Dopamine modulates the inhibition induced by GABA in rat cerebral cortex: an iontophoretic study

Etmpemt

205 ( 199 1) 225-231 Science Publishers B.V. All rights reserved (H114-~9YY/Yl/$l,-1.Sff

Jatwnal of ~~t~rtttuc~i~~,

6 1991 Elsevier

EJP 5.2156

Mario Beauregard and And& Ferron Cenlre de Recherche

en Sciences Nettrolagiqttes, D&partemettt de P/ly%alugiL:Factdth de M~drciw, Mottt&ai.

&tit .ersiti

de Monrr&l.

@ikbec, Cumzda H3C 357

Received 28 February 1991, revised MS received 30 August 1991, accepted IO September 1901

Effects of ~ntophoresed ~-aminobutyric acid @ABA) and two GABA agonists, 4,5,6,~-tetrahydrois~xa2oio-~~,4-c~pyridinc3-01 (THIP) and baclofen were quantitatively compared in the anterior cingulate, frontal, and parietal cortex of urethaneanesthetized intact rah after catecholamine (CA) depletion with a-methyl-p-tyrosine In-MPT) or selective dopamine (DA) denervation with 6-hydroxydopamine (6-OHDA). As assessed with to the ITso index, the postsynaptic sensitivity to GABA was significantly higher in anterior cinguiate than in frontal and parietal cortex. The responsiveness to GABA was also greater in frontal than in parietat cortex. Sensitivity to GABA was significantly reduced in both anterior cingulate and frontal cortex after CA depletion, and similarly, after DA dene~ation with &OHDA. The difference in Ihe sensitivity to GABA between the three cortical regions in intact rats as well as after CA depletion did not seem to be correlated with either GABA, or GABA, receptors since the responsiveness to both GABA agonists in every region examined was comparable in intact rats, and remained unchanged after a-MPT treatment. This finding raises the possibility that some GABA receptors in the cerebral cortex may be pharmacologically distinct from the two main subtypes of GABA receptors, GABA, and GABA,. When GABA was administered by iontophoresis in the anterior cingulate cortex during continuous applications of subthreshold currents of DA, the inhibition induced by GAJ3A was either increased or decreased. As DA innervation density is nearly two-fold greater in anterior cingulate than in frontal cortex, and 30-fold greater in anterior cingulate than in parietal cortex, these results suggest that responsiveness to GABA may be correlated with the regional density of DA innervation and that elevated levels of DA may enhance the sensitivity to GABA. Cerebral cortex; Dopamine; GABA (y-aminobutyric

In addition to its direct effects on neuronal firing, dopamine (DA) is known to modulate the efficacy of neurotransmitters. In particular, interactions between DA and y-aminobu~ric acid (GABA) containing neurons have been demonstrated. Thus, in the rat, extracellular single unit recordings of substantia nigra pars reticulata neurons have demonstrated that iontophoretically applied DA reduces the inhibition of firing caused by iontophoretic administration of GABA (Waszczak and Walters, 1983). On the contrary, Chiodo and Belger (1986) have showed that DA iontophoresed in the striatum enhances GABA-mediated inhibition of striatal neurons. A modulatory effect of DA on GABA inhibition in the globus pallidus has been reported,

Correspondence to: A. Ferron. Centre de Recherche en Sciences Neurologiques, Departement de Physiologic, FacultC de Midecine, UniversitC de Montrkal, CP 6128, Succursaie A, Montrial fQui), Canada H3C 357. Tel. 1214.343 6356, fax 1514.343 2111.

acid); Dopamine-GABA

interactions

whereby DA was found to excite pallidal neurons by attenuating GABA-induced inhibition (Bergstrom and Walters, 19841. Besides the fact that DA can interact with GABA at postsynaptic sites, several lines of evidence suggest that this biogenic amine may modulate GABA release. Thus, push-pull experiments using selective DA receptor agonists and antagonists have indicated that DA can exert either excitatory or inhibitor effects on GABA release in the striatum (Kernel et al., 1987). Rernath and Zigmond (1989) demonstrated that DA can either Lnhance striatal GABA release via D, receptor stimulation or decrease it via D, receptor activation. Little is known concerning DA-GABA interactions within the cerebral cortex. While there is no evidence of direct interrelationships or connectivi~ between DA afferents and intracortical CABA neurons (Vemey et al., 1990), current topographic data on the DA and GABA innervation of the rat cerebral cortex raise the possibility of such local interactions. Indeed radioautographic and immunoc~~hemical studies on the respective distribution of the DA terminals and SABA-

~~~~~ti~~~~~~~~ nettn~~ n~tit of the kttter bemg derived ~~~$~~ ~~~r~~~~~ neurons (Emson and Lindvall. JY?Y1, have ~~d~~~te~ &it b0th transmitters ill% fOUlId ~~~~~~~~~~t the mmrirw cingulatc rmd the fronta regfranz and in fayer VJ of the porietal cortex (Descnrries et &, !YSp. ~~~~~~~~ni rmd OerteJ. fY85; Vincent et al.. 1, ~~~~~~~~~~~~~~?~~I visualization of DA termi1 n&s by electnan microscopy showed that these cortical sffercnts in rat prefrontal cortex form synaptic contacts (S&gu& et al.. IQ&S:Van Eden et al., 19871, preferenic tree of the pyramidal neurons J. The CAMP- and DA-regulated ~~~~o~~tein, DARPP-32. which is enriched in neurons bearing D, relrptors. also has predilection for cmrticaf ~y~rnid~l eetts (Hemmings et al., lQS?J. Corti&XlGABA terminals also form synaptic contact onto pymnrida8 neurons (Houser et al,. 19S4: Se&la et 31.. JQSJ. In addition to these morphotogical data, current electrophy&logical and pharmacological findings also support the possibility of zm interaction between DA and GABA in the cerebral corms. Thus. in vitro elect~~h~~~Jog~~aJ studies have demonstrated that DA increases the firing rate of GABA interneurons in the rat prefrontal cortex (Penit-Soria et al., 1987). and that the activation of D, receptors induces inhibition of the eJectricafly evoked release of ‘H-GABA from prefrontai corticaI slices (Penit-Soria et al.. 1989). We have now used iontophoresis to detect possible ~n~~~c~ions between DA and GABA in anterior cinguJste, frontal and prtrietal cortex, three crrtical regions showing wide differences in DA innervation density (Descarries et af.. 198X The responsiveness to GABA and to GABA agonists was assessed quantitatively by measuring the amount of drug required to induced SK reduction of s~ntaneous firing. Data were obtained not only from intact urethane-anesthetized rats, but also after catecholamine (CA) depletion with LYm~thyJ-~~~ine Qu-MPTJ or selective denervation with 6hydroxydopamine t6-OHDA) in the presence of desipramine.

2. P. Iovttophoresis

Ad&t male Sprague-Dawley rats (body weight 225350 gJ anesthetized with urethane (1.5 g/kg i.p.J were used throughout these studies. For iontophoresis, fivebarrel micropipettes were used as previously described (Ferron et al., 1982). The central barrel contained 3 M NaCl for recording as did one outer barrel for current balancing. The remaining barrels were filled with either 0.05 M GAJ3A (Sigma), 0.05 M t&J-baclofen (RBII. 0.05 M 4,5,6,7-tetrahydroi~xazolo-[5,4-c]pyridine-3-ol hydrochloride UJ-JIP; RBIJ, or 0.05 M DA

hydrochloride (Sigma), ail in saline and prepared in a 0.1% ascorbic acid solution and adjusted to pH 4.00. Extracellular action potentials were obtained from spontaneously active neurons in frontal (A IG-11.5, L l-2.5 mm and 0- 1.J mm below cortical surface (b.c.s.11, anterior eingulate (A Q.5- II, L 0.5-1.0 and 2-3 mm b.c.s.J, and parietal cortex (A 6-S, L 2-4 and OS-l.5 mm b.c.s.1, according to Paxinos and Watson (1986). Neuronal responsiveness was assessed by means of JTTc,.This index (in nCJ is the current tin nAJ multiplied by the time tin s) required to obtain a 50% depression of the spontaneous firing rate (De Montigny and Aghajanian, 1977). IT,, provides a measure of postsynaptic sensitivity to iontophoretically applied drugs (Brunei and De Montigny, 1988). In the present study, units failing to show a 50% reduction of firing following drug ejection with a 50 nA current for 120 s were considered to be unresponsive. To facihtate comparisons, the same micropipette was generally used in the various regions from a given animal. In early experiments, three IT,, vahres were obtained with different currents on most responsive units. Few differences were then noted between ITS, s measured with currents ranging from 20 to 50 nA. Jn subsequent experimetrrs, the ejection current was kept consrant (40 or 50 nA for GABA, 30,40, or 50 nA for THIP and baclofen). L2. Treatv?lt?vIts

Three groups of rats were used: (1) intact rats (n = 7); (2) CA-depleted rats {n = 11) treated with the synthesis inhibitor, cw-MPT (Sigma, 200 mg/kg, i.p.1, 18 and 2 h before the recordings; (3) DA-denervated rats (n = 41 subjected to selective 6-OHDA lesioning or DA nerve cell bodies in the ipsilateral midbrain tegmenturn, 2-4 weeks prior to the recordings. Two microinjections of the drug (6-OHDA hydr~hloride, Sigma) were made in the ventral tegmental area and in the substantia nigra pars compacta, respectively, according to stereotaxic coordinates: A 3.8, L 0.8, H i-2.0 mm and A 3.8, L 2.4 and H + 1.8 mm (Paxinos and Watson, 1986); 8 Fg of the salt was dissolved in 2 ~1 of saline containing 0.1% ascorbic acid, and delivered over 5 min at each site, 1 h after desipramine pretreatment (Sigma, 25 mg/kg i.p.J to protect ascending noradrenaline (NA) pathways (Breese and Traylor, 1971). The resulting changes in cerebral CA content were measured by HPLC with ion-pairing and etectrochemical detection following established procedures (Reader et al., 1986). Compared to the controls, DA content was decreased by more than 95% after tu-MPT treatment, and by 70% after 6-OHDA denervation. NA was also decreased by 70 and 17% after a-MPT and 6OHDA, respectively, but the latter decrease was not statistically significant.

2.3. Co-administration of DA and GABA

CONTROL

A series of complementary experiments was carried out to determine possible interactions between DA and GABA in the anterior cingulate cortex. After the effect of GABA alone (with ejection currents of 40 nA) had been tested, the neurons were examined for their responsiveness to DA and, the DA current (l-3 nA) was then reduced until little or no effect was detectable. The subthreshoId DA current was then applied continuously for several minutes while GABA was retested with 10-20 s pulses applied at 30-40 s intervals, FoIlowing three or four such pulses, recovery of the control responsiveness to GABA alone was verified. interaction was postulated when DA application decreased or increased the GABA-induced inhibition by more than 30%. Here also, inhibition was defined as a reduction of the spontaneous firing rate by at least 50%. The decrease in firing was assessed by subtracting the number of spikes in the inhibition period from the average found before application of GABA, and was expressed as percent of control. Statistical comparisons between control and experimental groups were iarried out using an analysis of variance.

z;: I

3. Results 3.1. Spontaneous firing A total of 305 neurons were studied in 22 rats. The spontaneous discharge pattern was relatively stable in every region and condition tested, consisting of rather regular spikes in the parietal cortex, frequent bursts in anterior cingulate, and mostly individual spikes with occasional bursts in frontal cortex. The mean neuronal spontaneous firing rate, which was averaged over a period of 60 s, was not significantly different in the three cortical regions examined, even though there were considerable variations in the level of spontaneous unit activity within each region. Thus, the average baseline firing rate ranged from 2.5 to 38 spikes/s (s.P.s.) in parietal (mean: 14.1 & 5.7 (SD)), from 2 to 3.5 s.p.s. in anterior cingulate (mean: 13 + 6.51, and from 2.5 to 25 s.p.s. in frontal cortex (mean: 11.9 + 6.2). DA depletion or dene~ation had no apparent effects on these firing rates. 3.2. Responsir*eness to GABA und its agonists In every region and under all conditions tested, 100% of the responsive units were inhibited by GABA and the two GABA agonists, TI-IIP (GABA,) and baclofen (GABA,). With currents of 40 or 50 nA, the time necessary to reach 50% inhibition ranged between 5-8 s in anterior cingulate, 14-18 s in frontal, and

Fig. 1. Cumulative rate histograms illustrating the responsiveness to iontophoresed GARA in three cortical regions - anterior cingulate (ACg), frontal (FR), and partrtal (Par) cortex - in intact and rxMPT-depleted rats. In this and the following figures, bars with numbers show the duration of drug applications with the ejection currents given in nA. The frequency of discharge was integrated over 10 s intervals. NEte that in intact rats, the mean ITsa value for GABA was higher in parietal (Par1 than in fronta! WRt and anterior cingulate (ACg) cortex, and higher in frontal (FR) than in anterior cingulate (ACg) cortex. as indicated by the time of ejection. Also, a-MPT treatment decreased the responsiveness to GABA in both anterior cingulate fACg) and frontal (FR) cortex, but remained without effect in parietai (Par) cortex.

between 30-40 s in parieta1 cortex. The inhibition of firing induced by GABA in intact rats was frequently abrupt in onset. In these animals, the average amount of GABA required to induce 50% inhibition was consistently higher in parietal Cl577 nC f 214 (S.E.M.)) than in frontal (708 + 53) and anterior cingulate cortex (293 f 44) (P < 0.001). The mean IT,, value was also significantly higher in frontal than in anterior cingulate cortex (P < 0.001) (figs. 1 and 3). In contrast, postsynaptic sensitivity to THIP or baclofen was not significantly different between the three cortical regions (figs. 2 and 3). CA depletion with cw-MPT significantly decreased the postsynaptic sensitivity to GABA in anterior cingulate (813 it: 90; P < 0.001) and frontal cortex (953 f 95; P < 0.05) but remained without any effect in parietaf cortex. Moreover, after a-MPT treatment, the mean ITS0 value was similar in anterior cingulate and frontal

WWX, wh~~as a significant difference

remained beEkk%rqions snd I~~irtaI cortex ifigs. 1 and st. CA depletion did not affect the responsiveness to both GABA agonists
A

GABA zm5

n

CONTROL

I El a-MPT

f

BACLOFEN

8 2055

Interactions between DA and GABA in anterior em&ate cortex were examined by comparing the effects of GABA before and during DA application. In 35% of the 57 neurons studied, DA modified the magnitude of the inhibitory response to GABA (fig. 4). The ~-~ndu~d inhibition was increased during ~ecmcomnant ejection of DA in 26 neurons, whereas GABA inhibition was decreased in 17 ceils (fig. 4). No

1500

2 g

1055 555

0

c

CONTROL

$

2

Par

THIP

1555‘ 555.

5’

BA THIP

FA

Par

Fig. 3. Mean ITjo (kS.E.M.1 of responses to iontophoretically applied GABA, baclofen CBA) and THIP, in the three cortical IACg, FR, Par) regions examined in intact and a-MPT-depleted rats. The number of cells is shown at the bottom of each column. See legends of figs. 1 and 2 for main results. * Statistically significant difference between control and a-MPT groups for a given drug. ‘Significant difference between controls for a given drug.

50 25 0 li

interaction was detected for the remaining 14 GABAsensitive units which responded to DA (fig. 4).

4. Discussion emin Fig 2 Cumulative rate histograms demonstrating, the responsiveness to iontopboresed baclofen (BA) and THIP in anterior cingulate (Acg). frontal (FR), and parietal (Par) cortex of intact and a-MPTdepleted rats. In intact rats, the postsynaptic sensitivity to THIP and backfen was comparable between anterior cingulate (ACg), frontal @IO, and parietaf (Par) cortex. Also, the responsiveness to both GABAergic agonists remained unchanged after DA depletion with U-MPT.

The present study has demonstrated that, in intact rats, the amount of GABA required to induce 50% inhibition of spontaneous firing was significantly higher in the parietal than the anterior cingulate and frontal cortex, and in the frontal than the anterior cingdate cortex. This interregional difference in average IT,, was unexpected since the cortical regions examined have a similar density of GABA cell bodies and axon

A

DA

DA

Fii. 4. Cumulative rate histograms illustrating interactions between DA and GABA in anterior cingulate fACg1 cortex. (A) DA ejection enhanced the inhibitory effect of GABA. (B) A different cell in which DA decreased the inhibitory response to GABA. (0 A different cell in which DA had no apparent direct effect on the GABA induced inhibition.

terminals (for review, see Mugnaini and Oertel, 19851, and comparable endogenous contents of GABA (Peinado et al., 1984). This difference in postsynaptic sensitivity to GABA could reflect quantitative differences in the cortical distributjon of GABA receptors. This possibility is however unlikely as current evidence indicates that the rat cerebral cortex shows a fairly consistent concentration of high-affinity GABA receptors, pa~icularly in the outer four layers (Palacios et al., 1981; Bowery et al., 1987). A possible explanation for these observations is that, in the cerebral cortex, the presence of DA fibers (or DA itself) favours responsiveness to GABA. This hypothesis is supported by our results in the parietal, anterior cingulate and frontal cortex after a-MPT depletion and 6-OHDA denervation, which showed a si~ificant reduction of the postsynaptic sensitivity to GABA. The differences in GABA responsiveness between the three cortical areas seemed to be correlated with the regional density of DA innervation. Indeed, the responsiveness to GABA was higher in cortical areas with the densest DA innervation. The DA innervation density is 1.5 times greater in the anterior cingulate than the frontal cortex, whereas the DA innervation of the parietal cortex is scant (30-fold lower than in anterior cingulate cortex) and is confined to the bottom of its layer VI (Descarries et al., 1987).

An afternate explanation for the differences in GABA responsiveness between the three cortical regions might be the different GABA uptake capacities in the parietal versus the anterior cingulate and the frontal COrteX. The amount of ejected GABA that reaches recorded units could be lower in regions where GABA uptake is higher. Were this the case however, the decreased responsiveness to GABA in the anterior cingulate and frontal cortex following DA depletion or denervation wouid imply that, in these brain regions at least, the GABA uptake process is regulated by DA. According to our iontophoretic data, when GABA was ejected during continuous application of subthreshold currents of DA, the GABA-induced i&b& tion was most often enhanced. However, GABA responsiveness was attenuated in a significant number of cells. These opposite effects could be associated with postsynaptic differences in the DA concentration and in the subtypes of DA receptors activated. fnterestingly, electrophysiologicat studies in vitro have demonstrated that CA1 pyramidal neurons in the hippocampus are excited by low concentrations of DA and inhibited by higher doses of DA (S~alows~ and Bijak, 1987). Experiments with selective DA receptor agonists have suggested that the excitatory response to DA is mediated by D, receptors whereas depressant responses follow D, receptor activation (Smialowski and Bijak, 1987). Stimulation of D, and D, receptors also produces opposing effects on the spontaneous release of GABA in the striatum. While stimulation of D, facilitates GABA release, activation of D, receptors inhibits it (Girauh et al., 1986). In contrast to DA actions in the hippocampus, low concentrations of DA preferentially activate D, receptors in the striatum while high doses stimulate D2 receptors (Bemath and Zigmond, 1989). Thus, the different types of postsynaptic interactions between GABA and DA in cerebral cortex might be due to differences in the relative protlortions of D, and D, receptors activated. Confirmation of this hypothesis will require more extensive pharmacological analysis. It appears likely that under conditions of minimal dopaminergic tonus, GABA neurons might transmit less inhibitory information whereas during period of enhanced DA release, they could become more strongly inhibitory. A markedly diminished res~nsiveness to GABA followed either CA depletion with LY-MPT or DA denervation with 6-OHDA. Several studies have emphasized the possibility that a reduction in transmitter receptor sensit~~ would be the result of an ‘overstimulation’ of these receptors (Gnegy and Costa, 1980; Schwartz et al., 1978). DA administration has been found to decrease the refease of GABA in cerebral cortex in vitro (Flint :t al., 1985). In this context, it is plausible that the reduction of GABA responsiveness observed in the anterior cingulate and the frontal COT-

and h-QHDR treatments rcprcof CABA pnstsynnptic receptors foIlowing exposure tu higher levels ol’ GABA in the absence of DA. The fact that GABA responsiveness was reduced similarly after &OtiDA dcnervation and n-MPT dcplction suggested that this decrease was imputalrIc to the absence of DA itself rather than that of DA afferents. A DA depletion of 70% appcarcd sufficient to induct the decrease in GABA responsiveness. This decrease occurred rapidly and was seemingly permanent since it ws comparable I8 h after n-MP’I dcpktion and 2-4 weeks after 6-UHDA dcnervation. However, DA dcpletian had na significant effect on the responsiveness to THIP and baclofen in anterior cingulate and frontal cortex. Morcovcr, in intact rats, there was no difference in the sensitivity to these GABA, and GABAH agonists in the three cortical regions examined, as had been observed with GABA. This finding suggests that, in the cerebral cortex at least, &me GABA receptors may he pharmacologically distinct from the two main types of GABA receptors (GABA, and GABA,). An alternative expianation would bc that although THlP and baclofen may act at GABA receptors, their actions are not modulated in the same way as that of the natural transmitter and thus are not affected by DA. Our demonstration al the modulation of GABA responsiveness by DA leads to the conclusion that the function of GABA in the neocortex might partially depend on its interactions with the DA systems. Considering the numerous DA therapeutic agents related to DA that are used clinically, it is important to envisage the possibility that these drugs might also exert indirect effects on GABA function.

Acknowledgements This study was supporled by an operating grant (AI.)

and a

s;ludenrship (MB.) from the Medical Research CouncilOCCanada. Tt~t:authors thank Daniel Cyr. Giovanni Battisla Filosi and Claude Gauthier for photographic and graphic work. They are also grateful 10 Drs. Laurent Descarries. Robert W. Dykes and Jean-Claude Lacaille for constructive criticism of the manuscript.

nreest, Ci.R. and T.D. Traylor. 1971. Dcptctiun of brtiin norad~ rcnaline and dopaminc by 6-hydrmydopamine. Br. J. Pharmaeul. 42. xx. Brund. S. and C. Dr Monrigny, IYXH,Validation of Ihc ITqI meihud for assessing nr?uron:lt rcsponsivcncks Iu microionlophorelic applicatlons: a single cell recording study, 1. Phurmucol. Meth. 19. 21. Chiodo, L.A. and T.W. Berger, 1986. Inferaction I~ween dopaminc and amino-acid induced cxcirurion 2nd inhibition in the slriarum. Brain Res. 375, IYH. De Monligny, C. and G.K. Ayhajaniun, 1977, Preferenliul rction of 5-mcthclxydimc-thyltryptanline on presynapric serolonin receptors: A vtlmpar&ve iontophuretic study with LSD and 5-HT. Neurt)pharmticolo~ 16. XI 1. Dcscarrics, L., B. Lcmuy. C. Doucel and B. Berger, 19X7, Regional and tamintlr density of thr dopamine innervation in adult rat cerehrrl corw. Neuroscience 21,8W. Emscn, P.C. and 0. Lindvall. 1979. Dislrlbution of putalive neurotransmitters in Ihe neclcortcx, Ncuruscicncc 4, 1. Ferrrm, A.. 1. Descarrics and T.A. Reader, 1982, Altered neuronal respunsivenrss to biogenic amines in rat cerebral cortex after nerolonin denervation or depletion, Brain Res. 231, 93. Flint, R.S., J.M. Murphy, P.M. Calkin and W.J. McBride, 1985, Monoaminc aminn ecitl nnd cholinergic interactions in slices of rat ccrcbwi co&X. Brain Res. Bull. 15. 197. Girautt, J.A.. U. Spampinato. J. Glowinski and M.J. Besson, 1986. In vivo release of [“HI gamma-aminobutyric acid in Ihe rat nenstriaturn. II. Opposing effects of Dl and I), dopaminc receptor stlmulrtion in the dorsal caudak putzkmen, Neuroscience 19, 1109. Gnegy, ME. and E. Cosla, 1980, Calccholamine receptor supersen& tivily and subsensitivity in Ihc central nervous system, in: Essays in Neurochemistry and Ncuropharmacology, Vol. 4, eds. M.3.H. Youdim, W. Lovcnbcrg, P.F. Sharmnn and J.R. l.agnado (WiIey. Londun) p. 249. Hemmings, Jr., H.C., S,I, Walaas, CC. Ouimel and P. Greengard, 1987, Dopaminergic regulation of protein phosphorylation in the strialum: DARPP-32. Trends Neurosci. 10. 377. Houser, C.R., J.E. Vaughn, S.C,II. Hendry, E.G. Jones and A. Peters, 19&l, GABA neurons in Ihe cerebral cortex, in: Cerebral Cortex, Vol. 2. eds. E.G. Jones and A. Peters, (Plenum Press, New York) p. 63. Kernel, M.L., C. Gauchy, J.A. Girault, M.J. Besson and J. G&in&i. 1987, Relations entre neurones GABAergiques et doprlminergiques dans Ies ganglions de la base, Encephale 13, 16Y. Mugnaini. E. and W.H. Oertet, 1985, An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry, in: Handbook of Chemical Neuroanaromy. Vol. 4, eds. A. Bjljrklund and T. Hiikfelt (Elsevier. New York) p. 436. Palacios, J.M.. J.K. Wamslcy and M.J. Kuhar, 1981. High-affinity GABA receptors - Auloradiogwphic localization, Brain Rcs.

222, 2R5. Paxinus, G. and C. Watson,

lMxences Brrgslrom, P.A. tend J.R. Walters. 1984, Dopaminr attenuates the effect of GhBh on single unil activity in the globus pallidus, Brain Res. 310. 23.

Bernarh, S. and M.J. Zigmond. 1989, Dopamine may influence striatal GABA relcasc via three separate mechanisms, Brain Rcs. 476, 373. Bowcry, N.G., A.L. Hudson, irnd G.W. Price. 1987, GABA, and GABAn rccaptor site di&bution in the rat central nervous system, Neuruscicncc 20, 365.

1986. The Rat Brain in Srereomxic Coordinates (Academic Press, Auslralicr). Peinadu, J.M., .I.A. Gomee-Capilla and F. Morn, 1984, Cerebral cortex and aminn acid neutn~ransmi~bzrs:higher levels of asparlic acid but not GABA in Ihc frontal cortex of the rat, Brain Res. Bull. 12.625. Penit-Sorin. J., E. *Audinat and F. Crcpel, 1987. Excitation of rat prefrontal cortical neurons hy dopamine: an in vitro etectrophysiotogical study, Brain Rcs. 425, 2h3. P&it-Soria, J., S. Retaux and Y. Maurin, 1989, Effets dc la stimulation des rkepleurs D, et Dz dopaminergjques sur la lihktinn d’acide gamma-[%11 aminobutyrique induite &ctriquement dans le cortex prifrontal du rat, C.R. Acad. Sci. [Ill]? 309, WI.

Reader, T.A.. R. t3ribz. L. Grundin and A. Frrron, 1986. Effects of p-chlomphcnylulanine an cortical monurmines and on the wtivity uf nuradrenergic neurons. Ncur*clwn~. Res. 1 I. lO25. Schwurtz. J.C.. J. Costentin, M.P. Martrd~, P. Pccrtiis and M. Be:~ur.lry, 1978, Modulittitln CICrrccptar mechanisms in the CNS: hyperand hyposensi(ivity lo c;wchul;rmines, Ncurt)philt~ttdcnloFy 17. 665. SEguCla, P.. H. Gamrani. M. Crefbrd. A. Cala nnd M. 1-e hloal. 19% Ultrustructur:ll immunt)cybchcrrtislr)~ nf y-nminohutyratc in thl: cerebral and ccrehdlar cortex of the adult rat. Nuuroscienw 16, HhS. S&uila. I’.. KC. Watkins and L. Driscarrius. I’X%3. Ultrasrructural feaiurcs of dopamine axtIn tcrminnls in the antcru-mu&o1 unJ soprarhinnl cortex of adult rat, Brain Res. 131. I I. Smi*‘owski, A. and M. Bijak. 1987. Excitatory and inhihirtq :dm of dnpamine un hippwamp;~l neurons in vim. Inwlvemcnl rrf D2 and 0, receptors, Nturnscicnce 3. 95.