingand motor transmission.Anatomically, tency by 22% relative to the control level of inhibition achieved the pars reticulata lies at the junction of several intersecting prior to amphetamine administration. Amphetamine caused no neuronal pathways. It is infiltrated by a densenetwork of do- decreases in GABA’s effectiveness, however, in animals that pamine-containing dendrites (Bjorklund and Lindvall, 1975; had previously received treatments that depleted or destroyed Wilson et al., 1977), and it receives a major innervation, which nigral dopamine stores, i.e., in rats pretreated with reserpine is in part GABAergic, from the striatum (Fonnum et al., 1974; and a-methyl-p-tyrosine, or in rats with 6-hydroxydopamine le- Kim et al., 1971). Its projections to several premotor nuclei sions of the nigrostriatal dopamine pathway. In a second set of outside the basalganglia (Graybiel and Ragsdale, 1979) confer experiments, amphetamine or dopamine was delivered ionto- upon the pars reticulata the role of a strategic information- phoretically while monitoring the GABA-mediated (bicuculline- processingstation, which receives,integrates, and transmits dis- reversible) inhibition of reticulata neurons that can be elicited crete movement-relatedcommands to motor effector sites(Cools by striatal stimulation. Both iontophoretically applied amphet- et al., 1983; Olianas et al., 1978; &heel-Kruger, 1982). Thus, amine and dopamine significantly reduced the ability of striatal if releaseof dopamine within the ventral substantia nigra either stimulation to inhibit reticulata cell firing for 67% (n = 12) and directly influences pars reticulata neurons, or alters their re- 44% (n = 18) of reticulata cells tested, respectively. For am- sponsesto afferent inputs, then one physiological consequence phetamine, this attenuation was reflected by a S-fold increase of dendritic releasewould be modification of motor transmis- in the number of spikes occurring during the inhibitory interval sion from the basalganglia to effector nuclei. following delivery of each stimulus. These findings extend those We recently reported that dopamine, applied by microion- of our previous studies by demonstrating that the dopamine- tophoresis, can change the baseline firing rates as well as the GABA interaction can occur with physiological release of the responsesof parsreticulata neuronsto iontophoretically applied two transmitters within the nigra. They further confirm the role GABA (Waszczak and Walters, 1983). Specifically, baselinefir- of dopamine as a neuromodulator that can act downstream from ing rates of approximately half of all cells tested were increased the striatum, within the substantia nigra, to modify responses by 20% or more during dopamine application. Unrelated to the of pars reticulata output neurons to their striatonigral GABA- increasesin baselinefiring rates, dopamine also markedly and ergic innervation. consistently lessenedthe inhibition of reticulata cell firing elic- ited by GABA iontophoresis.The ability ofexogenously applied dopamineto modify responsesto appliedGABA raisesthe ques- Received Apr. 8, 1985; revised July 18, 1985; accepted July 19, 1985. tion of whether a similar modulatory interaction between these We wish to thank Mr. Raymond R. Vane for his expert technical assistance. transmitters might occur upon their endogenousrelease from Correspondence should be addressed to Dr. Judith R. Walters, NINCDS, Build- processeswithin the substantia nigra. The objective of the fol- ing IO-ACRF, Room 5C103, Bethesda, MD 20892. ’ Current address: Pharmacology Section, College of Pharmacy and Allied Health lowing studieswas to explore whether endogenousdopamine, Professions, Northeastern University, Boston, MA 02 I 15. released pharmacologically from dendrites of pars compacta Copyright 0 1986 Society for Neuroscience 0270-6474/86/010120-07$02.00/O dopamine neurons, might act as a neuromodulator that dimin-

120 The Journal of Neuroscience Dopamine Modulation of GABA Effects in SN Pars Reticulata 121 ishes responses of pars reticulata neurons to their striatal GABA- as applied here, involves expressing amphetamine-induced alterations ergic innervation. In support of this hypothesis, we report here in the response to GABA in terms of the change in numbers of spikes that intravenous or iontophoretically applied amphetamine, inhibited by GABA relative to the original, predrug baseline firing rate, which releases dopamine from dendrites in substantia nigra rather than the variable, and frequently higher, rate often attained after amphetamine administration. This method of evaluation (comparing (Cheramy et al., 198 1; Nieoullon et al., 1977), attenuates the changes in absolute numbers of spikes inhibited relative to the original GABA-mediated inhibition of pars reticulata cells evoked by baseline rate) frequently imposed a more stringent criterion for deter- either GABA iontophoresis or striatal stimulation. mination of modulatory interactions than would comparison of per- centage changes in firing based upon the original and amphetamine- Materials and Methods altered rates. Our use of this method implies no underlying mechanism of action of amphetamine in altering responses to GABA. Single-unit recording techniques The ability of i.v. amphetamine to modify responses of pars reticulata Extracellular, single-unit activity of substantia nigra pars reticulata neu- neurons to applied GABA was also evaluated in rats that had previously rons was monitored in male Sprague-Dawley rats, 2X1-350 gm, which received treatments that depleted or destroyed nigral dopamine stores. were anesthetized with chloral hydrate, 400 mg/kg, i.p. Animals were To this end, one group of rats received reserpine (1.5 mg/kg, i.p.) 16- mounted in a stereotaxic apparatus and a needle was inserted into a tail 24 hr before the recording experiment, followed by cu-methyl-p-tyrosine vein. Subsequent injections of chloral hydrate were administered intra- methyl ester hydrochloride (250 mg/kg, i.p.), administered 2-4 hr before venously, as needed. Body temperature was maintained at 36-38°C. recording. A second group of rats received ipsilateral 6-hydroxydopa- The techniques for extracellular, single-unit recording and microion- mine lesions of the ascending nigral dopaminergic neurons l-4 weeks tophoresis were the same as described previously (Waszczak and Wal- prior to the recording experiment. To test for the previously reported ters, 1983; Waszczak et al.. 1980)., A small burr hole was drilled in the modulatory response to applied dopamine, iontophoresis of the trans- skull at a site 2.0 mm lateral to the midline suture and 3.0 mm anterior mitter was initiated from 8-10 min after the amphetamine injection to the lambdoid suture. An electrode was lowered through the hole to and was continued for a 3-5 min period. the level of the substantia nigra with a hydraulic microdrive. At the end of recording experiments, a small amount of Pontamine All cells recorded in these studies were located in the pars reticulata blue dye was iontophoretically deposited in the brain by passing a 15 region of the substantia nigra and within the following stereotaxic co- PA current through the recording barrel of the electrode for approxi- ordinates: 1760-2580 pm anterior, 1.8-2.5 mm lateral, and - 1.5 to mately 15 min. The animal was then decapitated and the brain removed, -2.5 mm ventral, according to the atlas of Kiinig and Khppel(l970). fixed, sectioned, mounted, and stained. The location of a blue spot Identification of neurons as pars reticulata cells was made during the within the pars reticulata verified correct placement of the electrode recording period on the basis ofthe following criteria: (1) location ventral during the recording period. In all studies where rats received i.v. am- to the pars compacta dopamine neurons, whose electrophysioloaical phetamine injections, only one cell was monitored after the drug injec- characteristics are easily recognized and well documented (Bunney et tion. al., 1973: Guvenet and Aahaianian. 1978: Waszczak et al.. 19801 and (2) the distinguishing elect~ophysiological features of pars reticulataneu- &Hydroxydopamine lesion technique rons, including their smooth, sharp, biphasic action potentials of 0.5- Unilateral 6-hydroxydopamine lesions were placed in the left nigro- 0.7 msec duration and their firing rates (typically 10-40 spikes/set), as striatal pathway at a location just anterior to the substantia nigra, as previously described (Guyenet and Aghajanian, 1978; Waszczak et al., described previously (Waszczak et al., 1984b). Lesioned animals were 1980). Verification of correct placement of the electrode tip in the pars used in recording studies l-4 weeks after 6-hydroxydopamine injec- reticulata could be made histologically after recording experiments, as tions. After recording experiments, striata were removed and rapidly detailed below. frozen to -60°C. Dopamine levels were later determined in the left and The electrodes used in these studies were five barrel-glass micropi- right striata of each rat by high-pressure liquid chromatography with pettes (tip diameter, 4-6 pm) with a central recording barrel, and four eiectrochemical detection, according to the method of Wagner et al. outer barrels used for iontophoretic drug delivery and current-balancing. (1982). Animals were excluded from the study if dopamine levels in All barrels were prefilled with glass fibers prior to pulling the pipette, the lesioned striatum exceeded 10% of those on the unlesioned side. in order to facilitate filling with the appropriate solutions. The central barrel was filled with 2 M NaCl which contained 1% Pontamine skv Striatal-stimulation studies blue. One of the outer barrels was filled with 4 M NaCl and served as A second series of studies was conducted to determine the effects of a balance channel. The remaining outside barrels were filled with one iontophoretically applied dopamine, d-amphetamine, and bicuculline of the following combinations of three drug solutions: GABA (0.00 1 M methochloride on the inhibition of reticulata cell firing elicited by stria- in 0.2 M NaCI, pH 4), dopamine hydrochloride (0.2 M, pH 4), and NaCl tal stimulation. Stimuli consisted of square pulses, 100-300 PA in in- (0.2 M); or dopamine hydrochloride (as above), d-amphetamine (0.2 M, tensity and 300 psec in duration, which were delivered at a frequency pH 4), and bicuculline methochloride (5 mM in 165 mM NaCI). The of 1 Hz for 1 min to the left striatum, ipsilateral to the recording site, impedance of the recording barrel typically ranged from 2 to 7 Ma, and through a 2 x 2 array of four electrodes (approximately 2 mm tip that of the balance channel ranged from 15 to 50 MQ. Drug-containing separation; insulated to within 0.5 mm of the tips). The four electrodes barrels typically had impedances of 65-100 MQ. were stereotaxically positioned at 2.0 and 4.0 mm lateral, 9.0 and 11.0 mm anterior, and between 3 and 6 mm ventral. Stimuli were applied Studies with i.v. d-amphetamine to three electrodes and returned through the fourth. In one series of experiments, the effects of intravenously administered Poststimulus time histogram plots (1 msec/bin) of the effects of striatal d-amphetamine on responses of pars reticulata neurons to iontophoret- stimulation on pars reticulata cell firing were constructed using a Med- ically applied GABA were tested. For each cell, repeated 30 set ion- ical Systems Corp. Neurograph Model STA- 1 signal analvzer/averaaer. tophoretic pulses of GABA, separated by 30 set periods of baseline Corresponding oscilloscope traces of neuronal-activity during 1 &in activity, were delivered at an ejection current sufficient to inhibit retic- periods of stimulation were also stored and photographed. After com- ulata cell firing by at least 50%, but not totally. After establishing a pletion ofat least three control poststimulus time histograms, dopamine consistent response to GABA (at least three applications, yielding sim- (10 nA ejection current), d-amphetamine (5-10 nA ejection current), ilar degrees of inhibition at the same ejection current), amphetamine and bicuculline methochloride (5-20 nA ejection current) were applied (1.6 mg/kg) was administered i.v., and responses to GABA during the in turn by microiontophoresis for a period of 8-10 min each. During next 5-10 min interval were compared with those before amphetamine application of each drug, poststimulus time histograms and oscilloscope administration. The average inhibition elicited by GABA before am- records were obtained for at least three 1 min periods of striatal stim- phetamine injection (in numbers of spikes/5 set) was compared with ulation. Between drug applications, sufficient time was allowed for the the average inhibition produced by GABA after amphetamine admin- inhibitory response to striatal stimulation to recover to approximately istration for each neuron tested. In order to distinguish and separate that of the control period, recorded before drug application. Values for any amphetamine-induced changes in baseline firing from amphetamine the average number of spikes that occurred during the prefixed (control) effects on GABA’s inhibitory potency, we have used a method of cal- inhibitory interval were then determined for the histograms generated culation that was described in earlier reports concerning the modulatory during drug applications. Drug-induced changes in the inhibitory re- effects of dopamine (Waszczak and Walters, 1983, 1984). This method, sponse to striatal stimulation were calculated by comparing the number 122 Waszczak and Walters Vol. 6, No. 1, Jan. 1986

AMPH, 1.6mg/kg DA IOnA t4a+ 10 nA 0 GAEA ?,,A ------_ _ - - - a

Figure 2. Effectsof iontophoreticallyapplied dopamine (DA; 0.2 M, pH 4) on the inhibition of a substantianigra pars reticulata neuron by iontophoreticallyapplied GABA. Dopamine, but not an equimolar so- lution of NaCl, attenuated the inhibitory responses to applied GABA, as previously reported (Waszczak and Walters, 1983, 1984). AMPH, 1.6mglkg before, compared with after, amphetamine administration av- GABA (10nA) - - - - -‘- - - ______- eraged 15.6 & 3.7% of the baselinerates (n = 15; p < 0.001 by paired Student’s t test). This difference representsa net decrease in GABA’s potency by 22% relative to the control levels of inhibition achieved prior to amphetamine administration. In a parallel seriesof experiments, in which rats had been pretreated with reserpineand a-methyl-p-tyrosine, or had pre- viously received 6-hydroxydopamine lesions,i.v. amphetamine did not diminish responsesof pars reticulata neuronsto GABA. In the reserpine- and a-methyl-p-tyrosine-pretreated rats (Fig. 3), GABA inhibition of reticulata cell firing averaged 73.6 * 5.2% in the control trials before amphetamine injection. Fol- I 5 min I lowing amphetamine administration, GABA inhibited firing by 86.2 + 8.6% of the original baselinefiring rates (n = 7). This Figure 1. Ratemeterrecording of the effectsof i.v. administrationof change representsa slight, nonsignificant increase in GABA’s d-amphetamine(AMPS) on responsesof substantia nigra pars reticulata inhibitory potency, rather than the customary attenuation of neuronsto iontophoreticallyapplied GABA. Although amphethamine had variableeffects on baselinefiring ratesof someneurons, it signifi- GABA responsesobserved in animals not depleted of nigral cantly reducedthe inhibitory responsesof many cellsto iontophoreti- dopamine. Similarly, in rats that received 6-hydroxydopamine tally appliedGABA. lesions l-4 weeks before the recording experiment, responses to GABA were slightly, but not significantly, potentiated after amphetamine treatment (Fig. 4). In these animals, GABA in- of spikes/msecduring the control inhibitory periodwith the numberof hibited firing by 78.9 + 3.5% before amphetamine, and by spikes/msecduring periodsof drugapplication. Statistical significance 94.6 f 10.8%after amphetamineadministration (n = 12).Thus, (p < 0.05)of suchchanges was determined by Student’st test. in those situations in which dopamine stores were depleted, In striatal-stimulationexperiments, frequently more than onecell was either pharmacologically or by lesionsof the nigrostriatal do- monitoredper rat. The electrodeposition for the lastneuron was marked pamine neurons,amphetamine failed to lessenresponses of pars by ejectionof Pontamineblue dye. Verificationof correctplacement of reticulata neuronsto iontophoretically applied GABA. the electrodefor previouslyrecorded neurons was made by reconstruct- ing their location from stereotaxiccoordinates and histologicalexam- In both groupsof dopamine-depletedrats, i.e., thosepretreat- inationof sections. ed with reserpine and cu-methyl-p-tyrosine and those with 6- hydroxydopamine lesions, iontophoretically applied dopamine Results retained its ability to reduce responsesto applied GABA (Figs. 3 and 4) despite the inability of i.v. amphetamine to do so. In Effects of i.v. amphetamine on responses of pars reticulata fact, in animals lesioned with 6-hydroxydopamine 5-6 weeks neurons to iontophoretically applied GABA prior to recording experiments, iontophoresed dopamine has Intravenous administration of d-amphetamine (1.6 mg/kg) was beenfound to causea significantly greaterattenuation of GABA’s able to reduceresponses of reticulata neuronsto applied GABA effects than in unlesionedcontrol rats (Waszczak and Walters, in animals with intact nigral dopaminergic systems. The re- 1984). sponseto amphetamine (Fig. l), which was presumably due to releaseof dopamine from nearby dendrites, wasconsistent with Effects of iontophoretically applied amphetamine and dopamine and similar to the previously reported (Waszczak and Walters, on GABA-mediated inhibition of pars reticulata neurons 1983, 1984) attenuation of reticulata responsesto GABA ob- evoked by striatal stimulation served during dopamine application (Fig. 2). When the re- These studies assessedthe ability of both iontophoresed do- sponsesof all cells tested were considered together, amphet- pamine and amphetamine to modify the GABA-mediated in- amine was found to have caused a significant reduction in hibition of pars reticulata neurons that can be readily elicited GABA’s inhibitory potency. This reduction was reflected by an by striatal stimulation. Poststimulustime histogram plots, con- decreasein the number of spikes inhibited by GABA before, structed during 1 min periods of striatal stimulation at 1 Hz, versusafter, amphetamine administration. During the predrug revealed that firing of reticulata neurons was inhibited after a trials, GABA caused an average inhibition of reticulata cell short latency (average, 8.2 + 0.8 msec)and for a discreteinterval firing of 70.5 f 2.7%, while after amphetamine injection, the (average, 26.0 * 5.0 msec)following delivery of each stimulus. ability of GABA to inhibit firing was reduced to a value of This inhibitory responseto striatal stimulation could be attrib- 54.9 + 4.5% ofthe original baselinefiring rates.Thus, the overall uted to activation ofstriatonigral GABA pathways, sinceit could difference between the degreeof inhibition achieved by GABA be reduced by iontophoresisof the GABA antagonist, bicucul- The Journal of Neuroscience Dopamine Modulation of GABA Effects in SN Pars Reticulata 123

AMPH, 1.6mg/kg DA (10nA) GABA(InA) ______- - Figure 3. In ratspretreated with re- serpineand cY-methyl-p-tyrosine,i.v. 50 d-amphetamine(AMPS) failedto re- H duceresponses of parsreticulata neu- 9 rons to iontophoretically applied $ GABA. Despitethe inability of am- Y a phetamineto do so,iontophoretically applieddopamine (DA) retainedits lJJ 0 customaryability to lessenresponses 1I 1 to 5 min GABA. line methochloride(5-20 nA ejecting current; Fig. 5). In addition attenuate the inhibitory effects of iontophoresed GABA on rat to this consistently observed, short-latency inhibition of firing, substantianigra parsreticulata neurons(Waszczak and Walters, striatal stimulation also occasionally evoked short-latency ex- 1983). In those initial studies,both dopamine and GABA were citatory responses,as evidenced in Figure 5. However, our eval- applied by iontophoresis to simulate endogenousconditions uation of how dopamine or amphetamine modified responses under which the two transmitters might possibly interact. The to striatal stimulation was restricted, during this study, only to current experiments were directed toward establishingwhether the analysis of how these drugs altered the bicuculline-sensi- the reported dopamine-GABA interaction could occur when tive, striatal-evoked inhibitions ofreticulata cell-firing. Limiting the two transmitters were releasedfrom endogenousstorage sites the scopeof our analysisin this manner is not intended to imply within the substantia nigra parsreticulata, i.e., from dopamine- either an ability or inability of dopamine to alter excitatory containing dendrites and GABAergic striatonigral terminals. In responsesto striatal stimulation. support of this possibility, numerousprevious reports have ver- Iontophoresed dopamine significantly (p < 0.05) reduced the ified that dopamine is physiologically and pharmacologically ability of striatal stimulation to inhibit pars reticulata cell firing releasable,in vivo and in vitro, from dendrites within the pars for 8 out of 18 cells (44%) tested (Fig. 5). This attenuation by reticulata. Releasehas been shown to be induced by electrical applied dopamine was reflected by a lo-fold increase in the stimulation (Korf et al., 1976), intranigral infusions of veratri- number of spikesoccurring/msec bin during the previously de- dine (Tagerud and Cuello, 1979) or potassium (Cheramy et al., fined inhibitory interval (establishedduring control periods,when 198l), or by d-amphetamine, an indirectly acting sympatho- striatal stimulation was carried out prior to dopamine appli- mimetic drug (Cheramy et al., 1981; Nieoullon et al., 1977). In cation). The number of spikes/msecduring the control inhibi- addition, pars reticulata neurons have been shown to be con- tory period (beforedopamine application) averaged0.05 + 0.02, tacted by GABAergic terminals of striatonigral neurons(Oertel whereasthis value increasedto 0.5 f 0.11 spikes/msecduring et al., 1982). Electrical stimulation of this striatonigral pathway dopamine iontophoresis. When dopamine application was ter- has been reported to elicit a short-latency, monosynaptic in- minated, the inhibitory responseto striatal stimulation typically hibition of reticulata cell firing that can be blocked by ionto- recovered within several minutes (Fig. 5). phoretic application of bicuculline, a GABA antagonist (Col- In testing the abilty of iontophoretically applied amphetamine lingridge and Davies, 1981). to modify the inhibitory responsesof reticulata cells to striatal In thesestudies, amphetamine was usedas a tool for stimu- stimulation, a similar attenuating effect was observed. Am- lating releaseof dopamine from nigral dendrites, and striatal phetamine, ejected with 5-10 nA currents, significantly 0, < stimulation was usedas a meansof eliciting a GABA-mediated 0.05) reduced the striatal-evoked inhibition of 8 out of 12 neu- inhibition of reticulata cell firing. Amphetamine, either admin- rons (67%) studied (Fig. 5). This attenuation was reflected by a istered intravenously or applied iontophoretically, modified re- 5-fold increase in the number of spikes occurring during the sponsesof reticulata neuronsto GABA in a manner consistent predefined inhibitory interval: from 0.11 f 0.05 spikes/msec with our previously reported findings (Waszczak and Walters, in the control periods to 0.57 f 0.11 spikes/msecduring am- 1983). First, i.v. amphetamine, like iontophoretically applied phetamineiontophoresis. As before, when amphetamine deliv- dopamine, reduced responsesto applied GABA. In view of the ery was discontinued, the inhibitory responseto striatal stim- fact that amphetamine was unable to attenuate GABA’s effects ulation returned within several minutes. in animals that had previously received treatments known to deplete or destroy nigral dopamine stores, the attenuation ob- Discussion served in normal animals was most likely attributable to an Previous electrophysiologicalstudies from this laboratory dem- amphetamine-inducedrelease of dopaminefrom dendritesnear onstrated an ability of iontophoretically applied dopamine to the reticulata neuronsbeing monitored. Second,when amphet-

AMPH, 1.6mg/kg DA (10nA) GABA(8nA) ______- - - - _ ------

Figure 4. In rats that had previously received ipsilateral6-hydroxydopa- mine lesionsof the nigrostriataldo- pamineneurons, i.v. d-amphetamine (AMZVf) failedto attenuateresponses of parsreticulata neuronsto ionto- phoreticallyapplied GABA. Dopa- mine(DA), appliedby iontophoresis, retainedits customaryability to re- I 5min I duceresponses to GABA. 124 Waszczak and Walters Vol. 6, No. 1, Jan. 1986

STORED OSCILLOSCOPE TRACES POST STIMULUS TIME (60 Sweeps 1 HISTOGRAMS

Control S 1

During DA lontophoresis S 1

Recovery after DA

During AMPH lontophoresis i

Recovery after AMPH ;

During BIG-MeCI lontophoresis S

Figure 5. Effects of iontophoretical- ly applied dopamine (DA), d-am- phetamine (AMPS), and bicuculline methochloride (BIG-MeCf) on the in- hibition of pars reticulata cell firing evoked by striatal stimulation. Stim- ulation caused a short-latency inhib- Recovery after BIC-MeCI itory response which was, for this S neuron, preceded by a short-latency excitatory response. Application of each drug lessened the stimulus- evoked inhibition offiring. Shown are stored oscilloscope traces (left) and corresponding poststimulus time his- togram plots (right) generated during a 1 min period of striatal stimulation at 1 Hz. 50 msec The Journal of Neuroscience Dopamine Modulation of GABA Effects in SN Pars Reticulata 125 amine was iontophoretically applied in the vicinity of pars re- ionic mechanismat the membranelevel, remainsto be revealed. ticulata neurons, the majority of thesecells exhibited a signifi- The significanceof thesefindings with regard to basalganglia cantly diminished inhibitory responseto striatal stimulation. information-processingand transmissionare considerable.Do- Since it was possible to reduce or abolish the striatal-evoked pamine’s ability to lessenresponses of pars reticulata neurons inhibitions of firing by iontophoretic application of the GABA to their striatonigral GABAergic innervation may constitute an antagonist, bicuculline methochloride, our results suggestthat important and previously unrecognized mechanismby which the inhibitory responsesto striatal stimulation were mediated nigral dopamine neuronsregulate basalganglia output function by endogenouslyreleased GABA, and that their attenuation by without directly involving the striatum. In fact, deciphering how applied amphetamine was mediated by endogenouslyreleased dopamine ultimately influencespars reticulata output function dopamine. Thus, for more than half of reticulata cells tested, will now require recognition of both its indirect, striatally me- we could demonstrate that the dopamin&ABA interaction diated effects as well as its local, direct effectson reticulata cell occurred even under experimental conditions in which dopa- activity. Indeed, extending the range of dopamine’s functions mine releasewas presumably restricted to dendrites within the to include its direct, GABA-attenuating effects may help to re- circumference of amphetamine’sdiffusion from the pipette tip, solve certain discrepanciesthat have already developed in the and that GABA releaseonto the cell presumably had to occur literature with regard to the net effect of dopamine on basal from striatonigral terminals also located within the sphere of gangliaoutput transmission. For instance, it is widely believed amphetamine’sdiffusion and dopamine release. that activation of striatal dopamine receptors, as produced by In addition to the evidence cited here, we have previously systemic administration of drugs like apomorphine, causesan reported other findings that support the conclusion that dopa- inhibition of substantia nigra output function, presumably me- mine’sattenuation ofreticulata responsesto GABA occursphys- diated by activation of striatonigral GABAergic pathways to iologically. Specifically, reticulata neurons studied in rats given reticulata neurons(Childs and Gale, 1983; DiChiara et al., 1978; 6-hydroxydopamine lesions of nigral dopamine neurons 5-6 Kilpatrick et al., 1982; Olianas et al., 1978; Reavill et al., 1981; weeks earlier exhibited a greater sensitivity to dopamine’s ap- Redgrave et al., 1980; Scheel-Kruger, 1982). However, single- parent “modulatory” effect than did cells from unlesionedrats unit recording studies of the effects of i.v. apomorphine on (Waszczak and Walters, 1984). This development of supersen- reticulata neuronsrevealed that theseneurons are quite variably sitivity to the GABA-attenuating action of dopamine, together affected by the drug: 34% of the cells studiedexhibited increases with the current results, supports the view that the interaction in firing; 43% exhibited no changesin rate; and only 23% dis- between dopamine and GABA can and does occur to a physi- played decreasesin firing rate (Waszczak et al., 1984a, b). One ologically relevant degreewithin the pars reticulata. explanation for the variable responsesof reticulata neurons to Severalother investigators have reported data consistentwith apomorphine might be a variable balance achieved between the apparent neuromodulatory role of dopaminedescribed here. inhibitory, striatally mediated effects of the drug and its “dis- In two recent studies, dopamine was found to attenuate re- inhibitory” (GABA-attenuating) effects, which occur locally sponsesto GABA in other nuclei of the basalganglia. For in- within the pars reticulata. stance, comparable extracellular recording studiescarried out In recent years, the substantianigra parsreticulata hasgained in this laboratoryy revealed that applied dopamine reducesre- acceptance as a critical motor output center, which both re- sponsesof globus pallidus neurons to iontophoresed GABA ceives, and then transmits, striatal movement-related com- (Bergstrom and Walters, 1984). Similarly, Bemardi and co- mands to premotor nuclei outside the basalganglia. Discovery workers (1984) describedan ability of iontophoretically applied ofa new role for dopamineas a “modulator” of this transmission dopamine to reduce both the membrane hyperpolarization and processshould now reshapeand enlarge our view of the pars inhibition of firing causedby iontophoresisof GABA onto stria- reticulata beyond that of a simple relay center, to include in- tal neurons, recorded intracellularly. In addition to the above tegrative and “gain-adjusting” functions as well. studies,which closely parallel the current investigation, several earlier reports might now be reinterpreted as possibleexamples References of dopamine-mediatedmodulation of responsesto GABA. One Bergstrom,D. A., andJ. R. Walters. (1984) Dopamineattenuates the study, by Gonzalez-Vegasand Pardey (1979), describedan abil- effectsof GABA on singleunit activity in the globuspallidus. Brain Res.310: 23-33. ity of iontophoresed dopamineto lessenthe inhibition of nigral Bemardi,G., P. Calabresi,N. Mercuri, and P. Stanzione (1984) Evi- neurons, presumably pars reticulata neurons, elicited by stim- dencefor a neuromodulatoryrole of dopaminein rat striatalneurons. ulation of the globus pallidus. Another report, by Fisher et al. Clin. Neuropharmacol.7 (Suppl.1): 66-67. (1982), described an ability of i.p. amphetamine to diminish Bjorklund,A., and 0. Lindvall (1975) Dopaminein dendritesof sub- inhibitory responsesof nigral neurons,including pars reticulata stantianigra neurons:Suggestions for a role in dendritic terminals. cells, to caudate stimulation. In both of these cases,electrical Brain Res.83: 53l-537. stimulation evoked an inhibition of reticulata neurons, possibly Bunney,B. S., J. R. Walters,R. H. Roth, andG. K. Aghajanian(1973) mediated by GABA, which was diminished or blocked by do- Dopaminergicneurons: Effect of antipsychoticdrugs and amphet- pamine iontophoresisor amphetamine administration. amineon singlecell activity. J. Phannacol.Exp. Ther. 185:560-57 1. Cheramy,A., V. Leviel, and J. Glowinski (1981) Dendriticrelease of Thus, the resultsof this and previous studiesstrongly suggest dopaminein the substantianigra. Nature 289: 537-542. that dopamine can act as a neuromodulator at several sites Childs,J. A., and K. Gale (1983) Evidencethat the nigrotegmental within the CNS. The term “neuromodulator” is used here to GABAergicprojection mediates stereotypy induced by apomorphine describe the apparent ability of dopamine to modify respon- and intraniaralmuscimol. Life Sci.33: 1007-1010. sivenessof neuronsto a secondtransmitter, i.e., GABA. Other Collingridge,2. L., and J. Davies (1981) The influenceof striatal monoamine transmitters, including serotonin (McCall and stimulationand putative neurotransmitterson identifiedneurons in Aghajanian, 1979) and norepinephrine (Moises et al., 1979, the rat substantian&a. Brain Res.212: 345-359. 1981; Rogawskiand Aghajanian, 1980; Waterhouseet al., 1980) Cools,A. R., R. Jaspers,W. Kolasiewicz,K.-H. Sontag,and S. Wolfarth have been shown by similar techniques to act as neuromodu- (1983) Substantianigra as a stationthat not only transmits,but also transforms,incoming signals for its behaviouralexpression: Striatal lators at other sites in the brain. While our findings bring to dopamineand GABA-mediatedresponses of pars reticulata neurons. light a potentially important functional interaction between do- Behav.Brain Res. 7: 39-51. pamine and GABA within the substantia nigra, they do not DiChiara,G., M. Morelli, M. L. Porceddu,and G. L. Gessa(1978) elucidate the mechanismunderlying this interaction. Whether Evidencethat nigralGABA mediatesbehavioural responses elicited the transmitters interact at the receptor level, or through an by striataldopamine receptor stimulation. Life Sci. 23: 2045-2052. 126 Waszczak and Walters Vol. 6, No. 1, Jan. 1986

Fisher, R. S., M. S. Levine, C. D. Hull, and N. A. Buchwald (1982) Redgrave, P., P. Dean, T. P. Donohue, and S. G. Pope (1980) Superior Amphetamine alters evoked responses of nigral neurons in kittens colliculus lesions selectively attenuate apomorphine-induced oral and adult cats. Brain Res. 237: 415-427. stereotypy; a possible role for the nigrotectal pathway. Brain Res. 197: Fonnum, F., I. Grofova, E. Rinvik, J. Storm-Mathisen, and F. Walberg 54 l-546. (1974) Origin and distribution of glutamate decarboxylase in sub- Rogawski, M. A., and G. K. Aghajanian (1980) Modulation of lateral stantia nigra of the cat. Brain Res. 71: 77-92. geniculate neurone excitability by noradrenaline microiontophoresis Geffen, L. B., T. M. Jessell, A. C. Cuello, and L. L. Iversen (1976) or locus coeruleus stimulation. Nature 287: 73 l-733. Release of dopamine from dendrites in rat substantia nigra. Nature Ruffieux, A., and W. Schultz (1980) Dopaminergic activation of re- 260: 258-260. ticulata neurones in the substantia nigra. Nature 285: 240-24 1. Gonzalez-Vegas, J. A., and B. Pardey (1979) A presynaptic action of Scheel-Kruger, J. (1982) GABA: An essential moderator and mediator dopamine on globus pallidus afferents to substantia nigra in the rat. in the basal ganglia system of dopamine related functions. Acta Neu- Neurosci. Lett. 14: 77-80. rol. Stand. 65 (Suppl.) 90: 40-45. Graybiel, A. M., and C. W. Ragsdale (1979) Fiber connections of the Tagerud, S. E. O., and A. C. Cue110 (1979) Dopamine release from basal ganglia. Prog. Brain R&. 51: 239-283. the rat substantia nigra in vitro. Effect of raphe lesions and veratridine Guvenet. P. G.. and G. K. Aehaianian (1978)\ , Antidromic identifi- stimulation. Neuroscience 4: 202 l-2029. cation’of dopaminergic and other output neurons of the rat substantia Wagner, J., P. Vitali, M. G. Palfreyman, M. Zraika, and S. Huot (1982) nigra. Brain Res. 150: 69-84. Simultaneous determination of 3,4-dihydroxyphenylalanine,5-hy- Kilpatrick, I. C., G. L. Collingridge, and M. S. Starr (1982) Evidence droxytryptophan, dopamine, 4-hydroxy-3-methoxyphenylalanine, for the participation of nigrotectal gamma-aminobutyrate-containing norepinephrine, 3,4-dihydroxyphenylacetic acid, homovanillic acid, neurones in striatal and n&al-derived circling in the rat. Neuroscience serotonin, and 5-hydroxyindoleacetic acid in rat cerebrospinal fluid 7: 207-222. and brain by high-performance liquid chromatography with electro- Kim, J. S., I. J. Bak, R. Hassler, and Y. Okada (197 1) Role of r-amino- chemical detection. J. Neurochem. 38: 124 l-l 254. butyric acid (GABA) in the extrapyramidal motor system. 2. Some Waszczak, B. L., and J. R. Walters (1983) Dopamine modulation of evidence for the existence of a type of GABA rich strionigral neurons. the effects of -y-aminobutyric acid on substantia nigra pars reticulata Exp. Brain Res. 14: 95-104. neurons. Science 220: 2 18-22 1. Kiinig, J. F. R., and R. A. Klippel (1970) The Rat Brain: A Stereotaxic Waszczak, B. L., and J. R. Walters (1984) A physiological role for A&, R. E. .Krieger, Huntington; NY: dopamine as modulator of GABA effects in substantia nigra: Super- Korf. J.. M. Zieleman. and B. H. C. Westerink (1976) Donamine sensitivity in 6-hydroxydopamine lesioned rats. Eur. J. Pharmacol. release in substantia nigra. Nature 260: 257-258.‘ ’ - 105: 369-373. McCall, R. B., and G. K. Aghajanian (1979) Serotonergic facilitation Waszczak, B. L., N. Eng, and J. R. Walters (1980) Effects of muscimol of facial motoneuron excitation. Brain Res. 169: 11-27. and picrotoxin on single unit activity of substantia nigra neurons. Moises, H. C., B. D. Waterhouse, and D. J. Woodward (1981) Locus Brain Res. 188: 185-197. coeruleus stimulation potentiates Purkinje cell responses to afferent Waszczak, B. L., E. K. Lee, T. Ferraro, T. A. Hare, and J. R. Walters input: The climbing fiber system. Brain Res. 222: 43-64. (1984a) Single unit responses of substantia nigra pars reticulata neu- Moises, H. C., D. J. Woodward, B. J. Hoffer, and R. Freeman (1979) rons to apomorphine: Effects of striatal lesions and anesthesia. Brain Interactions of norepinephrine with Purkinje cell responses to puta- Res. 306: 307-3 18. tive amino acid neurotransmitters applied by microiontophoresis. Waszczak, B. L., E. K. Lee, C. A. Tamminga, and J. R. Walters (1984b) Exp. Neurol. 64: 493-5 15. Effect of dopamine system activation on substantia nigra pars retic- Nieoullon, A., A. Cheramy, and J. Glowinski (1977) Release of do- ulata output neurons: Variable single-unit responses in normal rats pamine in vivo from cat-substantia nigra. Nature 266: 375-377. and inhibition in 6-hydroxydopamine-lesioned rats. J. Neurosci. 4: Oertel, W. H., M. L. Tappaz, A. Berod, and E. Mugnaini (1982) Two- 2369-2375. color immunohistochemistry for dopamine and GABA neurons in Waterhouse, B. D., H. C. Moises, and D. J. Woodward (1980) Nor- rat substantia nigra and zona incerta. Brain Res. Bull. 9: 463-474. adrenergic modulation of somatosensory cortical neuronal responses Olianas. M. C.. G. M. de Montis. G. Mulas. and A. Taaliamonte (1978) to iontophoretically applied putative neurotransmitters. Exp. Neurol. The striatal dopaminergic function is mediated by-the inhibition of 69: 30-49. a n&al, non-dopaminergic neuronal system via a strio-nigral Wilson, C. J., P. M. Groves, and E. Fitkova (1977) Monoaminergic GABAergic pathway. Eur. J. Pharmacol. 49: 233-241. synapses including dendrodendritic synapses in the rat substantia Reavill, C., P. Jenner, N. Leigh, and C. D. Marsden (1981) The role nigra. Exp. Brain Res. 30: 161-174. ofnigral projections to the thalamus in drug-induced circling behavior in the rat. Life Sci. 28: 1457-1466.