disinhibition of the mediodorsal thalamus induces fos-like immunoreactivity in both pyramidal and...

Disinhibition of the Mediodorsal ThalamusInduces Fos-Like Immunoreactivity in BothPyramidal and GABA-Containing Neurons

in the Medial Prefrontal Cortex of Rats, ButDoes Not Affect Prefrontal Extracellular

GABA LevelsMICHAEL BUBSER,1 JOHANNA M. DE BRABANDER,1 WIA TIMMERMAN,2

MATTHIJS G.P. FEENSTRA,1* ERNA B.H.W. ERDTSIECK-ERNSTE,1 ALICE RINKENS,1JOHANNA F.M. VAN UUM,1 AND BEN H.C. WESTERINK2

1Graduate School Neurosciences Amsterdam, Netherlands Institute for Brain Research, 1105 AZAmsterdam, The Netherlands

2Department of Medicinal Chemistry, University Centre for Pharmacy, University of Groningen, 9713 AVGroningen, The Netherlands

KEY WORDS bicuculline; GABA; immediate early genes; immunohistochemistry;mediodorsal thalamus; microdialysis

ABSTRACT Stimulation of the mediodorsal and midline thalamic nuclei excitescortical neurons and induces c-fos expression in the prefrontal cortex. Data in theliterature data suggest that pyramidal neurons are the most likely cellular targets. Inorder to determine whether cortical interneurons are also impacted by activation ofmediodorsal/midline thalamic nuclei, we studied the effects of thalamic stimulation on(1) Fos protein expression in g-aminobutyric acid (GABA)-immunoreactive neurons andon (2) extracellular GABA levels in the prefrontal cortex of rats. Perfusion of the GABA-Areceptor antagonist bicuculline for 20 minutes through a dialysis probe implanted intothe mediodorsal thalamus induced Fos-like immunoreactivity (IR) approximately 1 hourlater in the thalamus and in the medial prefrontal cortex of freely moving rats.Immunohistochemical double-labeling for Fos-like IR and GABA-like IR showed thatabout 8% of Fos-like IR nuclei in the prelimbic and infralimbic areas were located inGABA-like IR neurons. Fos-like IR was detected in three major subsets of GABAergicneurons defined by calbindin, parvalbumin, or vasoactive intestinal peptide (VIP)-likeIR. Dual probe dialysis showed that the extracellular levels of GABA in the prefrontalcortex did not change in response to thalamic stimulation. These data indicate thatactivation of thalamocortical neurons indeed affects the activity of GABAergic neurons asshown by the induction of Fos-like IR but that these metabolic changes are not reflectedin changes of extracellular GABA levels that are sampled by microdialysis. Synapse30:156–165, 1998. r 1998 Wiley-Liss, Inc.

INTRODUCTION

The mediodorsal thalamic nucleus (MD) and adja-cent midline and intralaminar thalamic nuclei provideinput to the medial prefrontal cortex (PFC) of rats(Berendse and Groenewegen, 1991; Krettek and Price,1977; Uylings and van Eden, 1990). In anaesthetizedrats, electrical stimulation of the MD induces cell firingin the PFC that is blocked by non-N-methyl-D-aspar-tate glutamate receptor antagonists (Ferron et al.,1984; Gigg et al., 1992; Pirot et al., 1994; Thierry et al.,1990). Recent studies in awake, freely moving rats haveshown that pharmacological stimulation of the MD/

midline thalamic nuclei that contain both glutamateand aspartate-containing neurons (Bentivoglio et al.,

Contract grant sponsor: Van den Houten Foundation.

*Correspondence to: Dr. M.G.P. Feenstra, Netherlands Institute for BrainResearch, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands. E-mail:[email protected].

Michael Bubser’s present address is Vanderbilt University School of Medicine,Department of Psychiatry, Psychiatric Hospital at Vanderbilt, Suite 305A, 160123rd Avenue South, Nashville, TN 37212.

Erna B.H.W. Erdsieck-Ernste’s present address is Department of Anatomy andEmbryology, Faculty of Medicine, University of Amsterdam, Meibergdreef 15,1105 AZ Amsterdam, The Netherlands.

Received 29 July 1997; Accepted in revised form 3 December 1997

SYNAPSE 30:156–165 (1998)

r 1998 WILEY-LISS, INC.

1996) causes a transient increase of extracellular gluta-mate levels in the PFC (Feenstra et al., 1995) andinduces c-fos mRNA and Fos-like immunoreactivity(IR) in PFC neurons (Bubser et al., 1997; Erdtsieck-Ernste et al., 1995).

The cellular targets exhibiting c-fos in the PFC inresponse to stimulation of thalamic afferents have notyet been determined. Since MD neurons form asymmet-ric synapses with dendritic spines of pyramidal neu-rons of the PFC (Kuroda et al., 1995), and there are nostudies available that demonstrate synapses betweenthalamocortical neurons and prefrontal cortical inter-neurons, it is reasonable to assume that all of the c-fosexpressing neurons are pyramidal cells. However, thishas not yet been demonstrated and in other corticalareas, thalamic afferents have also been shown tocontact GABAergic interneurons (Freund et al., 1985).Given the involvement of both prefrontal GABAergicinterneurons (Wilson et al., 1994) and the MD (Markow-itsch, 1982) in PFC-dependent cognitive processes, andthe numerous indications that pathology of both PFCinterneurons and MD may be involved in schizophrenia(Akbarian et al., 1995; Andreasen et al., 1994; Davissand Lewis, 1995; Pakkenberg, 1992; Siegel et al., 1993),it would be of considerable interest whether prefrontalinterneurons are activated by MD thalamic input.

Cortical interneurons have been characterized exten-sively using immunohistochemical and electrophysi-ological techniques. It has been estimated that 16% ofall cortical neurons are interneurons, the vast majorityof which contain the inhibitory amino acid, g-aminobu-tyric acid (GABA) as transmitter (Beaulieu, 1993;Gabbott et al., 1997). Except for layer I, GABAergicneurons are fairly evenly distributed throughout thePFC (Beaulieu, 1993; Esclapez et al., 1987; Gabbott etal., 1997), but they form different subsets as defined bytheir morphology, electrophysiological properties, andtheir content of the calcium-binding proteins and neuro-peptides (Connors and Gutnick, 1990; DeFelipe, 1993;Hendry et al., 1989; Kawaguchi and Kubota, 1993;Kubota et al., 1994).

In the present study, we applied immunohistochemi-cal double staining in order to determine whetherthalamic disinhibition by local perfusion of the GABA-Areceptor antagonist, bicuculline, in the MD/midlinethalamus transsynaptically activates GABA-like immu-noreactive (IR) neurons of the PFC as shown by in-creased expression of Fos-like IR. Fos, the proteinproduct of the immediate early gene, c-fos, is expressedin many, but not all neural systems in response to avariety of physiological and pharmacological stimuli(Deutch et al., 1991, 1992; Morgan, 1990; Robertsonand Fibiger, 1992; Sagar et al., 1988). We determinedthe proportion of Fos-like IR nuclei that are expressedin GABA-like IR PFC neurons and we assessed whetherFos-like IR is induced in the major subsets of corticalGABAergic neurons as defined by calbindin, parvalbu-

min, or vasoactive intestinal peptide (VIP)-like IR.Finally, we applied dual probe microdialysis (Feenstraet al., 1993; Giovannini et al., 1994; Moor et al., 1994;Santiago and Westerink, 1991) to study the effects ofstimulation of the MD/midline thalamus on extracellu-lar levels of GABA in the PFC.

MATERIALS AND METHODSSubjects and housing

Male Wistar rats (Harlan, Zeist, The Netherlands)were housed in groups of four to six animals and theywere maintained on a 12:12 hour light/dark schedule(lights on at 07:00 a.m.). A minimum of 2 days beforesurgery (experiment 1) or on the day of surgery (experi-ment 2), rats were transferred to Perspex cages (25 325 3 32 cm) where they were housed individually withfood and water available ad libitum.

Experiment 1: Thalamic stimulation and Fosexpression in prefrontal Gabaergic neurons

Surgery

Rats weighing 240–315 g were implanted underchloral hydrate anaesthesia (400 mg/kg, i.p.) with aconcentric dialysis probe into the MD in the righthemisphere (AP -2.5 mm [bregma], L 11.5 mm, V -6.5mm with a medio-lateral angle of 12°; incisor bar 2.5mm below the interaural line) (Paxinos and Watson,1986). The effective dialysis area of the probe (length1.5 mm, diameter 0.33 mm) was made from Hospaldialysis tubing (Erdtsieck-Ernste et al., 1995).

Microdialysis

Two days after surgery, the dialysis probe was at-tached to the perfusion system via a dual-channelliquid swivel (Instech 375, Plymouth Meeting, PA,USA) and perfused at a flow rate of 3.0 µl/min with amodified Ringer solution (Ringer: 145 mM NaCl, 2.7mM KCl, 1.2 mM CaCl2, and 1.0 mM MgCl2, pH5.5–6.5). Following 4 hours of Ringer perfusion, 0.1 mMbicuculline methyl chloride (Research BiochemicalsInternational, Natick, MA) dissolved in Ringer wasperfused for 20 minutes; thereafter, Ringer perfusionwas continued. Rats that were continuously perfusedwith Ringer served as non-stimulated controls.

Perfusion and immunohistochemistry

Seventy-five minutes after bicuculline had reachedthe brain, rats were deeply anaesthetized with Nembu-tal (60 mg/kg, i.p.) and intracardially perfused with 100ml physiological saline followed by 500 ml of aldehydesdissolved in 0.1 M phosphate buffer (pH 7.4). A mixtureof 1% glutaraldehyde/2.5% paraformaldehyde was usedfor Fos-like IR and GABA-like IR staining whereas 4%paraformaldehyde was used when staining for Fos-like

157THALAMIC REGULATION OF PREFRONTAL GABA NEURONS

IR and calbindin-IR, parvalbumin-IR or VIP-like IR.Following post-fixation, 50 µm sections were cut on avibratome, collected in 50 mM Tris-buffered saline(TBS, pH 7.4), thoroughly washed in TBS, and pre-treated for 20 minutes with TBS containing 1% H2O2

and 0.5 % TritonX-100. For all subsequent incubations,antisera were diluted in TBS containing 0.25 % gelatineand 0.5 % TritonX-100. In a first step, sections wereincubated overnight at 4°C in a sheep polyclonal antise-rum against Fos and Fos-related antigens (CambridgeResearch Biochemicals, Cambridge, U.K., OA-11–823,lot 09046) in a dilution of 1:10,000 (Bubser et al., 1997;Erdtsieck-Ernste et al., 1995). Subsequently, sectionswere incubated at room temperature for 1.5 hours witha biotinylated rabbit anti-sheep antibody (Vector Labo-ratories, Burlingame, CA, 1:400) and thereafter for 1.75hour with an avidin-biotin complex (Vectastain Elite,Vector Laboratories, 1:800). Fos-like IR was visualizedby incubating the sections for 10 minutes with 0.05%diaminobenzidine, 0.01% H2O2, and 0.2% ammonium-nickelsulfate dissolved in 50 mM Tris-HCl (pH 7.4)which produced a black precipitate located in thenucleus. For double-staining experiments, the remain-ing peroxidase was inactivated by washes in methanol/H2O2. Thereafter, sections were incubated overnight inmouse monoclonal antibodies raised against eithercalbindin (clone no. CL-300, Sigma, St. Louis, MO,1:500) or parvalbumin (clone no. PA235, Sigma,1:10,000) or in rabbit polyclonal antisera raised againstVIP (Netherlands Institute for Brain Research [NIBR],1:1,000) or GABA (NIBR, 1:2,000; van Eden et al.,1989). Incubations with secondary and tertiary antibod-ies were carried out for 1 hour at room temperature andimmunoreactivity for the second antigen was visual-ized by reacting the sections with 0.05% diaminobenzi-dine and 0.01% H2O2 dissolved in 50 mM Tris-HCl (pH7.4), which produced a brown reaction product in thecytoplasm. Following each incubation step, sectionswere washed for 4 x 15 minutes with TBS.

Cell counting and charting

In four bicuculline-stimulated rats and one Ringer-perfused rat, the percentage of Fos-like IR nuclei thatwere located in GABA-like IR cells was estimated. Forthis purpose, layers III and V of the ventral PFC(prelimbic [PL] and infralimbic [IL] subareas) werestudied in the section 2.5 mm anterior to bregma. Witha Zeiss (Weesp, The Netherlands) microscope withPLAN 40x (oil immersion) objective and PLAN 12.5xoculars, the PL and IL areas were divided into squaresof 95 x 95 µm. Divided in this way, PL consisted of 12rows and 10––12 columns and IL of 6 rows and 10columns, with column 1 always bordering on the pialsurface. In order to investigate layer III, counting wasperformed in squares alternating in columns 3 and 4 inthe subsequent rows, and to investigate layer V, alter-nating columns 7 and 8 were counted. This was done

both ipsilateral and contralateral to the side of probeplacement in the thalamus.

The percentage of Fos-like IR nuclei located in GABA-like IR neurons of layers III and V was calculated asfollows: % double labelled cells 5 100 x number ofdouble labelled cells/total number of Fos-like IR cells.

In addition to the cell counts, drawings of the ipsilat-eral PFC of a Ringer-perfused rat and a bicuculline-stimulated rat were made at two different levels of thePFC in order to show the distribution of both singlelabelled Fos-like IR cells and cells double labelled forFos-like IR and GABA-like IR.

Experiment 2: Thalamic stimulation andprefrontal GABA levels

Surgery

Under chloral hydrate anaesthesia (400 mg/kg i.p.),rats weighing 260–325 g were implanted in the righthemisphere with dialysis probes aimed at the MD (AP-2.5 mm [bregma], L 11.5 mm, V -6.5 mm with amedio-lateral angle of 12°) and at the medial PFC (AP13.0 mm [bregma], L 11.8 mm, V -5.5 mm with amedio-lateral angle of 12°; incisor bar 2.5 mm below theinteraural line) (Paxinos and Watson, 1986). The effec-tive exchange lengths of the thalamic and corticalprobes were 1.5 and 3.0 mm, respectively.

Microdialysis

On the evening of the same day, the PFC probe wasconnected to the perfusion pump via a dual-channelliquid swivel (Instech 375) and Ringer solution (seeabove) was perfused overnight at a flow rate of 3.0µl/min. This procedure results in stable basal levels ofexcitatory amino acids in the PFC that respond tothalamic disinhibition (Feenstra et al., unpublishedobservations). On the following morning, the thalamicprobe was also connected to the perfusion system.Following 3 to 4 hours of Ringer perfusion through thethalamus 0.1 mM bicuculline methyl chloride dissolvedin Ringer was perfused for 20 minutes; thereafterRinger perfusion was continued.

GABA analysis

GABA was quantified by an automated on-line HPLCmethod employing pre-column derivatization withophthaldialdehyde (OPA) (Sayin et al., 1995; Westerinkand de Vries, 1989). Briefly, a T-piece was used to mixthe dialysate (flow 3 µl/min) on-line with the derivatiza-tion agent (flow 1 µl/min) and this mixture was injectedevery 10 minutes by a timer-controlled valve with a50-µl-loop onto the HPLC column (S3 ODS2 Spher-isorb, 4.2 x 100 mm, 3 µm particle size) that was kept ata temperature of 30°C. The mobile phase (final pH 5.95)consisted of an aqueous solution of 0.05 mol/l Na2HPO4

and 0.01 mmol/l EDTA containing 40% (v/v) methanoland 0.6% (v/v) tetrahydrofuran and was delivered at a

158 M. BUBSER ET AL.

flow rate of 1 ml/min by a LKB 2150 HPLC pump. Afterthe recording of the GABA peak, the column wasautomatically flushed with 600 µl 90% ethanol/H2O(v/v) to remove late eluting compounds. OPA-deriva-tized compounds were detected by means of a Shi-madzu RF-10A(Shimadzu, Hertogenbosch, The Nether-lands) spectrofluorimetric detector (excitation 340 nm,emission 460 nm) and chromatograms were recorded ona Kipp & Zonen (Delft, The Netherlands) strip chartrecorder. The derivatization agent was freshly pre-pared each day by dissolving 5 mg o-phthaldialdehydein 50 µl methanol and adding 4.95 ml 0.5mol/l NaHCO3

(adjusted to pH 9.5 with NaOH) containing 10 µl2-mercaptoethanol. Chromatographic data were cor-rected for dead volume of the perfusion system and areexpressed as percentage of baseline control (mean ofthe four samples preceding bicuculline perfusion). Afterthe experiment, rats were sacrificed by an overdose ofchloral hydrate and the brains were fixed by immersionin 4 % paraformaldehyde. Probe locations were verifiedin Nissl-stained sections.

RESULTSProbe placements

Both in Experiment 1 and 2, the dialysis area of eachthalamic probe was located in the MD with the tips ofthe probes also entering adjacent thalamic nuclei.Probe locations were comparable to those in a previousstudy (Bubser et al., 1997). All probes aimed at the PFC(Exp. 2) were located either in the deep (n 5 5) orsuperficial (n 5 2) layers of IL and/or PL.

Experiment 1: Thalamic stimulation and Fosexpression in prefrontal GABAergic neurons

Thalamus

Thalamic perfusion of bicuculline resulted in theappearance of Fos-like IR cells in the MD and midlinethalamic nuclei. Fos-like IR was limited to the ipsilat-eral MD, but it encompassed almost completely themidline paraventricular, intermediodorsal, and centro-medial thalamic nuclei. A much lower degree of Fos-likeIR expression was seen in the same areas in response toRinger perfusion (data not presented; for representa-tive figures see Bubser et al., 1997; Erdtsieck-Ernste etal., 1995).

Prefrontal cortex

The distribution of Fos-like IR cells in the ipsilateralPFC in response to thalamic perfusion of Ringer orbicuculline is depicted in Figure 1.

Fos-like IR cells were present in the PL, IL, anteriorcingulate cortex, and area 2 of the frontal cortex (Fr2,Paxinos and Watson, 1986). Outside of these corticalareas, the number of Fos-like IR nuclei decreasedsharply (data not presented). Fos-like IR was alsoobserved contralateral to the side of thalamic perfusion,

but the number of Fos-like IR cells was higher on theipsilateral side (Table I).

Fos-Like IR in GABA-like IR neurons

Immunohistochemical staining for GABA-like IR vi-sualized neurons that were homogeneously spread overthe different cortical layers (except layer I) and over thesubareas of the PFC. GABA-like IR neurons expressingFos-like IR could be clearly recognized by a blacknuclear precipitate surrounded by a brownish cytoplas-mic precipitate (Fig. 2A). In the bicuculline-perfusedrats, Fos-like IR nuclei that were located in GABA-likeIR cells were on average found to be equally distributedover the investigated areas (Table II). However, inindividual rats, the level of double-labeling varied andthe example in Figure1B had more double labeled cellsin the IL than in the PL. Although there was somevariability in the frequency with which double labeledcells were encountered, the percentage of double la-beled cells was always low. The mean percentage 6S.E.M of double labeled cells was 8.2% 6 1.9 (n 5 4) inthe ipsilateral PFC and 7.7 % 6 2.3 (n 5 4) of the totalnumber of cells with Fos-like IR in the contralateralPFC. In the Ringer-perfused rat, hardly any doublelabeled neurons were present (see Fig. 1.A).

Fos-Like IR in subsets of GABAergic neurons

Double staining for Fos-like IR and calbindin IR,parvalbumin IR, or VIP-like IR showed that Fos-like IRnuclei were present in some neurons of each cellpopulation defined by these cytoplasmatic markers(Fig. 2B–D).

Experiment 2: Thalamic stimulation andprefrontal GABA levels

During Ringer perfusion of the thalamus, the extra-cellular levels of GABA in the PFC averaged 19.7 6 3.2fmol/min (mean 6 S.E.M, n 5 7). One-way repeatedmeasures ANOVA revealed that 20 minutes perfusionof 0.1 mM bicuculline through the MD did not affectextracellular GABA levels in 10-minute fractions ofdialysate from the PFC (F11,73 5 0.49, P 5 0.90, Fig. 3).

DISCUSSION

The aim of the present study was to determine theeffects of thalamic stimulation on the activity ofGABAergic neurons in the PFC. Our data show thatpharmacological stimulation of the mediodorsal/midlinethalamus induces Fos-like IR in a large number of neuronsin the PFC, including GABA-like IR neurons, withoutaffecting extracellular levels of GABA in the PFC.

In accordance with previous studies, relieving thethalamus of its GABAergic inhibition induced Fos-likeIR in the MD/ midline thalamus and in the PFC(Bubser et al., 1997; Erdtsieck-Ernste et al., 1995). Thenumber of Fos-like IR nuclei was higher in the PFC


Fig. 1. Chartings of single-labelled Fos-like IR cells (closed circles) or cells dual la-belled (open circles) for Fos-like IR/GABA-likeIR in two rostro-caudal levels of the medialprefrontal cortex following Ringer perfusion(A1/A2) or 20 minutes perfusion of bicuculline(0.1 mM) (B1/B2) through the mediodorsalnucleus of the thalamus. Note the sparseexpression of Fos-LI and the virtual absence ofdouble labelled cells in the Ringer-perfusedrat and the increased number of Fos-LI cellsand double-labelled cells in the bicuculline-perfused rat.


ipsilateral to the thalamic dialysis probe than in thecontralateral PFC. This effect was most prominent inthe prelimbic cortex, which receives thalamic inputpredominantly from the MD. In the infralimbic cortex,

the difference between ipsilateral vs. contralateral acti-vation was less pronounced, which was most likely dueto the ‘‘bilateral’’ activation of the midline thalamicnuclei providing input to right and left infralimbic

TABLE I. Number of Fos-like IR nuclei in the ipsilateral and contralateral medial prefrontal cortex ofrats after bicuculline perfusion of the mediodorsal/midline thalamus1

Area Layer

Fos-like immunoreactive nuclei

Ipsilateral Contralateral

Mean Range % Mean Range %

Prelimbic area III 51.8 24–68 34.7 8.0 2–19 13.7V 33.3 30–37 22.3 11.5 3–16 19.7

Infralimbic area III 39.3 32–44 26.2 24.0 9–18 41.0V 25.0 18–33 16.8 15.0 14–34 25.6

Total 149.4 117–163 100 58.5 36–86 1001Fos-like IR nuclei were counted as described in Materials and Methods. Mean represents four bicuculline-perfused rats.

Fig. 2. High-power photomicrographs of sections through the prefrontal cortex that were processed forFos-like IR in combination with GABA-like IR (A), calbindin IR (B), parvalbumin IR (C), or VIP-like IR(D). All sections were derived from rats receiving thalamic perfusion of 0.1 mM bicuculline for 20 minutes.Double-labelling of Fos-LI and the cytoplasmic marker is indicated by an arrow.


cortex. The ‘‘bilateral’’ activation of the unpaired mid-line thalamic nuclei indicates that there was somediffusion of bicuculline to the side contralateral to thedialysis probe implantation (see also fig. 5 in Erdtsieck-Ernste et al., 1995). However, this spread appears tohave been limited to the midline nuclei because therewas no Fos-like IR induction in the contralateral MD.As one of the main mechanisms involved in c-fosinduction is neuronal excitation and subsequent cal-cium influx (Sheng et al., 1990), the transsynapticexpression of Fos-like IR in the PFC is in line with datashowing that stimulation of mediodorsal/midline tha-lamic nuclei increases extracellular levels of glutamatein the PFC and excites PFC neurons (Feenstra et al.,1995; Ferron et al., 1984; Gigg et al., 1992; Thierry etal., 1990).

Using immunohistochemical double labelling, weshowed that in bicuculline-stimulated rats about 8% ofFos-like IR nuclei are located in GABA-like IR neurons.

According to recent quantitative studies, < 16% of allneurons in various frontal cortical areas of the rat areGABAergic (Beaulieu, 1993; Gabbott et al., 1997). Evenif one takes into account that some weakly stainedFos-like IR nuclei may not have been detected in darklystained GABA-like IR cells, our data suggest thatFos-like IR is less frequently expressed in GABA-likeIR cells than in non-GABAergic neurons. As the vastmajority of cortical interneurons use GABAas transmit-ter, our data indicate that pyramidal neurons are thepredominant, albeit not the sole neuronal target beingactivated by thalamic stimulation.Apreferential activa-tion of pyramidal cells in the PFC has also been shownin response to other pharmacological challenges (Deutchand Duman, 1996). On the basis of our data, it cannotbe conclusively determined whether GABAergic neu-rons are less likely to be activated by thalamic stimula-tion or whether they are less prone to express Fos thanpyramidal cells.

As GABAergic interneurons do not form a homoge-neous class of neurons, we examined whether thalamicstimulation induces Fos-like IR in distinct subsets ofGABAergic interneurons. Calbindin, parvalbumin, andVIP (which is colocalized with calretinin) are usefulmarkers for defining largely non-overlapping subsets ofGABAergic interneurons (DeFelipe, 1993; Hendry etal., 1989; Kubota et al., 1994) that also differ in theirelectrophysiological properties and in their morphology(Connors and Gutnick, 1990; Hendry et al., 1989;Kawaguchi and Kubota, 1993; Kawaguchi, 1995). Ourdata show that members of each of these subsets ofinterneurons express Fos-like IR in response to tha-lamic stimulation.

The activation of pyramidal neurons is most likelydue to direct synaptic input from thalamic afferentsforming asymmetric synapses with dendritic spines ofpyramidal cells (Kuroda et al., 1993, 1995), but themechanisms that may underlie the thalamically-induced activation of cortical GABAergic neurons re-main elusive. To the best of our knowledge, synapticcontacts between thalamic afferents of the PFC andprefrontal interneurons have not been reported inneuroanatomical studies, as it has been shown in thevisual cortex (Freund et al., 1985). Monosynaptic tha-lamic input to neuronal elements with interneuroncharacteristics, i.e., fast spiking neurons (Connors andGutnick, 1990, Kawaguchi and Kubota, 1993) wererevealed in an electrophysiological study in the visualcortex (Swadlow, 1988). Electrophysiological studiesand neuronal network analysis of thalamocortical sys-tems have led to the hypothesis that, in addition topyramidal neurons, inhibitory interneurons are animportant target of thalamocortical projections (Doug-las and Martin, 1991). Another possibility that has to betaken into account is that the activation of GABAergicneurons is caused indirectly via the collaterals ofthalamically-activated pyramidal cells (Buhl et al.,

TABLE II. Percentage of Fos-like IR nuclei located in GABA-like IRneurons in the ipsilateral and contralateral medial prefrontal cortex

of rats after bicuculline perfusion of the mediodorsal/midlinethalamus1

Area Layer

Percentage of double labelled neurons

Ipsilateral Contralateral

Mean Range Mean Range

Prelimbic area III 6.33 1.5–13.7 0V 9.0 2.9–16.2 6.5 0–15.4

Infralimbic area III 8.9 2.3–18.8 7.3 2.9–15.4V 9.0 5.3–12.1 13.3 0–22.2

1Mean represents four bicuculline-perfused rats.

Fig. 3. GABA levels in 10-minute fractions of dialysate from themedial prefrontal cortex following 20 minutes perfusion of bicuculline(0.1 mM) through a dialysis probe implanted into the mediodorsalnucleus of the thalamus. Data are means 6 S.E.M. of 7 rats. One-wayrepeated measures ANOVA revealed no effect of bicuculline perfusionon prefrontal GABA levels (F11,73 5 0.49, P 5 0.90).


1997; Deuchars and Thomson, 1995; McGuire et al.,1991; White, 1989). Finally, it is conceivable that theincrease in dopaminergic activity in the PFC that isinduced by thalamic stimulation (Feenstra et al., 1993;Jones et al., 1988) leads to an activation of GABAergicneurons (Bourdelais and Deutch, 1994; Retaux et al.,1990).

The microdialysis experiments were carried out inorder to determine whether, in addition to the inductionof Fos-like IR in GABAergic neurons, thalamic stimula-tion affects extracellular GABA levels. Our microdialy-sis experiments failed to show any stimulation-inducedchanges in extracellular GABA levels. We had expecteda stimulation-induced increase in prefrontal GABA,because thalamic stimulation excites cortical neurons(Ferron et al., 1984; Thierry et al., 1990) includingputative cortical interneurons (Swadlow, 1988) andbecause stimulation of glutamate receptors increasescortical and cerebellar GABA release in vitro and invivo (Castro-Alamancos and Torres-Aleman, 1993; Dre-jer et al., 1987; Dunlop et al. 1991; Moghaddam et al.,1996; for review see Ruzicka and Jhamandas, 1993).

There are several points that may explain this unex-pected finding. On the one hand, it is possible that theexcitatory projections from the MD/midline thalamusto the PFC do not control prefrontal GABA release.Such a conclusion would imply that there is a dissocia-tion between the thalamically-mediated activation ofGABAergic neurons, as shown by the induction ofFos-like IR, and the synaptic release of GABA in thePFC. However, as GABAergic neurons do not onlysynapse onto pyramidal cells (Cowan et al., 1994;DeFelipe and Farinas, 1992) but also onto otherGABAergic neurons (Kisvarday et al., 1990), it is alsopossible that an initial increase in prefrontal GABArelease resulted in the inhibition of ‘‘secondary’’GABAergic neurons. Taking into account that the stimu-lus parameters used in the present study produce onlya short-lasting increase of glutamate in the PFC (Feen-stra et al., 1995) and that microdialysis samples werecollected over a 10-minute period, it is conceivable thatan initial increase in GABA may have been obscured bya subsequent reduction of GABA release from ‘‘second-ary’’ GABAergic neurons. Evidence for short-lastingincreases of extracelluar GABA after afferent stimula-tion has been presented by Timmerman and Westerink(1997a), using 3-minute sampling times. Finally, it hasto be mentioned that the basal levels of GABA indialysate appear not to be solely derived from synapticsources (Timmerman and Westerink, 1997b; Unger-stedt, 1991), because dialysate levels of GABA are not(Campbell et al., 1993; Westerink and de Vries, 1989) orare only in part reduced by tetrodotoxin or Ca21 deple-tion (Bourdelais and Kalivas, 1992; Bourdelais andDeutch, 1994; Osborne et al., 1991). The presence ofsuch a ‘‘non-responsive,’’ non-synaptic pool of extracellu-

lar GABA may have blunted changes in synaptically-released GABA.

In summary, our study shows that a small percentageof the neurons that are metabolically activated bystimulation of the midline/mediodorsal thalamus areGABAergic and belong to several distinct subsets ofinterneurons defined by their content of calcium-binding proteins or VIP. The question as to whether themetabolic activation of prefrontal GABAergic neuronsis accompanied by an increased synaptic release ofGABA cannot be answered conclusively. Nevertheless,the finding that GABAergic neurons are impacted bythalamocortical neurons has several functional implica-tions. For example, the activation of GABAergic inter-neurons, which are known to have multiple contactswith pyramidal cells (Cowan et al., 1994; DeFelipe andFarinas, 1992; Kisvarday et al., 1990; Tamas et al.,1997), may serve to protect the cortex from excessiveexcitation (see also Ruzicka and Jhamandas, 1993).Furthermore, it is tentatively suggested that thalamo-cortical neurons may be one of the driving forces ofprefrontal GABAergic neurons that have recently beenshown to be involved in cognitive processes (Wilson etal., 1994).

ACKNOWLEDGMENTS

We thank Dr. Corbert G. van Eden, Dr. Holly Moore,and Prof. Dr. Harry B. M. Uylings for helpful discus-sions and their comments on this manuscript. Theexpert technical assistance of Paul Evers is gratefullyacknowledged. We thank Henk Stoffels for drawing thefigures and Gerben van der Meulen for photography.This work was supported by the Van den HoutenFoundation.

REFERENCES

Akbarian, S., Kim, J.J., Potkin, J.O., Tafazzoli, A., Bunney Jr., W.E.,and Jones, E.G. (1995) Gene expression for glutamic acid decarbox-ylase is reduced without loss of neurons in prefrontal cortex ofschizophrenics. Arch. Gen. Psychiatry, 52:258–266.

Andreasen, N.C., Arndt, S., Swayze II, V., Cizadlo, T., Flaum, M.,O’Leary, D., Ehrhardt, J.C., and Yuh, W.T.c. (1994) Thalamicabnormalities in schizophrenia visualized through magnetic reso-nance image averaging. Science, 266:294–298.

Beaulieu, M. (1993) Numerical data on neocortical neurons in adultrat, with special reference to the GABA population. Brain Res.,609:284–292.

Bentivoglio, M., Frassoni, C., and Spreafico, R. (1996) Glutamate,aspartate and calbindin in the medial thalamus: An immunohisto-chemical study in the rat. Soc. Neurosci. Abstr., 22:357.

Berendse, H.W., and Groenewegen, H.J. (1991) Restricted corticalterminal fields of the midline and intralaminar thalamic nuclei inthe rat. Neuroscience, 42:73–102.

Bourdelais, A.J. and Deutch, A.Y. (1994) The effects of haloperidol andclozapine on extracellular GABA levels in the prefrontal cortex ofthe rat. An in vivo microdialysis study. Cerebr. Cortex, 4:69–77.

Bourdelais, A.J., and Ka1ivas, P.W. (1992) Modulation of extracellularg-aminobutyric acid in the ventral pallidum using in vivo microdialy-sis. J. Neurochem., 58:2311–2320.

Bubser, M., Feenstra, M.G.P., Erdtsieck-Ernste, E.B.H.W., Botter-blom, M.H.A., van Uum, H.F.M., and Pool, C.W. (1997) Modulatoryrole of catecholamines in the transsynaptic expression of c-fos in therat medial prefrontal cortex by disinhibition of the mediodorsalthalamus: A study employing microdialysis and immunohistochem-istry. Brain Res., 749:214–225.


Buhl, E.H., Tamas, G., Szilagyi, T., Stricker, C., Paulsen, O., andSomogyi, P. (1997) Effect, number and location of synapses made bysingle pyramidal cells onto aspiny interneurones of cat visual cortex.J. Physiol., 500:689–713.

Campbell, K., Kalen, P., Lundberg, C., Wictorin, K., Rosengren, E., andBjorklund, A. (1993) Extracellular g-aminobutyric acid levels in therat caudate-putamen: Monitoring the neuronal and glial contribu-tion by intracerebral microdialysis. Brain Res., 614:241–250.

Castro-Alamancos, M.A., and Torres-Aleman, I. (1993) Long-termdepression of glutamate-induced gamma-aminobutyric acid releasein cerebellum by insulin-like growth factor I. Proc. Natl. Acad. Sci.U.S.A., 90:7386–7390.

Connors, B.W., and Gutnick, M.J. (1990) Intrinsic firing patterns ofdiverse neocortical neurons. Trends Neurosci., 13:99–104.

Cowan, R.L., Sesack, S.R., van Bockstaele, E.J., Branchereau, P.,Chan, J., and Pickel, V.M. (1994) Analysis of synaptic inputs andtargets of physiologically characterized neurons in rat frontalcortex: Combined in vivo intracellular recording and immunolabel-ing. Synapse 17:101––114.

Daviss, S.R., and Lewis, D.A. (1995) Local circuit neurons of theprefrontal cortex in schizophrenia: Selective increase in the densityof calbindin-immunoreactive neurons. Psychiat. Res., 59:81–96.

DeFelipe, J. (1993) Neocortical neuronal diversity: Chemical heteroge-neity revealed by colocalization studies of classic neurotransmitters,neuropeptides, calcium-binding proteins, and cell surface mol-ecules. Cerebr. Cortex, 3:273–289.

DeFelipe, J., and Farinas, I. (1992) The pyramidal neuron of thecerebral cortex: Morphological and chemical characteristics of thesynaptic inputs. Prog. Neurobiol., 39:563–607.

Deuchars, J., and Thomson, A.M. (1995) Innervation of burst firingspiny interneurons by pyramidal cells in deep layers of rat somato-motor cortex: Paired intracellular recordings with biocytin filling.Neuroscience, 69:739–755.

Deutch, A.Y., and Duman, R.S. (1996) The effects of antipsychoticdrugs on Fos protein expression in the prefrontal cortex: Cellularlocalization and pharmacological characterization. Neuroscience70:377–389.

Deutch, A.Y., Lee, M.C., Gillham, M.H., Cameron, D.A., Goldstein, M.,and Iadarola, M.J. (1991) Stress selectively increases Fos protein indopamine neurons innervating the prefrontal cortex. Cerebr. Cor-tex, 1:273–292.

Deutch, A.Y., Lee, M.C., and Iadarola, M.J. (1992) Regionally specificeffects of atypical antipsychotic drugs on striatal Fos expression:The nucleus accumbens shell as a locus of antipsychotic action. Mol.Cell. Neurosci., 3:332–341.

Douglas, R.J., and Martin, K.A.C. (1991) A functional microcircuit forcat visual cortex. J. Physiol. (London), 440:735–769.

Drejer, J., Honore, T., and Schousboe, A. (1987) Excitatory aminoacid-induced release of 3H-GABA from cultured mouse cerebralcortex interneurons. J. Neurosci., 7:2910–2916.

Dunlop, J., Grieve, A., Schousboe, A., and Griffiths R. (1991) Stimula-tion of gamma-[3H]aminobutyric acid release from cultured mousecerebral cortex neurons by sulphur-containing excitatory amino acidtransmitter candidates: Receptor activation mediates two distinctmechanisms of release. J. Neurochem., 57:1388–1397.

Erdtsieck-Ernste, E.B.H.W., Feenstra, M.G.P., Botterblom, M.H.A.,Van Uum, H.F.M., Sluiter, A.A., and Heinsbroek, R.P.W. (1995) C-fosexpression in the rat brain after pharmacological stimulation of therat ‘‘mediodorsal’’ thalamus by means of microdialysis. Neurosci-ence, 66:115–131.

Esclapez, M., Campistron, G., and Trottier, S. (1987) Immunocyto-chemical localization and morphology of GABA-containing neuronsin the prefrontal and frontoparietal cortex of the rat. Neurosci. Lett.,77:131–136.

Feenstra, M.G.P., van der Wei, W., Hamstra, J.J., Botterblom, M.H.A.,and Buijs, R.M. (1993) In vivo dopamine release in the rat prefrontalcortex is increased by stimulation of the mediodorsal thalamicnucleus. Soc. Neurosci. Abstr., 19:1382.

Feenstra, M.G.P., Bubser, M., Erdtsieck-Ernste, E.B.H.W., and Botter-blom, M.A. (1995) Stimulation of the mediodorsal thalamus-prefrontal cortex pathway: In vivo transmitter release, fos expres-sion and behaviour. Eur. J. Neurosci. (Suppl.), 8:102.

Ferron, A., Thierry, A.M., Le Douarin, C., and Glowinski, J. (1984)Inhibitory influence of the mesocortical dopaminergic system onspontaneous activity or excitatory response induced from the tha-lamic mediodorsal nucleus in the rat medial prefrontal cortex. BrainRes., 302:257–265.

Freund, T.F., Martin, K.A.C., Somogyi, P., and Whitteridge, D. (1985).Innervation of cat visual areas 17 and 18 by physiologically identi-fied X- and Y-type afferents. II. Identification of postsynaptic targetsby GABA immunocytochemistry and Golgi impregnation. J. Comp.Neurol. 242:275–291.

Gabbott, P.L.A., Dickie, B.G.M., Vaid, R.R., Headlam, A.J.N., andBacon, S.J. (1997) Local-circuit neurones in the medial prefrontalcortex (areas 25, 32 and 24b) in the rat: Morphology and quantita-tive distribution. J. Comp. Neurol. 377:465–499.

Gigg, J., Tan, A.M., and Finch, D.M. (1992) Glutamatergic excitatoryresponses of anterior cingulate neurons to stimulation of themediodorsal thalamus and their regulation by GABA: An in vivoiontophoretic study. Cerebr. Cortex, 2:477–484.

Giovannini, M.G., Mutolo, D., Bianchi, L., Michelassi, A., and Pepeu,G. (1994) NMDA receptor antagonists decrease GABA outflow fromthe septum and increase acetylcholine outflow from the hippocam-pus: A microdialysis study. J. Neurosci., 14:1358–1365.

Hendry, S.H.C., Jones, E.G., Emson, P.C., Lawson, D.E.M., Heizmann,C.W., and Streit, P. (1989) Two classes of cortical GABA neuronsdefined by differential calcium binding protein immunoreactivities.Exp. Brain Res., 76:467–472.

Jones, M.W., Kilpatrick, I.C., and Phillipson, O.T. (1988) Dopaminefunction in the prefrontal cortex of the rat is sensitive to a reductionof tonic GABA-mediated inhibition in the thalamic mediodorsalnucleus. Exp. Brain Res., 69:623–634.

Kawaguchi, Y. (1995) Physiological subgroups of nonpyramidal cellswith specific morphological characteristics in layer II/III of ratfrontal cortex. J. Neurosci., 15:2638–2655.

Kawaguchi, Y., and Kubota, Y. (1993) Correlation of physiologicalsubgroupings of nonpyramidal cells with parvalbumin- andcalbindinD28K-immunoreactive neurons in layer V of rat frontalcortex. J. Neurophysiol., 70:387–396.

Kisvarday, Z.F., Gulyas, A., Beroukas, D., North, J.B., Chubb, I.W.,and Somogyi, P. (1990) Synapses, axonal and dendritic patterns ofGABA-immunoreactive neurons in human cerebral cortex. Brain113:793–812.

Krettek, J.E., and Price, J.L. (1977) The projections of the mediodorsalnucleus and adjacent thalamic nuclei in the rat. J. Comp. Neurol.,171:157–192.

Kubota, Y., Hattori, R., and Yui, Y. (1994) Three distinct subpopula-tions of GABAergic neurons in rat frontal agranular cortex. BrainRes., 649:159–173.

Kuroda, M., Murakami, K., Oda, S., Shinkai, M., and Kishi, K. (1993)Direct synaptic connections between thalamocortical axon termi-nals from the mediodorsal thalamic nucleus (MD) and corticotha-lamic neurons to MD in the prefrontal cortex. Brain Res., 612:339–344.

Kuroda, M., Murakami, K., Kishi, K., and Price, J.L. (1995) Thalamo-cortical synapses between axons from the mediodorsal thalamicnucleus and pyramidal cells in the prelimbic cortex of the rat. J.Comp. Neurol., 356:143–151.

Markowitsch, H. (1982) Thalamic mediodorsal nucleus and memory: Acritical evaluation of studies in animals and man. Neurosci. Biobe-hav. Rev., 6:351–380.

McGuire, B.A., Gilbert, C.D., Rivlin, P.K., and Wiesel, T.N. (1991)Targets of horizontal connections in macaque primary visual cortex.J. Comp. Neurol., 305:370–392.

Moghaddam, B., Bagley, J., and Bruins, H. (1996) Glutamatergicregulation of GABA release in the prefrontal cortex during basal andstress-activated conditions. Soc. Neurosci. Abstr. 22:74.

Moor, E., de Boer, P., Beldhuis, H.J.A., and Westerink, B.H.C. (1994) Anovel approach for studying septo-hippocampal cholinergic neuronsin freely moving rats: A microdialysis study with dual probe design.Brain Res., 648:32–38.

Morgan, J.I. (1990) Proto-oncogene expression in the nervous system.Disc. Neurosci., 7:7–49.

Osborne, P.G., O’Connor, W.T., Kehr, J., and Ungerstedt, U. (1991) Invivo characterisation of extracellular dopamine, GABA and acetyl-choline from the dorsolateral striattum of awake freely moving ratsby chronic microdialysis. J. Neurosci. Methods, 37:93–102.

Pakkenberg, B. (1992) The volume of the mediodorsal thalamicnucleus in treated and untreated schizophrenics. SchizophreniaRes., 7:95–100.

Paxinos G., and Watson, C. (1986) The Rat Brain in StereotaxicCoordinates, 2nd ed., Academic Press, London.

Pirot, S., Jay, T.M., Glowinski, J., and Thierry, A.-M. (1994) Anatomi-cal and electrophysiological evidence for an excitatory amino acidpathway from the thalamic mediodorsal nucleus to the prefrontalcortex in the rat. Europ. J. Neurosci., 6:1225–1234.

Retaux, S., Besson, M.J., and Penit-Soria, J. (1990) D2 dopaminergicreceptor activation enhances the spontaneous release of 3H-GABAin the rat prefrontal cortex, in vitro. C.R. Acad. Sci. Paris, 311(SerieIII):295–300.

Robertson, G.S., and Fibiger, H.C. (1992) Neuroleptics increase c-fosexpression in the forebrain: contrasting effects of haloperidol andclozapine. Neuroscience, 46:315–328.

Ruzicka, B.B., and Jhamandas, K.H. (1993) Excitatory amino acid


action on the release of brain neurotransmitters and neuromodula-tors: Biochemical studies. Prog. Neurobiol., 40:223–247.

Sagar, S.M., Sharp, F.R., and Curran, T. (1988) Expression of c-fos inbrain: Metabolic mapping at the cellular level. Science, 240:1328–1331.

Santiago, M., and Westerink, B.H.C. (1991) The regulation of dopa-mine release from nigrostriatal neurons in conscious rats: The roleof somatodendritic autoreceptors. Eur. J. Pharmacol., 204:79–85.

Sayin, U., Timmerman, W., and Westerink, B.H.C. (1995) The signifi-cance of extracellular GABA in the substantia nigra of the ratduring seizures and anticonvulsant treatments. Brain Res., 669:67–72.

Sheng, M., McFadden, G., and Greenberg, M.E. (1990) Membranedepolarization and calcium induce c-fos transcription via phosphor-ylation of transcription factor CREB. Neuron, 4:571–582.

Siegel Jr., B.V., Buchsbaum, M.S., Bunney Jr., W.E., Gottschalk, L.A.,Haier, R.J., Lohr, J.B., Lottenberg, S., Najafi, A., Nuechterlein, K.H.,Potkin, S.G., and Wu, J.C. (1993) Cortical-striatal-thalamic circuitsand brain glucose metabolic activity in 70 unmedicated maleschizophrenic patients. Am. J. Psychiatry, 150:1325–1336.

Swadlow, H.A. (1988) Efferent neurons and suspected interneurons inbinocular vision cortex of the awake rabbit: Receptive fields andbinocular properties. J. Neurophysiol., 59:1162–1187.

Tamas, G., Buhl, E.H., and Somogyi, P. (1997) Fast IPSP’s elicited viamultiple synaptic release sites by different types of GABAergicneurone in the cat visual cortex. J. Physiol., 500:715–738.

Thierry, A.M., Godbout, R., Mantz, J., and Glowinski, J. (1990)Influence of the ascending monoaminergic systems on the activity ofthe rat prefrontal cortex. In: The Prefrontal Cortex: Its Structure,Function and Pathology. Progress in Brain Research, Vol. 85.H.B.M. Uylings, C.G. van Eden, J.P.C. de Bruin, M.A. Corner, andM.G.P. Feenstra, eds. Elsevier, Amsterdam, pp. 357–365.

Timmerman, W., and Westerink, B.H.C. (1997a) Electrical stimulationof the substantia nigra reticulata: detection of neuronal extracellu-lar GABA in the ventromedial thalamus and its regulatory mecha-nism using microdialysis in awake rats. Synapse, 26:62–71.

Timmerman, W., and Westerink, B.H.C. (1997b) Brain microdialysis ofGABA and glutamate: what does it signify? Synapse, 27:242–261.

Ungerstedt, U. (1991) Introduction to intracerebral microdialysis. In:Microdialysis in the Neurosciences. Techniques in the Behavioraland Neural Sciences, Vol. 7. T.E. Robinson and J.B. Justice, Jr., eds.Elsevier, Amsterdam, pp. 3–22.

Uylings, H.B.M, and van Eden, C.G. (1990) Qualitative and quantita-tive comparison of the prefrontal cortex in rat and primates,including humans. In: The Prefrontal Cortex: Its Structure, Func-tion and Pathology. Progress in Brain Research, Vol. 85. H.B.M.Uylings, C.G. van Eden, J.P.C. de Bruin, M.A. Corner, and M.G.P.Feenstra, eds. Elsevier, Amsterdam, pp. 31–62.

van Eden, C.G., Mrzljak, L., Voorn, P., and Uylings, H.B.M. (1989)Prenatal development of GABAergic neurons in the neocortex of therat. J. Comp. Neurol., 289:213–227.

Westerink, B.H.C., and de Vries, J.B. (1989) On the origin of extracel-lular GABA collected by brain microdialysis and assayed by asimplified on-line method. Naunyn-Schmiedeberg’s Arch. Pharma-col., 339:603–607

White, E.L. (1989) Cortical Circuits: Synaptic Organization of theCerebral Cortex: Structure, Function and Theory. Birkhauser, Bos-ton.

Wilson, F.A.W., O Scalaidhe, S.P., and Goldman-Rakic, P.S. (1994)Functional synergism between putative g-aminobutyrate-contain-ing neurons and pyramidal neurons in prefrontal cortex. Proc. Natl.Acad. Sci. U.S.A., 91:4009–4013.


disinhibition of the mediodorsal thalamus induces fos-like immunoreactivity in both pyramidal and...

Documents