A0D-AI07 453 NORWEGIAN DEFENCE RESEARCH ESTABLISHMENT KJELLER F/ B A lbNEURDTHANSMITTER MECHANISMS IN THE NUCLEUS ACCUMRENS SEPT! AND -ETCIU)JUN 81 1 WALAAS

UNCLASSIFIED NORE/PUBL-81/1SS1 NL, 2EEEEEE

JIN 1112.5

1 .8

125 1111 1.6____

MI(CFOPCl¥y RISOLLIIION I CSI CHART

% At

LLffL 0

DNEUROTRANSMITTER MECHANISMS IN THErNUCLEUS ACCUMBENS SEPTI AND RELATEDel REGIONS IN THE RAT BRAIN

By

IVAR WALAAS

NDRE/PUBL-81/1001

ISSN 0085-4301 DTICF!ELECTEftNov 1 6 1981

A

FORSVARETS FORSKNINGSINSTITUTTNORWEGIAN DEFENCE RESEARCH ESTABLISHMENTP 0 Box 25 - N-2007 Kjeiler, Norway

This doaumot hamsi een,,-_j jL605for public release and sale itsU distribution I. unlimited.

NEUROTRANSMITTER MECHANISMS IN THE NUCLEUSACCUMBENS SEPTI AND RELATED REGIONS IN THERAT BRAIN

by

Ivar Walaas

NORE/PUBL-81/1001ISSN 0085-4301

FORSVARETS FORSKNINGSINSTITUTTNORWEGIAN DEFENCE RESEARCH ESTABLISHMENTP 0 Box 25 -N-2007 Kieller. Norway

April 1981

4-

S, lI1, , i' ,° I

NORWEGIAN DEFENCE RESEARCH ESTBLSHEN _______UNLASSIFIE

FORSVARETS FORSKNINGSINSTITUTT (FF1) SECURITY CLASSIFICATION OF THIS PAGE

POST OFFICE BOX 25 wtrda need

N-2007 ICJELLEA, NORWAY

REPORT DOCUMENTATION PAGE

11 PUB,- HEFOCRI NUMBER 21 SECURITY CLASSIFICATION 3) NUMBER OF

NI)RE/PUBL-8 1/1001 UNCLA..SSIFIEDPAEIJUH NEFF fENCF 2j) OECLASSIFiCATION IOWNGRADING SCHEDULE 116

41 TITLE

NEU RO'rRANSML1TTER MIECHIANISMS IN THIE NUCLEUS ACCUMIBENS SEPT) ANDRELATED REGIONS IN THE RATf BRAIN

5) NAMES OF AUINUNIS) IN FULL Isurname titsII

WVALAAS Ivar

6) DISTRIBUTION S1 ATEMENT

Approved for public release. Distribution unlimited(Offcntlig tilgjengelig)

71 INDEXING TENMS

IN ENGLISHI IN NORWEGIAN

a -aminobutx'rate a) yaiouva

b~guant __ ___b) glutamat W__ ____

Lnucleus accumbens_ c) nucleus accumbenis

dneostrialu u)rnsratr

e) hippocampus e) hippocampus

I HE SAL' RUSFF-L NE NCE.

RI ABSTRACT coryvItwe on ,everse side ifncsiy

'The present investigation compares the localization of different transmitter candidates,particularly the amino acids *1aminobutvrate (GAB.\) and glutamnate (GLU), in limbic andbasal ganglia regions in the rat brain. In particular, the characteristics of nucleus accumbenssepti have been studied in some detail. (GABA neurons have been found in nucleusaccumbens, and (;ABA projections from this nucleus have been identified in restricted basalforebrain and mesencephalic regions. GLU p~rojections from the neo- or allocortex have beenfound to terminate in nucleus accumbens and other forebrain and hypothalamic nuclei.Neurotransmitters in local neurons have been identified in the hippocampus, nucleusaccumbens, septum and caudatoputamen by means of local kainic acid injections, whileneurons in thceniediobasal hypothalamus have been studied after systemic treatment ofnewborn animals with monosodium glutamate. The results are discussed as a basis for abetter understanding of limbic-basal ganglia interactions.

9) DATE TIIS PAGE AUTHORIZED BY POSITION[30 June 1981 1Superintendent

UNCLASSI FlEDFFI-8-22. 1980 SECURITY CLASSIFICATION OF THIS PAGE

(whten data entered)

PREFACE

The present investigation has becn carried out at the Norwegian Defence ResearchEstablishment, Division for Toxicology, Kjeller, in the period 1976 - 1979. It repre-sents part of a continuing research program undertaken b' tf-e Ins-titute, aimed at theelucidation of biochemical aspects of brain organization and function at the synapticlevel. Increasing our knowledge in this area is a prerequisite for understanding drugactions, intoxications and specific disorders of the brain, and for devising rationaltreatments for these conditions.

I wish to express my gratitude to I)r philos F Fonnum, Head of the Division forToxicology, for introducing me to the excitement of modern neuroscience, and forhis continuing and inspiring support during the investigation. I also wish to thank theDirector of the Institute, Mr Finn Lied, for providing research facilities. I am furtherindebted to Dr R Lund Karlsen for discussions and suggestions during the initialinvestigations. Mrs K Holte and Mr PJ Karlsen have given me technical help, Mrs EIversen has instructed me in some of the microchemical techniques used, and togetherwith Mrs L Eliassen given me technical assistance during parts of the study. I am alsograteful to my colleagues at the NI)RE for the stimulating milieu they have provided.

I am particularly grateful to my father and my late mother, who by their ownexample showed me the attractions of a life in biomedical science, and to my wifefor her support throughout the period of this work.

New Haven, Connecticut, June 1981

Ivar Walaas

' . k - - " -Ia '

CONTENTS

Page

I OBjECT OF INVESTIGATION 7

2 GENERAL SUMMARY 10

3 DISCUSSION OF METHODS 113.1 Dissection of samples 11

3.2 lcsioning techniques 11

3.3 Biochemical methods 133.3.1 )etermination of enzyme activities 133.3.2 .kcidic amino acids 14

4 GENERAL DISCUSSION 144.1 Characteristics of' the nucleus accumbens 14

4.2 Neurochemical characteristics 15

4.2.1 Introduction 154.2.2 Cortical fibers 154.2.3 Local neurons in GLU-innervated regions 16

4.3 Nucleus accumbens efferents 174.3.1 Projections to the basal forebrain 174.3.2 Projections to the mescncephalon 18

References 20

iw

7

NEUROTRANSMITTER MECHANISMS IN THE NUCLEUS ACCUMBENS SEPTIAND RELATED REGIONS IN THE RAT BRAIN

SUMMARY

The present investigation compares the localization of different transmitter candida-tes, particularly the amino acids '-aminobutyratc (GABA) and glutamate (GLU), inlimbic and basal ganglia regions in the rat brain. In particular, the characteristics ofnucleus accumbens septi have been studied in some detail. GABA neurons have beenfound in nucleus accumbens. and GABA projections from this nucleus have beenidentified in restricted basal forcbrain and mesencephalic regions. GLU projectionsfrom the neo- or allocortex have been found to terminate in nucleus accumbens andother forebrain and hypothalamic nuclei. Neurotransmitters in local neurons havebcen identified in the hippocampus, nucleus accumbens, septum and caudatoputamenby rnians of loctl kainic acid injections, while neurons in the mediobasal hypothala-rouis have been studied after systemic treatment of newborn animals with monoso-dium glutamate. The results are discussed as a basis for a better understanding oflimbic-basal ganglia interactions.

1 OBJECT OF INVESTIGATION

The present study was initiated in order to gain basic information on some aspects ofncurotransmitter mechanisms in the mammalian brain. Analysis of regions with dop-aminergic innervation seemed of particular iml)ortance, due to the suggested role ofsuch regions in movement disorders, endocrinological and mental dysfunction, andantipsychotic drug action (Costa and Gessa 1977, Roberts et al 1978). Furthermore,anatomical and functional reclassifications of the limbic system-basal ganglia interface,particularly the nucleus accumbens septi, have recently been proposed (Heimer andWilson 1975, lleimcr 1978, Nauta ctal 1978, Nauta 1979, Mogenson etal 1980).Ncurochcmical data relevant to these hypotheses have, however, been scarce (Wilson1972, Nauta 1979). Therefore, the major brain regions chosen for study were relatedto the 'limbic system' and/or the basal ganglia, and all had a well-defined dopaminer-gic innervation. l)uring the study, it also became clear that the neurons containing theamino acid transmitter candidates y-aminobutyrate (GABA) and glutamate (GLU) inthese rcgions were largely unidentified. These transmitter systems have therefore re-ceived particular attention.

Initial work analysed the detailed topographical distribution of the monoamine, ace-tylcholine (ACh) and (;AB.\ neurons in the regions by means of enzymatic trans-mitter marker analysis (l'onnum 1975a) on microdissected tissue samples (Papers I,VI, VIII, Xi). The role of the excitatory amino acid GLU was studied by means ofamino acid analysis and high affinity GLU uptake activity measurements. In two suchstudies, the GIl] input to nucleus accumbens was identified and compared to that ofcaudatoptutanen (Papers II, lII), while two studies analysed allocortical GLU projec-tions from the hippocampal formation to the septum, basal forebrain and hypothala-mus (Papers IV, V).

Local transtnittcr mechanisms were studied by means of 'excitotoxic' amino acidlesions. The hypothalamic toxicity of systemically administered GIU to newbornanimals was analysed with morphological, histochcmical and microchemical methods

.. . . . .. . . .. I w i

8

(Paper VI). The toxicity and specificity of local application of the GLU-analoguekainic acid was studied in the hippocampus, a region with well-defined monoamine,ACh and GABA fibers (Paper IV). Later work used this compound in studies onnucleus accumbens, caudatoputamen and septum, where local neurons were identified(Papers II, VII). The cellular and subcellular localization of guanylate cyclase, anenzyme suggested to be involved in ACh and GLU neurotransmission (Greengard1978, Biggio et al 1978) and in kainate neurotoxicity (Honegger and Richelson 1977),was also investigated in the latter nuclei (Paper VII).

Finally, the organization of GABA neurons in the nucleus accumbens was comparedto the caudatoputamen, and the participation of such neurons in the accumbensefferents was investigated. Regions in the basal forebrain and mesencephalon receivingaccumbens fibers were characterized morphologically, histochemically and biochcmi-cally, and compared to regions receiving GABA cfferents from the caudattoputamcn(Papers I, VIII, IX).

The results have been described in nine separate publications:

Fonnum F, Walaas I, Iversen E (1977): Localization of GAB.Xergic. choli-nergic and aminergic structures in the mesolimbic system. J Neurochern29, 221-230.

11 Walaas I, Fonnum F (1979): The effects of surgical and chemical lesionson neurotransmittcr candidates in the nucleus accumbens of the rat..Vcu-roscience 4, 209-216.

III Walaas 1 (1981): Biochemical evidence for overlapping neocortical andallocortical glutamate projections to the nucleus accumbens arid rostral

caudatoputamen in the rat brain. Neuroscience 6, 399-405.

IV Fonnum F, Walaas I (1978): The effect of intrahippocampal kainic acid

injections and surgical lesions on neurotransmitters in hippocampus andseptum. JNeurochem 31, 1173-1181.

V Walaas 1, Fonnum F (1980): Biochemical evidence for glutamate as atransmitter in hippocampal efferents to the basal forebrain and hypothal-amus in the rat brain. Neuroscience 5, 1691-1698.

VI Walaas I, Fonnum F (1978): The effect of parenteral glutamate treatmenton the localization of neurotransmitters in the mediobasal hypothalamus.Brain Res 153, 549-562.

VII Walaas I (1981): '[he effects of kainic acid injections on guanylate cvcla-se activity in the rat caudatoputamen, nucleus accumbens and septum.JNeurochem 36, 233-241.

VIII Walaas I, Fonnum F (1979): The distribution and origin of glutamatedecarboxylase and choline acetyltransferase in ventral pallidum and otherbasal forebrain regions. Brain Res 177, 325-336.

IX Walaas I, Fonnum F (1980): Biochemical evidence for y-aminobutyratecontaining fibers from the nucleus accumbens to the substantia nigra andventral tegmental area in the rat. Neuroscience 5, 63-72.

r __3

9

Abbreviations used

ACh - acetylcholine

ACHE - acetyicholinesterase

,WI) -~ aromatic L-amino acid decarboxylase

ASP -~ L-aspartic acid

ChATr - choline acetyltransferase

GABA -r -aminobutvric acid

GAD) L-gTlutamate decarboxylaseG LU -L.glutamic acid

10

2 GENERAL SUMMARY

The origin, distribution and density of neurons containing different neurotransmittercandidates have been analysed in rat brain regions usually considered to belong toeither the 'limbic system' or to the basal ganglia. Particular emphasis has been placedon the neurons connected with the nucleus accumbens septi, and comparisons havebeen made between the afferent, local and efferent neurons in this region and thosein the septum and caudatoputamen (neostriatum). After surgical interruption of allo-cortical fibers from the hippocampal formation running in the fimbria/fornix bundles,specific decreases were seen in both the activity of the high affinity GLU uptake andin the concentration of endogenous GLU in the lateral septum, the nucleus accum-bens, the rostral diagonal band nucleus, the bed nucleus of the stria terminalis, themediobasal hypothalamus, and the mammillary body. Similar results were also foundin the lateral septum after specific chemical destruction of pyramidal cells in thehippocampal formation by means of local kainic acid injection. Thus, most of thoseallocortical excitatory fibers from the hippocampal formation which run in the fomixbundle appear to use GLU as neurotransmitter.

The origin and distribution of GLU fibers in different parts of the nucleus accumbensand the caudatoputamen were further compared. Fibers from the frontal neocortexwere found in all parts of the caudatoputamen and the nucleus accumbens, whilefibers from the caudal neocortex were found in the dorsal caudatoputamen only.Allocortical fibers disrupted by fornix lesions were seen in the ipsilateral nucleusaccumbens and the ventral caudatoputamen. Thus, the nucleus accumbens and theventral caudatoputamen are innervated by both neocortical and allocortical GLU fi-bers, while the dorsal caudatoputamen is reached by neocortical GLU fibers only.

Monoamine fibers were studied by analysis of aromatic L-amino acid decarboxylase(AAD) activity. High density of such fibers was seen in the caudatoputamen, nucleusaccumbens and olfactory tubercle, with considerably lower density in the septum. Adecrease in AAD activity in all forebrain regions after brain hemitransections confirm-ed that these fibers are ascending. High AAD activity was also found in the mesen-

cephalic ventral tegmental area and the substantia nigra, pars compacta, which consti-tute the origins of the major ascending dopaminergic systems, and in the hypothala-mic median eminence and arcuate nucleus, regions with a local doparnine neuronalsystem.

Microinjection of kainic acid into the hippocampus or the nucleus accumbens led toextensive destruction of local neurons without affecting afferent projections. Systemicadministration of GLU to newborn animals induced similar local neuronal necrosis inthe hypothalamic arcuate nucleus. By these techniques, local cholinergic neurons asmeasured by choline acetyltransferase (ChAT) activity were found to be highly con-centrated in the caudatoputamen, nucleus accumbens and probably also in the olfac-tory tubercle. Lower density of such neurons was found in the septum, and the latterwere insensitive to local kainate injections. In the hypothalamus, ACh neurons, proba-bly originating in the arcuate nucleus, were concentrated in the median eminence.Local GABA neurons, analysed by glutamate decarboxylase (GAD) activity measure-ments, were highly concentrated in the lateral septum, dorsal olfactory tubercle andnucleus accumbens, with lower density found in the caudatoputamen. The medianeminence contained only few GABA terminals, but some of them appeared to belongto a local, tubero-infundibular system.

Kainate injections were also used to analyse the cellular localization of guanylatecyclase, an enzyme suggested to be involved in ACh or GIl transmission. Thisenzyme was highly concentrated in local neurons in the nucleus accumbens and

caudatoputamen, while the intermediate activity in the septum was insensitive tokainate.

Finally, the descending efferent projections from the nucleus accumbens were analy-sed. A major GABA projection was found to terminate in the rostral substantiainnominata and, less importantly, in the ventral globus pallidus. The characteristics ofthe substantia innominata supported the designation of this region as a 'ventral palli-dum'. A significant GABA pathway was also found to terminate in the rostromedialsubstantia nigra. These projections further support the "striatal" nature of the nucleusaccumbens. In contrast, only a small number of GABA terminals were destroyed in arestricted part of the ventral tegmental area after accumbens lesions. Thus, no major'feed-back' GABA pathway was identified in the mesolimbic system.

3 DISCUSSION OF METHODS

3.1 Dissection of samples

The present study has tried to identify neurotransmitter mechanisms by either corre-lating the distribution of neurochemical markers with the distribution of specificneuronal populations, or by following the changes in these markers induced by de-generation of different neurons (Fonnum 1975a, Storm-Mathisen 1977a). In principle,sufficiently sensitive assay methods should allow neurochemical studies on very smallpopulations of neurons. Histochemical, particularly immunocytochemical methods arepresently the most sensitive and specific tools for such studies (Hkfelt etal 1980).However, these methods are not generally available for the study of small-molecularweight compounds, and they also do not give quantitative results. This study hastherefore used chemical analysis of microdissected samples, following anatomical de-scriptions of the neuronal populations. In detailed distribution studies of small regi-ons, microdissection on freeze-dried sections was employed (Lowry 1953). This ap-proach allows analysis of submicrogram samples (eg Paper IX), but precludes studyof labile activities such as uptake mechanism. However, it allows simultaneous inspec-tion of the experimental sample and morphological controls in the dissection micro-scope, and facilitates later dissection of fresh tissue slices or microwave-inactivatedbrain tissue. Furthermore, the present work has consistently studied samples with'normal' anatomical outlines, in contrast to the widely used 'punching' technique(Palkovits 1973). The advantage of this approach was particularly evident in thehypothalamus and mesencephalon (Papers I, VII, IX).

3.2 Lesioning techniques

.lost nerve fiber projections in the brain terminate in a very precise and restrictedpattern The small, restricted lesions used by anatomists to induce anterograde de-generation therefore usually do not lead to measurable chemical changes in the samp-les analysed by neurochemists, except in certain cases, eg the septohippocampal andstriatonigral projections (Fonnum et al 1974, Lewis et al 1967). Surgical interruptionof projections must therefore include most of the fibers under study (Storm-Mathisen1977a), which introduces problems of specificity. These lesions, whether electrolyticor crude transections, will usually encroach upon other regions, destroy fibers 'enpassage' and possibly damage the blood supply to different regions. I have approachedthese problems both by making crude lesions, eg brain hemitransections, and analys-ing chemical changes in restricted regions, and by making more restricted lesions andanalysing both the target regions and adjacent nuclei. Most of the conclusions drawnin this work are based on a combination of these approaches. Morphological controls

L.A

* I12

of the surgical lesions were also carried out, either by staining sections (Paper VIIl,IX), or by inspecting the brains during the sample preparation. Other chemical activi-ties were also analysed in or near the target areas, to exclude unspecific effects.

Surgical lesions are unable to identify the local neurons present in most brain regions,even if retrograde changes are sometimes observed (e g Quik et at 1979). Specificneurotoxins have been described for monoamine neurons, eg 6-hydroxydopamine(Ungerstedt 1971). but other specific toxins have not been described. A major ad-vance in neurobiology was therefore made when the lesioning properties of certainacidic amino acids were discovered. Initial observations (Lucas and Newhouse 1957)demonstrated that GLU when systemically applied to newborn rodents was toxic tocertain neurons in the retina. The GLU effects were further explored in detail byOlney and coworkers (Olney 1974, 1979), who showed that most or all local neuronwere destroyed in those regions where GLU penetrated the blood-brain barrier, i e thehyp-othalamic median eminence and arcuate nucleus. Afferent fibers or glial cells werenlot destroyed. ihe potet (;lA'-analogue kainic acid (Olney et al 1974) was thereforetested in the caudatoputamcn, and similar results were seen, i e a destruction of localneurons, no major effects on afferent fibers, and a reactive gliosis (Covle andSchwarcz 1976, McGceer and McGeer 1976, 1978, Schwarcz and Co\lc 1977). Themechanism of action of this compound remains unlefined (Nadler 1979), the effectseens to be linked to intact excitatory inputs (McGeer et a[ 1978) which howevermay use different transmitters (Nadler 1981). Guanosine 3',5'-monophosphatc (cyclicGMP) has also been suggested as an intermediary in the effect (tloneggcr andRichelson 1977).

We extended the results from caudatoputamen to the hippocampus, chosen becauseof its well-defined transmitter chemistry (Storm-.Mathisen 1977a) which is very diffe-rent from that of the basal ganglia regions (Paper IV). After infusion of high (loses ofkainate into the hippocamnpus we observed similar results, i e no effect on neuro-chemical or histochemical markers for afferent fibers, but substantial decreases inmarkers for local neurons. Similar results were reported simultaneously (Schwarczet at 1978). Morphological controls indicated little or no damage to other brain ri-ons (Paper IV). This was surprising, as later studies have shown that kainatc, especial-

ly when injected into the hippocampus, usually induces major necrotic changes inboth adjacent and remote brain regions (\Vurthclc etat 1978, Zaczek ctal 1980).Thus, both the subiculum, entorhinal cortex and contralateral hippocampus have beenreported to be affected by this treatment, in contrast to our results. Similarly, thepiriform cortex is often destroyed by intrastriatal kainate (Zaczek et at 1980), whilemy injections of kainate into the nucleus accumbens usualiy left this cortical regionuntouched (Paper II). The reasons for these more specific lesions arc not clear, butare probably a result of my use of diazeplam in the anesthetic used for lesioning theanimals. Thus, both benzodiazepincs and long-acting barbiturates have been reportedto prevent the remote effects of kainatc without changing the local toxicity (Zaczeket at 1978, Ben-Ari et at 1979).

In conclusion, the studies of kainate effects in nucleus accumbens, septum and cauda-toputamen (Papers II, VI) are probalvl valid as identifications of local transmittermechanisms. However, the effects on very short-axoned proJe'tions (eg from the bednucleus of the stria terminalis to the nucleus accumbens (Svanson and Cowan 1979)cannot be ascertained from these studies.

13

3.3 Biochemical methods

The identification of neurotransmitters relies on the fulfilment of the criteria ofpresence, release and identity of action (Werman 1966, Storm-Mathisen 1977a). Ofthese, the presynaptic criteria (presence, release) are crucial for the biochemical iden-tification of a transmitter system, while the postsynaptic action is electro-physiologi-cally defined (Werman 1966). Biochemical studies on receptor-binding or second mes-senger generation are not conclusive as these activities may be present outside therelevant synapses (Storm-Mathisen 1977a). The present study has, however, used bothpre- and postsynaptic activities to study the characteristics of the regions. The pre-synaptic markers include enzyme activities, high affinity uptake systems and analysisof endogenous amino acids.

3.3.1 Determination of enzyme activities

ChAT is known to be a specific marker for cholinergic neurons, while AChE is not(Fonnum 1975a). The methods used for these enzymes have been developed in thislaboratory (Fonnum 1969, 1975b), and are known as rapid, simple and extremelysensitive, suitable for assays of submicrogram samples (Fonnum 1970, Storm-Mathisen1970). Triton X-100 has always been included in these assays, to release occludedactivity (Fonnum 1966).

GAD is also a specific marker, in this case for GABA neurons, with only very lowactivity found outside these neurons (Fonnum and Storm-Mathisen 1978). GAD wasanalysed with a CO 2 -trapping micromethod (Fonnum et al 1970). The possibility ofnon-GAD CO 2 production from GLU was controlled by inclusion of Triton X-100 inthe incubation medium (MacDonnell and Greengard 1975).

AAD is concentrated in both dopamine, noradrenaline and serotonin neurons, but isalso present in non-neuronal structures (Fonnum 1975a, Kellogg etal 1973). It is anacceptable marker for monoamines in regions with high activity, e g, substantia nigra,caudatoputamen, while results from regions with low enzyme activity must be evalu-ated more carefully. The analytical technique used was based on liquid cation ex-change (Broch and Fonnum 1972), similar to the ChAT method.

Carnitine acetyltransferase is known as a mitochondrial enzyme (McCaman etal1966), and was analysed both as a subcellular marker (Paper lI) and as a generaltissue marker (Paper VI) with a liquid cation-exchange technique (Sterri and Fonnum1980). 5'-nucleotidase is known as a membrane marker, possibly concentrated in glialcells (Kreutzberg et al 1978), and was assayed by a sensitive radioenzymatic method(Avruch and Wallach 1971).

Guanylate cyclase was analysed by a radioenzymatic method based on that describedby Krishna and Krishnan (1975). Due to the controversy described in Paper VI, thisactivity was characterized in more detail by the use of activators and extractionprocedures, and by identification of the product by chromatography and by treat-ment with phosphodiesterase (Paper VI). Some experiments were also done with[3HI-GTP as substrate and an alumina-anion exchange purification technique (Maoand Guidotti 1974). This method was less sensitive, but gave the same subcellular andtopographical distribution as that found with the method based on [a- 3 2 PI GTP (un-published). The specific activities found in the soluble fractions were higher thanthose reported by other workers (Hofmann et al 1977). The reason for this is un-known, but large variations in brain guanylate cyclase activity have previously beenpresented in separate publications by the same investigator (Bartfai et al 1978,Tjornhammar et al 1979). The enzyme is therefore probably sensitive to a number ofunidentified factors.

14

3.3.2 Acidic amino acids

'he lack of specific, sensitive markers for the excitatory amino acid transmittercandidates was a major obstacle in the investigation of the role these compoundsplayed in the brain (review, )avidson 1976). The demonstration of a sodium-depend-ent, high affinity uptake of GilU and aspartate (ASP) into nerve terminals (Logan andSnyder 1971) was a major step forward. lowever, the lack of discrimination be-tween GLU and ASP, and between other similar compunds (Balcar and Johnston1972), made it clear that additional parameters would have to be studied. A seminalpaper from Snyder and coworkers (Young et at 1974) indicated that analysis of freeamino acids might be combined with uptake analysis in studies on putative GLU orASP fibers. Later studies supported this strongly (Divac et at 1977, McGeer et at1977a, Kim et at 1977, Storin-Mathisen 1977b). This approach was therefore adopted(review, Fonnum et at, 1979, see, however Cotman et at 1981).

Amino acid analysis was initially done (Papers 11, IV, VII) with a double isotopedansyl-chloride technique (Karlscn and Fonnum 1976). The samples were dissectedfrom animals killed by focussed microwave irradiation of the brain, a techniquedeveloped to minimize post-mortem changes in labile compunds (Guidotti et at 1974).Such changes are known to afflict GABA (Alderman and Shellenberger 1974, Balcomet al 1975), but not GLU (Balcom et al 1976). lowever, the mediobasal hypothala-mus and basal forebrain were difficult to dissect after this procedure, and rapidcooling and freezedrying of the samples were adopted for a later study (Paper V), inwhich an automic amino acid analyser was used for quantitation. Results from nu-cleus accumbens and septum were comparable when analysed with either of the twomethods (Papers I1, IV, V).

High affinity uptake mechanisms were studied in sucrose homogenates containingnerve terminals (Fonnum ct at 1979). Experiments showed that the accumulation of[-'Ill I-GLU in this preparation occured into synaptosomes, and was linear withamount of tissue (Papcr ll). Thus, further purification of the nerve terminals (cgMcGeer et at, 1977a, Zaczek et at, 1979) was deemed unnecessary. Lesion-inducedchanges in this uptake have been shown to represent changes i~i the number of uptakesites and not in the affinity for the substrate, in a number of systems (Young ctal1974, Storm-Mathisen 1978, Zaczek et at 1979). The question of homo-exchange oflabelled GLU with the intracellular CIU (Levi and Raiteri 1974) is not important inthis context, as my studies were aimed only at studying the number of uptake sitespresent. lowever, the possibility of different putative GLU nerve terminals havingdifferent affinity of GLU, as has been found for GABA (Storm-Mathisen 1975) andsuggested for GLU in hippocampus (Storm-Mathisen 1978), is important. Thus, theregional distribution of the uptake may not necessarily indicate the true density ofGLU or ASP terminals (see, e g Paper 11I). Also, the GLU uptake into glial cells(Hlenn et al 1974) may have a regional variation (Schousboc and I)ivac 1979). [low-ever, this glial uptake does not appear to interfere significantly in my uptake condi-tions (Fonnum et at 1979).

4 GENERAl. DISCUSSION

4.1 Characteristics of the nucleus accumbens

Most early workers related the nucleus accuinbens to the basal ganglia (.\riensKappers and Theunissen 1907). particularly as a part of the "olfactostriatum"(Ilerrick 1926). The exact relationships of the region remained unknown, however,until inputs from the hippocampal formation and stria terininalis were demonstrated(Fox 1943, Raisman et at 1966, de ()lmos and Ingram 1972). These results linked thenucleus closely with the "limbic system". The cytoarchitectonics of the nucleus were

S,__t ' ,

15 -

described in different species (eg from mouse to elephant) (Koikegami etal 1967),and reported to be similar to those of the caudatoputamen (Wilson 1972). Thenucleus began to attract major interest, however, only after the dense dopaminergicinnervation of the nucleus, originating in the socalled AIO cells in the mesencephalicventral tegmental area was reported (Andern et al 1966, Ungerstedt 1971). The simul-taneously developed "dopamine theory of schizophrenia", ie that psychotic symp-toms in amphetamine-induced psychosis and possibly in schizophrenia occured toget-her with hyperactivity in dopaminergic neurotransmission (Snyder 1973, Hbkfelt etal1974), led to a great interest in the so-called mesolimbic dopamine systems, particu-larly that of the Al0-nucleus accumbens fibers (reviews, Costa and Gessa 1977).Other anatomical characteristics were soon described, particularly the efferent projec-tions to the substantia innominata (Wilson 1972, Williams et al 1977, Nauta et al1978), and to the globus pallidus, the ventral tegmental area and the substantia nigra(Swanson and Cowan 1975, Conrad and Pfaff 1976, Nauta etal 1978). The nucleuswas therefore reclassified (Heimer and Wilson 1975) as being a part of the "ventralstriatum", innervated by the limbic cortex but projecting most densely to basalganglia-related regions (Heimer 1978).

4.2 Neurochemical characteristics

4.2.1 Introduction

When this study was initiated, several neuronal links had been chemically defined inthe basal ganglia (reviews, Brownstein 1979, Fonnum and Walaas 1979). Initial workwas therefore aimed at a comparison between the nucleus accumbens, the olfactorytubercle, the septum and the caudatoputamen. Striking similarities were seen in thedensities of the monoamine, ACh and GABA systems in the nucleus accumbens andolfactory tubercle when compared with the candatoputamen. No evidence for a de-scending GABA projection from these regions to the AlO-cells was found, however, incontrast to the major striatonigral GABA tract (Paper I). Thus, biochemical differen-ces appeared to exist between the region.

4.2.2 Cortical fibers

Preliminary reports indicating that GLU could be transmitter of the excitatory corti-costriatal (1)ivac et al 1977, McGeer et al 1977a, Kim et al 1977) and hippocamposep-tal fibers (Storm-Nlathisen and Opsahl 1978) led to further studies on nucleus accum-bens and septum. Active high affinity GLU uptake was found in both regions, and adetailed study in nucleus accumbens showed that this activity appeared to be locali-zed in certain nerve terminals (Paper 11). Further, lesions which involved the fornixwere able to decrease the activity in the ipsilateral nucleus accumbens. Similarly,fomix transections and destructions of hippocampal neurons with kainic acid decreas-ecl the GLU uptake in the lateral septum, while the medial septum was unaffected(Paper IV). Amino acid analysis after these lesions showed that GLU was the onlyfree amino acid which decreased in concentration after the lesion (Papers II, IV, V).Thus, the reductions in GLU uptake were apparently a result of degeneration ofGLU, as opposed to ASP terminals (Fonnum et al 1979).

These results all conformed to the anatomical studies which were published while thiswork was in progress (Swanson and Cowan 1977, Meibach and Siegel 1977). Thus,the fibers to nucleus accumbens were shown to arise in the subicular part of thehippocampal formation and to travel through the fimbria to the ipsilateral nucleusaccumbcns, while most of the fibers to the lateral septum arose in the hippocampusproper and terminated on both sides of the lateral, but not the medial septum.Further, all these fibers were reported to be excitatory (McLennan and Miller 1974.l)eFrance and Yoshihara 1975).

16

The suggestion that GLU might be the major transmitter in the allocortical fibers tothe nucleus accumbens and septum has later been supported by other workers whichhave employed both similar surgical/chemical lesions (Zaczek et al 1979, Wood et al1979) and total hippocampal extirpations (Nitsch etal 1979). Therefore, the highaffinity GLU uptake was used in a study of the distribution of different corticofugalfibers in the rostral caudatoputamen and nucleus accumbens (Paper III). Surprisingly,also the ventral caudatoputamen had some GLU-accumulating terminals destroyed bya fornixlesion, and only the dorsal caudatoputamen appeared to be connected onl\with the neocortex. Similarly, the nucleus accumbens was found to be innervatedfrom the ipsilateral frontal cortex. Anatomical studies support the latter findings(Beckstead 1979), while the proposed fornix fibers to the ventral caudatoputamen arcunconfirmed (Swanson and Cowan 1977).

The combined amino acid and GLU uptake analysis was also used in a study on otherfibers in the fornix bundle to the basal forebrain and hypothalamus (Paper V). Thepreviously suggested GLUT fibers to the mammillary body (Storm-Mathisen and Opsahl1978, Nitsch etal 1979) were confirmed. In addition, inputs to the anterior diagonalband nucleus, the bed nucleus of the stria terminalis and the mediobasal hypothala-mus, all probably arising in the subiculum (Chronister et al 1976, Swanson andiCowan 1977) were identified as possible GLU projections (Paper V). lowever, thelesions also encroached upon the stria terminalis, and the possible contribution of thisfiber bundle to the observed decrease in G(LU uptake and concentration in the striaterminalis nucleus and mediobasal hypothalamus (de Olmos and Ingrai 1972) re-mains to be defined. It is, however, clear that the GLU system in the hypothalamus isextrinsic, as a hypothalamic deafferentation has been reported to decrease the (;l"concentration in hypothalamus considerably (Racagni et al 1981).

4.2.3 Local neurons in (;LU-innervated regions

The demonstration of a specific neurotoxic action of kainic and glutamic acid, whichwere suggested to destroy only neurons intrinsic to the regions treated with theneurotoxins (Coylc and Schwarcz 1976, Mc(ceer and McGeer 1976), led to a stud' ofthe effects of kainic acid in the hippocampus (Paper IV). The results supported theclaims for specificity and toxicity of kainate actions. This neurotoxin was thereforeused in the nucleus accumbens, septum and caudatoputamen (Paper 11, VI) and theresults compared to previous reports from the caudatoputamen (Schwarcz and Coyle1977, McGeer and McGeer 1978). Most. and probably all cholinergic terminals in thenucleus accumbens and caudatoputamen were found to belong to local, kainate-sensitive neurons, while the moderate amount of such fibers in the septum probablyarrives partly from the diagonal band (Emson 1978) and partly belong to local,kainate-insensitive cells in the medial septum (Malthe-Sorenssen et a! 1980). Thus, thecholinergic system in nucleus accumbens and caudatoputamen was clearly different

from that in the septum.

In contrast, all regions had apparently only local GABA cells. These wvere mostconcentrated in the lateral septum, dorsal oliactory tubercle and nucleus accumbens(Papers I, IV), and most of them disapperared in all regions after local kainate injec-tions (Papers II, VI). Thus, intrinsic GABA neurons are present in both the nucleusaccumbens, septum and caudatoputamen.

A decrease in the GIlU uptake in nucleus accumbens was also observed after localkainate injection (Paper 11). Thus, some intrinsic GLU-cells may exist in this region.However, it appears more likely that the kainate-sensitive (;IAU fibers in the nucleusare derived from neighbouring cortical regions destroyed by kainate (Wurthele et al1978, Zaczek et al 1980).

17

A functional link between cholinergic or GLU neurotransmission and cyclic GMPsynthesis has been proposed from studies in both the peripheral and central nervoussystems (review, Greengard 1978, Ferrendelli 1978). The distribution of guanylatecyclase was therefore also analysed (Paper VI). The enzyme was found in both parti-culate and soluble fractions in both the nucleus accumbens, caudatoputamen andseptum, with particularly intense soluble activity in the nucleus accumbens and cauda-toputamen. The particulate enzyme was evenly distributed in all regions. Contrary tothe results of other workers (Tjornhammar et al, 1979) both enzyme forms werekainate-sensitive in the caudatoputamen and nucleus accumbens, but not in the sep-tum. Thus, this "second messenger" system also emphasizes the similarities betweenthe nucleus accumbens and the caudatoputamen.

The mediobasal hypothalamus has also a well-defined dopamine system, with cellbodies located in the arcuate nucleus and terminals in the median eminence andpituitary (Lindvall and Bj6rklund 1978). The organization of ACh and GABA neuronsis less well known, however (Brownstein et al 1976), As nucleus arcuatus and itsfibers to the median eminence are the major targets of systemically administered GLU(Olnev 1979), the neurochemical results of this lesions were analysed (Paper VII).Initial studies identified a cholinergic system, probably originating in the arcuatenucleus and terminating in the median eminence (Karlsen et al 1977) similarly to theresults of Carson et al (1977). AChE-staining was also seen on arcuate neurons and inthe median eminence, findings which also were confirmed by Carson etal (1978). Inour material, the stain disappeared in the GLU-lesioned animals. This was also observ-ed by others in the arcuate nucleus, but apparently not in the median eminence(Carson et al, 1978).

Similarly, a decrease in GAD was seen in the median eminence after GLU-treatment,indicating either a small tuberoinfundibular GABA system from the arcuate nucleus,or some intrinsic GABA cells in the median eminence. GABA terminals have laterbeen demonstrated in the median eminence by autoradiogaaphy (Tappaz et als, 1980),and the decrease in GAD followinz GLU treatment has been confirmed by some(Tappaz et al, 1981) but not by other workers (Nemeroff et al 1977). The dissectionprocedure seems crucial in this respect (Paper VII). In conclusion, GABA terminalsare probably present in the hypothalamic-pituitary interface, where they may playimportant functional roles (Racagni et al, 1981, Olnev and Price 1981).

In contrast to the ACh and GABA markers, no effects were seen in the markers usedfor the local dopamine cells, i e AAD activity and high affinity uptake of dopamine.These results are not conclusive, however, as AAD is an unspecific marker (section3.3.1), and the uptake studies were done on more crudely dissected samples. Also, arecent studh has questioned the very existence of high affinity dopamine uptake inthe median eminence (Demarest and Moore 1979), even if this activity had been usedwith apparent success as a marker by others in this region (Cuello et al 1973). Thus,the question of dopamine neurons and GLU toxicity in the arcuate nucleus should bereinvestigated with more specific methods. Indeed, other workers have reported adecrease in tyrosine hydroxylase in the arcuate nucleus and median eminence afteradministration of very high GLU doses (Nemeroff et al 1977).

4.3 Nucleus accumbens efferents

4.3.1 Projections to the basal forebrain

The major efferent projection of the nucleus accumbens was first identified by Wilson(1972) to terminate in the rostral substantia innominate. The characteristics and eventhe outline of this region were until then largely unknown, but it was soon character-

a \i

18

ized as a 'ventral pallidum' (Heimer and Wilson 1975, Heimer 1978). Simultaneousstudies by us (Paper I) showed that this region had the highest density of GABAterminals in the rat brain. In a more detailed study (Paper VIII), more than 50% ofthese GABA fibers were found to be destroyed after accumbens destructions. A lessimpressive GABA projection was also found to terminate in the anteroventral globuspallidus. Our results thus confirmed the detailed anatomical mapping of the accum-bens efferents published by Nauta et al (1978), and the physiological studies of Drayand Oakley (1978) and Jones and Mogenson (1980a).The importance of these GABAprojections has not been defined, but they appear to be involved in locomotion(Pycock and Horton 1976, Jones and Mogenson 1980b, Mogenson et al 1980).

Our results thus supported the concept of a 'ventral pallidum' represented by therostral part of the substantia innominata in the rat brain (Swanson 1976). This regioncan apparently be histochemically defined by iron staining with Peri's method (Hilland Switzer 1979). By this method, the region is seen to extend rostrally as fingerlikeextensions, each of which is 'capped' by one of the insulae Callejae (Hill and Switzer1979). Our results appear to support this, as we also found high GAD activity andlow AChE staining in the region rostral to the commissura anterior and situatedbetween the nucleus accumbens and the olfactory tubercle (Papers 1, VIII). Theseresults on GAD were also confirmed and extended by Okada et al (1977).

4.3.2 Projections to the mesencephalon

The caudatoputamen is known to send a major projection to the substantia nigra(Grofova' and Rinvik 1970). These fibers were early demonstrated to be inhibitory(Yoshida and Precht 1971, Precht and Yoshida 1971), probably GABAergic (Kimet al 1971, Fonnum et al 1974), and surmised to represent feed-back fibers respon-sible for reulating the activity of the nigrostriatal dopamine neurons (Anden et al1964, Anden and Stock 1973, Cattabeni etal 1979, Bunney and Aghajanian 1978).The possibility of a similar 'mesolimbic' feed-back system therefore attracted greatinterest as a possible link in the neuronal system which was believed to be involved inpsychotic phenomena (Stevens 1979). We therefore compared the AIO region withthe substantia nigra, and found that the major difference consisted of a low densityof GABA terminals in the AIO region, while the substantia nigra, pars reticulatadisplayed a very dense population of such fibers (Paper 1). Furthermore, no evidencefor a GABA projection from the forebrain to the AIO region could be found afterlesions which both transected the striatonigral GABA fibers in the capsula interna andthe ascending monoamine fibers in the medial forebrain bundle. Similar results werereported by McGeer etal (1977b). Thus, the transmitter of the accumbens-mesencephalic projection (Swanson and Cowan 1975, Conrad and Pfaff 1976) remain-ed unknown.

However, later reports stated that most of the striatonigral GABA fibers probably didnot represent feed-back fibers, and that they rather might constitute an output sys-tem (Arbuthnott 1978, Scheel-Krdiger et al 1977, Di Chiara et al 1979) which probab-ly terminated in the pars reticulata on interneurons and output neurons projecting tothe tectum, thalamus or reticular substance (Domesick 1977. Nauta and Domesick1979, Grace and Bunney 1979, Waszc7.ak etal, 1980). Nauta and coworkers (Nautaetal 1978) demonstrated that this also might be the case for the accumbens projec-tion, which terminated most densely in the substantia nigra, not in the A1O region.Usunoff etal (1979) confirmed this at the electron microscopic level. We thereforereinvestigated the distribution of GAD in the mesencephalon after electrolytic des-tructions of the nucleus accumbens (Paper IX), and found that a considerable numberof GABA terminals had disappeared in the rostromedial substantia nigra after thislesion. About half of the nigral GABA terminals in this restricted region appeared tooriginate in the nucleus accumbens. However, the GABA fibers in the rest of the

19

substantia nigra were unaffected, which explains the previously negative findings ofl)ray and Oakley (1978).

Thus, a descending accumbens GABA projection terminates in a nigral region withfew or no dopamine cells projecting back to the nucleus accumbens (Nauta etal1978). This GABA projection might therefore be a part of the striatonigral GABAoutput pathway (Paper IX). The question of a feed-back projection to the dopamineneurons was still unsolved, however, as both physiological and neurochemical studieshad presented evidence for such a pathway to the A10-cells (Wolf etal 1978, Wad-dington and Cross 1978). Our new results again demonstrated much lower density ofGABA terminals in the AIO region than in the substantia nigra. Furthermore, nochange was seen in GAD in the caudal AlO after accumbens lesions (Paper IX), andOnly a small decrease in GAD was seen in the rostral A10. The demonstration of aGABA-mediated inhibition from nucleus accumbens to some AIO dopamine cellswhich projected back to nucleus accumbens was therefore surprising (Yim and Mogen-son 1980). The AIO region has recently been shown to be more complex thanoriginally believed, with a multitude of different inputs converging on the local cells(Phillipson 1979). My studies of the region, performed with precise microdissectionand extensive morphological controls (Papers I, IX), do not exclude that some ofthese inputs could bc GABAergic, but indicate that such projections must be quanti-tatively vcry small compared to the striatonigral and accumbens nigral "output"GABA pathways.

Other neuronal systems not analysed in the present study are also present in the basalganglia-mesencephalic circuits, and some will be briefly mentioned. Substance P-fibersare present in the striatonigral projection (Brownstein etal 1977, Jessell etal 1978),and these fibers do apparently terminate on both dopamine and non-dopamine cells(Michelot et al 1979). No such fibers have been seen in the accumbens efferents(Kanazawa cl al 1980). Substance P in the A10 region is derived from the habenula,not from the basal ganglia (Cuello etal 1978). The distribution of the cholinergicsyIstem (Paper I) is in general agreement with other workers, i e extremely dense inthc interpeduncular nucleus, and very sparse in the substantia nigra (Kataoka et al1973, Fonnum etal 1974). No change was observed in this transmitter after rostralhemitransections, in contrast to a recent study which suggested that part of ChAT inthe interpeduncular nucleus derives from the diagonal band region (Gottesfeld andJ acobowitz 1978). However, the bilateral termination of the fasciculus retroflexus inthe interpeduncular nucleus (fierkenham and Nauta 1979) may have obscured smallchanges. Also, some of our lesions may have spared the initial, ventral part of thestria medullaris which convey the fibers from the diagonal band (Swanson and Cowan1979). However, the results in Paper I still indicate that most of the ACh fibers in theinterpeduncular nucleus originate in regions caudal to the globus pallidus, eg, in thehabenula (Kataoka et al 1973, Cuello et al 1978).

In conclusion, the present study has demonstrated that the major cortical input tothc limbic forebrain nuclei, among them the nucleus accumbens, appears to be GLUfibers coming through the fornix similar to the innervation of other limbic andhypothalamic nuclei. ilowever. the nucleus accumbens also shares a neocortical GLU)rojection with the caudatoputanen. [he local transmitter systems and neurochemical

characteristics of nucleus accumbens are strikingly similar to those in the caudato-putamen, and the major GABA efferent are also organized similar to, but not identi-cal with, the output from the caudatoputamen. Thus, this study supports the sugges-ted dual limbic-basal ganglia role of the nucleus accumbens.

L a ...... .. . .....

20

References

Alderman J I, Shelleniberger M K (1974): -r-aminobutvric acid tG.-BA) in the ratbrain: re-evaluation of sampling procedures and the post-mortem increase, J Neuro-chern 22, 93 7-40.

Ande'n N E, Stock G (1973): Inhibitory effect of gammahydroxybutyric acid andgammaam inobuityric acid on the dopamine cells in the substantia nigra. Naunvn-Schmiedeberg 's.4rch Pharmacot 2 79, 89-92.

And5n N E, Carlsson AX, lDahlstr6m A, Fuxe K, llillarp N A, Larsson K (1964):D~emonstration and mapping out of nigroneostriatal dopamnine neurons, Life, Sci 3.523-30.

Anden N E, I)ahlstrbm A, Fuxe K, Larsson K, Olsson 1. and Ungerstedt U (1966):Ascending monoamine neurons to the telencephalon and diencephalon, ;Icta PhysiolScand 67, 3 13-26.

Arbuthnott Gi %V (1978): Studies on the interactions between the substantia nigra andthe neostriatumn, A-dv fi cherm Psychopharmacol 19, 235-48.

Ariens Kappers C U, Thecunissen %V F (1907): IDic Phvlogencse dles Rhine ncep)h alons,(des Corpus Striatumn und (ter Vorderhirnkonimissurcn, Foia A'curobiol I, 1 73--288.

Avruchi J,. Wallach 1) F H1 (1971): Preparation and p~ropecrties of plasmia membrane andendoplasmic reticulum fragments from isolated rat fat cells, Bioc/zim Biophvs .1cta233, 334-47.

lacr\*J, Johnston Gi A R (1972): The structural specificity of' the hi fitvUp~take of 1.-glutamnate and L-aspartate by rat brain slices, J Ncuroczem 19, 2657-6.

Balcom G J, Lenox R 11, Meyerhoff ., (1975): Regional -y-aminolbutyric acidl levelsin rat brain (determined after microwave fixation, J Neurochem 24, 609-13

Balcom Gj, Lenox R fl, Meverhoff ~j 1, (1976): Regional lutamate levels in rat brain(determined after microwave fixation, J Ncurvehem 26. 423-5.

Bartfai T, Study R E, Greengard P (1978): Nluscarinic stimulation an(1 cGMP synthesisin the nervous system. In: Cholincr~ric Mechanisms and Psy chop harm acolo0g), (IEd1) Jjendlen), Plenum Press, New York, 285 -95.

Beckstead R NI (1979): An autoradiographic examination of corticocortical andl sub-cortical projections of the mediodlorsal-projection (prefrontal) cortex in the rat,j comp Neurol 184, 43-62.

Ben-Ani V, Tremblay E, Ottersen O P. Naqluet R (1979): Evidence suggesting secon-dary epileptogenic lesions after kainic acid: pretreatment with diaizepamn reduces dis-tant but not local damage, Brain Res 165. 362 -5.

Biggio G, Corda MI G, Casu M, Salis M, Gessa G 1, (1978): Disappearance of cerebellarcyclic (iMP induced lby kainic acid, Brain Res 154, 203 8.

lBroch 0jj.Fnu F (1972): The regional and subcellular (list ribhLt ion of cate-chol-O-methyltransferasc in the rat brain, J .Vcurochem 19, 2049 55a.

21

Brownstein M J (1979): Biochemical anatomy of the extrapyramidal system. In:Dopaminergic Ergot Derivatives and Motor Function (Eds K Fuxe, D B Calne), Perga-mon Press, New York, 3343.

Brownstein MJ, Palkovits N, Tappaz M L, Saavedra J M, Kizer J S (1976): Effect ofsurgical isolation of the hypothalamus on its neurotransmitter content, Brain Res 117,287-95.

Brownstein MJ, Mroz E A, Tappaz M L, Leeman S E (1977): On the origin of sub-stance P and glutamic acid decarboxylase (GAD) in the substantia nigra, Brain Res135, 315-23.

Bunney B S, Aghajanian G K (1978): D-amphetamine-induced depression of centraldopamine neurons: evidence for mediation by both autoreceptors and a striatonigralfeedback pathway, Naunyn-Schmiedeberg's Arch Pharmacol 304, 255-61.

Carson K A, Nemeroff C B, Rone M S, Young-Blood W W, Prange A J jr, Hanker J S,Kizer J S (1977): Biochemical and histochemical evidence for the existence of atuberoinfundibular cholinergic pathway in the rat, Brain Res 129, 169-73.

Carson K A, Nemeroff C B, Rone M S, Nicholson G F, Kizer J S, Hanker J S (1978):Experimental studies on the ultrastructural localization of acetylcholinesterase in themediobasal hypothalamus of the rat, J comp Neurol 182, 201-20.

Cattabeni F, Bugatti A, Groppetti A, Maggi A, Parenti M, Racagni G (1979): GABAand dopamine: their mutual regulation in the nigrostriatal system. In: GABA-Neuro-transmitters (Eds P Krogsgaard-Larsen, J Scheel-Kriiger, H Kofod), Munksgaard,Copenhagen, 107-17.

Chronister R B, Sikes R W, White L E jr (1976): The Septo-Hippocampal System:Significance of the subiculum. In: rhe Septal Nucleis, (Ed J F DeFrance), PlenumPress, New York, 115-32.

Conrad L C A, Pfaff D W (1976): Autoradiographic tracing of nucleus accumbensefferents in the rat, Brain Res 113, 589-96.

Costa E, Gessa G L (1977): Nonstriatal Dopaminergic Neurons, Adv Biochem Psycho-pharmacol 16, Raven Press, New York.

Cotman C W, Foster A, Lanthorn T (1981): An overview of glutamate as a neuro-transmitter, Adv Biochem Psychopharmacol 27, 1-27.

Coyle J T, Schwarcz R (1976): Lesion of striatal neurons with kainic acid provides amodel of Huntington's chorea, Nature (Lond) 263, 244-6.

Cuello A C, Horn A S, MacKay A V P, Iversen L L (1973): Catecholamines in themedian eminence: new evidence for a major noradrenergic input, Nature (Lond) 243,465-7.

Cuello A C, Emson P C, Paxinos G, Jessell T (1978): Substance P containing andcholinergic projections from the habenula, Brain Res 149, 413-29.

Davison N (1976): Neurotransmitter Amino Acids, Academic Press, London.

De Olmos J S, Ingram W R (1972): The projection field of the stria terminalis in therat brain - An experimental study, J comp Neurol 146, 303-34.

22

DeFrance J F, Yoshihara H (1975): Fimbria input to the nucleus accumbcns septi,Brain Res 90, 159-63.

Demarest K T, Moore K E (1979): Lack of a high affinity transport system for dopa-mine in the median eminence and posterior pituitary, Brain Res 171, 545-51.

Di Chiara G, Porceddu MI L, M~orelli M, Niulas MI L, Gessa G 1, (1979): Strio-nigral andni,,ro-thalamic GABA-ergic neurons as output pathways for striatal resp~onses. In:GA BA -Neu trotransmitters (Eds P Krogsgaard- Larsen, .1Schell-Krdger, If Kofod),Munksgaard, Copenhagen, 465-81.

Divac 1, Fonnum F, Storm-Mathisen J (1977): Hligh affinity up~take of glutamatc interminals of corticostriatal axons, Nature (Lond) 266, 377-8.

Doinesick V B (1977): The topographical organization of the striatonigral connectionin the rat, Anat Rec 187, 567.

Dray A, Oakley N R (19 78): Projections from nucleus accumbens to globu1) alliduSand substantia. nigra in the rat, Experientia (Basel) 34, 68-70.

Emson P C (1978): Complementary distribution of dopamine, substance P, and acet\vl-choline in the rat prefrontal cortex and septum, Adi' Biochcm Psychophairiacol 19,39 7-400.

Ferrendelli 1 ^(197D': D~istribution and regulation of cyclic GMIP in the central nier-vous system, ldv Cyclic Nucleotide Res 9, 453-64.

Fonnur, V -y: A radiochemical method for the estimation of choline .tcet~trans-ferase, i,.:ochern / 100, 479-84

Fonnum F- NO6): Radiochemnical microassays for the determination of' choline ace-tyltrai1sferase and acetyicholinesterase activities, Bioc/zern]j 115, 465-72.

Fonnum F (1970): Topogrraphical and subcellular localization of choline acctyltrans-ferase in rat hippocampal region, j Neurochemn 17, 1029--37.

Fonnum F (1 975a): The localization of glutamate decarboxylase, choline acetyltrans-ferase and aromatic amino acid dccarboxylase in mammalian and invertebrate nervoustissue. In: Metabolic Compartrnentation and Neurotransmnission (Eds S lierl, 1) 1)Clarke, D) Schneider), Plenum Press, 99-122.

Fonnum F (1975b): A rapid radiochemnical method for the determination of'cholinleacetyltransferase, J Neurochem 24, 40 7-9.

Fonnum F, Storm-Mathisen j (1978): Localization of GABAergic neurons in theCNS - Handbook of Psychopharmacol 9, Plenum Press, New York, 357-401.

Fonnum F, WValaas 1 (1979): Localization of neurotransmitter candidates in neostria-turn. In: The Neostriatum (Eds I Divac, G E Obergr), Pecrgyamon Press, 53 -69.

Foninum F, Storm-Mathisen J,' Walberg F (1970): Glutatmate decarboxvlase inl inhibi-toryv neurons -- A\ study- of the enzyme in Purkinje cell axons and boutons in the cat,Brain Res 20, 259 - 75.

1-onnum F, Grofova 1, Rinvik E, Storm-Ma~thisen j1, Walbcrg F (1974): Origin anddistribution of glutamate decarhoxvlase in substantia nigra of the cat, Brain Res 7177 92.

23

Fonnum F, Karisen R L, Malthe-Sorenssen D, Skrede K K, Walaas 1 (1979): Localiza-tion of neurotransmitters, particularly glutamate, in hippocampus, septum, nucleusaccumnbens and superior colliculus, Progr Brain Res 51, 167-91.

Fox C A (1943): The stria terminalis, longitudinal association bundle and precommis-sural fornix fibers in the cat, I comp Neurol 79, 277-95.

Gottesfelci Z, Jacobowitz D (1978): Cholinergic projection of the diagonal band tothe interpeduncular nucleus of the rat brain, Brain Res 156, 329-32.

Grace AXA, BunneN B1 S (1979): Paradoxical GABA excitation of nigral dopamninergiccells: indirect mediation through reticulata inhibitory neurons, Europ IJPharnacol 59,211-8.

Greengard P (1978): Cvclic Nucleot ides, Phosphorylated Proteins, and Neuronal Func-tion. Raven Press. New York.

Grofova 1, Rinvik E (1970): An experimental electronmicroscopic study on the stria-ton iral projection in the cat. Exp Brain Res 11, 249-62.

Guidotti A, Cheney 1) L, Trabucchi M, Doteuchi M, Wang C, Hawkins R A (1974):Focussed microwave radiation: a technique to minimize postmortem changes in cyclicnucleotides, dopa and choline and to preserve brain morphology, Neuropharmacology13, l11522

Ilcimer 1, (1978): The olfactory cortex and the ventral striatumn. In: Limbic Mfecha-nisins (Eds K E Livingston, 0 Hornykiewicz), Plenum Press, 95-187.

1Ileimer I., Wilson R 1) ( 19 75): The subcortical projections of the allocortex: similari-ties in the neural associations of the hippocampus, the piriform cortex, and theneocorte x. In: (Jo 1qi Centennial Symposium: Perspectives in Neurobiology (EdMI Santini), Raven Press, 177-93.

llcnin F A, Goldstein M, lamberger A (1974): Uptake of the neurotransmitter candi-dlate glutamnate by glia, N'ature (Lond) 249, 663-4.

Ilerke~nhain M, Nauta WJ 11 (1979): Efferent connections of the habenular nuclei inthe rat, J (omp .Vcurol 187, 19--48.

I lerrick CJ (1( )26): 11w Brain of Rats anti Men, University of Chicalgo Press.

I lill, j1 M, Switzer R C (1979): The relationship b~etween the islands of Calleja and tilestriatopalli(dil cortex, Neurosci .1bstr 5 72.

llofniann M, Spano P F, I'rabucchi M, Kumakura K (1977): Guanvlate cyclase activi-ty in various rat lbrain areas,]J Neurochem 29, 395-6.

Ilokfelt T, [Jungdahl A, Fuxe K, Johansson 0 (1974): IDopamine nerve terminals inthe rat limbic cortex: aspects or the dopamine hypothesis of schizophrenia, Science184 177 9.

1l6kfelt TI, Johansson 0, Ljungdahl A, Lundberg, J M, Schultzberg NI (1980): Pep-tidergic neurons, Nature (Lond) 284, 51. -21.

lone-,'er P, Richelson E (1977): Kainic acid alters neurochemical development infe-tal rat brain aggregating cell cultures, Brain Res 138, 580-4.

24

Jessell T M, Emson PC, Paxinos G, Cuello A C (1978): Topographic projections ofsubstance P and GABA pathways in the striato- and pallidonigral system: a biochemi-cal and immunohistochemical study, Brain Res 152, 487-98.

Jones D L, Mogenson GJ (1980a): Nucleus accumbens to globus pallidus GABAprojection: electrophysiological and iontophoretic investigations, Brain Res 188,93-105.

Jones, D L, Mogenson GJ (1980b): Nucleus accumbens to globus pallidus GABAprojection subserving ambulatory activity, Am J Physiol 238, R63-9.

Kanazawa I, Mogaki S, Muramoto 0, Kuzuhara S (1980): On the origin of substanceP-containing fibers in the entopeduncular nucleus and the substantia nigra of the rat,Brain Res 184, 481-5.

Karlsen R L, Fonnum F (1976): The toxic effect of sodium glutamate on rat retina:changes in putative transmitters and their corresponding enzymes, JNeurochem 27,1437-41.

Karlsen R L, Fonnum F, Walaas 1 (1977): The toxic effect of L-glutamate on neuro-transmitters in rat retina and hypothalamic nuclei, Proc Int Soc Neurochem 6, 200.

Kataoka K, Nakamura Y, Hassler R (1973): Habenulo-interpeduncular tract: a possiblecholinergic neuron in the rat brain, Brain Res 62, 264-7.

Kellogg C, Lundborg P, Ramstedt L (1973): Analysis of capillary and parenchymalaromatic L-amino acid decarboxylase activity in regional brain areas during ontogene-tic development in the rat, Brain Res 50, 369-78,

Kim J-S, Bak I J, Hassler R, Okada Y (1971): Role of - -aminobutyric acid (GABA)

in the extrapyramidal motor system - 2. Some evidence for the existence of a typeof GABA-rich strionigral neurons, Exp Brain Res 14, 95-104.

Kim J-S, Hassler R, Haug P, Paik K-S (1977): Effect of frontal cortex ablation onstriatal glutamic acid level in rat, Brain Res 132, 370-4.

Koikegami H, Hirata Y, Oguma J (1967): Studies on the paralimbic brain structures -I Definition and delimitation of the paralimbic brain structures and some experimentson the nucleus accumbens, Folia Psychiat Neurol Jap 21, 151-80.

Kreutzberg G, Barron K D, Schubert P (1978): Cytochemical localization of 5'-nucleo-tidase in glial plasma membranes, Brain Res 158, 247-57.

Krishna G, Krishnan N (1975): A rapid method for the assay of guanylate cyclase,J Cyclic Nucl Res 1, 293-302.

Levi G, Raiteri M (1974): Exchange of neurotransmitter amino acids at nerve endingscan simulate high affinity uptake, Nature (Lond) 250, 735-7.

Lewis P R, Shute C C D, Silver A (1967): Confirmation from choline acetylase ana-lyses of a massive cholinergic innervation to the rat hippocampus, J Physiol (Lond)191, 215-24.

Lindvall 0, Bjorklund A (1978): Anatomy of the dopaminergic neuron system in therat brain, Adv Biochem Psychopharmacol 19, 1-23.

25

Logan WJ, Snyder S H (1971): Unique high affinity uptake systems for glycine,glutamic acid and aspartic acid in central nervous tissue of the rat, Nature (Lond)234, 297-9.

Lowry 0 H (1953): The quantitative histochemistry of the brain, IHistochem Cyto-chem 1, 420-8.

Lucas D R, Newhouse J P (1957): The toxic effect of sodium-L-glutamate on theinner layers of the retina, Arch Ophtalmol 58, 193-201.

MacDonnell P, Greengard 0 (1975): The distribution of glutamate decarboxylase inrat tissues, isotopic vs fluroimetric assays, J Neurochem 24, 615-8.

Malthe-Sorenssen D, Odden E, Walaas I (1980): Selective destruction by kainic acid ofneurons innervated by putative glutamergic afferents in septum and nucleus of thediagonal band, Brain Res 182, 461-5.

Mao CC, Guidotti A (1974): Simultaneous isolation of adenosine 3',5'-cyclic mono-phosphate (cAMP) and guanosine 3',5'-cyclic monophosphate (cGMP) in small tissuesamples, Anal Biochem 59, 63-8.

McCaman R E, McCaman M W, Stafford M L (1966): Carnitine acetyltransferase innervous tissue, J biol Chem 241, 930-4.

McGeer E G, McGeer P L (1976): Duplication of biochemical changes of Huntington'schorea by intrastriatal injections of glutamic and kainic acid, Nature (Lond) 263,517-9.

McGeer E G, McGeer P L (1978): Some factors influencing the neurotoxicity of intra-striatal injections of kainic acid, Neurochem Res 3, 501-17.

McGeer P L, McGeer E G, Scherer U, Singh K (1977a): A glutarnatergic corticostriatalpath? Brain Res 128, 369-73.

McGeer P L, McGeer E G, Hattori T (1977b): Dopamine-acetylcholine-GABA neuronallinkages in the extrapyramidal and limbic systems, Adv Biochem Psychopharmacol 16,397-402.

McGeer E G, McGeer P L, Singh K A (1978): Kainate-induced degeneration of neo-striatal neurons: dependency on the corticostriatal tract, Brain Res 139, 381-3.

McLennan H, Miller JJ (1974): The hippocampal control of neuronal discharges inthe septum of the rat, J Physiol (Lond) 237, 607-24.

Meibach R C, Siegel A (1977): Efferent connections of the hippocampal formation inthe rat, Brain Res 124, 197-224.

Michelot R, Leviel V, Giorguieff-Chesselet, M F, Cheramy A, Glowinski J (1979):Effects of the unilateral nigral modulation of substance P transmission on the activityof the two nigro-striatal dopaminergic pathways, Life Sci 24, 715-24.

Mogenson G J, Jones D L, Yim C H (1980): From motivation to action: functionalinterface between the limbic system and the motor system, Progr Neurobiol 14,69-97.

Nadler J V (1979): Kainic acid: neurophysiological and neurotoxic actions, Life Sci24, 289-300.

____ ____ ___ ____ -~- ---..--- -

26

Nadler J V (1981): Role of excitatory pathways in the hippocampal damage producedby kainic acid, Adv Biochem Psychopharmacol 27, 395-402.

Nauta HJ W (1979): A proposed conceptual reorganization of the basal ganglia andtelencephalon, Neuroscience 4, 1875-81.

Nauta WJ 11, Domesick V B (1979): The anatomy of the extrapyramidal system. In:Dopaminergic Ergot Derivatives and Motor Functions (Eds K Fuxe, 1) B Calne), Perga-mon Press, 3-22.

Nauta WJ H, Smith G P, Faull R LNI, Domesick V B (1978): Efferent connectionsand nigral afferents of the nucleus accumbens septi in the rat, Neuroscience 3,385-401.

Nemeroff C B, Konkol R J, Bissette G, Young-Blood W W, Martin J B, Brazell P,Rone MS, Prange jr AJ, Breese G R, Kizer J S (1977): Analysis of the disruption inhypothalamic-pituitary regulation in rats treated neonatallv with NISG - Evidence forthe involvement of tuberoinfundibular cholinergic and dopaminergic systems in neuro-endocrine regulation, Endocrinology 101, 613 -22.

Nitsch C, Kim J-K, Shimada C, Okada Y (1979): Effect of hippocampus extirpationin the rat on glutamate levels in target structures of hippocampal efferents, NeurosciLetts 11, 295-9.

Okada Y, Hosoya Y, Kurosawa F (1977): High concentration of -aminobutvric acid(GABA) and glutamate decarboxylase (GAD) activity in the regions of medial fore-brain bundle, diagonal band of Broca and the third layer of the olfactory tubercle,Proc Japan Acad 53, 235-40.

Olney J W (1974): Toxic effects of glutamate and related amino acids on the develop-ing central nervous system. In: Heritable Disorders of Amino Acid Metabolism (EdW L Nyhan), John Wiley & Sons, 501-12.

Olney J W (1979): Excitotoxic amino acids: research application and safety impli-cations. In: Glutamic Acid: Advances in biochemistry and physiology (Eds 1.. Filerjr, S Garrattini, M R Kare, W A Reynold, RJ Wurtman), Raven Press, 287- 320.

Olney, J W, Price M T (1981): Neuroendocrine effects of excitotoxic amin(o acids, .dvBiochem Psychopharmacol 27, 423-32.

Olney J W, Rhee V, Ho 0 1. (1974): Kainic acid: a powerful neurotoxic analogue ofglutamate, Brain Res 77, 507-12.

Palkovits XI (1973): Isolated removal of hypothalamic or other brain nuclei of the rat,Brain Res 59, 449-50.

Phillipson O T (1979): Afferent projections to the ventral tcgmental area of Tsai andinterfascicular nucleus: a horseradish peroxidase study in the rat, J comp Neurol187, 117-44.

Precht W, Yoshida M (1971): Blockade of caudate-evoked inhibition of neurons in thesubstantia nigra by picrotoxin, Brain Res 32, 229 -33.

Pycock C, Horton R (1976): Evidence for an accumbens-pallidal pathway in the ratand its possible GABA-minergic control, Brain Res 110, 629 34.

27

Quick M, Emson P C, Joyce E (1979): Dissociation between the presynaptic dopa-mine-senisitive adenylate cyclase and ' 1--spiperone binding sites in rat substantia nigra,Brain Res 167, 355-65.

Racagni G, Apud J A, Civati C, Cocchi D, Casanueva F. Locatelli V, Nistico E, WillerE (1981):.Neurochcmical aspects of GABA and glutamnate in the hypothalamo-p~ituitary system, Adv Biochem Psychopharmacol 26, 261-71.

Raisman G, Cowan W NI, Powell I P s (1966): An experimental analysis of the effe-rent projections of the hippocampus, Brain 89, 83-108.

Roberts P J, Woodruff G N, Iversen 1, L, (1978): Dopamine, Adv Biochem Psycho-pharmacol 19. Raven Press, New York.

Scheel-Kriiger J, Arnt J, Magelund G (1977): Behavioural stimulation induced bymuscimol and other GABA agonists injected into the substantia nigra, Neurosci Letts4, 351-5.

Schousboc A, Oivac 1 (1979): Differences in glutamate uptake in astrocytes culturedfrom different brain regyions, Brain Re's 177, 407-9.

Schwaxcz R, Coyle J TF (1977): Striatal lesions with kainic acid: neurochemical charac-teristics. Brain Res 127, 235-49.

Schwarcz R, Zaczck R, Coyle J T (1978): Microinjection of kainic acid into the rathippocamnpus, Eur J Pharmacol 50, 209-20.

Snyder S 11 (1973): AXmphetamine psychosis: a "model" schizophrenia mediated bycatecholamines, A-m JPsvchiat 130, 61-7.

Sterni S, Foninum F (1980): .Xcetvl-CoA svitthesizing enzymes in cholinergic nerveterminals, J Neurochem 35, 249-54.

Stevens J R (19 79): Schizophrenia and dopamine regulation in the mesolimbic system,rrcnds Ncurosci 2, '102-5.

Storm-Math isen J (1970): Quantitative histochemnistry of acetylcholinesterase in therat hippocampal region correlated to histochemical staining, J .eurochem 17,739 -50.

Storm-Mathisenl J (1975): High affinity uptake of GABA in presumed GABAergicnerve endlings in rat brain, Brain Res 84, 409 -27.

Storm-Mathisen J(1 977a): Localization of transmitter candidates in the brain: thlehippocampal formation as a model. Progr Neurobiol 8, 119-81.

Storm-Niathisen J (197 7b): Glutamic acid an(' excitatory nerve endings: reduction ofglutamic acid uptake after axotomy, Brain Rc's 120, 379-86.

Storm-Nlathisen J (1978): Localization of putativ'e transmitters in the hippocampalformation. In: Functions of thc Septo-flippocam pal System, Ciba Foundation Sym-posium 58 (nlew series), Elsevier/North-Iliolland, 49- 79.

Storm-Niathisen J, Opsahl MI W (1978): Aspartate and/or glutamate may be transmit-ters in hippocampal cfferents to septum and hypothalamus, Neurosci Letts 9s,.65-70.

28

Swanson L W (1976): An autoradiographic study of the efferent connections of theprecoptic regions in the rat, J comp Neurol 167, 227-56.

Swanson L W, Cowan W M (1975): A note on the connections and development ofthe nucleus accumbens, Brain Res 92, 324-30.

Swanson L W, Cowan W M (1977): An autoradiographic study of the organization ofthe efferent connections of the hippocampal formation in the rat, J comp Neurol172, 49-84.

Swanson LW, Cowan W M (1979): The connections of the septal region in the rat,J comp Neurol 186, 621-56.

Tappaz M, Aguera M, Belin M F, Pujol J F (1980): Autoradiography of GABA in therat hypothalamic median eminence, Brain Res 186, 379-91.

Tappas M L, Aguera M, Belin M F, Oertel W t1, Schmechel D E, Kopin I J, Pujol J F(1981): GABA markers in the hypothalamic median eminence, Adv Biochem Psycho-pharmacol 26, 229-36.

Tj6rnhammar M-L, Schwarcz R, Bartfai T, Fuxe K (1979): Guanylate cyclase activityincreases after kainic acid lesion of rat striatum, Brain Res 171, 567-72.

Ungerstedt U (1971): Stereotaxic mapping of the monoamine pathways in the ratbrain, Acta physiol scand Suppl 367, 1-48.

Usunoff K G, Romansky K V, Ivanov D P, Blagov Z A (1979): Electron microscopicstudy of the projection of fundus striati (nucleus accumbens septi) to substantia nigraof the cat, with a note on the topographic distribution of striatonigral fibres, ComptRend Acad buig Sci 32, 1125-8.

t Waddington J L, Cross AJ (1978): Neurochemical changes following kainic acidlesions of the nucleus accumbens; implications for a GABAergic accumbal-ventraltegmental pathway, Life Sci 22, 1011-4.

Waszczak B L, Eng N, Walters J R (1980): Effects of muscimol and picrotoxin onsingle unit activity of substantia nigra neurons, Brain Res 188, 185-97

Werman R (1966): A review - criteria for identification of a centi-: nervous systemtransmitter, Comp Biochem Physiol 18, 745-66.

Wilson R D (1972): The neural associations of nucleus accumbens septi in the albinorat, Thesis, Massachusetts Institute of Technology.

Williams DJ, Crossman A R, Slater P (1977): The efferent projections of the nucleusaccumbens in the rat, Brain Res 130, 217-27.

Wolf P, Olpe H-R, Avrith D, Haas H L (1978): GABAergic inhibition of neurons inthe ventral tegmental area, Experientia (Basel) 34, 73-4.

Wood P L, Peralta E, Cheney D L, Costa E (1979): The turnover rate of ACh in thehippocampus after lesion of hippocampal pyramidal cells with kainic acid, Neuro-pharmacology 18, 519-23.

Wurthele S M, Lovell K L, Jones M Z, Moore K E (1978): A histological study ofkainic acid-induced lesions in the rat brain, Brain Res 149, 489-97.

w

29

Yim C Y, Mogenson G J (1980: Effect of picrotoxin and nipecotic acid on inhibitoryresponse of doparninergic neurons in the ventral tegmental area to stimulation of thenucleus accumbens, Brain Res 199, 466-72.

Yoshida M, Precht W (1971): Monosynaptic inhibition of neurons of the substantianigra by caudatonigral fibers, Brain Res 32, 225-8.

Young A, Oster-Granite M L, lerndon R M, Snyder S H (1974): Glutamic acid: selec-tive depletion by viral induced granule cell loss in hamster cerebellum, Brain Res 73,1-13.

Zaczek R, Nelson M F, Coyle J T (1978): Effects of anaesthetics and anticonvulsantson the action of kainic acid in the rat hippocampus, Eur J Pharmacol 52, 323-7.

Zaczek R, Hedreen J C, Coyle J T (1979): Evidence for a hippocampal-septal gluta-matergic pathway in the rat, Exp Neurol 65, 145-56.

Zaczek R, Simonton S, Coyle J 1 (1980): Local and distant neuronal degenerationfollowing intrastriatal injection of kainic acid, J Neuropathol Exp Neurol 39, 245-64.

.. . ... .. . . .I. r i I-- . . .

31 MlR I

J,. .~..,I '.i r . I977 Vol 29. pp 221-230 Perpmoe Pien Prntd in Gres lnumn,

LOCALIZATION OF GABAERGIC, CHOLINERGICAND AMINERGIC STRUCTURES IN THE

MESOLIMBIC SYSTEM

F. FONNUM. I. WALAAS and E. IvuasENNorwegian Defence Research Establishment, Division for Toxicology. N-2007 Kjefler. Norway

(Received 5 January 1977. Revised 7 March 1977. Accepted 21 March 1977)

Abstract-The localization of cholinergic. GABAergic and aminergic structures in the "mesolimbic'system has been discussed from studies on the topographical distribution of choline acetyltransferase.glutamate decarboxylase and aromatic amino acid decarboxylase in normal rat brain and in brainshemitransected at the level of globus pallidus. The structures analysed included nucleus accumbens.olfactory tubercle, septum, medial forebrain bundle. striatum. substantia nigra ventral tegmental areaand nucleus interpeduncularis.

Choline acetyltranferase was highly concentrated in the nucleus interpeduncularis. but it did alsoexhibit considerable activity in the nucleus accumbens, the olfactory tubercle and the striatum. Theactivities did not change after hemitransection. Aromatic amino acid decarboxylase was highly concen-trated in the ventral tegmental area, but high activities were also found in the striatum. the nucleusaccumbens. the olfactory tubercle and the pars compacta of the substantia nigra. The activity decreasedin all areas rostra) to the hemitransection. Glutamate decarboxylase was highly concentrated in thedopamine innervated regions, moreso in the limbic structures than in the striatum. Much higher activitywas found in the substantia nigra than in thc ventral tegmental area. After hemitransection the activityin the substantia nigra was decreased whereas in the ventral tegmental area it was unchanged. Ourresults thus suggest that dopaminergic cells in the ventral tegmental area do not receive GABAergicfibres from the terminal regions of the ascending dopaminergic fibres. In addition, we found a veryhigh concentration of glutamate decarboxylase in a region traversed by the rostral medial forebrainbundle. Here the activity was mainly confined to the particulate fraction, probably the synaptosomes.This fraction also displayed a very active high affinity uptake of y-aminobutyric acid.

t IT HAS been shown that two of the major dopaminer- inhibitory GABAergic fibres which terminate in thegic fibre systems in mammalian brain, the nigrostria- substantia nigra arise in the nucleus caudatustal and the 'mesolimbic', display several strikingly putamen and/or globus pallidus (FoNNum et al., 1974;similar features. They both originate in the mesence- HATTORI et al., 1973; KnI et al., 1971; PRECHT &phalon, where the fibres to striatum arise from the YOSHIDA, 1971), while the nucleus accumbens projectspars compacta of the substantia nigra (the socalled to the ventral tegmental area and the substantia nigraA9 cell group), and the fibres to the nucleus (CoNRAD & Pl.A&. 1976; SWANSON & COWAN, 1975).accumbens, the olfactory tubercle and the septum The transmitter in this pathway, however, remainsarise from the ventral tegmental area of Tsai, which unknown.corresponds to the AI0 cell group (ANDtN et al., 1964. The aim of the present investigation was to com-DAHLST1t6M & Fuxz, 1964; UNGERSTEDT. 1971; LND- pare the localization of neurotransmitters in the sub-VALL & Bj6RKLUND. 1974; SIMON et al., 1976). The stantia nigra, in the neostriatum and in some 'meso-neostriatum and the above mentioned limbic regions limbic' structures, e.g. the nucleus accumbens. thehave similar composition of transmitter candidates. olfactory tubercle, the septum, the nucleus inerpe-They contain a high level of acetylcholinesterase duncularis, and the ventral tegmental area. The topo-(AChE, EC 3.1.1.7) (JACoNOWITZ & PALKOVTrS, 1974), graphical distribution of glutamate decarboxylasecholine acetyltransferase (ChAT, EC 2.3.1.6) (PALKO- (GAD, EC'4.1.1.15) and ChAT in these regons wasvtrs et al., 1974; KATAOiKA et al., 1975), and further- therefore studied in microdiuected freeze-driedmore the dopamine receptors in striatum and nucleus samples from normal brains, and from unoperatedaccumbens behave similarly with respect to synthesis and operated sides of rat brains which had been hemi-of cyclic AMP (CLEMENT-CORMIER et al.. 1974). Neur- transected at the level of the globus pallidus. As the,rhysiological studies indicate that both dopamine distribution of the ascending aminergic fibres arei, I , -aminobutyric acid (GABA) are inhibitory trans- strictly ipsilateral (LINDVALL & BJtKLUND. 1974;,i,,.r in these regions (WooDRUFF et al., 1976). UNOmISTsDT, 1971), aromatic amino acid decarboxy-

iilly, also the efferent connections of the above- lae (AAD, EC 4.1.1.26) was studied in operated, tioned regions show some similarities. Thus the brains using the unoperated side as control. These

221

,FriLCL.G I-,iA bLAW-WMQ JL&MF6D

32

222 F. FoNNuK 1. WALAAS and E. IVUMN

studies allowed a more detailed mapping of the distri- brains running from A 9500 to A 5900 (Fig. 2) and particu-bution than those reported by other workers lar emphasis was placed on the sections at A 7200, where

(KATAOKA et al., 1975; PALKovrrs et al., 1974, TAPPAZ samples from lateral and medial preoptic area, nucleus

et al.. 1976). In the operated brains, a decrease in tractus diagonalis and MFB could be distinguished (Fig.

enzyme activities was taken as evidence of degener- 3).atio ofascndig ordesendng ibre cotaiingthe For subecuhtar fractionation and uptake studies samplesation of ascending or descending fibres containing the from normal brains were used. Striatum, globus pallidusrespective transmitters. During this study, we made and MFB were dissected out from 300um sections pre.the interesting observation that parts of the region pared by a Sorvall tissue chopper at levels between A 7600traversed by the rostral medial forebrain bundle and A 6700. For MFB a sample corresponding to samples(MFB) contained a very high level of GAD which 3 and 4 in Fig. 3 was dissected.merited further investigation. Part of this work has In mesencephalon structures were dissected at levelbeen presented in a preliminary form (FONNUM et al., A 2000-A 1500 as indicated in Fig. 4. The ventral tegmen-in press (a)). tal area was divided in three parts, the ventral, middle

and dorsal. Substantia nigra was divided into pars corn-

MATERIALS AND METHODS pacta and pars reticulata.Biochemical methods. DL-3,4-Dihydroxyf2-"C]phenyla-

Surgical operation. Albino rats of both sexes weighing lanine, L-[t-"4C]glutamic acid and [1-"C]acety-CoA150-250g were anaesthetised by administration i.p. with were obtained from the Radiochemical Centre, Amersham;Nembutal Vet (Abbott) or Hypnorm Vet (Janssen Pharm. [2,3-3H(N)]y-amino-butyric acid was from New EnglandComp.. A transverse cut was made in the left part of the Nuclear, Boston.skull 0.5-1 mm in front of the bregma, the dura mater was ChAT was assayed according to FONNUM (1975). GADsplit with a scalpel and a hemitransection of the brain was determined by the method of Ataaus & BRADY (1959)was performed with a blunt spatula placed at 90 to theskull surface. The transection went from the midline and4-5 mm laterally, and in most cases passed through thebrain at the level of the rostral part of the globus pallidus,thus interrupting all ascending and descending fibres atthis level (UNGERSTEDT, 1971; CONRAD & PFAFF, 1976;FONNUM et al, 1974; LINDVALL & BJi:RKLUND, 1974).

Tissue preparation. Normal and operated animals werekilled by decapitation, the lesioned animals 6-14 days afteroperation. The brain was rapidly removed and frozen ona microtome chuck with a CO2 jet, and 40 An serial sec-tions from the telencephalic and mesencephalic regions A amunder study were cut in a cryostate at about - 16'C. Thesections were freeze-dried and collected as previously de-scribed (FoNNUM et al., 1970). At regular intervals sectionswere collected on microscopic slides, allowed to dry and A 7890stained with methylene blue or for AChE as described bySTORM-MATHISEN (1970). During the preparation of thesections, the hemitransection was examined under themicroscope.

Dissection. Corresponding freeze-dried and stained sec-tions were compared simultaneously under a stereo micro-scope. The different regions were identified from the atlasof K6NIG & KLIPITL (1963) and the parts to be analysed A 6790were dissected by the use of razor blade splints andweighed on a fish-pole quartz fibre balance (FoNNuM etal, 1970). In the telencephalon, we used sections from thelevel A 9400-A 8400 (KbNIG & KLIPPE., 1963) to dissect A 6000

the olfactory tubercle. the nucleus accumbens. the septumand the striatum (Fig. I). The olfactory tubercle wasdivided in two parts; a ventral part containing the olfac-tory tubercle proper and a dorsal part being contaminated Flo. 2. Localization of samples taken out for analysis ofby the islands of Calleja and perhaps by some fibres from medial forebrain bundle (MFB) in rostrocaudal direction.the MFB. Drawinp modified after KoNio & KumiJ. (1963). Abbre-

The nucleus accumbens was divided into a medial and viations: CC - corpus callosum v - lateral ventricle;a lateral part by placing a vertical cut at the ventral tip cp - nucleus caudatus; a - nucleus accumbens; tu -

of the lateral ventricle. The sample from septum was taken tuberculum olfactorium; pi - cortex piriformis;from the area adjacent to the medial boundary of the nu- TOL - tractus olfactorius lateralis; se - septum;clew accumbens. Samples from striatum were taken out CA - commissura anterior; st - nucleus interstitialias indicated in Fig. I. striae terminalis; pom - medial preoptic area; pol - Is-

Samples from the region traversed by MFB were in- teral preoptic area; td - nucleus tractus diagonalis;itially dimected from sections between A 8400 and A 7900. F - fornix; CO - chiasms opticum; ha - nucleus anter-Later this region was studied in sections from unoperated ior hypothalami; vill - third ventricle.

33

FiG. 1. Right side: Section from A 8620 IK6N4iG & KLIPPEL. 19631 stained for AChE. Left side: Localiza-tion of dissected samples. Abbreviations: cp = nucleus caudatus putamen: am =nucleus accumbens,medial part; al =nucleus accumbens, lateral part se = septal sample tu'.= tuberculum olfactoriuns,ventral part; tud =tuberculum olfactorium. dorsal part; CA = commissura anterior; CC corpus cal-

losum; TOL =tractus olfactorius lateralis.

34

FiG. 4. Left side: Section from A 1800 (Ki6NIG & KLIPPai.. 1963) stained for AChE. Right side: Localiza-tion of diaaected samples. Abbreviations: CS - colliculus superior; cgm = corpus geniculatum mediale;sgc - substantia griaea centralia. r - nucleus ruher; LM - lemniacus medialis; hi - hippocampus;ip - nucleus interpeduncularis; Yta = ventral Iegmental area; snc - substantia nigra, pars compacia.snr =substantia nips,. pars reticulata; CC = crus cerebri.

35

Neurotransmitters in mesolimbic system 225

cantly higher level of activity than in the lateral part(P < O.0li. A high level of activity was also foundin the region of the rostral MFB. After hemisection.the enzyme activity on the lesioned side decreasedin all regions studied (Table I). due to the degener-

CID ation of the rostral parts of the ascending aminergic"' fibres. In striatum. nucleus accumbens and olfactory

tubercle. the AAD-activity of the lesion side wasreduced more than 5000.

SThe ChAT activity (Table 2) was found to be veryco ;' high in all regions examined, particularly in structures

A 7190 known to contain a high level of dopamine (striatum,nucleus accumbens and olfactory tubercle). The

1-to. 3 L.oaluation of samples dissected for study of medial part of nucleus accumbns again containedmedial forebrain bundle (MFB) in the frontal plane. Draw- a higher activity than the lateral part. Hernisectioning modified after Kic, & KLIPPEL (1963). -6: Samples of the brain did not significantly reduce the activitydissected. Abbresiations: CAI = capsula interna; else as

in Fig. 2. in any of the structures studied (Table 2.AChE-staining of the telencephalon (Fig. I) resulted

in a picture comparable to that described by JACO-as described by HPKf & FONNUM (1976L except that the DOWITZ & PALKOVITS (1974). The highest AChE ac-incubation temperature was 37 C. In samples dissected tivity was located in the medial nucleus accumbensfrom fresh tissue (Table 51. the final glutamate concen- and the ventral olfactory tubercle, in good agreementtration was reduced to lOmaM and the absolute activity with the high ChAT activity found in these structures.found cannot he compared directly with those in Tables AChE-staining did not change after hemitransection.3 and 4. AAD was assayed as described by BocH & FON-NIH 197) uingKalgnot exracion(FONUM 199). The GAD activity (Table 3) was high in all limbic%t'm. 11972) using Kalignost extraction jFONNUm, 19)9.

The final concentration of L-DOPA was 0.3 mam. Protein structures, particularly the medial part of nucleuswas determined as described by LowiY et at. (1951). accumbens, olfactory tubercle and septum. There was

High affinitq uptake was studied in normal brain, no loss of activity after hemitransection. The activitySamples were dissected from wet tissue and homogenized in the region of the rostra] MFB, however, was veryat 800res. min in a glassTeflon homogenizer in 1000pl impressive, being comparable to the activity of the0.32 m-sucrose iFo4NVM et at.. 1970). 10/p1 of the homo- pars reticulata of the substantia nigra, which hithertogenate (5-30 pg proteini were added'directly to 0.5ml oxy- has been acknowledged to be the region with thegenated incubation medium, which consisted of highest GAD activity in mammalian brain (ALBERS15 mm-Tris- HCI buffer. pH 7.4, 140 mm-NaCL 5 mm-KCI. & BRADY, 1959; FONNUM et at.. 1974; TAPPAZ et al,1.2 mx-CaCI 2. 1.2 mu-MgSO 4. 1.2 mm-Na 2-HPO 4 ,pH 7.4. and 10 M-D-glucose. The tissue was preincubated 1976). A rostrocaudal survey of GAD in the MFB-for 3 m in a shaking incubator at 25'C, and the uptake region lFig. 2) disclosed that this activity peaked instarted by addition of radioactive GABA (final concen- the sections between A 7900 and A 7000 (Fig. 5). Heretration I x 10-' Mi The incubation was stopped after the MiB emerging from the lateral hypothalamus.three minutes by rapid filtration through Millipore filters traverses the lateral preoptic area, and, passing the10.45 psm pore size). followed by washing with 0.9". NaCI. commissura anterior, enters the socalled substantiaThe radioactivity was extracted in 100,. Triton X-100 and innominata. In the middle of this region, the GADafter addition of scintillant. counted in a scintillation distribution in the frontal plane was further analysedcounter. Blanks were run similarly but using a sodium free (Fig. 3). The activity was found to be concentratedincubation medium, where 280rmM-sucrose and 1.2mm- in a region immediately ventral to the commissuraTris phosphate buffer had replaced the sodium chlorideand sodium phosphate. anterior, and to be clearly separated from the nucleus

The proportion of particulate GAD in normal brain was tractus diagonalis, the nucleus interstitialis striae ter-determined on samples dissected and homogenized as for minalis and the medial preoptic area. all of which'high affinity uptake' The homogenate was centrifuged at also contained high GAD activites (Table 4). Dissec-17o0Mg for h GAD and protein was compared in super- lion of samples from fresh tissue confirmed that thenatani and pellet. GAD activity and GABA uptake were comparable

to those in substantia nigra and higher than in globus

RESULTS pallidus (Table 5). Homogenisation and subsequentcentrifugation of a sample from this region (parts 3

7elhncephaon and 4. Fig. 3) demonstrated most of the enzyme to

The AAD activity (Table I) was highly concen- be localized in the particulate fraction (Table 5).trated in the striatum. the nucleus accumbens andthe ventral part of the olfactory tubercle. In the Mesencephaon

medial part of nucleus accumbens. reported to con- AAD was found to display a high activity in thetain more cells fPoiiti & LEMAN. 1976) and staining middle and dorsal parts of the ventral tegmental areamore intensely for AChE lFig. I). we found a signifi- whereas a significantly lower activity was found in

SI

36

226 F. FONNUM, 1. WALAAs and E. IvEasEN

TABLE 1. REGIONAL DISTRIBUTION OF AROMATIC AMINO ACID DECARBOXYLASE IN NORMAL AND TRANSECTEDSIDE OF RAT BRAIN

Aromatic amino acid decarboxylsUnoperated Transected

Region (Mmrol/h/g dry wt) (Percent of normal brain)

Olfactory tubercle V 10.2 ± 2.2 (7,2) 37 ± 7(13,9)0Olfactory tubercle D 4.4 ± 1.0(4.2) 50 + 7 (22N. Accumbens media] 16.6 ±1.2(10,3) 39 ± 3(ll.8)0N. Accumbens lateral 10.4 ±0.8(14,3) 40 ± 6(108)0Septum 6.2 ±0.6(5.2) 33 ± 3 (Z.1)*Neostriatum 14.2 ±1.0(15.3) 14 ±3(ll.8)sMedial forebrain bundle 14.0± 2.6(7.3) 18 5 (Z2)0N. Interpedunicularis 5.1 ± 0.8 (19.7) 102 ±12(21,7)Ventral tegmental 24.4 ± 2.2(00.3) 86 ±15(10.3)

area DVentral tegmentall 28.6 ± 4.6110.3) 95 ±18(8.3)

area MVentral tegmental 13.2 ± 2.646.3) 132 ±14(7.2)

area VSubstantia nigra 15.0 + 2.6(5.2) 99 ± 1(7,2)

compactaSubstantia, nigra 4.0 ± 0,6(19,2) 74 ±15(11,7)

reticulata

The results are expressed as mean ±S.E.M. (number of samples, number of animals). D =dorsal.

M - middle. V - ventral.*Operated side significant different from unoperated side. P < 0.05 (Student t-test).

Since we routinely measured the activity in n. accumbens, olfactory tubercle V and neostriatumnto examine the extent of the lesion more data appear on the transected side than on the normalside.

the ventral part of this region (P < 0.05. ftest). fibres belonging to the cholinergic habenulo-interpe-Higher activity was found in the pars compacta than duncular tract (KATAOKA et al.. 1973), and this prob-in pars reticulata of substantia nigra, and an inter- ably accounts for the fac that this region has a sig-mediate activity was found in the interpeduncular nu- nificantly (P < 0.05) higher activity than the ventralclews (Table 1). After hemisection, no statistically sig- par. No significant changes occurred in these struc-nificant changes occurred in any of these regions. tures after hemnisection (Table 2).IChAT was highly concentrated in nucleus interpe- AChE-staining of the mesencephalon (Fig. 4)duncularis. The activity was low in the ventral teg- showed an intense staining of nucleus interpeduncu-mental area and in substantia nigra (Table 2). The laris in agreement with the high ChAT activity alsodorsal part of the ventral tegmental area contains present. The pars compacta of suhsantia nigra and

TABL 2. REGIONAL DISTRIBUTION OF CHOLINE ACETYLTRANSFERASE IN NORMAL AND HEMITRANsEeTED RAT BRAIN

Choline acetyltransferase activityNormal brain Operated brain

Unoperated side Transected sideRegion (umnol/h/g dry wt) (Per cent of normal brain)

Olfactory tubercle V 96 ± 6(14,4) 106 ± 9(14,5) 114 ± 8 (10,3)Olfactory tubercle D 71 ± 4(7,4) 109 ± 13(7,4) 104 ± 22(6.3)N. Accumbens medial 84 ±17(9,3) 106 ± 6(21,7) 118 ± 14(14,6)N. Accumbens lateral 41 ± 38,3) 120 ± 10(21,7) 117 ±10(13.6)Septum 45 ±6(11,3) 112 ± 14(12.5) 96% 22(g.3)Neostriatum 83 ± 6(7,3) 98 ± 5(25.6) 102 ±9(17.5)Medial forebrain bundle 52 9 9(9,4) 125 ± 17(11,4) 117 ±15(10.4)N. Interpeduncularis 311 ±26(5,3) 109 ± 9(15,5) 122 ±10015.5)Ventral tegnental

area D 18 ±4(83) 91 + 13(8,4) 834 04i VVentral tegmental

area M 10 + 217.3) 1 t0 2011,4) 112 ±20(10,4)Ventral tegmental

area V 7 ± 2 (7.3) 109 ±28 (8,4) 103 ±11(9,4)Substantia nigra

reticulata + compacts 3 ± 1(0.3) 110 ±25(10,2) 120 ±28 (9,2)

The results are expressed as mean ±s.E.m. (number of samples. number of animals). Abbreviations as in Table 1.

37

Neurotransmitters in mesolimbic system 227

TAREiE 3 Ri GIONAL D)ISTRIBUTION OF GLUTAMATE DECAXBOXYLASE IN NORMAL AND HEMITRANSECTED RAT BRAIN

Glutamate decarboxylase activityNormal brain Unoperated side Transected side

Region (,pmol/hig dry wt) (Per cent of normal brain)

0liaictor v tiibercic V 220 ± 21 (9.3) 103 ± 9012,5) 120 ± 11(17,4)O)lt "Ctor% tuibercle D 274 + 5(9,3) 108 ± 10(10,5) 93 ± 8(12.51N. A.ccuinbens medial 220 ±11(11.3) 103 + 15(20.5) 110 ± 8(095)N Accumbe.ns lateral 163 ±48(9,3) 88 ± 21(15.3) 92 ± 10013.5)Septum 187 ±6(9,3) 117 ± 16 (12Z4) 112 ± 5(1223)Neostriatumn 90 ±11(9,3) 114 9 9(18,5) 118 ±12(11.4)Medial forebrain bundle 600 ±88(6.2) 110 ±11(5,3) 108 ±17053)N. Inierpedunicularis 144 ±10(6.3) 99 ± 5(18,7) 96 ±5(16,5)Ventral tegmental

airea D 91 ±8(6.3) 97 + 8(19.5) 104 ±11(15,5)Ventral tegmental

area M 94-± 8(6.3) 96 ± 9(16.5) 98 ±12(13,5)Ventral tegmental

area V 103 9 9(14,5) 94 ± 10)14,5) 91 ±8112,7)Substantia nigra

reticulata 615 ±43 (6.3) 91 ± 6 (39,6) 19 ± 4 (30.6)*Subsiantia nigra

Compacta 234 ±41 (43) 109 ± 12(19,3) 57 ± 5 (13.3)*

The results are expressed as mean ±S.FN. (number of samples, number of animals). Abbreviations as in Table 1.*Significantl% different from normal brain P < 0.01 (Student f-test).

the ventral tegmental area displayed intermediate stantia nigra contained much higher activity than anystaining, which is in contrast to (he low ChAT activity part of (he ventral tegmental area and nucleus inter-in these regions and further demonstrates that the peduncularis. Hemnisection was accompanied by a.. ChE is unreliable as a cholinergic marker in this dramatic loss of GAD in both parts of substantiaregion. in agreement with other reports (FoNNUM.1973a. FONNum et al.. 1974).

Higher activity of GAD was found in pars reticu- 1

htta than in pars compacta of substantia nigra. Sub-

TAft i 4. DiSTRIBLUTION OF GAD ACTIVITY IN THE REGION

TRA'.TRSEO BY THE MEDIAL FOREBRAIN BUNDLE75

Sample numberaccording to GAD activity"-o

Fig. 3 (pmol/h/g dry wt)I

1 612±32252 715±1083 1042+76*4 757±9 %____7_______6

5502±46 8 7 66 508 ±98 Oistanice(mm) from intrural Ine (Mung Kl S1pe..E6)

the results are expressed as mean ± s.EM. and are taken FIG. 5. Distribution of GAD activity in the M FB in rostro-from 6 8 samples. caudal direction. Each point is the mean ± S.E.M. of 4-14

*Sample 3 significantly higher than the other samples analyses. If less samples were analyzed, individual valuesP til)5. are given.

TARi 1 5. ToTAL GAD ACTIVITY, PARTICULATE GAD ACTIVITY AND HIGH AFFINITY GADA UPTAKE OF HOMOGENATE FROM

SOME REGIONS RICH IN GABA

Total GAD activity Particulate GAD activity GABA uptake)Mmol!h/g protein) (Per cent of total) )pmol/min/mg protein)

Medial 715 ±110 79 +4 15.2 ±1.2forcbrain bundle

(,lohus pallidus 361 ±28 79 1 6.3±1.2Striatum 161 ±32 79 ±2 4.3 0.5Siibsitintia nigra 591 ±82 85 ±5 12.1 ±0.8

The results ire mean %F m from 3 experiments.

38

2,28 F. FONNUM, 1. WALAAS and E. IvEasEN

nigra on the lesioned side, but it was without effect the olfactory tubercle since our lesion destroyed aton GAD activities in the ventral tegmental area or least the caudal part of the preoptic area.in nucleus interpeduncularis (Table 3). The distribution of AAD, GAD and ChAT have

been previously studied in detail in neostriatum (cau-DISCUSSION datus-putamen) (FONNUM et al. in press (b)). The level

of AAD and ChAT was similar to the levels in theIn the present report we have studied the levels mesolimbicareas, whereasGAD was significantly lower

of transmitter synthesising enzymes as markers for in the dorsal striatal sample than in the mesolimbiccholinergic. GABAergic and aminergic structures. The areas. ChAT and GAD did not decrease after hemi-advantage of studying ChAT and GAD instead of transection. The findings in striatum after hemitran-ACh and GABA has been discussed elsewhere (FON- section are in good agreement with the suggestedN4UM. 1973b). AAD is. however, involved in the syn- ascending dopaminergic fibres (ANDN et al., 1964),thesis of both dopamine, noradrenaline and serotonin cholinergic interneurons (GUYENET et al., 1975;(KUNTZMAN et al., 1961). Immunofluorescence shows, McGEER et al., 1974) and descending GABAergichowever, particular intense staining of AAD in dopa- fibres (KiM et al., 1971: FONNUM et al., 1974, 1977b).minergic terminals, rather than in serotonergic or The localization of ChAT and GAD in differentnoradrenergic neurons (HOKFELT et al., 1975). Since parts of septum have been studied previously (PALKO-the present study involves regions where the level of viTs et al., 1974; TAPPAZ et al., 1976)_ We includeddopamine predominates over noradrenaline and sero- septum in our study since it also receives dopaminer-tonin (BRowNsTEIN et at., 1974; HORN et al., 1974; gic fibres from the AI0 group (LINDVALL, 1975). TheSAAVEDRA et al., 1974), the level of AAD thus prob- sample which included the nucleus of the diagonalably mainly reflects dopaminergic structures. It is well band and the medial septum contained less AAD andestablished that the dopaminergic fibres are derived ChAT activities than the other regions examined, butfrom the socalled A9 and AI0 cell groups in the relatively high GAD activity. Interestingly, ChAT andmesencephalon (DAHLSThOM & FuxF, 1964; UNGER- GAD. unlike AAD, did not decrease after hemisec-STEDT. 1971), and the decrease of AAD in the telence- tion.phalon was therefore taken as evidence for a success- Several interesting results were found in theful lesion. In all operated animals, AAD was de- samples from the rostral part of the mdial forebraincreased by more than 50*/ on the lesion side. bundle. This bundle consists in part of serotonergic

Previous studies have revealed that the nucleus dopaminergic and noradrenergic fibres (MOORE et al..accumbens contains a high concentration of dopamine 1965; UNGERSTEDT, 1971), and we found a high level(BRowNSTEIN er al., 1974; HORN et at., 1974), tyrosine of AAD in the region, in keeping with the role ofhydroxylase (SAAVEDRA & ZIVIN, 1976), GAD (TAP- the enzyme in the synthesis of both catecholaminesPAZ et at., 1976) and ChAT (PALKOVTS et al., 1974). and serotonin. After hemisection, the activity de-These findings have been substantiated and expanded. creased as expected. The level of ChAT was inter-The highest concentration of AAD, GAD and ChAT mediate, in agreement with the diffuse AChE staining.was localized to the medial part, which is particularly Since ChAT was unchanged after transection, therecell rich, of the nucleus (POWELL & LEMAN, 1976). are probably no cholinergic fibres in the bundle. TheThis part also stained more intensely for AChE and most interesting result in this region, however, waswith catecholamine fluorescence (JACOBOWITZ & PAL- the unexpectedly high level of GAD. which in theKOVITS. 1974). Neither ChAT nor GAD changed sig- freeze-dried sections was well above the activity ofnificantly after hemisection, excluding the possibility GAD in the substantia nigra. In the fresh tissue sec-that these elements derived from ascending fibres. tions (Table 5). the dissection encompassed a larger

Also the olfactory tubercle contains a high level of region, but the activity was still well above the leveldopanine (BROWNSTEIN et al.. 1974; HORN et al., both in substantia nigra and in globus pallidus.1974) and a dense distribution of dopaminergic ter- regions with high GAD activity (FoN' tM et al.. inminals (UNGERSTEDT, 1971). GAD and ChAT have press (b)). Several research groups have found highalso been found in considerable concentrations enzyme activities in neighbouring regions such a.s thefKAl AOKA et al., 1975; PALKOVIs et al., 1974; TAPPAZ lateral and medial preoptic areas (TAPPAL Vr jl, 1976;et al.. 1976). Whereas the highest level of AAD was KATAOKA et al., 1975), or the bed nucleus of the striaconfined to the ventral part, only moderate differences terminalis (BFN-AE et al., 1976). These results havewere found for ChAT and for GAD slightly more been confirmed, but they do not account for the highwas found in the dorsal part. The latter may be due activity in the MFB. Our results demonstrate thatto slight contamination of the dorsal part with the this maximum is located in a region immediately ven-MFB. Previous work has suggested that the choliner- tral to the commissura anterior and traversed by thegic inpult to the olfactory tubercle may be derived MFB. The most caudal part corresponds to the ros-from cells in the lateral preoptic area (LEwis & tral part of the lateral preoptic area. This is in agree-SHUTF. 1967) If a substantial part of the cholinergic ment with the distribution of GABA in the lateralinput had been derived from this area. we would have hypothalamus, which was found to be highly concen-expected to find a substantial decrease of ChAT in trated at the A 7000 level (KIMuRA & KtRIYAMA,

39

Neurotransmitters in mesolimbic system 229

1975?. The region of high GAD activity is not limited ANDtN N. E.. CARLSSON A.. FIxE K., HiLLARP N. A. Soto it single nucleus delineated by classical anatomical LAR5SSON K. (1964) Lile Soc. 3. 5.14-530.

methods. Tice vlosem may be the substanti a i .nno- BEN-Axi Y., KANAZAWA L. & ZhGWND RL E. (1976) J.zrnnta hic. hweve. i il-deinedin odets.The Neurochem. 26. 1279- 1283

lat regih. h ee sgglest ned o repreent The e- 6~RKLuND A. & LINDVALL 0. 11977 in The Continuinglattr rgio ha bee sugesed o reresnt he en- Evolution of the Limnbic Siystem Con~tsr (LIVINGSTON K.

tral part of globus pallidus (HMER & WILSON. 1975). & HoRNYKiEwics 0.. eds.). Plenum. New York. in press.Our findings on the rostral part of MFB seem to be BROCH 0. J., Jr. & FoNNum F. (19720 J. Neurochem. 19,in agreement with such a suggestion since both MFB 2049-2055..iiid globus pallidus contain high GAD activities. BROWNSTEIN M.. SAAVEDRA J, & PAtKOViTS M. (1974)actise high aflinity uptake of GABA and moderate Brain Re-s. 79. 431-437.ChAT aicti% ties iTable 5: FONNUm et al., in press (b)). CLEMENT-CORMiER Y.. KEBABIAN J.. PFTZOLO G. & GREEN-

Fiaainiion of GAD activity in more rostral and GARD P. (1974) Proc. natn. A cad. Sci.. U.S.A. 71.caucdal parts of the MFB demonstrated a distinct de- 1113 1117.

lin indcts t. i areeentwit TAPAZetal. CONRAD L. C, A. & PFAFF D. W. (1976) Brain Re-s. 113,clin inact~itN inagremet wih TPP~ ef 589-596.

I 191t. %%ho onl% found an intermediate level of GAD DAHLSTRbm A. & FuXE K. (1964) Act a phrvswl. scand. 62.in the MI-B dissected from hypothalamus. Thus the Suppl. 232, 1 55.high acti'it% found %as probably not due to MFB FoNNLM,- F. (19691 Biochern. J. 113, 291-298.fibres. In aercement with this. no change of activity FoNNULm F. (19731a) Brain Res. 62. 497-507.wacs Found after hemnitransection. indicating that the FoNNIum F. (1973h) Adr. Biochem. Psi-chopharmac. 6,Mi-B does not contain GABAergic ascending fibres. 75- 88.

In the mesencephalon. the ventral tegniental area FoNNUm F. (1975) J Neurcechem. 24. 407-A09.contains the dopaminergic cell bodies which give rise FONNMM F.. STOR54-MATHiSFN1 J. & WALHERG F (1970)

to the 'mesolimbic* fibres (UNGERSThDT. 1971:- BJoRK- Brain Res. 20. 2.59-275.FoNNium F.. GROFOVA L. RiNviic E., STRsx-MATHISEN J.

Lt Ni) & INIS.LL. in press: SIMON et al.. 1976). Since & WALBERG F. (1974) Brain Res. 71. 77-92.the functional subgroups of these cells are differently FoNNusi F.. IVESEN E. & WALAAS 1. in Adi. Biochem.topographically organized. we divided the area into Pslvchopharniac. In press (a).three. i.e. the dorsal part which projects mainly to FONNUM F.. Go-rESFELD Z. & GROFOVA 1. Brain Rex. Inseptum. the middle part which mainly projects to nu- press )b).cleus accumbens. and the ventral part which mainly GL'YENET P. G.. AGiD Y., JAvoy F.. BF.ALTICAN J. C..projects to olfactory tuberculum (BJiRKLUND & Rossieft J. C. & Gt.owtiss J. (1975) Brain Res. U4

LINVI ALL. in press). The three parts of the ventral 227-244.tegmnental region did not differ markedly. particularly HATTrORI T.. McGEER P. L.. FiRIGER H. C. & MCGEER

whenoneremmber tht te vetra saple lsoin- E. G. (1973) Brain Re-s. 54. 103-114.whenoneremmber tht te vetra saple lsoin- HEIMER L. & WILSON It D. (1975) in Goloi Centennial Svm-

eluded the area near the brain surface which contains f'(SIUM (SANTINI M.. ed.). pp. 1'77- (93. Raven Press. Newfew dopaminergic cells (HOKFEI.T et al.. 1975). In the York.substantia nigra. as expected, we found the dopa- HENKF. H. & IcONNI'm F 119761 J. Neurochen. 27. 387 391,minergic cells to be concentrated in the pars COM- HORN A, S.. CUFLLo A. C. & MILLER R. J. t0974) J N-un,-pacta. The ('hAT lcsels in both these dopaminergic ciem. 22. 265-270.cell nucli were low. The GAD activity, however, was HOKFFLT T.. FvXF K. & GOLDSTEIN Mi 11975) A4nn N.Y.

differently distributed. The activity in the substantia Acad. Sd. 254, 407 -432.nigra iwas 5 it, 10 times higher than in the ventral JA(OBOWITz D. & PALKOVIrS Mi. (19741 J. romp. Ni-urc!.

tcc'mental area Fiirthermore. whereas the GAD level 157. 13.28.susata KATAOKA K.. NAKAMURA Y. & HASSLER R. (1973) Brain

on the lesioned side %%as reduced in the Ri-s 2. 26t.26nigra after hemnitransection. the activity in the ventral KATAOKA K.. SORIMACHI M.. OKUNO S. & Mizuso Ntegmentail area was unchanged We therefore con- (1975) Brain Res. 88. 513- 517.elude that the mesolimbic dopamninergie cells do not Kim J. S., BlAK I. J_ HASSLER R. &o OKADA Y. (1971) Exp.rccei'.e . icsce iding (iABAergic innervation, as do Brain Res. 14. 95 104.

the doparmergic cells in the substantia nigra. KIMI RA H, & KvRIVAMA K 01975) J Neuriichem. 24.In conclusion there were many similarities between 903 -907.

the distribution of GA D. ChAT and AAD in the lrn- KhNi(i J. F R. & KLIPPEi. R. A. (0963) The Rat hrain.bie and neostriatal structures examined, but there .4 Streitartic Mtas ofi ihe Frehrain and Lower Parts

were considerable differences in both the origin as of the Brain Stem. Williams & Wilkins. Baltimore.,Acl I Ilc l~clof AB~rgi firesin he wo KuNIZMAN R. SHORF P. A.. Booo-,ANKi D. & BRoniF B.

rellon .is te i dxlo (inflergivcel f ibes. nte w , (196f) J1. V'ikh,'m. 6. 220-2regins ont~n~e dpamneric ellbodes.Lk.wis P. R. & SHUTT C. C. D. (1967) Exr. Brain Re.

90. 521 540.LIN~vAi L 0. (1975) Brain Res 87. 89- 95

I I RFNCVS LINDvAIA. 0. & B16RKII'Nto A, (19741 .4ctec phvsiol. scand.suppl. 412. 1 41)

.% l it s Rt W A, Bt siiy R I ('S159 .1, Not. C'hem. 234. Lowity 0. H.. RosrRRC0Ui4 N. J.. FARR A. L & RANDALL*i'bc'sR 1. (1951) J. hiot. Che'm 19.1. 265 275

40

230 F. Fotu, 1. WAL.As and E. IV SEN

McGEU P. L., GREwA.AL D. S. & McGEER E. G. (1974) SAAVEDRA J. & ZIVIN J. (1976) Brain Res. 105, 517-524.

Brain Res. 80, 211-217. SIMON H.. LE Mo . M., GALEY D. & CARDO B. (1976)

MOORE R. Y.. WONG S.L. R. & HELLER A. (1965) Ar'hs. Brain Res. 115, 215-231.Neurol. 13, 346-354. STORIM-MATHISEN J. (1970) J, Neurochem. 17, 739- 750.

PALKOVITS M., SAAVEDRA J, KOBAYASHI R. M. & BROWN- SWANSON L. W. & COWAN W. M. (1975) Brain Res. 92,

STEIN M. (1974) Brain Res. 79, 443-450. 324-330.POWELL E. W. & LEMAN R. B. (1976) Brain Res. 105, TAPPAZ M. L., BROWNSTEIN M. & PALKOVrrS M. (1976)

389-406. Brain Res. 10, 371-379.

PRECHT W. & YOSHIDA M. (1971) Brain Res. 32, 229- UNGEsmtEDT U. (1971) Acta physiol. scand. 82, Supplt 367,

233. 1-48.SAAVLDRA J., BROWNSTEIN M. & PALKOVITS M. (1974) WOODRuwF G. N.. MCCARTHY P. S. & WALKER R. J. (1)76)

Brain Res. 79, 437--441. Brain Res. 115, 233-242.

I

41

\,.,*',,5 1r ",'~ 1"'." Ii d'' ~ 0. PAPER If

THE EFFECTS OF SURGICAL AND CHEMICALLESIONS ON NEUROTRANSMITTER CANDIDATES

IN THE NUCLEUS ACCUMBENS OF THE RAT

1. WALAAs and F. FONNI'M

Nor~cgian Defence Research Establishment. Disision for Toxicologs. Box 25. N-2(K)7 KjellerNorwa\

Abstract The origin of fibets containing different neurotraim Itter candidates in the nucleus

accumbens of rat brain has been studied %%ith surgical and chemical lesion techniques. Destruction

of the medial forebrain bundle decreased the actiwil of aromatic amino acid decarbosx use b\ SO",.in the nuIcleus Cutting ol the fornis or a hemnitransection decreased the high afimii uptake of glutamnate

bs 45;". and the ctndogenous lesel of glutamnate h\ 3.l. The high affint' uptake of glutamate ssasconcentrated in the , \naioomal fraction and the decrease after the lesion %kas most pronounced

in (hi, fraction Restricted lesions indicated that fiber, in the fimrbria fornis coming from the suibiculumii crc responsible lor t his pall oft the gluLitnate uptake in the niucleu s. Local injection of kainic acid

into the nucleus usis accompanied bx a 75", decrease in choline acetxltransferase and a 35", decreasein ictsIhl eteaeact isities. a 70-_, decrease in glutarmate decar bosslase aett it\ and a 60",, decrease

in the high atf1it uptake of ;-aminobUt , rate, a 45' , decrease in high alffni t\ glutamnate uptake.

aind not chainge in aromatic amino acid decarboss lase aclits. Performing a lesion of the forni\ afterkainic acid iniection led ito an 05" , decrease in high atfinit.\ glutamate uptake. us ithout further affecting

the other neilron a mark ers.T-he results itidicaie that all arninergic fibers ito tlie nucleus accumben, are ascending in the medial

forebrain biuid le. that t he ii h"ilr w uii currihen, fibet ' are 'glutamergic'. and] that the n ucleus als0

contains intrinsic glutamrergic or *spairtergic cells. (holinergic and *aiminobut. rate-containing cells

are us holls intrinsic to lihe nucleus

Ti ti 'ctit s accumbhenN sept i has in recent \ ears & 1Is t-SI \. 1 9771 It k as showhn that a hemnitran-been the subject of a numbller of morphological and section caudal to the nucleus destrosed the amninergicI' dekelopmental studies uslitch has e Indicated that the fihers %%hich trais l in the medial forebrain bundlenucleus andi its cottlecttions has e characteristics in lij ~\t sI & IlJot)[ ,t), 1974: PALKOX1 t 1 77 1.

commoni both m th 'linihic' strIcItres andi ssith the but left T he cliolinergic and (3A BAergic struc-

basal gatnglia tMi Sit & WttS0i1. 1975 ' Sss A\Sui\ & tures intact I iwt 5 et i.. 19771. In the present

C'iWSN. 1 9'S: ((),RAI) & PFAI'i. 1976. Puuwmti & Stud%, thle precise origin of these *h, -Itnergic,

li-SIAN. 1976. Wti ttsst%-. ('uSssts, & St sit 1t977: (;ABAergtc attd als.o possible 'glutamer- . oers has

LA\WSIIN. MANi & WtiLt [,\%I. 19771, Also, the electro- been insestigated ss ith Surgical an' l 'esions.

pbs siologtcal and ticurochemnical interactions between A lureliminars account of some :se - esiilts has

seseral transmitter candidates present in the nucleus. beeti giseit IWAAAS & Fi)NNI M0 ~

eg. doparmne. acet~ '%choline, glutamate andi ;-amino-

hUtsNrc acid (iABAI. base attracted attention (Wuxto- I \PI RlIME Th L PR(I Ii DI) RIHI FF. M( ( ARiii & WAL JR. 1976. M( (ARtIM.

WA1 is RI & W(XiliRt Iti. 1 977: * ARt 1101.t\1, 1977: ' Iirul( w.sonio. LAIDnN.SK . J)A--1 )I & (iH5jz7,j. 1977: MAO. Kainic ,acid iskis fron Sigma Chemicali Co.. St Louis.

CHIiiNON. MIAH(lo. Ri-S itTA & ('OStA. 1977: PIi t' SA White Wistar rats of both seseN. \keighing

Mt LA NI)RA & Fti \i. 1977). How4eser. except for the FO 1i 041g. \\cre frosm Dyrlaege Mollergaard-Harissensascending ~ ~ ~ ~ ~ ~ ~ ~ 5

. Islnegc ies(iS t1 B6K k ahoratoriumn. D)enmark R adiochem icals sAereascedin dopminrgiclibrs ~t~il~tt. &BjiK ,)btlamied from Nen% -ngland Nuclear. Boston. V S A and

itt). 19741 our ktoss ledge of thL natomnical orgait- The Raduocheinical ( ettre. -\mersham. t* K;,atlon of the fibers containintg these transmitters tsstill incomplete.

We base pres iousl\ shoss i that itt the rat. the Fhe aitnil, Issicrc ai est het id %%it h a miture of diac-

rlucleus contains high densities of putaftue aniinergtc pant aid f? 'pfnorm icteriflarti tJ,inssen Pharmacological

tmist I\ d opaininergic and serotoriergic (I10K N. ( o tatid plaiced in a K opl itereotlasic apparatus The skull

II'liii & Mvit Ii H . 1974. SAMS I tORA. BHgi iWSFI IN & \uas opened a it h a detlal il, andi k nou i affetents to

Pst i ist t s. 1 974i ]. cholitiergi: atnd (iABAergic -;true- ( lie nuJcus a-urniimbeiusIDi Ot Sills & 14I Ai I~'2.li

(tircescpectalI. titti Oleedial part lhiN-st Si.- WAt A ISIIIti&. MiN15KA\%A & N sio. 197~4. Puit t & Li-sIA..~. ~ .9ht6. S\ss Aswi & (iio%%.\., 19,71 acre lestotied

thhrnitautiotis A(tIt. accis Icholuneslerase (hAT. cho. iI i 'X (i'iiiuida,, rion usas, done suith a blunt spatula ait

lilc wiets lrinsfetuse ai-I ~-iitinohuturite the les el ol thle biegmia. is pres ousls describeil iti %si

209

42

210 1, WALAAS afld F. Foossa w

et, al.. 1977). This lesion destroyed all ipsilateral ascending the fimrbria. was done with a cannula in the Name %aN.fibers to the nucleus hut moving it rostrallN onls from A 3f)1 to A 6.0 (Fig, li.

02) The mcedial forebrain handile was cut b% lowering a l101 A dorsal parusawiial Ic ccion which comprised thethin spatula at tire level of the bregma to the hase cif the docrsal tiria terininalis and lateral timbriai was done withbrain 1.5 3 mm lateral to the midline. This lesion cut a scalpel 2.5 mm lateral to the midline. lowered 6 mml fromthrough globus pallidlus and the median forebrain bundle the skull surface and moved from the level of the bregmafibers, but left the fornix bundle intact, to the level of the lambda, terminating 4mm from the

13) Fibers from the thalamuis were destro~ocd bN plIecing skull surface. This lesion left the niediail and cccud.il partan electrode insulated to the tip at coordinates A' 4.11. of the hippocampal formation iiiictL, 1.0. V: -0.2 IKosimG & KLIPPt-L, 19631, and making K tornit a(cid %was dissolIed in 0 9' -Nil( 1 (4 mg c~anti

an electrocoagulation with 2 mA for 20 s. This led to a the pH adjusted to 7.2 with NaOH A hocle was madelesion centered in the parafascicular and centromedian in the skull, a Hamilton microsNringe was lowered i,, Lo-nuclei, ordinates A 9(0. L. 1 0. D) - 0.9tKiioi & KiiPi1. 14611

141 The dorsal cieiicicrris was rpmoi'.iJ i ocic sidle ccv ico ccc ,, crc. icicid - ;~:c. - V I' ,nmsuction down to the white matter under visual guidance. %%c~:.~.cc performed by iniu~ing o9N N't. I In I (liereaching within 2 3 mm from the fronctal pole and inclid- sanic mianner rhe kainie acid lesion was, ciniried sining the whole cingular cortex. sections fromt separatc brains stained for cells wih May)

(5) The caudal pircform~ cortev was exposed after remo- Grunwald Giemsa stainsal of the zygomatic arch aid ventral skull (PO)WELL. To studs the effect of combined chemical and surgicalCOWAN & RAISMAN. 19651 and lesioneci icN suction under lesions. .inimals were injected with kainic acid, allowed tovisual guidance. This lesion destroyed the cortex between recover for I week and thecn operated %kith ic hiliotr.ilthe rhinal fissure and the base of the brain. but left the tornix lescociamygdalotd nuclei intact.

(6) 'Fornix :esion'. which comprised the cinguluim Ii55iii prepacraicni

bundle. fornix and fornix superior and stria terminalis. was For ei Po iii ii t o cr ioptciki ruictis. normnal or lesioneddone with a scalpel 2 mm behind the bregma. lowered animals were decapitated, the brain taken out and cut in6mm from the skull surface and mosed 4mmn from the 40)01 cX~m thick slices with .i Sorv~cll tisue chopper inmidline on both sides lFig. 1). To differentiate between (he cold roomn Alt lesions were inspected, and the brainsthe fiber bundles destro~ed b.N this lesionitia series of more discarded i necessary The nucleus accurnbens. and. in nor-restricted lesions were performed' tial brains, the olfactor' tubercle. lateral septum. cauidaic

171 The ciciquli handcle asr cut bilaterallN with at scalpel pittamen rineostriaturnt and glohu, pailltduv were dissectedjust in front of the bregma. lowering it 5tmn from the following ire atlas of KomoR & Ki.it L 119631. Theskull surface (Fig. I. s amples, were homogenized in glass Tetisin homogenizers

(8) A Icing parcisciiil leivi which comprised the stria at SWt rev. min in cold (.12 oij sucrose .w % L so as toterminalis. the lateral alseus-fimbria sts1cm and also the preserve sxnapioscimcs [ oN~ i. 1971(1. kept on ice andventral amygdalofugal pathwaN was done with a blunt can- ,inal /cd for uptake ictis il% wrthirn1 itt F or aiiiic Ic iinula lowered to the base of the brain 3 mmn lateral to the incai %is. the animal, %%ere killed by inicrovsti irradiationmidline, moved in the sagit ta plane fromn A 1.it to A fctised on (ile skuall for I (;I tuict i. I.11i I N. TRA viii6.0. and then laterally to the side of the bratin (Fig I I , tit, l)(11ci1 Iil. "vs %sN(, & H-4 vv ss d incd tile ucleic'

(9) A short parcccc;-: .i" cc, vo- i coccprised the in~ acecuitchens dissected oiut I oi dccc 11,J cccc icci cIl ue

itial part of tie tina ermninalIts, the senlti a IN 1,i lfUgi I n uc eic iccuinbeim I apprc vli tg wet we cgi was, (tonfibers and also a small part cif the .ini ricyeni r.cI pari 1i1 cc ccied in (5 iml ticmior, siicrcc'c a nd separated int nicc

.-pellet and iiicchondriacl. s% iapccscnal. nicln acnd

cc Isle fraiocns as prey icuskI desciibed it I % K vs si N

& Fo v i, sN. I) Ifi Ne ~napiocrcl mar ker 0I0cu1 citace-/ ivti:,iistec~ise lh-I. It 3 I (,I the ncitcehccndrcal market

7 Aciiiiii .cectiictiI f(U 2 'I.') tie glUtincaic

,c1 7' 1cdieccII ccntent were nicasicred in) Iiinst iS

F ccr Pcii cciiccui 5cl cof bomccgenaie were: cadded ico

Itiml IIIi K rebs oediscm. neic bated with rad icaci iseligacid fccr 3 mcc ,it 25 I.and the uptake terminated boo

M illipc utlrat ciii as prey icctis described IFciNi t %I cc l.N-)' As ligand cither 1 211'-H Iglutanate 00 Il sit cit

AMIritistc of I 'H 1glitarm 10 Ii v, ItInd [ 1 4( G-BcIm ". vi wvere used I riter these oncdmiocca. glutamnate and(rNH~A dcc cci interee wstb each thber's tracnspcirt uBsI

FIi, I Schematic drawing showing site' ctil lesioins ccl dil- . v & JitNi .I')-'j Separate especriments cdemosn-ferent protections to the nukcleus accui hen as viecc frcmu "u -di ,

1ct I he ccicaiotoivt aocuci clacted in nceleus

above 1. hemitranvection. 6. fiurni s leviocm. '. ci ngi tin a1wccii hen (I, 1c-cna ctes was prciport tonal to the amountbundle lesion. ii. Icing paraiiagittal lesion. 9. shcrl pair&. cf issue vs ((cci iIe r.ingi: used iii this stusfIn i a fewsagittal lesion .Ilihreiciccccnc A( ( . nucleus iccimbeis experimenis. c-I . il laspartate wI% used as..I ncsn-mcta-AM. am.%gdala. HE1 hippos-ampis. S. suhiculuirn . ST sIrici hu'11 ligitid (IIvis & jcmIis,i. I ]Q'(t tic exclude the

irrmtnalis. 1. tmbrca pissihiltil ccl nictihclio. influence in the ciptitke

43

Neurotransmitter, in the nucleus .ccumbens 211

rsmi., I tiiI Il- I( I iii In %1il I RA\%5I to'(% 01% \l L ROT RAW411TER LIARk IR% I-, rHt %.1 ([I11 S A( CL Mt M%

t'noperated side Operated side Differencetpmol ht g proteini lpimol h g protein II-

Glutamate uiptake 2.6 1 + 0.13 i14i 1.42 + 0+16 (141 -5GABA uptake 5.48 , 0.36(7) 4.87 0156 (71 - 1Aromatic arnino aicidl decarbox~ use 35.6 + 6.5 (91 7.1 2.3 191 *. SW1('holine actItransferase I1 If, 1 (91 1(XM + 20 181 - 15Glutamate decarhoxss usc 310 - 31 (61 296 + 5! (61 - 5

Results preseitted as mleant +- (..snumber of unimalsi.P < 0151 (Witcos~on paired comparison testi.

'('r "": It"' anl" i". I Ile 1horniouenmi %e re (red ted %kith opera ted and LInoperated animals (datLa not sho~kn I.Iton \-IO d~Iiinil :onc. 0_2.2 u Prcs iiis described This hats also been shown to be the case for aromatic

nctho'ds %%ere used or deternminatinofa ( hAT iI-oi\t St. amino acid decarboss lase. (hAT and glutamate

j B R5 -i & F o i \ I . g l intht flOr b o a o %eu b s J s I C l a s e 6 I K l s d 'i e ca!hr f o e u e a s c o n t r o l i n t h e r e s t4.mn 1. 1 -vss 111F \t I l 977 luandat \Ihletrs

4.1.1.15P1 3.+ 71ss xi et \I. 19691 ('riiet trans- of this stud),. except where bilateral lesions were per-

ferase acoxii% \%its anal' ed using Kalignot extractioni at frmd

.acid pl ito isolate the product (Li \i) KARL5L\ & Fissi %I Further anal~sis of the glutamate uptake demon-197sx v strated that the decrease following hemitransection

UM101i0'iiii noo k~~sere anal.,zed b~ the doubte was almost maximallN developed after 7 days (Fig. 2).isotope dans'.lation techn]iLue as prcxiousi~l described Subcellular fractionation demonstrated that the up-Iliist .I & Wsi %\xs. 19 7xi Protein \ads measured accord. take \%a concentrated in the s~naptosomal fraction.ing to LimwkN. Roxi lROIi (;Ii. FARRk & RAMliALt 119511. %%here most of the decrease after lesion was also found

.%tiqiu a tri'atmn (Fig. 3). In contrast, the mitoehondrial marker carni-

(*omparisons betaeen the side, %ith and Aithout lesions 10or brains s% ith and a ithouit lesions %%ere performed withthe Wilcoson paired comparison test or the Wilcoxon I\&osample test as, appropriate it-F iii,; & Li lIMANs. 19641.

I ESILTSv

To stud\ al1rent transmitters in the nucleus0accumbens. stirgical lesion techniique-; Acre emplo ed.Sex en das s alter I Iteinit ran sect ion thIirough t hcglobtiS pallidLiS. thle h11ih afimI1 it. tarnate Liptake 14kka.s decreased b\ 45',, and the aromatic amino acid Daysdecarbo\y lase tctis it.\ sits decreased by SO".. xx htle Ftu;. 2. Time course of changes in high attinit% glutarnategltitantate decarhoxxi lase and (hAT activities and uptake in the nucleus accumbhens after hemnitransectionthe high atlittit iuptaike of GABA %kcrc unchanged Each point is mean + %.ix of live to nine operated ai(Table 11. Samples from the Uinoperated side displayed mals expressed as per cent of simultancouI\s anal\7ed tin-the same gkltamlate and GABA uptakes ats shamn operated animals.

CHOLINE ACETYLTRANSFERASEO UNOPERATEC oHEMlTRANSECTED HIGH AFFINIIY GLU UPTAKECARNiTiNE ACETYLTRANSFERASE A UNOPERATED8 UNOPERATED OHEMITRANSECTED a HEMITRANSECTEM4

a- 2

0. 6 12 14 010.4SUCROSE MOLARITY SUCROSE MOLARITY

I it, 3. Suc~rose densit% gradient centriftigation sho%4ing the subeellular distrbution of choline .lce I-IanleraIS. carihitine accil hranslerase and htgb atjfinit glutamate uptake in the nucleus acerimbcnl,

ot inoperated and heiniutrinsected rat braitt. Fiach point is tmean + %. i~i. from three animals expressedas, relaiixe to the actiS t,% in P,-pellet from unoperated animals isce Fxperimental Procedurotl

44

212 I. WALAAS and F. FONNUM

tine acetyltransferase (MCCAMAN, MCCAMAN & STAF- TABLE 3. THE EFFECT OF RESTRICTED LESIONS ON THE UPTAKE

FORD, 1966) and the synaptosomal marker ChAT did Of GLUTAMATE AND ON THE AClVITN Of ARO'.JATIC AMINO

not display decreased activities in any of the fractions ACI) DfCARI)OXNLAS iN Till NUi'LIUS ACCMRENS

studied. Topographical distribution studies demon- Aromatic aminostrated that glutamate was most active in the neo- Glutamate acidstriatum, slightly less so in the lateral septum and uptake decarboxylasenucleus accumbens, rather low in the olfactory tubercle Region with lesion (per cent of control brain)

and very low in the globus pallidus (Table 2). Dorsal cortex 99 + 7 (7)Selective lesions of known afferents to the nucleus Thalamic nuclei 105 + 13(8)

accumbens (Fig. 1) revealed that lesions of the medial Medial forebrain bundle 86 ± 15(8) 21 + 7(8)0forebrain bundle, the thalamic nuclei, the dorsal neo- Pyriform cortex III + 13(4)cortex, the caudal pyriform cortex and the cingulum Bilateral fornix 4 + 4 (11)"

Cingulum bundle 93 + I1(5)bundle had no effect on glutamate uptake in the Dorsal parasagittal lesion 97 ± 4 (41 91 + 6(4)nucleus (Table 3). A short parasagittal lesion, which Short parasagittal lesion 88 + 13(4) 87 ± 8(4)destroyed the stria terminalis and the ventral amygda- Long parasagittal lesion 67 + 16(5)0* 97 + 9(5)lofugal pathway, or a dorsal parasagittal lesion, whichdestroyed the stria terminalis and the lateral fimbria, Result presented as mean ± s.i.t. Inumber of animals).

but left the ventrocaudal part of the hippocampal for- P < 0.001, * P < 0.05 Wilcoxon two-sample test).

mation intact, also did not decrease the glutamateuptake. A long parasagittal lesion, however, which cutthrough the whole hippocampal region, led to a sig- 5- 6 days after injection (not shown) demonstrated anificant decrease in glutamate uptake. A fornix lesion, heavy loss of cells with large nuclei and basophil cyto-designed to destroy all fibers coming from the hippo- plasm while the remaining cells were characterizedcampal formation, led to a decrease comparable to by small dark nuclei, thus probably being of a glialthat found after hemitransection. These results indi- nature. This lesion was not completely confined tocate that part of the glutamate uptake in the nucleus the nucleus accumbens. but in most animals alsoaccumbens is dependent on fibers which originate in encroached on adjoining regions of the septum. neo-the hippocampal formation and travel through the striatum and olfactory tubercle. Biochemical analysisfornix. Aromatic amino acid decarboxylase activity after this lesion demonstrated approximatet a 70,analysis after some of these lesions confirmed that decrease in the activities of ChAT and glutamatethe decrease in glutamate uptake was independent of decarboxylase. a 60, decrease in GABA uptake andthe aminergic innnervation, as a median forebrain a 35",, decrease in AChE activit,. while the aromaticbundle lesion decreased the aromatic amino acid amino acid decarboxylase actiit) was unchanged.decarboxylase actiity b) the same amount as a The glutamate uptake was found to be decreased byhemitransection without affecting the glutamate approximatel) 45",, (Table 5). In sham injected ani-uptake (Table 3). mals. no decrease in these parameters was found.

Amino acid analysis after a lesion of the fornix Performing both kainic acid injection and fornixdemonstrated a significant decrease in the concen- transection in the same animals led to approxiniatel'tration of endogenous glutamate. and also a slight an 85",, decrease in the glutamate uptake. This treat-decrease in aspartate. which, however. was barely ment had no additional effect on ChAT. glutamatesignificant (P = 0.05 with the Wilcoxon two-sample decarboxylase or aromatic amino acid decarboxylasetest, P > 0.10 with the Student's r-test). The levels of activities compared to kainic acid alone (Table 6).GABA. glycine and glutamine were esscntially un- Thus. the decreases in glutamate uptake after thesechanged (Table 4). two lesions were found to be additine.

To stud) the intrinsic transmitters in the nucleusaccumbens. the neurotoxic glutamate analogue kainicacid was injected stereotacticallh into the nucleus. TAR[ i 4 THi IFi (- i t) |ORgIX IRA ( tio0% o's iII CO%-

Histological staining of sections from these animals (iNIRAIiO%S Oi SOMI AM'tO AlID IN lilt Nt(LFI'S

A((I itI N

TABLI 2 THI DISTRItiI I,'s ,I H:i t PTiAK[ f 10 - st ot i - Unoperaled side Lesioned sideI&MATt iN lilt BS l ,AI N(I 1 AND *', 4 tIA MRIt RitilONS 11 i 6 i i

Ifmol g protein) (pniol g protein) A",.Glutamale uptake _ __ - __ JIpmol h g proteini (ilutamite 1034 + 7 7 694 + t , 33

- Aspartale 256 + 21 20.0 -1 2 1 22.Neostriatum 4 14 + 018(271 (1scine 141 + 19 155 + 44 +I)Globus pallidus 0.46 + 0,07151 -- amlnobulsrate 446 + 36 3140 - 3 S 13Lateral septum 262 ± 0 271111 Glutamine 94.5 + 58 60 + 99 9Nucleus accumbens 2 67 + 0 161151Olfaclor tihbercle 1 42 + 1 121121 Results presented its mean 4 st M-

S- I P < 002. ** P - 005 lWilcoxon tIo-,ample test). orResults presented ,is mean + s% m (number of animals) 0 11) < P < 1)20 (Student', -tet I

45

Neurotran-,mitters in the nuceleus accurnhens 213

TABI 1 5. Tmtt IF I I fX UAL KAIN IC ACI IiiNJIl t10%(ON NIA ROt IRANS.MlI tIX R tA I ,1Is IN I Hit NI tl S M ( 1 SIBP NS

LUninjected side injected side Difference(limol h g protein) i

(ilutamnate uptatke 2.49 + (0.25(9) 1.32 + 0.21 (9) - 47WGABA uptake 4.00 + (0.41 (5) 1.52 + 0).25(5) --62*Aromatic amino acid decarboxslase 39.4 + 1.8 (9) 41.6 + 2.1 (91 +6Acet v cholinest erase 7341 + 440 (6) 4760 ± Sx) (6) - 35*Choline acetsltransferase 129 + 7 (131 31 + 5 (13)- 6-Glutamate decarhosx (use 245 ± 20 (10) 75 + 24 (10) - 70"

Results pi~esented as mean + s.1 .s. (number orf aiiailsf.*P 0 (.03. - P < (002. ** P < R).K[ (Wileoxon paired comparison test).

TA11-1 6. TIt t I I OF t %ILA rtRAL KAtNIC AC]i1) INJiItlIN 1lOILIEW~t) RNIBit-Atl RAt. LI St0Of (iHI FORNtX 0(INR NRO-RA\5t1 iIR tARKt RS IN I It- NI (Lit S ACCL&5IHINS

tinlesioned animals tmninjecied side Change Injected side Change(lmol h g proteIn) (limol h g protein) (11") )pmol h g protein) ("'.)

iltitamnate upiake 3.17 + (0.2)) 1.87 + ((.22 - 41 0.54 ± 0.16 - 83*Aromatic aioiio acid dccarbox.%laxe 4(0.7 , 2.4 42.8 + 2.4 + 5 41.7 + 3.2 + 3Choline acets Iransferase 10(2 + 9 96 + 7 -6 22 + 7 -78Glutamnate dcarbtixlase. 256 + 3(0 259 + V( + 1 73 + 27 - 72*

Results presented ifts mean + s.t.s. from SIX linlesioned and six operated animals.P ,0.0011. ** P (10.02 (compared with unlesioned animals. Wilcoxon two-samlple test)

DISCUSSION These resltts have been confirmed. but in addition

Meh'ioo()ocica1 coinsiderationis we found that the high affinity glutamnate uptake de-creased b% 40 50',, after (his lesion, while the high

The sltitabilit% of aromatic amino acid decarbox%- alnt GB .llewa rcagdlase. (hAT atnd gluitimate decarbos ae ats markers Gfu"IIII(itrji Ifterents. The possibilitN of a glula-for catecholamninergic indolaminergic. cholinergic and mergic innersation of the nucleus accumbens %asGABAergic fihers respectiseK has previoIusl been therefore studied in more detail. Firstl%, a topographi-discussed (FoN\,i %t. 1 975h: F0NNiM %i etia.. 1977l. Itt cal distribution studx of glutarnate 'uptake demotn-qcontrast. it marker for plttatise glutamergic fibers has strated great differences in uptake actisitN in different

been difficut to fintd. Hoeerecent reports indicatte regions. e.g. it high uiptake %%a found in 'the necostria-t hat the so-called 'high at(un it\ uptake* of gI Itttmate flm. slight[\ less in the lateral septum and nucleusis specificalls located in stich termitnals (Yt NG. aceumbhens. lower still tn the olfactor\ tubercle. atndOSrI R-GRANI~II. HI R-,i)EN & SNN DIR. 1974: STORM- \,er\ low in the globuts pallidus. Of these regions. bothMAtots N. 1977: Dtix M. FoNNt m & StORM-MA1-ttt- the ileostriatumn. the lateratl septlum and the olfactori.si:N. (977: MCI(;tR MCGI IR. SCOFZRLR & SINoLI.- tubercle have been suggested to recetve glutamergic(977: lt-Nt, KARt.SI N & FoNxt-t. 1978: FE)NNLUt & fibers (Di\AC Vl (IL, 1977: Kim. HASSLER. HAUG &WAt AAS. 197814 This utake might therefore be a stllt- PAIK. 1977: FoNNUI. & WALAAS. 1978: HARM.~able marker for suich iletirons. but at least in I tr Su~tiIt LD. GRAHAM & APRISON. 1975). Secondlh.the carrier does not seem to be able to distinguish subcellular fractionation of homogenates from the nu-between gluttamiate and aspartate. atnother tratnsmitter dlells atccumnbens showed that glutamate uptake wascididate (BAt( At) & JittNSTONs. 1972). In this sitild '. tros[ active in the s nilpiosome-enriched fraction.wec have therefore ulsed glutatmate uiptake its matrker Third[%. thle lime course of the decrease fin uptakefor both pItttive gllitamergic an~d aspariergic fibers. after placement of a lesion was gecrallN comparableattd performed amino atcid analsses to differetitte to that Itsuall\ found for other transm'iltcr markersbetween them. following iixot'om% (Ruis. GtLAD. PICIT & J011. 1978).

I ral~plilvr% n alt-epitfibr~sAlso, tile decreatse in uptake was clearl\ not it resultI rois~tiI('rsUt iI~&c~tifih'rsof changes in amtino acid metabolisnm in the nucleus

Juidging from ticiiroanatotiical stuldies. most fibers following lesions. as pilot experitments with the meta-reachtng the nucleus accumbhens in rat braIin seem to boltcall\ tnert ligatnd tu-itspilrate. which travels b\ theoriginate fromn more caudatl regions (SWA\soN & same utptake miechaInisin as i.-glutamate aOld t.-aspitr-0OWAN. (975) We therefore started these studies %%ith tate (DA\S t & .101INSiON. 1976). demonstrated ap-it COMplele hemnilratnsccttot placed caudail to the nli- proximtel\ the same decrease atfter lesion (35 40",,cletts. This has pres ousl * been shown to be xx itholit not three animals).effect oin glutlamate decatrbox~ lase and C hAT r ;thi- These reslts thuls indicated that aIlmost 5(Y',, of thetes. hot to llecremIe the aromnatc t m DO aIcid decatr- hiigh I(liniit\ ghtiflatnate utake tn the It lcletis

ho\\ lse act i It siths tan tia llvk fl-osNNf st atii. t 977 3) iccibcits w~as spcci lica I h loca li/ed in tiersc lctr-

46

214 I. WALAAs and F. FONNUM

minals of an ascending projection to the nucleus, acid injection may be interpreted mainly as a conse-More restricted lesions were then performed to find quence of the loss of intrinsic cells.the origin of this projection. Following injection of kainic acid, massive losses

No significant difference in glutamate uptake was in both glutamate decarboxylase and ChAT activitiesdemonstrated after removal of the cingular or the were found. Indeed, in some animals the nucleus waspyriform cortex. Electrocoagulation of intralaminar found to be almost totally devoid of these enzymes.thalamic nuclei also left the uptake unchanged. Tran- Acetylcholine and GABA therefore seem to be presentsection of the medial forebrain bundle destroyed the almost exclusively in intrinsic neurons. GABA uptakeaminergic innervation but had no effect on the uptake decreased slightly less than the fall in glutamateof glutamate. Thus fibers from the ventral tegmental decarboxylase activity and the activity of AChE de-area and brain stem, thalamus. and cingular or pyri- creased markedly less than that of ChAT. A smallform cortex did not possess a detectable part of this part of the GABA uptake is therefore probably local-uptake capacity. Bilateral lesion of the fimbriaifornix, ized outside kainic acid-sensitive neurons, possibly inhowever, led to a decrease in glutamate uptake com- glial cells, while a substantial part of the AChEparable to that found after hemitransection. Lesions activity could be present in glial cells (HEMMINIKI,designed to differentiate between the fibers cut by this HEMMINIKI & GIAcoStNI. 19731 or in afferent non-lesion indicated that the glutamate-accumulating cholinergic fibers (LEHMANN & FIBIGER. 1978). Thefibers innervating the nucleus accumbens through the aminergic marker aromatic amino acid decarboxylasefornix originate in regions located medial and caudal was unaffected by kainic acid, demonstrating that thein the hippocampal formation. The excitatory fibers ascending dopaminergic and serotonergi" fibers werein the fimbria 'fornix (DE FRANCE & YOSHIHARA. 1975) intact.which probably come from the subiculum and ter- In contrast, the high affinity uptake of glutamateminate in the nucleus accumbens (SWANSON & was decreased by 40- 500,, after injection of kainicCOWAN, 1977) seem to fit this description reasonably acid, the same decrease as found after lesion of thewell. Also, finding the same decrease in uptake after fornix. This result would seem to indicate either thehemitransection and bilateral lesion of the fornix indi- existence of intrinsic glutamate/aspartate neurons incated that the projection was unilateral, as has been the nucleus accumbens, or alternatively that the affer-shown to be the case for the subicular fibers (SWAN- ent glutamergic fibers had been destroyed by kainicSON & COWAN, 1977). Amino acid analysis after place- acid. To test these possibilities, we combined thement of the lesion demonstrated a significant decrease lesion of the fornix and the injection of kainic acid.in the level of glutamate but not in aspartate. Thus, and found that the decreases in glutamate uptakethe subiculum-accumbens fibers are excitatory, they were additive. Thus. the afferent fibers were notpossess high affinity glutamate uptake, and have a lesioned by kainic acid, and destruction of some localhigh concentration of glutamate in their terminals, cells, either neurons or glia. which also can accumu-We therefore propose that these fibers are glutamer- late glutamate in a high affinitN manner (HtNn,gic. GOLDSTEIN & HAMBUiRGER. 1974), must have been re-

sponsible. Since glial cells are left unchanged morpho-logicallN after the application of kainic acid (OLNENTransmitters in loc'al neuronset al.. 1974: HATlTORI & MCGFFR. 1977: S('HWAR(Z

With the introduction of the selective neurotoxic & Coyi.i. 1977). the decrease in glutamate uptakecompound kainic acid, which in several brain regions probably reflects a degeneration of intrinsic glutamer-has been shown to destroy local neurons but to leave gic or aspartergic neurons.afferent fibers intact (OLNEY. RHEY & Ho. 1974;MCGEER & McGEER. 1976: COYLE & SCHWARCZ. Conclusions1976: HATTORI & McGEER. 1977: HEaNDoN & Our results indicate the existence of a subiculum -COYLE. 1977: FONNUM & WALAAS, 1978). a direct accumbens projection which probably uses glutamatelesioning technique has been made available for as transmitter, ascending aminergic fibers from theanalysis of the transmitter composition of local mesencephalon. and populations of intrinsic choliner-neurons. However. microscopical control of the type gic. GABAergic and glutamergic or aspartergic cells.and the extent of the lesion will be necessary. In the The identification of these systems should help inpresent study, the lesion exhibited the same light- clarifying the functional interplay of the differentmicroscopic characteristics as those described by transmitter systems present in the nucleus, and alsoSCHWARCz & COYLE (1977) in the neostriatum. with the functional organization of the different "Iimbic"a selective loss of neuronal cells. However. the and "extrapyramidal' regions connected with thechanges were also found to be present in adjoining nucleus accumbens.regions. Thus, short projections to the nucleus Acknowledgements We thank DR IRFNA GROFOVA foraccumbens (PoWFI.t. & LEMAN. 1976) would probably help with the thalamic lesions, and Ms E, IltRsEN andalso be affected. However. such projections are prob- Lv ELA1seN for technical assistance. I. WALAAS is a fellowably quantitatively insignificant (SWANSON & COWAN. of the Norwegian Research ('ouncil for Science and the19751. and we therefore think that the results of kainic Humanities.

47

\curotraistniiters in the nucleus accumbelis 215

REV ERENCES

HAI( 'i %R' J. & JiII\AIjiv G. A. R. (19721 The structural Npeiccficit% of the high affint% uptake of i-glu~tmatc andI -aspartaic b.% rat brain slici:5. J. tieurmocon, 19. 2657 2666.

B ARmotiII\i 0 09I71 In icracI on of dopainine. acei.0choline and --aminobut~ric acid in timbic and strial itructure,.relat ion to ss e thcrap of schi/ophrem a. fin Rinepi, phil. .\',,tannsl Irc P5, i'sis, led, Di AR RI lu1 I KRAj. & tss R.I. pp, 241 21Masson. Paris.

BiIli 0) J.. JR & Uosm5 %t F (19721 The regional and subcellular distribution of catechol-O0nieth\l transferame int he rat bra in . \curoli'n 19. 21149 24455.

Co,,RI) L I. A. & PisAll 1) W, j19761 Autoradiographic tracing of nucleus accuinbens efferents in ihe rat. BranR,-\ 113. ,51,9 ,590,

0%s\'oi i S.. LMi.~\IIBismm.ii S & Gill ,i D, i 19771 Apparent lack of a dopaminer-gic cholinergic link in therait nLuCcUs a1ccumb11etscpiti-twherculum olfactoriurn. Brain Rt,'. 135. 225 263

(tim J T & Si I.sK. /~ R. t l96i Lesion o1 tlrialal neurone, A ith katntc: acid pros ides a model for Huntington'scliorc-i S isi, 1,ui. 263. 244 2146,

DA)Ws i .L & .loii~stii\ Ui A. R kjq7ti LUptake and release of i)- and i.-aspartaic h% rat brain slices. J. \041,001oiM26. ti11i' 10i14.

DI V-K A\I I1 1 & 'Ii silifix-AR * H 41 9 -5 1 unibria input to the nucleus accumbens septi. Brain Res 90. 1,;,)I 03[) it lii .1. S & l\,RAi W, R. i19721 The projection field of the stria termindils in the rat brain. Ali ex~perlitental

si ~iS.1 su \,Ui't146. 1ii1 334DI\ V 1, 1 1\\i M U- & S[iiRAIAm~isi\ .1 (19"1r Hisgh aftirm t u piake of gIat amate in terminals of con icost ri.i i1

J\0111 'taurn Lonl 266. 3-~73.Io m i I ili I ~diohiiaIn coa~\o thle delernmition (if cho line acet \It ra nsferase and acetilcliollnesterase

Jciii tics Jr/lt / 115. 465 472.Vi\i \ A I iW fi Topograipicail and subcellular localization of' choline acet transferase in rat hippocampal regiotn.

J45W e 17. M4 ~I i)-\iA 1 I I) \~i r.i pid rad ioclietii m et hod for the determintation of choline acethtransferase.e rol ltl'iiil.

24. 4o1' 4449Vo- i\\4 AlIF li he locali/ation of glutarme decarbosx lase, choline acetillransferase. and aromatic amino acid

dc.rboA laie iniii maiait aind In'ertebrate tiersous t issue. In Metihra,,i ( - i1ipartmei'itiillian ',i,/ ernran liin, sit

leds HI RI S .(i.A-. 1) D & St ilimiR D.i. pp. 99 122. NASI Life Science Series. Plenum Press. NesYork.I (,\\i \t I & ),% %I \% s I. It Ti hc effect of iniraltippocainpal kaitic acid injections and surigical lesions on neurolrIns-

Initters III iippocafiipii and septui,. J \Sen" 1honf .1. 11- 1 I'%-s'I At \k*\ ,\Si s I & I\IRHNl\ U t197' Locai/iation of (jABAeritc. cholinerL'ic antd aitnergic structure, in the

itesohiiic .5sen.1 5 ,uro hens. 29. 221 230i(t mit nit A . ( itim 1IN ) I.. TRAlM (i (i M\1. Dill III I Ms.. W.'As(, U & HAWKi\s R. A. 119,741 Focussed niuicrosaste

radihitioni. a techitquie to iiiliie postmortem change, of c\,dijs Iiucleottdes. dopa and cholhne and to presebraiit isorptiolog.\ S sphiisioa . 11llS 1122

HARMI J A . Si MIMI il liC (iRAiiASI L T. JR & A'RItil\ M4 H. 4'51 Putaime transmitters in denersated'Ilictors orte\ .1 S ,iro h,ns 24. 445 441)

IfI 141RI F& *VI (i i III 1 6 09--1 1-ine struCtUral change, in the rat stirlau ,IIler local injections o1 kainic acid.libum R, 129. 1 '4 1 NO

bimi kl R. N. & 'An 5 s Ri I) 0I)->1 I-le utbcomtcal protection, of the hippocampus. the p' riformn corcA aind the neocories.III I',rs,-; s Ss-hh' 'P 6,11,s ( 011ls,111,i sp 1,ssim ted. SA\S\ tiiM.4. pp 177 193. Ra'.eti Press. %c% YorkItl lllk K . Ill Al\li\iKi I* & (it ki oilim I i19"11 t Xtisits of ctimttes relaled to lieuiroitransiiism in in neurotial

id 01.it It1iftolis it .4 \, ir, 1 5. S_ 90iI N\ IIt. (i(I sII I-, % I & Himl 55111 R .. A119741 1'ptake of the neurotrantiitter candidale glulamnale t,, glia

S ,ou. . .iI249. 66 664

iI R\IitI\i R mI & (I. I r tI igf't sceetise destruction of neuronecs hb% a transmitter aizonis V ten, . \ Y 198.

HI Lii I . 4H & LI li,%\\ F F L 0964t 811- ('iil% Wt Praailin anid Srtmsti(s Holdeit-Da' San Fraincisco1109 NR . I I ~I I 10 AC & %,III IlIKI R 1 14741 Dopatiltt in tife mcsolilc systen of the rat brain. endogenous

IC\Ci elad i lie effects of d rugs tit) the ifiplik e mechanism and st imutal ion of aden.%late c.\clase actis lt J ttlen, A hen22. 265 17to

11i) N . ISiliti KI.Sis. \11 \kA V & NAtili H 119741 .Microcirode stiJ.% of projection,, from the amsgvdaoisl conipleAto lie ttictlis .icciilIIbII1,I itle Cait Snout Rv 67. Its 341.

Ki Al IS. i ssiit R . UIA C. P & PAiI, K -S t I9771 Jfi~ect of fronttal corteA ablation lilt trial~al glutamic alcid leseliii ral Biasii R1' 132. 37 17t 4

Kwo's. J F R & KvliiiiiiI R. H 963t l/te Rat Braim I Swrircvu Va it thte foorehram ii Limo I."'r 0aIs f4it,

Braiii .StIn Williams & Wil ILins. Baltimore.S % , Mto %i51 K & WiIt I Ikls I H] t I9"7 Prenal nellrogenests in the %eptal region iif the ral. Brum~

Re, 129. 14' 151IIl lmiAAS J & F- nlta~ Kt Hi ( 114711 Acei\ lcholineslerae tit the suhstania nigra and caudate putanen ilf the rat*

pnilperlic asn1ld local/atlil fit dopa nilnergic netiroi s .1 \eourIoi/Iep. .30. 615 624Lisi' AllI () & Bili)ks I \it A. t49741 The organiratiiim of the asc:titng calecholanine neurotn \qcmili til the rat

brain Ifi it hivl '11411,.tSipl 4M 1. 484

it -- ---

48

216 I. WALAAS and F. FONNUM

LOWRY 0. H.. RosEBRoucut N. J.. FARR A. L. & RANDALL R. J. (1951) Protein measurement with the Folin phenolreagent. J. bioI. Chem. 193. 265-275.

LUND KARLSEN R. & FONNUm F. (1978) Evidence for glutamate as neurotransmitter in the corticofugal fibers to thedorsal lateral geniculate body and the superior colliculus in rats. Brain Res. 151, 457-467.

MCCAMAN R. E., MCCAMAN M. W. & STAFFORD M. L. (1966) Carnitine acetyltransferase in nervous tissue. J. biol.Chem. 241. 930-934.

MCCARTHY P. S.. WALKER R. J. & WOODRUFF G. N. (1977) On the depressant action of dopamne in rat caudatenucleus and nucleus accumbens. Br. J. Pltarmac. 59, 469P-470P-

McGEER E. G. & McGEER P. L. (1976) Duplication of biochemical changes of Huntingdon's chorea by intrrastriatalinjections of glutamic and kainic acid. Nature. Land. 263, 517-519,

McGEER P. L., McGEER E. G.. SCHERER U. & SINGHt K. (1977) A glutainatergic corticostriatal path! Brain Re's. 128.369- 373.

MAO C. C.. CHENEv D. L.. MARCO E.. REVUELTA A. & COSTA E. (1977) Turnover times of gamma-aminobutyric acidand acetylcholine in nucleus caudatus. nucleus accumbens. globus pallidus and substantia nigra: effects of repeatedadministration of haloperidol. Brain Res. 132, 375-379.

OLNEY J., RHEE V. & Ho 0. L. (1974) Kainic acid: a powerful neurotoxic analogue of glutamate. Brain Res. 77.507-5 12.

PALKOVITS M. (1977) Serotonin in the rhinecephalon. In Rhins'ncephalei. Neuroiranvime;:eurs et Pst-choses (eds DE AJURiA-GUERRA J. & TissoT R.). pp. 115 133 Masson. Paris.

PEREZ DE LA MORA M. & FLLJxE K. (1977) Brain GABA. dopamine and acetylcholine interactions-I. Studies withoxotremorine. Brain Res. 135, 107 122.

POWELL E. W. & LEMAN R. B. (1976) Connections of the nucleus asccumbens. Brain Res. 105. 389-406.POWELL T. P. S.. COWAN W. M. & RAISMAN G. (1965) The central olfactorv connexions. J. Ana(. "9. 791 -813.REIS D. J.. GILAD G.. PICKEL V. M. & JOH T. H. (1978) Reversible changes in the activities and amounts of tyrosine

hydroxylase in dopamine neurons of the substantia nigra in response to axonal injury as studied by immunochemicaland immunocytochemnical methods. Brain Res. 144. 325- 342.

SAAVEDRA J. M.. BROWNSTEIN M. & PALKOVITS M. (1974) Serotonin distribution in the limbic system of the rat. BrainRes. 79. 437-441.

SCHWARcz R. & COYLE J. T. (1977) Striatal lesions with kainic acid: neurochemical characteristics. Brain Re's. 127.235-249.

STORM-MATHISEN J. (1977) Glutamic acid and excitatory nerve endings: reduction of glutamic acid uptake after axotom.Brain Res. 120. 379-386.

SWANSON L. W. & COWAN W. M. (1975) A note on the connections and development of the nucleus accumbens.Brain Re's. 92. 324-330.

SWANSON L. W. & COWAN W. M. (1977) An autoradiographic study of the organizaton of the efferent connectionsof the hippocampal formation in the rat. J. conp. Neurol 172. 49- 84.

WALAAS I. & FONNIIM F. (1977) Neurotransmitter candidates in nucleus accumbens. Proc. Int. Soc. Neurochem. 6.181WILLIAMS D. J.. CROS55IAN A. R. & SLATFR P. 11977) The effereni projections of the nucleus accumbens in the rat.

Brain Re's. 130, 217 227.WOODRUFF G. N.. MCCARTHN P. S. & WALKER R. J. (1976) Studies on the pharmacology of neurones in the nucleus

accumbens of the rat. Brain Re.,. 115. 233-242.YOUNG A. B.. OSTER-GRANtTF M. L.. HERNDON R. M. & SNYDER S. H. (1974) Glutamic acid: selective depletion bN

viral induced granule cell loss in hamster cerebellum. Brain Res. 73. 1 13,

(Accepted 28 August 1978)

L

49 PAPER III

6 .V i 6 N , ' I P P 1 4 l 4 4 ) 5 I M IP d in (irci Brui.,, 0306-4522 81/030399-07502.0,O

Perpmon Prm Lid0 1981 IBRO

BIOCHEMICAL EVIDENCE FOR OVERLAPPING

NEOCORTICAL AND ALLOCORTICAL GLUTAMATEPROJECTIONS TO THE NUCLEUS ACCUMBENS ANDROSTRAL CAUDATOPUTAMEN IN THE RAT BRAIN

IVAR WALAAS'

Norwegian Defence Research Establishment. Division for Toxicology, N-2007 Kjeller, Norway

Abstract The high affinity uptake of L-glutamate has been used to investigate the origin and distribu-

iJon of putati'.e glutamate fibers in restricted parts of the rostral caudatoputamen and the nucleusaccumbens of the rat brain. Ablation of the frontal cortex reduced the glutamate uptake heavily (- 77")in the dorsal part of the ipsilateral caudatoputamen, but also led to significant decreases in the ventralparts of the ipsilateral caudatoputamen J -62'. and - 53'.). in the ipsilateral nucleus accumbens (-250%and - IK",,) and in the contralateral dorsal part of the caudatoputamen (-21,). Lesion of the caudalneocortex reduced the glutamate uptake in the dorsal part of the ipsdh, ,.iiiudatoputamen only( - 23.j Lesions of the fimbria fornix reduced the glutamate uptake in t , of the ipsilateralnucleus accumbens I - 46",, and - 34%o, and by approximately 20', in the 'A h .rsoventral extent ofthe anterior caudatoputamen.

The results indicate that the frontal neocortex distributes fibers which may use glutamate as neuro-transmitter both to the whole ipsilateral caudatoputamen and to the nucleus accumbens, and also to thedorsal parts of the contralateral caudatoputamen. The caudal neocortex probably sends such fibers tothe dorsal ipslateral caudatoputamen, and the caudal allocortex sends such fibers through the fimbria!forni, to the nucleus accumbens and the ventral part of the ipsilateral caudatoputamen. The results thusiorroborate presious suggestions of close similarities between the nucleus accumbens and the ventralcaudatoputamen

REC'EN[ anatomical and neurochemical evidence indi- the dorsal, middle or ventral parts of the rostral cau-cates that the nucleus accumbens septi possibly might datoputamen and in the medial and lateral nucleusbe regarded as an integral part of the rostral neostria- accumbens after separate or combined lesions of theturn (HEtItR & Wit soN. 1975: SWANSON & COWAN, afferent cortical projections. The effects on the chol-

1975. FoNNt'M. WAt.AAS & IVERsEN. 1977; NAUTA, inergic marker enzyme choline acetyltransferase

SMITH. FArI t & DOMISICK. 1978; WALAAS & FON- (EC 2.3.1.6) and the monoamine marker enzyme aro-Nt'm. 1979a) Both these nuclei receive major cortical matic amino acid decarboxylase (EC4.1.1.26) (FoN-projections. and anatomical inve, ;nations have NUM er al., 1977) have also been investigated, in order

shown that the neostriatum ... ., putamen) to control the density of local neurons and mesence-

rce:elvs most of these fibers from the ne, urtex (KN- phalic afferents after the different lesions. Some pre-

ILF. 1975. HF t)RFN. 19771. while the nucleus accum- liminary results have been reported (WALAAS & FoN-

hens receives most of its cortical input from the allo- Num. 1979b; WALAAS. 1980).

cortical hippocampal formation (HEIMER & WILSON.1975. SW AsON & ('OWA'. 1977) We have previously EXPERIMENTAL PROCEDURES

demonstrated that these two projections contain Male Wistar rats, 180-200g body weight, were subjectedfibers which probabl.o use glutamic acid as neuro- to 4 types of lesions under anesthesia with fentanyl/fluani-transmitter IDIvA(. FONNUM & STORM-MATHISEN. sone (Hypnorm vet., Mekos) and Valium (Roche). The1977: WAI AAS & FONNI'i, 1979a). results which have frontal neocortex anterior to the bregma was removed uni-

been confirmed bN other workeTs tMcGEER. MCGEE. laterally by suction down to the white matter under visual

S('HIRFR & SINGH. 1977: ZA(-ZFK. HEDREEN & COYLE. guidance. with the lesion extending ventrally to the rhinal

1979). The present study was designed to further fissure and the olfactory bulb (Fig. I). In some of these

investigate some aspects of the distribution pattern of animals a small part of the cortex anteriomedial to thethese glutamnate fibres. and in particular to study forceps minor of the corpus callosum was spared (Fig. 2).

The fibers from the hippocampal formation were lesionedwhether they terminate in separate fields or whether either by a unilateral or bilateral knife cut through thethey overlap. The high affinity uptake of glutamate fimbria and fornix I mm behind the bregina and 6mmpresent in these fibres has therefore been analysed in deep to the scull surface (Fig. I) (WALAAS & FowNuM,

1979a). by a unilateral suction lesion through the dorsal

Present address Department of Pharmacology. Yale parietal cortex which destroyed the septal part of the hip-

University School of Medicine. New Haven CT 06510. pocampus and the fimbria under visual guidance, or by aUSA hemitransection of the brain at the level of the globus pal-

399

50

400 1. WALAAS

+1 +1 +1 +1 41 0

-- -4

CPC

2 0z

14 o

E . g

VFIG. 1. Schematic outline of lesions of the corticofugal pro- : .

jections. viewed from above. 1. anterior decortication; 2 .. 0fornix transection; 3. hemitransection; 4. posterior decorti Z C:-

cation. ,4bbreriations; A. nucleus accumbens; cp. caudato- E +

putamen; HI, hippocampus. LZ

lidus (WALAAS & FoNNUM. 1979a). A third group of ant-mals received both a frontal decortication and a bilmeral +1 +1 + +fornix cut and a fourth group received a suction lesion of N e ,

the caudal neocortex and corpus callosum overlying the -- 2dorsal hippocampu% Fig. 1) The latter lesion was designedto control for possible rieocortical fibres being destroyed .

by the fimbria fornix denersations.- . , .

Six to nine days after surgery. the animals were decapi- 4

tated and the nucleus accumbens and the anterior caudato- I - '

putamen were dissected and divided as shown in Fig. 2from frontal brain slices. The samples were homogenized icold 0.32 m sucrose. pH 7.4. in glass-Teflon homogenizers ~2 ~ cto preserve nerve terminals (WALAAS & FONNLIM. 1979if) 2 -J C N

and preincubated in Krebs-tris buffer in a shaking water ' .-bath at 25 C for 15 min. L-(2.3- 'H) glutamate 122.5 0iz 2mmol, New England Nuclear. Boston) was added (final -

conc. 10' ml. the incubation was terminated after 3 min : :by rapid filtration, and the amount of labelled glutamate Z -Eaccumulated in the tissue was measured by liquid scintilla.- ~ -

tion counting IFoNINUM er W)., 19771. 6For enzyme analysis, the homogenates were treated with _ EN E

Triton X-100 (final cone. 0.2".. v ) to release maximalactivity, and choline acetyltransferase and aromatic amino-acid decarboxylasc were analysed with previously de- -

scribed method, (BROCH & Fo%-NI M. 1972: FoNNum. 1975l. Z -

All results were related to protein content (Lo%'Rv. ROSE- E ~BRot~ut. FARR & RANDAlL. 1951). with bovine serum 2 Ei s1 +,albumine used as standard. 7+1N- ~ 4

RESULTS 16

In unoperated animals, the highest density of gluta- --

mate-accumulating structures was found in the vets-CBEE6.tral part of the rostral caudatoputamen, with dis- Z M

tinctly lower uptake activity in the medial nucleusM 0j ' 7;

accumbens (Table 1). In contrast. cholinergic ter- .

minals as demonstrated by choline acetyltransferase > J ~ ~activity were concentrated in the dorsal caudatoputa-men and medial nucleus accumbens. while the amtn-

51

('ortical pirojections (o nucleus accumbens and neostriatum 401

A 90

i"'"' \ AcCA

--- ~'LOT

c--- A 820

Ac CA

Ac - FR

LOT

Fici 2 Schematic drawing of frontal brain sections modified from KoNIG & KIIPPEL (l%3). Outline ofdissected samples indicated on the left side. 1, medial nucleus accumbens; 11, lateral nucleus accumbens:IlI, ventral caudatoputamen; IV. central caudat0Pi. amen; V. dorsal caudlatoputanien. Abbreviations:Ac. nucleus accumbens; cp. caudatoputamen; CC. corpus callosum; CA. commissura anterior; FR,rhinal fissure. LOT, lateral olfactory tract; SE. septum. Numbers indicate distance in front of interatural

line in Mm. Outline of anterior neocortical lesion indicated on upper panel, right side.

ergic terminals containing aromatic amino acid decar- in the central sample (- 120,, not significant). Second,boxylase were more evenly distributed (Table 2). ablation of the caudal dorsal neocortex and corpus

The denervatton procedures led to significant callosum overlying the dorsal hippocaxnpus decreasedeffects on the glutamate uptake in all regions the glutamnate uptake significantly in the ipsilateral(Table 11 Unilateral rostral neocortical ablation con- dorsal caudatoputamen ( -23*,) Third. unilateralsistently removed all cortical tissue on the frontal dor- destruction of the allocorticall fibres in the fimbria/solateral part of the hemisphere. but usually left a fornix by means of three different procedures allsmall part of the cingular. suprarhinsl and anterome- reduced the glutamnate uptake in both parts of thedial cortices intact. This lesion reduced the glutamate ipsilateral nucleus accumbens ( -450% and -35%) aMduptake significantly in all ipsilateral regions, most in the ipsilateral ventral and central parts of the cau-heavily in the dorsal caudatoputamten I-77"%) and datoputamen (24'. and - 2 1%). Two of these pro-least in the medial accumbens (-I",.Contralater- cedures. i.e.. the brain hemitransections and the forinix

ally, the glutamate uptake decreased significantly in transections. also destroyed afferents to the nucleusthe dorsal caudatoputamencr I - 21".) and also slightly accumbens running in the stria terminalis. while no

3,...- .--

52

402 I. WALAAS

TA9LE 2. AROMATIC AMINO ACID DECARBOXYLASE AND CHOLINE ACETYLTRANSFERASE IN NUCLEUS ACCUMBENS AND CAUDATO-

PUTAMEN FOLLOWING CORTICAL LESIONS. FORNIX LESIONS OR HEMITRANSECTIONS

Ipsilateral frontalUnoperated brain decortication, per Fornix lesion, per Hemitransection per

Region jumol/h/g protein cent of unoperated cent of unoperated cent of unoperated

Nucleus accumbens AAD 49.6 + 3.5 (6) 103 t 5 (5) 106 ± 10 (II) 23 ± 5 (4)*Medial part ChAT 216 ± 33 (5) 96 ± 3(6) 91 ± 3(5) -Lateral part AAD 47.9 + 4.4(6) 91 + 9(5) 104 + 11 (11) 12 + I (4)0

ChAT 157 + 19(5) 91 + 3(6) 83 ± 13(5) 104 ± 11(4)Caudatoputamen AAD 58.9 + 3.3(6) 95 + 7(5) 94 + 10(11) 13 1 1 (4)0

Ventral part ChAT 172 ± 5(10) 96 ± 6(5) 109 + 7(5) 116 ± 13(41Central part AAD 54.5 + 3.6(6) 101 + 9(5) 94 + 12(61 18 ± 3(4)*

ChAT 232 + 9(10) 91 ± 5(5) 98 + 7(5) 110 ± 10(4)

Dorsal part AAD 56.1 ± 2.2(5) 88 ± 3(5) 17 + 5(4)0ChAT 255 ± 33(5) 86 t 6(5) 98 + 351

Results presented as mean + SEM (number of animals). *P < 0.01 (Wilcoxon's 2 sample test).Abbreviations: AAD, aromatic amino acid decarboxylase: ChAT, choline acetyltransferase.

visible damage could be seen in this fiber bundle after ing that only these parts of the rostral caudatoputa-the suction lesions. After all these operations, the dor- men receive inputs from the contralateral frontal cor-sal caudatoputamen displayed approximately the tex (Fig. 3). Analysis of the activities of choline acetyl-same reduction in uptake activity as after a caudal transferase and aromatic amino acid decarboxylaseneocortical lesion. No significant effects were found in revealed that the densities of cholininergic and amin-any regions on the contralateral side after the caudal ergic terminals were unchanged following both neo-lesions (Table 1). More extensive lesions substantiated cortical and allocortical lesions in all regions studiedthese findings. Combining a unilateral frontal decorti- (Table 2). In the hemitransected animals, the aromaticcation with a bilateral fornix lesion demonstrated ad- amino ocid decarboxylase activity decreased byditive effects on the glutamate uptake in all regions 75 880. in both caudatoputamen and nucleus accum-ipsilateral to the frontal cortex ablation (Table 1). bens. These results agree well with our previousContralaterally, such additive effects were only found studies (FONNUM et al., 1977: WALAAS & FONNUM,in the 2 dorsal parts of the caudatoputamen, confirm- 1979 .

CAUDATOPUTAMENCONTRALATE1RAE ' I

ALLOCORTEX

NUCLEUS ACCUMBENSFIG. 3. Schematic outline of origin and distribution of putatie glutamate-using cortex fibres to the

caudatoputamen and nucleus accumbens in the rat brain.

53

Cortical projections to nucleus accumbens and neostriatum 403

DISCUSSION density of neocortical than allocortical glutamate

The present study confirms that lesions of cortical fibres (glutamate uptake: approximately 140 pmol/mgersein the rat brain decrease the glutamate protein/3 min vs 55 pmol/mg protein/3 min, seeneurons in the aidatop utam te above). Thus, the terminal field of the neocorticaluptake capacity in the caudatoputamen and nucleus fibres in the caudatoputamen does exhibit some sortaccumbens without destroying intrinsic cholinergic of boundary towards the nucleus accumbens, even ifcells or aminergic fibres from the mesencephalon such fibres also penetrate into the accumbens proper,(DivAc et al., 1977; McGEER et al., 1977; WALAAS & c lusionlterefre, the accumbens hasFONNUM, 1979a). Thus, this glutamate uptake ca- In conclusion, therefore, the nucleus accumbens has

pacity is probably concentrated in the terminals of the the same qualitative pattern of cortiual glutamate-cortico-striatal and cortico-accumbens fibres. It fibre innervation as the ventral caudatoputamen, butwould therefore appear to be well suited as a bio- the allocortical input dominates quantitatively in thechemical 'marker' in studies on both the topography nucleus accumbens while the neocortical input

and terminal density in addition to investigations on dominates in the ventral caudatoputamen. The distri-the transmitter identity of these projections. bution of cholinergic terminals found in this study

also does not distinguish the nucleus accumbensclearly from the caudatoputamen, and the pattern of

The nucleus accumbens and the ventral caudatoputamen afferent dopaminergic innervation and acetylcholin-esterase staining is also very similar (FALLON &

The major question investigated in the present MOORE, 1978; LINDVALL & BJORKLUND, 1978; FON-work is whether the allocortical fibres to the nucleus NUM et al. 1977). However, the detailed distributionaccumbens distribute in a terminal field clearly separ- of catecholamines is different, with highest concen-ated from the fundus of the anterior caudatoputamen. tration of noradrenaline found in the medial accum-Recent anatomical studies using either sensitive stain- bens and negligible amounts in the cadatoputamening methods after anterograde degeneration (HEIMER (OPSTAD & WALAAS unpublished). The two nuclei& WILSON, 1975) or autoradiographic studies after also exhibit differences in the density and projectionanterograde transport of tritiated protein (SWANSON pattern of y-aminobutyrate neurons (FONNUM et al..& COWAN, 1977) have clearly demonstrated the exist- 1977; WALAAS & FONNUM. 1979c. 1980). Therefore.ence of these fibres, but not defined their terminal despite the anatomical similarities between the 2 nu-fields unequivocally. The present results appear to clei (HEIMER & WILSON. 1975; HEIMER, 1978; NAUTAanswer this question: those allocortical fibres coming et al., 1978), the bulk of the neurochemical evidencethrough the fimbria/fornix and displaying glutamate indicates that the nucleus accumbens probably shoulduptake clearly extend outside the nucleus accumbens be regarded as a distinct part of the 'ventral striatum"as usually defined (K6NtG & KLIPPEL, 1963: NAUTA et (HEIMER & WILSON, 1975; HEIMEi. 1978) and not asal., 1978). Furthermore. assuming that these allocorti- an integral part of the caudatoputamen.cal fibres all have approximately the same glutamateuptake capacity, the results indicate that the terminal The dorsal caudatoputamendensity of this projection probably is greatest in the Two aspects of the glutamate-projection to the cau-

medial nucleus accumbens (where these fibres are datoputamen should also be briefly mentioned. First.responsible for the uptake of approximately the dorsal part of the caudatoputamen appears to75 pmol/mg protein/3 min of the labelled glutamate, receive an input of such fibres from the contralateralcalculated from Table I. However, the density is only frontal neocortex. This confirms previous results fromslightly lower in the lateral nucleus accumbens (ap- this laboratory (DivAc et al., 1977), and is in agree-proximately 65 pmol/mg protein/3 min) and in the ment with anatomical studies in other species (CAR-ventral caudatoputamen (approximately 55 pmol/mg MAN, COWAN, POWELL & WEBSTER, 1965; KUNZLE,protein, 3 min). Thus, no sharp boundary separates 1975). Second. a significant part of the glutamatethe nucleus accumbens and the ventral caudatoputa- uptake in the dorsal caudatoputamen appears to orig-men in this respect, inate in the caudal neocortex. These fibres were not

The present study also demonstrates that the ipsi- found when the whole caudatoputamen was testedlateral frontal neocortex distributes fibres to the nu- (FONNUM, STORM-MATHISEN & DIVAC, 1981) andcleus accumbens, thus confirming recent anatomical probably represents a restricted input to the dorsalstudies (BECKSTEAD. 1979). The terminal densit) of caudatoputamen only. This interesting projectionthis putative glutamate-using projection is more diffi- might possibly be involved in transferring informationcult to estimate, as some of the fibres may originate in from the visual cortex to the basal ganglia (HOL-regions in the anteriomedial cortex (BECKSTEAD, 1979) LANDER, TIETZE & DISTEL, 1979). Lastly, the dorsalwhich partly may have survived the ablation pro- striatal regions had considerably less residual gluta-cedure (Fig. 2). However. my results indicate that they mate uptake activity than the ventral samples afterprobably distribute terminal boutons in the nucleus combined cortical denervations (30-40 pmol/mg pro-accumbens less densely than do the allocortical fibres tein/3 min vs 80-100 pmol/mg protein/3 mi. respect-(Table I). In contrast, the ventral caudatoputamen ively, calculated from Table 1). These ventral regionslateral to the nucleus accumbens has a much higher may therefore contain more dense populations of

i,\NI

54

404 1. WALAAS

either local glutamnate neurons. glial cells accumulat- neurons. However. some cortical neurons may alsoing glutamate (HENN, GOLDSTEIN & HANIBERGER. have been destroyed by this lesion (FRIEDLE. KELLY &1974) or glutamnate fibres arriving from more ventral MOORE. 1978: WURTHtELE, LOVELL, JONES & MOORE.cortical regions (HEIMFR & WILSON. 1975). In the nu- 1978). The origin of this glutamate uptake will there-cleus accumbens. injections of the selective neurotoxic fore require further study.compound kainic acid, which probably destroys localneurons but leaves afferent fibres and glial cells intact,decreased the glutamate uptake significantly (WALAAS A4knowh'dernments I thank Liv ELIASSEN for expert tech-& FONNUM. 1979a). Thus, this nucleus may contain at nical assistance and Dr J. WAL KER for reading the manu-least some local glutamate (or aspartate-using) script.

REFERENCES

BECKSTEAD R. MI. (19791 Convergent prefrontal and nigral projections to the striatumn of the rat. Neureosci Let! 12. 59-64BRoct 0. 1. JR. & FoNNuM F. 0 972) The regional and subeellular distribution of catechl-O.-methy Itransferase in the rat

brain. J. Neurochem. 19, 2049 2055.CARM.AN J. B.. COWAN W. M., POWELL T. P S. & WEasTlt K. E ( 965) A bilateral corticostriate projection J Neiurrl

Neurosurg. ParVchiat. 28, 71 -77.DiVAc L. FONNum F. & STORM-NIATHISEN J. (1977) High affinity uptake of glutamate in terminals of corticostriatal axons

Nature, Lond. 266, 377-378.FALLON J. H. & MxoRk R. Y. t1978) Catecholamine innerivation of the basal forebrain IV. Topography. of the dopamine

projection to the basal forebrain and neostriatum. J. comp. Neural, 180, 545 580FONNum F. 11975) A rapid radiochemical method for the determination of choline aceil transferase J Nr'uriu hem 24.

Q07409.FONNum F., STORM-MATHiSF\ J. & DiAC I. (1981) Biochemical evidence for glutamate as neurotransmttter in cortico-

striatal and corticothalamic fibres in rat brain. Nerscoc 6. in pressFoNNUM F.. WALAAS 1. & IVERSEN E. (19771 Localization of GABAergic. cholinergic and aminergic structureS in the

mesolimbic system. J. Neurochem. 29. 221-230,FRIEDLE N. M.. Kr.iis P. H. & MisiftE- K. E. (1978) Regional brain atrophy and reductions in glutamate release and

uptake after ititrastriatal kainic acid. Br. J1. Pharmac. 63. 151 158HEDREEN J. C. (1977) Corticostriatal cells identified by the peroxidase method Neurostieni r Lriiers 4. 1HEimER L. (1978) The olfactory cortex and the ventral striatum. In Lirnfi M~echanisms teds Liit,xrrr\N K & HrRok.

iEWIcZ 0.) pp. 95 187. Plenum Press. NY.HEIMER L. & Wii soN R D, 11975)The subcortical protections of the hippocampus. the pyriform cortex, and the neocortex

In Perspecrircs ini Neurohaioloqi. (iolgi Centennial Symposium (ed. SAN,1tisi M. pp 177 193 Raseti Press. NNYHENN F. A.. GoLDtF-rIN M. & HAVIIRGFi A (1974) Uptake of the neurotransmitter candidate glutamate b., glia NatureLind. 249. 663 664.

HOLLANDER H.. TITIr/ J. & Disrti 11 (1979) An autoradiographic study of the subcortical projection, of the rabbitstriate cortex in the adult and during postnatal development. J , om N\erol 184. 7X3 744

K6NiG J. F. R. & K LIPPEL R. A. 11963) The Rat Brain. 4 Siwrerra si( 4r las oft he Forehrain anid 1. ie r Parr, ol the BrainStem. Williams & Wilkins. Baltimore

KUNZLE H. (1975) Bilateral projections from precentral motor cortex to the putamen and other parts of the basal gangliaAn autoradiographic study in Macaca fascicularis. Brain Res. 88. 195 209

LINDvAL. 0. & Bibi[LNi A. (1978) Organisation of catecholamine neurons in the rat central nersous sy- stem. InHandbook i'Psvchophairmtatioq .. (eds ivE.RsuE L. L., Iv~ttNiS. D. & SN~tLR S H (vol. 9 pp 139 231 Plenum Press.NY.

LOWRY 0. H.. ROSEBROUGH N. J.. FARR A. L. & RANt)AtL R. J. (19511 Protein determination vsith the folin phenolreagent. J. hiol. Chrem. 1931, 265 275.

NIcGFER P. L.. McGrEER E. G. SCHERE.R U & Si'-riH K. (19771 A glutamatergic corticostriatal path' Brain Res 128.369- 373.NAUTA W. J. H.. SMITHj G. P.. FAuilt. R. L. M. & DoMFSi(k V B (1978) Ufferent connections and nigral .ilferenis of the

nucleus accumbens septi in the rat. Neurosvience 3. 385 401SWANSON L. W. & COWAN W. NI. (1975) A note on the connections and deselopment of the nucleus accumbens Brain

Res. 92. 324 330.SWANSON L. W. & COWAN W. MI. (1977) An autoradiographic study oif thle organization of the etferent connections of thehippocampal formation in the rat. J comp. Neurol. 172, 49 X4WALAAS I. (1980) The localization of neurons probably using amino acid% as transmitters In I I Bel.. tm u iedTissor R.t Georg Masson. Paris. in press.WAI AAS 1,& FONNUM F. 119 7

9a) The effects of surgical and chemical lesions on neurotransmitter candidates in thenucleus accumbens of the rat. Neuroscience. 4. 209 216WALAAS 1. & Fosist'm F t1979ht Amino acid transmitters linking nucleus accumhefts Ito the basal ganglia .\eroit i Lei,.Suppl. 3. S 229.WALAAS I. & FosNsdrm F 119 79c) The distribution and origin of glutamate ilecarboxylase and choline acet~ltransferase iventral palladum and other basal forebrain regions Brain Re% 177. 325 136

Cortical projections to nucleus accumbens and neostriatum 405

WALAAS I. & FONNLM F. (1980) Biochemical evidence for -aminobutyrate containing fibres from the nucleus accumbensto the substantia nigra and ventral tegmental area in the rat. Neuroscience, 5, 63-72.

WtRTHEtIU S M.. Lov-LL K. L.. JoNEs M. Z. & MOORE K. E. (1978) A histological stud) of kainic acid-induced lesions inthe rat brain Brain Re.s. 149. 489 497.

ZACZEK R., HEDREEN J. C. & COYLE I. T. (1979) Evidence for a hippocampal-septal glutamatergic pathway in the rat. ExplNeur(l. 65, 145 156

(.4ccepted II September 1980)

ii

55 PAPER IV

THE rFFI[C r OF IN rRAIIPCcAMtI 5 1. KAINIC ACIDINJECTIONS AND) St'ROK. AL Hl SIONS ON

NET'ROTRANSMITTERS IN HIPPO(AMPI. S AND SEPTI,'M

F FOi%i1 M Aiid1 I A~ 41 Ak',

\rseoan Dlrinct: Research I stibishmnt t),,',-.n ir -i. . og

ohII 1 1-11 inJtIjin ki katnic acid 12 jgi vAa. d~ornpani,:, hi de'trtktion of iittimn'k TICLrIi,11

)in- rlirt tv hiprocampki% The lesion %js a.-iminid hi% a reductio.n in giuianitci

a'3,,t.%sa 1 i Nist .i Ni ItCdIitiori in the high ittitN uptakc ,I i -glut.iiiaiea1 411 (4) tcd~liontithe end setteow, lesel, ,Isanae glutamate and ( ABA and no Lhirge, in thL aitis lies of

shoIitck i's10ta1110Lra'C o it.'mori amiuno aciJ decarhoxilase in the dor,ii Ii~ppo-atifpl I tilAtCralde'r;ri1110 ICneir..n, in the dorsal hipptocampkis %,a% follov~ed h\ a 20 411" reduction in the high.,thtil, Iitake of ttiltamatv it lateral, hut not in medial septum, tin both ide, There vAa, no reducLtion

h.i, ,t\lnt i 91,ita'trie decoirhos\ iae or aromatic 11mit11 dod dec:arhosslase acti\itiC,

tic h, .. i or fn,diil pirt ot the septum transection of ttmria and uflriwr tornix is aciompaniedh% I !C-u,' - in lholifne acet ti atiserase and aromatic a mmlit acid decarbo\%, iw acltii

int tipporcinit in the high affinitN uptake (if gluitamnate and in the cndlogenious kid o f glutain tiein the la.dii septim The tcsults are consistent \hitli th, concpit hit :t the itipp-,aipiu Iitic

1,I~d d,etio\iriinsic tictions and not afferent fibres It seem, therefore thit all t .AB'serin I. )re%

Ini tie hippo:.itpii belong to intrinsic neurons ishereas glitaimergc aind aspartergic neurons helong

;,.rtk to 1, ~A%: neurons the ottrieion fro m the h ipposa mpa to ithe latieral ept-son pr 'hal his esQlutiateI J1 a transmitter

ltiii %w s5i irtcludtngl the area detata is aI rela- moss> fibres andi the Schaffer collaterals I( tSLA-ii)t

titel\ simple laminated structure \%ith restricted affer- & ( oNsNoR. 197.1. Siiuttsl-M 4,oijsI N. l9,71

cont fibres atnd tess Itirinsic cell t\ pes In recent work Recently it has been suggested that local injectionIthe localt/ation of putalte neurotransmitters in affer- of kainic'acid, A po\%crful glutamate analogue. inent tibres has been Aell documented. The cholinergic other brain regionis destros intrinsic neurons hut notstrulctuires are distributed in lamtnae lFiiNt M. 19)701 afferent fibres 1101NIN et at.. 197-4. S(-i\AAR?' &itid are denised Iroin cells in the medial septum and (o j)t. 1977a & hi. HAkijoii & MICC4+R. 1977 '

nucleus of the diagonal band iji wis ,-t ait. 1967. WAI AAS & Fo%\~ m. 197S. in pressi We baseliii~~~ RHi Ni i i.,a. 141741 The serotonergic and therefore studied the effect of kiainte ac,dl injection

niiradrenergit: fibres are (if estrahlippocampal origin onl the neurotransmitter candidates in hippocanipusStIR--Mstiii \ & I; 1tiiR6i. 144k and are de- Since th- pxramidlal cells in hippocampus protect torised from cells in the median raphe nucleus iLoRiNS tre Se'"um iRAIMA1 et a11. 19661. \%e base compared& (it iiuHF Hi, 14-4. A/Miti, & SF(ooo. in press) ar,' the cl."ect of katnic acid in dorsal hippocampus \kithlocus coeruleus tRoss & Ruts. 19"41. respectiselN simjaical transection of the hippocampo septa! fibres

Recent espciriments Also indicate that glutamate on neurotransmitters in hippocampus and septumand aspartate tnia% function ias ttansmitters in some There are on!> limited data ott l'calion of-,iflerent fibrv %> sts in hippocaimpus. e g the perfot - neur,,transM it ter, in sept umn (Iwo i\"I. I -f'Smint path itld ommissutral fibres iNADtii 0 i at file actis ties oif choline acets 'ransicrase ChAT.1,)-t, SIO HA I i1 iii . l4-1 I( _' I I (,i glutamnate decarho\%sa ise C;AD[. FIC

It is more dillicitltit) obtain direct esidence ito idci- 4 1 1 151 aind aromatic I -amino acid siccarhosIlaseitls the twor,,traitsnttters in intrinsic neurons. There (A,\[). H 4 1 1 281 bawe been used Is markers foris stronig circutnstanii 11 esidenceo that (jABA in hip- chollitergic. (iA.Aeri and citecholainergic fibresp i catfpss is present in local neurons (STORM-MATH- respectisel\ (FoN1 Mw 1 l. vbilc the so-called highi-! %. 1 'f'21 and also that glutamnate or aspartate may\ alfinit>\ uptake of I -glutamate has been used As Abe the triansinitter of intrinsic s!stems such as the marker fort glittamergic or aspartergic fibtes IN(t %(,

,I it .I974. STi ,pst-M'kijiist 1 9'. D-N (1, ai t

14'" t,i siKA~i sii' & [(,\~i m. 14-)io Sin,e u *glu-I Ah,, ..... 5.. . hNI choine aLCisltr.,ttra~e tamal: and Iaspartate are taiken tip into %snapto-

(,AD). gliitainaic dk- ,rb,'sslie AAR ar).it,,i i -airtin sorties hb'> the same mechanism iBAtI( AR & liisti.

cii de oh *i IseiqI'21. arutto icid ,itals si. has h,-et, isd ito ditteren-

- 7-7777.. -

1174 F FoNt M and I %AlAAs

tiate between the effect on aspartergic and glutamer- 1974. The samples were dissected out and homogenized

gic neurons (LuND KARLSIN & FONNt M. 1978). in 200 mm-bicarbonate acetone I I I Amino acid contentin the supernatant was determined by the dans lation tech-

MATERIALS AND METHODS nique using [-'Hidansyl chloride and 'C-labelled aminoacids as internal standards as described b. Lt',%o KARLSN

Albino male rats, weighing ca 180g. were obtained from & Fo',;s'm (1976). The dansyl amino acids were elutedDr Mollergaard-Hansens Asislaboratorium. Denmark. with 0.75 ml acetone:acetic acid 13:2 v'v). and counted inKainic acid was obtained from Sigma Chem Co.. St. Louis, a Packard scintillation counter Model 3380 fitted with anU S.A absolute activity analyser Model 544 set for simultaneous

determination of 'H and "C.Treatment ol a siiij.t For histological stain. transverse sections (40 Mm) were

The animals were anesthetized with a mixture of diaze- cut in a cryostat. allowed to dry on microscope slides andpam-Hypnorm and placed in a stereotaxic apparatus, stained either with thionin or for acetylcholinesteraseKainic acid was dissolved in 0.9,, NaCI l4mg/ml) and (AChE. EC 31.1.7) with ethopropazine as inhibitor forthe pH adjusted with NaOH to 7.2. Over a period of I butyrylcholnesterase (SToRM-MATHISEN. 1970jmin 0.5 Ml was infused through a hole in the scull into Staristical signilicarne of difference between dorsal andthe dorsal hippocampal region at coordinates A = + 3.5. ventral hippocampus. left and right hippocampus or leftL = 2.5. D = + 2.5 according to KONiG & KLIPPEL (1963). and right septum was assessed by the Wilcoxon pairedSham-injected animals received the same amount of 0.9'. comparison test Comparison between lesioned andNaCI. In pilot experiments, this was found to be without unlesioned animals was performed with the Wilcoxon twoeffect on the neurochemical markers analysed, and unle- sample test iHot(irs & LEHMANN. 1964)stoned animals were therefore routinely used as controls.

The fornix lesion was performed with a small scalpel2mm behind the bregma. lowered 6mm from the scull RESULTS

surface and moved 4 mm laterally from the midline Previous work has shown that biochemical changes

Preparation of tissue after fimbrial transection (STORM-MATHISEN. 19721and neuronal necrosis after kainic acid injections

The animals were decapitated after 5 7 days. The brains and & cosis after hi are fl eeloe

were carefully taken out and transverse slices (400 ,um pre-

pared with a Sorvail tissue chopper. The injection site was within 2 . days. We therefore limited our investiga-

inspected macroscopically. and the animal used if the injec- lions to a study of the effect of these factors on bio-

tion was located within the middle third of the dorsal hip- chemical parameters after 5-7 days. Five daNs afterpocampus. Samples for anal.sis %,ere taken from slices injection of kainic acid into the dorsal hippocampuswithin I mm from the injection site in the rostrocaudal a massive loss of pyramidal and granule cells wasdirection while samples from ventral hippocampus in the found. Figure I shows adjacent sections stained forinjected animals were taken Irom the same or the adjacent cells and for AChE from an animal where the injec-slice, tion was confined to the lateral hippocampus. In regio

For distribution studies, hippocampus from normal ant- inferior and lateral parts of regio superior and areamals was dissected out, freed from subiculum and fimbria.and divided in three equal parts. The dorsal and ventral dentata an almost complete loss of cells was found.

samples were analysed separately, while apparently normal neurons coulo be seen in

The samples from septum were taken from slices the medial part. In sections from two other animals8.6-7.2mm in front of the interaural line and dissected (not shown) a complete loss of pyramidal and granuleaccording to KONIG & KLIPPit. 11963). The medial sample cells was found. We assume that the more diffuselywas taken dorsal to the level of the commissura anterior. localized Golgi II and basket cells were probably alsoIt thus contained the medial septal nucleus, and a part affected although Ibis cannot be properly seen in ourof the nucleus of the diagonal band section. In contrast, the distribution of AChE staining

Biochemical methods did not change on the injected side The laminated

The samples were homogenized in (I.j2 m-siicroc 120' structure and intensity of the staining were well pre-

w'v) in such a way as to preserve synaptosomes (FONNt 1. served (Fig. Ic)1975al Since our lesion was restricted to the dorsal hippo-

The uptakC of 10 M-L-glutamate wa, studied by incu- campus we were also interested in studying the effectbating a 5pl sample (0 150 Pg protein in 05 ml Tris this lesion had on the ventral hippocampus. SeveralKrebs medium for 3 mm at 25 C as previously described of the parameters differed between the normal dorsalIFONNIJM el at, 1977) The uptake varied linearly with the and ventral hippocampus. The activities of AAD andweight of tissue under these conditions ('hAT were found to be significantly higher in ventral

For en:vrt, a vsaas the homogenate was treated with hippocampus, while GAD and glutamate uptake were0.2, 1v v. final cone.) of Triton X-1(X) ChAT was deter-mined as described byi FoNNiM (1975/,. AAD as describedby B1tocN & FoNS.V'm 11972) and GAD by the method Kainic acid injection led to a 740% decrease in GAD

ALB$ & BRtAI)Y t1959t as modified by li't ver al ctivit. and 63",, decrease in glutamate uptake underOf7t ProFtei wa BRDetrie b159 s thifemhd bIofom LeWra(19771. Protein was determined bv the method of Ux)w4 , high affinity conditions hut left ChAT and AAD ac-et a) 11951) tivilttes intact (Table 2). This lesion affected these par-

For amino acid analtsis the animals were killed h ameters neither in the contralaleral hippocampusfocussed microwave irradiation for 2 5 s (it tIit-i it at. (Talle 2t nor in the ventral hippx.ampus on the in-

57

i

411, .2 i

Lc

58

lKaiinic Acid injections on hippocampal neurotransmitters 1177

TA~t I I Disstt 4L11% OFi NI-i t 1TRANSMITTER MARKIRS IN HIPPOAMPLIS

Dorsal hippocampus Ventral hippocampus)5pmol hIit protein) (pumol,h~g protein)

Gis uptake 266 + 0)23(8) 2.45 t 0. 16 (K)ChAT S7 ± Its) 74 ± 3(8)0

(;AD 185 ± 1616) 206 ± 21)(6)AAD 3.5 ± 0.2 (6) 5.6 + 0.3 t6)

The results represent mean + s.i .m. (number of animals).*P < 0.02.

TAIILI 2. Tin I141(1 OSF INTRAHIIPPOCAMPAL INJECTION OF KAINIC ACID ONNi I ROI RANSMITU it MAR"KS IN HIPPOCAMPt S

tjnoperaicd Uninjected side Injected)pumolh it protein)l (". or unoperated)

Glu uptake 2.67 + 0.18)61 93 ±8(7) 37 + 400)GAD 192 + 1401 102 ±(7) 26 ± 4(7)-(hAT 62 ± 6)61 101 ±6)7) 90 ±7(7)AAD 3.7 + 016) 103 ±2(s) 97 + 3)5)

The results represent mean ± sAtm. (number of antmals).*P < 0.01 compared to unoperated animals.

jected side (('hAT: 93 ± 12', GAD: 85 + 7), A AD: demonstrated a high glutamate concentration in92 + 12" _ and glutamate uptake: 80 ± I15%. results lateral septum of unoperated animals (Table 6). Aexpressed as mean + s.F.m. and are in percent of ac- bilateral fornix lesion led to a significant decrease intivitics in ventral hippocampus on uninjected side), the concentration of glutamate. but not of aspartate.Amino acid analysis showed that the endogenous glycine. GABA or glutamine (Table 6).lesels of aispartate. glutamate and GABA were signifi-cantly reduced, while alanine. glycine and glutamine DSUSOwcre unchanged (Table 3). DSUSO

In the septum. the glutamate uptake under high The concept that kainic acid, the toxic glutamateaffinity conditions and GAD activity were concen- analogue, destroys cell bodies and not afferent fibrestrated in the lateral part, the ChAT activity was con- receives strong support in this study. Thus ChAT andcentrated in the medial part. while AAD activity was AAD in the dorsal hippocampus were unaffected byevenly distributed (Table 4). Following kainic acid in- kainic acid injection (Table 2). but were markedly de-jection into hippocampus. the glutamate uptake was creased after fimbria/fornix transection (LEWIS et al.,significantly reduced in the lateral septum on both 1967: Table 4). On the other hand, GAD which didthe ipsilateral side (350 decrease) and the contrala- not decrease after a series of lesions of afferents toteral side (23', decrease). The other neuronal markers hippocampus (STRoM- M AmIEN, 1972), was almostwere unaffected by this lesion both in lateral and completely lost after kainic acid injection (Table 2).medial septum (Table 4). Glutamnergic/aspartergic fibres which have been sug-

The effects of destruction of cells in hippocampus gested to be partly of extrahippocampal originby kamnic acid injection was compared with the effect (STORM -M ATitiSEN. 1977 ' NADLER et al., 1976) wereof cutting the afferents and efferents by surgical Iran- partly destroyed (Tables 2 and 3).section of fornix'fimbria. Following fornix,'fimhrtalesion. ChAT activity was decreased by 90', and TABLE .3 AMINO ACIiDS IN HIPPOCAMPtL 5 AFTLUIiNmT.itiPPO-

AAD actisity was decreased by 45'. in hippocampus CAMPAIL INJECTION Of KAiiC ACID

(Table 5). GAD activity and high affinity glutamate Uninjected side Injected sideuptake were unchanged by this lesion (Table 5). In (prnol,'pig protein) (*, of uninjected)the medial septum, the fornix/fimbria lesion was with-out effect on the neuronal markers (Table 5). In the Glu 92.1 ± (0.3 63 t 7*

Asp 14.0 ± 2.0 59 ± Slateral septum. a unilateral fornix lesion led to a 40%, Gly 10.3 ± 2.1 76 10tdecrease in glutamate uptake on the ipsilaecral side. GABA 25.3 +34 SI 9twhile a bilateral lesion led to a 70", decrease in gluta- Gin 46.3 ±6.3 83 ± 5mate uptake (Table 5) This lesion led to a slight hut Ala 7.3 ±1.5 82 t 5

sigifiantinceas inChA aciviy i th laera hu -_Results represent mean + s Fm from 6 animalsnot medial septum. while GAD and AAI) activities , P <o00were unchanged (Table it Amino acid analysis t P < 001

59

117KF 1'. % mtN%5 and I IlkALAAs

TIi 4 Tmu IItI( I III rsfttAHItPPIK AMPAI t%Jlt t( 105 Ki F-.41. A( 11) O\S %H) kItkA%S-

slIrITIR MIARK IRS 1%. %1 Ilt %I

Lateral q(~twniLnoperated tI IIcIe:Id sIde Iniected jd,:

litimil h g protein) 1.111perated I

UiIl ipralkc 266 0 11IS -7 8 (51* 65 61*(PIIl) 26' 151) 1211 + 11141 9)I ?S 5 1( 11 Nr 4i 2i61 94 -t 10151 X56 10 liti

AAI ' _045) 1013 , 151 ?5 (51

Medial septumn

(ILfI Upftake 0,95 + 0114(4) 96 X14) S6 10 (u41( iA ) 93 1 - I7 1116 914) 11.2 , 4( hA T 120 -16051 11.; (1o (4 05 , 1

\D)4~ 0.4051 it I-97 + 20

I lie tce-iil i cent meani I st %Iinfu mber ,I ainuis i* - 11 till L riipa red to un operated ani11mals.

I '.1) I TillI ItI I I Of FtH%iS t-I5IHRiA II S10% 1\) \1 'I RotitA\ t 1r1 It NSIARKI RS INs

HIPit( AI40)51 *\\[) %I Pi I %

Hippoi~ampus Medial sept urr Lateral %eptumI unops-ra td, controkli

hIk plk 'it lesion 101, _ b S1 91 91 310 4 4t61( upae Urflilit lesiOn 106t 4(5 ii I II 1' 59 5 (151+

It\Ibiat lesion s , I* 103 11~ 25 I IN I 1t61+uflulat lesion 16 11115

(.i)D bilar lesionl 92 11 i( 41 1 1 4 10141D hL ilir lesioni611 s- s (.3 1P 11 141

I hecsi wpircsent mean Isi %i numhei ol airinil', id are espressed asoIf unrcr.litd conirols i muIta neou sI an a Ised.

/P 3 coprdt nperated animals

The findings on cholinergic structures in hippo- acid and glutamic acid JOLNFN et al.. 1974) Iwere notcampus deserse some special comments. In other restricted to cell hodies. The present results demon-regions which have heen submitted to kainic acid in- strate that choltnergic axonal terminals are insensitiveJection. e.g. striatum ICGi I R & MCGFER. 1976. to kainic acid The small decrease in ChAT (10% inSCHWARC7 & (tON i14l77a). retina ISCHIAARC/ & Table 1) corresponded to the proportion of ChAT

(~li.19 77(-. nucleus accumbenIS IWVt AAS & Fo's- resistant to fimbrial fornix transection (8-16%, in

Nt,%t. in press). or in the mediohusal h ' pothalamus Table 51 and may be due to a small proportion of(WAi AAS & I'ONt"It 197S) and retina ILtND cholinergic cells %kith strong AChE staining in 'areaKARLSFN & FoNsst m. 19761 of glutamlate-treated ant- 31* of the molecular laser in rego superior (SREBROmals. there has al~ass been a dramatic decrease in & Mi LI.GRlFN,. 1974). An additional finding in our in-ChAT. One could therefore be led to think that in sestigation A~as the higher ChAT activity in the ven-cholinergic neuron% the excitotosic effects of kainic tral than the dorsal part of hippoicampus (Table 0.

This agrees "ith the more intense AChE stainingTAP 1 AINO1( 01 1\I II Al'-IPII MAFT RFOR[\ found in the \entral part IMALO a' al., 19771TA~ i X~t'tt ill is IsriItS S Pusi ~T~t ~tRtS AAD wshich did not decrease after katnic acid injec-

- ___ _______________ tions. is present in ierotonergic and noradrenergic('rlesionied brain I esionted bratin fibres which are of extrahippocampall origin (AZ.MITIAIpmtil iig proteini C., of unlestonedi & SIAF. 1979: LINiwALt. & B)tORKLUNDt. 1974). After

,~it .. h transection of these afferenit s~stcms, however, a sig-

Glti 121 I 4 9 h7 I nificant AAD actisiti, was left in the hippocampusAisp 200111 I (, 91 8 (Table 4, SrORM-MArIMst' & GtLDRERG. 1974). In6ily 6 7 + 01 96 + 2o cortex a substanutial part of AAD has been suggestedGABA -14 1 - 4 .2+tobhe present in capillarN walls (KELLOGG et al., 1973).

G~n ~ l +52 16 IThe present results show, that in hippocampus this

Results presented as imein s i presumed non-neuonal AAD aetixtt) is resistant toI' 0 W2) katIlic aicid

60

Kainic acid injections on hippocampal neurotransmitters 1179

Topographical studies of GAD activity in the hip- the dorsal hippocampus by kainic acid injection

pocampal region have demonstrated that the enzyme reduced the high affinity glutamate uptake in lateralactivity is concentrated in the neuronal cell body septum, but not medial septum, on both sides. Similarlayers and in the molecular layers (FoNNIm & STORM- results were obtained by unilateral fimbrial transec-

MATHISFN. 1969. STORM-MATHISEN & FONNUM, 1971). tion. As expected bilateral fimbria/fornix transection

There is apparently no significant difference between led to a more pronounced decrease. The effect ondorsal and ventral parts (Table I). The present work surgical transection were similar to the resultsshows a large loss (Table 4) of GAD and GABA con- obtained by STORM-MATHISEN & WOXEN OPSAHL (incomitant with the loss of hippocampal neurons after press).kainic acid. Since the loss of GAD was almost total Since the reduction in glutamate uptake was onlyin some cases, we assume that the activity in both slightly higher after unilateral transection than aftercell body layers and molecular layers has been destruction of pyramidal cells in the dorsal part, wereduced. Taken together with the previously reported think that the dorsal hippocampus must be the mostresistance of GAD to hippocampal deafferentation important source of these fibres. After bilateral tran-(STORM-MATHISFN, 1972). these results indicate that section of finbria/fornix there was a significant lossalmost all GAD is present in intrinsic neurons in hip- of glutamate but not of the other amino acids includ-pocampus. ing aspartate suggesting that the hippocampal-lateral

The fall in high affinity uptake of glutamate and septum connection uses glutamate as its transmitter.specific loss in endogenous glutamate and aspartate The regional differences found for GAD and ChATlevels after kainic acid treatment are consistent with in septum are in general agreement with resultsthe possibility that intrinsic neurons could use gluta- obtained by the 'punching technique' (TAPPAZ et al.,mate or aspartate as transmitters. This has also been 1976, PALKOVITS et al., 1974). Thus cholinergic ele-suggested after autoradiographic studies on the ments seem to be concentrated in the medial part,uptake of D-aspartate into hippocampal slices (TAXT while GABAergic elements are concentrated in theef a., 1977) and measurement of glutamate uptake lateral part (Table 4). The increase in ChAT activityinto homogenates from different regions of hippo- in lateral septum after fornix/fimbria transectioncampus (STORM-MATHISEN, 1977), NADLER et al., probably represents a 'pile up' of enzyme in the proxi-1976). These studies together suggest that the putative mal parts of the severed axons which arise fromglutamergic or aspartergic neurons which are des- medial septum as first described by Lwts ei al.troyed by kainic acid could be the granule cells which (1967). Our results indicate that GAD in septum isgive rise to the mossy fibres (BLA( KSTAD & KJ*RHEIM. resistant to caudal deafferentations. e.g. intrahippo-1961 ) and the CA 3 pyramidal cells which give rise campal kainic acid (Table 4) fornix transectionto the Schaffer collaterals (HJORT-SIMONSEN. 1973). (Table 5) or complete hemitransection through theBoth cell types were seen to degenerate after intrahip- globus pallidus (FONNUM et al., 1977). Thus GABApocampal kainic acid injection (Fig. 1). The uptake seems to be present mainly in intrinsic neurons. Thisof glutamate which remained after this treatment, is in agreement with electrophysiological studiescould be due to the terminals of the perforant path (McLENNAN & MILLER, 1974h). which indicate thatwhich arises from cells in entorhinal cortex (NADLER septum contains inhibitory interneurons. Some ofet al., 1976. STORM-MATHISEN. 1977). A part of the these neurons could use GABA as transmitter.uptake may also be into glial elements (HENN et al..1974). It has also been suggested that the commissuralfibres from the contralateral hippocampus use aspar-tate or glutamate as transmitters (STORM-MATHISEN, REFERENCES1977, NADLER et a., 1976). However, rostral deaffer-entation of hippocampus did not produce a signifi- ALEat R. W. , BRADy R. 0. (1959) The distributiom ofcant loss in the glutamate uptake when the dorsal glutamate decarboxylase in the nervous system of thehipocapuswastesed Tabe I Tale ) Tus he rhesus monkey. J. biol. Chem. 23.., 926-928.hippocampus was tested (Table 1, Table 4). Thus the AzmrrIA E. C. & SEGAL M. An autoradiographic analysiscommissural fibres in this part seem to represent only of the differential ascending projections of the dorsal anda small part of the total part of glutamergic asparter- median raphe nuclei in the rat. J. vomp. Neurol. (ingic fibres. press).

The pyramidal cells in the hippocampal formation BALCAR V. J. & JOHNsTON G. A. R. (1972) The structuralproject through fimbria and fornix superior to the specificity of the high affinity uptake of L-glutamate andlateral part of the septum on both sides (RAISMAN L-aspartate by rat brain slices. J. Neurochem. 19.et al., 1966, SWANSON & COWAN. 1977). These fibres 2657-2666.have been reported to be excitatory (DEFRANCE et BLACKSTAD T. W. & KJ-RHEIM A. (1961) Special axoden-

a!., 1973, MCLENNAN & MILLER. 1974a). Our results dritic synapses in the hippocampal cortex: Electron andtthat glutamate could be the excitatory trans- light microscopic studies on the layer of mossy fibers.

suggest tJ comp. Neurol. 117. 133-160.mitter released from these fibres. Thus we found that BaOm 0. J., JR. & FONNuM F. (1972) The regional andhigh affinity glutamate uptake was concentrated in subcellular distribution of catechot-O-methyl transferasethe lateral septum. Destruction of pyramidal cells in in the rat brain. J. Neurcwhem 19. 2049-2055.',, ii i,

61

1180 F. FoNNium. and 1. WALAAS

CRAWFORD I. L & COsNOR J. D ((973) Localization and LoRENs S. A. &GtiL)FRG H C 11974) Regional 5-hyd.release of glutamic acid in relation to the hippocamnpal roxytryptamine following selective miudbrain raphemossy fibre pathway Nature. Lond. 244. 442-443. lesions in the rat. Brain Rei 78. 45-5h

DEFRANCF1 J F, KITAI S& T & SEHIMONu T. (1973) Electro- LWwRN 0. H.. RilsERRnOL GH N 3.. FARR A. L & RANDALLphysiological analysis of the hippocampal-septai projec- R. J (1951) Protein measurement with the Folin phenoltions: 11. Functional characteristics. Expi Brain Res. 17. reagent. J. bio. Chemn. 193. 265-275.463 -476. LUND KARt.SEN R. & FONNIm F. (1976) The toxic effect

DIVAC L.. FoNNiss F & SrORM-MATMISEN 1. (1977) High of sodium glutimate on rat retina: changes in putativeaffinity uptake of glutamate in terminals of cortico--stria- transmitters and their corresponding enzymes. J. Ne-uro-ial axons Nature. Land 266. 377- 378. chemn. 27, 1437 1441.

FoNNum F. 11970) Topographical and subcellular localiza- LuND KARLSFN R. & FoN-4u. F (19781 Evidence for gluts-lion of choline acetyltransferase in rat hippocampal mate as a neurotransmitter in the corticofugal fibers toregion. J ,Veurochen 17. 1029- 1037 the dorsal lateral geniculate body and the superior

FONNUM F. (1975al A rapid radiochemical method for the colliculus in rats. Brain Res. 151, 457 467.determination of choline acetyllransferase. J. Neurochem. MALD P.. LAViNA M. D., RoDREQLLz EcHANDIA E. L. &24.,407-409 MACHADO A. (1977) Distribution of acetylcholinesterase

FoNNIm F 119751') The localization of glutamate decar- along the dorsoventral axis of the hippocampal forms-bosylase. choline acet',ltransferase and aromatic amino tian in the rabbit. J. Neurochem. 29. 729 733.acid decarbox ,lase in mammalian and invertebrate ncr- McGLER E. G. & McGiiEx P L. (1976) Duplication ofsous tissue, in kiferaholic Cornpartmentation and Neuro- biochemical changes of Huntington's chorea by intras-transmission (BtRi S. CLARKE D D & ScHNEirDER D.. Iriatal injections of glutamic and kainic acid Nature.eds 1, pp 99 122 Plenum Press. New York Land. 263. 517-519

FosNa m F (197K) Acetylcholine and amino acids:' trans- McLENNAN H & MILLER J. J. (1974a) The hippocampalmitter candidates in rhinencephalon. in Rhinencephale control of neuronal discharges in the septum of the rat.Neu,tramnmerieur. i Pstchises Sympiosium Bel-Air V. J Phvsiol.. Lond. 237. 60' 624.(PC AJttRIAftS'CRRA I & Tissoi R. edst pp 61-77 McLENNAN H. & MILLER J. J. (1974b) y-Aminobutyric acidMasson. Pari, and inhibition in the septal nuclei of the rat. J. Phy'siol..

FoNNum F & STORss-MATHISrN J 11969) GABA syn- L4ind. 237. 62S--633.thesis in rat hippocampus correlated to the distribution NADiER J. V., VACA K. V., WHITE W F.. LYNcH G. S,of inhibitory neuron% 4cta phisiol scand. 76. 35A & COTMAN C, W, (1976) Aspartate and glutamnate as37A. possible transmitters of excitatory hippocampal affer.

FOs%'ss F.. WALAAS I & lviRSFN F. 11971) Localization ents Nature. Lond 260. 539 540i iAl3kergic, cholinergic and aminergic structures in OtC t I)~l- NOWAK B. NkRKIEWIC-7 0. BIALOWAS I,.It tt.e rnesolmbic -vsem J V.-uroc~him 29. 221 230, DARROWS(A J .WIt RAKo A & GRAOKOWSKA M. (1974)

GUIDOI 7 A., ( Hi C% D L.. TRASI C(-Hi M . DOrrtc-"t M.. The influence of septal nuclei lesions on activity of ace.WANG C & HAWstis R. A 0l974) Focussed micro- tvlcholinesterase and choline actyltransferase in the hip-wave radiation. A technique to minimize postmortem pocampus of the rat. 4cta neurobiol. esvp. 34. 583-601.changes of cyclic nucleoitides. dopa and choline and OLNtN J W_ RHECE V & Ho 0. L. (1974) Kainic acid:to preserve I- rain morpholog Neurnpharmacologis 13. A powerful neurotosic analogue of glutamate. Brain Res-IllS 1122. 77. 507 r512

HATTORI T & %4(,ffrR F G (197') Fine structural PALI(OVITS M.. SAAVEDRA J M. KORAYAsmi R. M.&changes in the rat striatum aifter lalintections of kainic BROWNSTEIN~ M ( 19741 Choline acetyltransferase contentacid Brain Ri's 129, 1-4 lX) of limbic nuclei of the rat Brain Res 79. 443-450

HFt.i F A. Got :iiN M & HAMBRacwR A (1974) RAISMAN G. COWAN W M. & POWF.L T. P, S. 11966)Vptsj1, of the neurotranstiffe candidate glutamate bN An experimental analysis of the efferent projection ofglia NVature. Lo.id 249. 661 h.64 the hippocampus Brain 19, 83 l1tK

HuwsT.Ssni%,s% A 1971) Sonme intrinsic connections of Roms R A & R~is D 1 (1974) Effects of lesions of locusthe hippocampus in the rat an experimental analysis. coeruleus on regional distribution of dopamine-p-J comp Neurol 147, 145 162 hydroxylase activity in rat brain. Brain Res. 73. 161-166.

HiODtI J . JRA L& IoIMAN%i F L j(1964) Basic Concepts SCKWAR(-z R. At COYLE J. T (197 7a) Striatal lesions with

ot Pro1'ahiliti and Statistics Holden-Day. San Francisco katnic acid. Neurochemical characteritic. Brain ARnktit oir~ (. ,i LNsoBom P & RAM'.rf[OT L (1973) Analysis 127. 235 249

of capillary and parenchymal aromatic 1-amiuno acid SCHWARC7 R & COYLE J T (19776) Neurochemnical seqe-decarhoslase activity' in regional brain areas during Ise of kainate injections in corpus striatum and substan-ontogenctic dcselopment in the rat Brain Res 50. ias nigra of the rat. Lift Set. 20. 431-436369 179 !k"HWARCz R, & COYLE J T 11

977c) Kainic acid: neuro-

Ko%ieo J F R & Ki ipt't t.R A (1963) The Rat Brain toxic effects after intraocular injection Invest. Ophtalmol.41 Stereotaiusc 4tlas of the Forebrain and Lower Part~s 16. 141 148.

of the Brain Stern Williams & Wilkins. Baltimore SRUESRO B. & MELLrREN S. J. (19741 Changes in postnatalLEis P R . SHire C C D At SILvER A (1967) Confirma. development of acetylcholtnesterase in the hippocampal

tion from choline acetylase analyses of a massive cho- region after early %eptal lesions in the rat. Brain Re.linergic innervation to the rat hippocampus. J. Physiol.. 79. 119 1311Lond 191. 215S 224 STORM-MArss13EN 1 (1970) Quantitative histochemistry of

LNist 0 & BJsaKU Nty A (1974) The organization acetylcholinesterase in rat hippocampal region corre-of the ascending calecholamine neuron systems in the lated to histochemnical staining J Newrochen,. 17.

A Lr a t b r a in 4 c ia p h s-s ui l .sr a n d S u p p l . 4 1 2 .1 .4 8

7 3 9 7 5 0

62

Kainic acid injections on hippocampal neurotransmlitters 1191

STORM.-MATHisiN J. (1972) Glutamnate decarboxylase in the tions of the hippocampal formation in the rat. J. conmp.rat hippocampal region after lesions of the afferent fibre Neurol. 172. 49-84.sy stems. Evidence that the enzyme is localized in intrin- TAPPAZ M., BROWNSTEI% M. & PALKoVIrS M. (1976) Dis-sic nieurons. Brain Re-.%. 40. 215 235. tribution of glutamate decarboxylase in discrete brain

StIORM-MAriusts J. 19771 Glutamic acid and excitatorN nuclei. Brain Re. lOS. 371 379.nerve ending, reduction of glutamic acid uptake after TAXT T.. STORM-MATHISUN J.. FoNNIM F. & lvtsLsr L. L.axotom'. Brain R., 129. 379 386. (19771 Glutamate aspartate in three defined systems of

Siiiui-MMriisti's .1 & Fii m F. (19711 Quantitative his- excitatory nerve endings in the hippocampal formation.tvoclicilisir ' of glotaimnite decarboxylase in the rat hippo- Absivr. Comnnun. 6th mti Met. Int. Soc. Neurochem..caipal region. J .\,'ir.,ihem. 18. 1105 1111 Copenhagen. Denmark 647.

SriiRM-MAIIIIsl % I & Gi t imi R(. H C (1974) S-Hydroxy- WALAAS 1. & FONNum F. The effects of surgical andirvptarmne and noradrenaline in the hippocampal chemical lesions on neurotransmzitter candidates inregion: Effect of transection 01' afferent pathwa ' s on nucleus accumbens. Neuro~cience. in press.endogenous levels, high atfinit " uptake and some trans- WALAAS 1. & FONNLMt F. 01978) The effect of parenteralmutter-related enz " mes J. Neur,hein. 22. 793 W03. glutamnate treatment on the localization of neurotrans-

SIORNI-MArH's% J & WOMi% OPSAHL M. Aspariate mitters in the mediobasal hypothalamus. Brain Res. inand or glutamate may be transmitters in htppocampal press.efferenis to septum and hvpotlialamus. Neuros Lett.. Yot N-c A.. OSTER-GRANITE M4. L.. rRIMON R. M. &in press. S%'IW)R S. H. (1974) Glutamic acid: Selective depletion

SWANSON L. W & COWAN W. M. (1977) An autoradio- by viral induced granule cell loss in hamster cerebellum.graphic study of the organization of the efferent connec- Brain Res. 73. 1 13.

63 PAPER V

BIOCHEMICAL EVIDENCE FOR GLUTAMATE ASA TRANSMITTER IN HIPPOCAMPAL EFFERENTS

TO THE BASAL FOREBRAIN AND HYPOTHALAMUSIN THE RAT BRAIN

1. "'AAA and F. FoNN- siNorss egian Defence Research Establishment. Dics in for Toxicolog).

Box 25. N-2007 Kieffer. Norwka%

Alvatract The effect, of bilateral transection of the fornix bundle on the high affinitN uptake of gluta-mate and oin the amtno acid content in seseral nuclei oif rat forebrain and h~ptbfalamuLS were studied inorder it) inetigate the possible role of glutamate as a transmitter of these fibres. This lesion decreasedthe high affinity uptake of i -glutamate by 60 70",, in the mammillary bod% and lateral septum. and b\4t0 S0'' in the interior diagonal hand nucleus, the bed nucleus of the stria terminali. the mediohasalhy pothalanimus and thle nuICcus accumbhens The content of endlogenous glutamate in samples dissectedfront free/ec-di ted tissue also) decreased sig!ntficantl ' in these regions. Frndogenous aspartate \%as slightl\decreased in thle anterior diagonal hand nucleusand the mammillary, bodN, but( unchanged tn the otherteiztons No wifiicatnt changes ssere seen in the leselN of Nerine. ' -aminobutsric acid. glutamine andtaurine. ececpt for in increase in gliitamnine and taWinne in the bed nucleus of the stria termtnalis. Thehigh atfinit\ Uptake of I-aminobuts nc acid. tested in the bed nucleus of the stria terminalis. the medio-basal by pot halanw inn nd the mainumllary body" . ss as unchanged after the lesion.

The results idicate that allocoitical efferent, innersating subcortical nuclei through the forms mightus~e glutimnute ;as a tanmer. The study further supports the concept that glutamate pla\s an itnpor-I ant role as t rianstimitt of seSera I di iferent con ucofugal ibre )5stems in mammalian brain

Rtw itN anatomical tudics tn rodent brain base be the transmitter released by these fibres lFo(,,t % &demonstrated that most of the corticoutgal fiblres in Wx A*IALSs. 1 978: WA I AAS & FoN'ct wI 1 979a. Stmilarthe fornix buindle ortcinaite in the piramidal cells in results hawe been obtained bN other \Norker% (SIOR\I-

the Nubiculium and terminate ipsilaterally, in the sep. Mxr ,HtSN & Ot'SAIII 1978. NITS( H. KtiM, SHi%I)A &

tum. basal forebrain nuclei and the hypothalamus OKADA. 1979 7 A(Itlk. Ht t)RiIN & CoN IA. 19791.S" ANSON & Cuss \N. 1977. Nt tHACtt & Sit (itt.. Furthermore. release of t)-[iH4jaspartalc after specific:1977). In addition, some fibres t.%hteh distribute bila- stimulation of the htjpocampo septal tibreN also sup-terally ii) thle lateral septuma arise in (the pyramdal port the %iess that glutamate is the transmitter in thecells itt the (A3-CA i reions of the hippocampus latter fibres (MAI.THtI-SiRFNSSrN. SK~trDI: & FoNNIM.wproper IlANtIt IteSt\. i %\t) & Dt DAR. 1971: Sss S\NsiN 1980%

& 0tW, AN, 11771. File malorityi of both types of fornitt The present studti hats extended these obs~ersattons.libres appear to release an excitatory transmitter, ats In an effort to identif\ the pu~tatise transmitter oif thefound in both the septum 00( Lt\AN & Mti tiR. fibres in the fornix. the high atfinityi uptake mechan-l9"4. Dt It~~t. Ki iii & Sittinm~. 19731. the nu- tsms of glutamnate and ;-amlnohut~rat: IGABAi andcleus accumbens OfIt FRAM t & Y'(IS1HAA. 19)751 and the concentration of seseral endogenous amino acidsthe rostral part of thc diagonal band nucleus lilt ta- \%ecre anal)sed in some of the other target nuclei ofi theitt-Ri. Mitts & SIItiii. 19791. In agreemient \%it thts. forni\ bundle. tn unoperated rat brains and in brainsstimuilattin of thle hippocampus or the subiculumn tn- with a bilateral fornix transectton. The regions arla-creases the metabolic acti' it in the nuclei itners ated 1\-sed included the anterior part of the nucleuis of theby the fornix fibrei IKi tot & PotiIn. 1979). diagonal band. the bed nucleus of the stria terminalis.

The identitN of the neutrotransmitter respons;ible for the mediobasal hvpothalaius and the mammillarythewe effects hats lotte remained iinknossn. tio%%eler. body. as w4ell as the lateral septum and the nucleus%%c bawe recently, demionstrated that destruction of aceumbens. Some preliminary results have beenhippocampal pyraamidal cells or their axons in the presented IFoNNINt . La.ND-KARLSUN, MALI T:-S)R-fornix bundle selectively decreased both the high affi- I NSSI N~. SKRI DI & W-it NA%.. 14791.nit y uptake oif glutamnate and the content of endoge-nots glutamate in the lateral septum and the nucleusaccumbens. thtis indicating that this excitatory amino I \PI-RIMI N-I \1. PROCI i RII S

acid 10t Rt'IS. 1979: CIRTis & JOH~NSTON. 1974) could Maftertils__- - __ ------ i -[ 2. L't] ]gliutatnic acid a 22 5 (u iniol I \%,is Itom Neit

*lhhrownit G \lI. -aminobuiyrate I ngland Nuclear. Boston. and 1'it (1 imnht i

64

1692 1. WALA~S and F. FONNUrM

acid (GABA) (224 Ci/mol) was from the Radiochemical with water-saturated diethylether. and the samples wereCentre. Amersham. England. Male Wistar rats. 180-210g analysed in an automatic amino acid analyser (Kontronbody wt, were from Mdllergaard, Denmark. Other chemi- Liquimat 1ll) as described (FONNUSE et al.. 1980).cats were commercially obtained. Comparison between operated and unlesioned animals

was performed with the non-parametric Wilcoson two-Surgical operation sample test.

The rats were anaesthetized with diazepam and a mix-ture of fentanyl citrate and fluanisone (Hypnorm vet.Mekos: WALAAS & FONNUM. 1980). and placed in a stereo- RESU.LTStactic frame after KONtO & KLIPPEL (1963). A coronal slit Efc flsoso ihafnt paepoe~ owas made in the skull with a dental drill 6.4 mm in front of ELec of -guaaesin ohihifiivutk prc frthe interaural line, and the fornix bundle was transected L-ltmtbilaterally with a small scalpel which was lowered 6 mm In unlesioned brains, the capacity for high affinityfrom the skull surface and moved 2 mm laterally from the uptake L-glutamnale was concentrated in the lateralmidline (WALAAS & FONNUSI. 1979a). septum and nucleus accumbens. with lower aclisiiy

Sample analysis found in the mammillary body, the bed nucleus of ihe

For uptake studies, the operated animals together with sti emnl.thatrordgnlbnd ucts

unoperated controls were decapitated 7 10 days afierlesion, and 0.4-0.6 mm frontal slices were prepared in the ________________

cold room with a Sorvall tissue chopper. The extent of ihe -

lesions were controlled visually, and the regions under cstudy were dissected with razor blade splints under micro- CPA 8600scopic guidance (lFig. 1). The sample from the bed nucleusof the stria terminalis contained both the dorsal and Nen-tral parts of the nucleus and also some tissue from the Aanterior commissure-bundle. The mediobasal hypothala- . /CA

mus sample. taken between the optic chiasm And the pitui- C3 0tary stalk contained the wkhole arcuiate nucleus, the median - TOL

eminence and a medial part of the sentromedial nucleus. F

The mammillary body was taken caudal to the origin ofthe pituitary stalk. The samples were homogenized(2*o w, v) in cold 0 32 mi sucrose. pH 7 4. with ten strokes in LS Cglass-Teflon homogenizers. and preincubated in buffer IpMFcNiiiiam, WAt AAS & IvIt,,,, I1I'll at 21 C for l0 min ) A 7200

[3 Hlglutamate (final conc t0 ' ml and, in some expert- <

ments. [ 'CGAUA (final conc t0 ' m) were added, and kCA ST -. /the incubation %as stopped after Amin by means if rapid C >

filtration as described (Fo''sI. SiuiR~MAtiAMN & .

Dtvm. 10801 The amounts of laibelled glutamnate AndGABA accumulated in ihe tissue %%ere anali~ed hb liquiscintillation specir omet r in aPatk ant I r iar b I81 li td..J.with in absolute acti i% it naltset set lit simid itancous de- ,C

termination (if ['H] and [ r'itnder the .tinttiions used.,r \-

uptake acti itN was propori onal1 to imouni t (i il'SLueIWALAAS & F0%4%1 M. t"JuThe content 4 protein in thehomogenates was analNsed by the I ,wsMethod it OARSN _PRissBROUt's, FARit & R Atinsit i. JI i I i

For amino v id anal %si.. operated and u nlesioned ini - 01 A~ 3200-

mals were decapitated. anid the brain% rapidlv remirsed andfr izren on a microtomei chuck wtith au ( O.-let Time intersalbetween death and iompletion or the frecting was usualtsI 14 min Frontal 40 14m sections from the tel- and dien. i I Skhemaiic drawing of coronal sections from ratcephalon were then cut in a crvostat it l6i C, hophilized brair modified After Koiisa, & Ki iptt t046f31. Outline ofand stored as; described io%%i -A. SriiiMtiss& dissected samples indicated on left side bN hatched region,,WALBERG. 1970 The samples were dissected under a On right side is apprioximate outline of brain nucleistereomicroscope as indicated in F-ig 1. except for the Number"s indicate distance in front of interneural line insample from the bed nucleus oif the siria terminalis. which i n Abbtesiatiiin% Ac. nucleus ,uccumbt-ns. CA. anteriorconsisted of the dorsal and posterior parts of the nucleus commissure. BS. bed nucleus (if the stria terminalis. Ar,only The dissected tissue frrom onie nucleus w*as pooled nucleuis arcuatus. ('(. corpus callosum: ,(. caudatoputa-into 2 Ssamples and weighed on a Mettler mucrobalance men. F-. frms. HI. hippoicampal formation: MS. medialIf0)5 ig dry weight per pooled samplet The aminio acids septum. L.S. lateral septum, OiT, olfactory tubercle: TOL.

were extracted by hoimogenization in 2 V', trichloroacetic lateral olfactory tract, TI) diagonal band nucleus. TO.acid In addition the homogenate was frozen and thawed tractiis opticus. VM. nucleus senitriimedialis hypothalami.once Norleucine ior tit .aminoadipuc acid was added as MU. median emunen,:e. M-R. mamrmlarv body. Mi. main-internal standards Trichloroaceuc acid was then extracted milloihalamuc tract. PC, cerebral peduncle

65

(ilatanate it, . transmitter in happocampal elerents

I -sa\1 I 1iat sIi\iN s\a'v 1 P Alta I 1\5k SI t IN 1,s 9 1 RAt\ Sk( It lITIA itow IM, tOKIa\t IRKA5.St tI0'

(ilutaniate uptake CiABA uptakeI iaoperaied hin' I iaopeuated brains

paiole IIIII O perated brains nimole min Operated brain,Iag protean of Latoperated mg protein ,,of unoperated

\11!Crior diagonal '"0 4 S 49 Ot,hand nu etu,NLIcleus ,ccunabeaist 43) X 1,95) 56 _4*01

I aieral septumI ',I I 4 3 6) 314 5* 12)Bed nuc1dlu Af TIT, 1 341121 56 _* 2i1i 3'

5 1I S 96 154)

ma termainalis\tcdaluas"a 228 4i7 4*1111 31.5 *5.1 TQ 4 15 141

\tanarailar\ hod% 29 7 sS 33 9*t4; 29.1 _3.3 S5t 9(0 1415 1

Rcsults prc'caated is raiaxai- SlUMN niambr of anamal, each a.sayecd ar t arlacatet.Maiistieal olntcace P' tt

antd the nitidaobasoi hy poth laant us Table 11I. A minto DISC( SSIONacid a nal,tl ia ndica ted athata the con tent of hot h glut a-

nMe and 52lutaMI)tie \%a's haghest in the nucleus The results support the use of the hagh affinatyacciamnbern and laltal septuan. follossed b\ the dorsal uptake mechanasm of i.-glutamate and analysts of

part of the bed nukcleus, of the itlta tcriinals. thle endogenous amino acid levels as means of themeduobastal hy pothalaria-. the mlammanjllars bod\ and adentafRing nerve fibres using [-glutamate or i.-aspar-

le anritor dIIA , k0Il bhand i Table 21. Aspa title \&it, tate as, transmitters lYotNGi. OSTFR-GRANt-ii. HtzRN-

mnore cxenly dlit(uited. (i \A %%,as lughl.\ concenl- D01, & SNNDaR. 1974;, Di\A. Fos;\(m & STORM-

trued In ihe dluagotial band nutcleuis and the lateral MAMItsiN. 1977: LLN1D-KARLSF\N & FoNI., 197S.septWliI okmed h\1 thre auircua accuambehns and bed [a iNlI i & WAI *xA5. I 9'78 WAI. AS & FoNNa M.

ii icleti Af tile traa ici niunlali. flo%\ ei. these high 1 919a. Faa'si %it ata.. 1980). Furthermore, they eon-.ltues ittigt bie too lumfa as at result ot the high aetavaty firm Our prevaous suggestion that glutamate as the

thie sA nthetuc etrexi me lttanate decarbox~lase To main neurotransmitter in the neo-and allo-cortacofu-be present I these nuclei uT~stt's,. BttOa'ANStt t\ & gal fibres (FONNI M 1a il , 1979). Some further points

sa aax t/ I1)56. i l ,(~ 11ata. 1 F:IN7. & ,also deserse comments.Va i s s I ( ouaidiab~le aiuitt ol tauraine anad

setII tie C .11rc l0 touunkf inific hsicvions Table 21. Altaldoloqi,ala atsdaiaiai

\tier bilteral forII 1. sct ilota. tlae aicnt of the O ur pres tous studies has e demonstrated that theIIIla0 atharaty *ruait u 111111CIptake decre~ased by\ 60) 70'', amnino acatd uptake mechanisms present in crude su-IT tile iti.1111t111llry body and thle lateral septumn and crose homogenates predominantl\ represents neuro-s 401 'iitaI thle taujletasL a1ccumbens. the antertor nal uaptake Ires aew FoNNI M t t a.. 1979). Ho~hcser. the

111ol thle daaarotaal band naicleus. the hed nucleus Of Uptake mechanism for i-glutamate does not das-lte stria acirntirih nanlstd in the mediobasal h ypotha ha- t atgatash betwseern puatatis e glutamate and aspartate-

Tim, i [able Ii. The uptake of (i-BA wshich waas using fibres (BAICstt & JOHNiSTON. 1972: Lum1-a-sayed itniultaneouasl\ ssitfa the LItatnate uptake ant K ARL.SIN & Fovsi mi. 1978; iNI Fo.ate ti.. 1979f.

soatie espratnents did not decrease significantly in Selectase changes an aspartate glutamate levels hase.

AMa of1 tfae reattOta staudied. how~eser. been found in both the neostriatumn and tha-

4 \"01 .? "ifw o' C111(40114. WIin acds lamic nuclei as well as in the lateral septum and thep a s a,' i t at a atal saaoias tatitatatid nuacleus accumbens after deners at ions 1Kitwl H ASSLI-R.

I lie lesiot alsa tiduced changes lin the endogenouas HAtIG & PAtK. 1977: LLVNi-KAR1Sl-s & FoNMMx.

a111a11 acuds Glutantl concentration decreased sig- 1978; FtaO' t \ & WALAAS. 1978: WALAAS & Fo'sit mi.aiicaintl\ In all reagiotn,. most heastily in the septuam 1979a: FoNN~t \it ohi.. 19801.

i 4o4$'.id feast pronouanced an the diagonal band In sonic studies 'AAA5 & Fo\N't 5$, N79,11,IS \spirtate oncetatratton decreased signifi- focuased microwase irradiation was used to achtese

-aitly itt the diagotnal banid and mammillary body. rapad tertitnation of possible post-mortem changes tnhilit not at the cepturn. bypothalamats. naucleuas actamn- the endogenous amino acid concentrations. Howe-.er.betas Atnd Tilie heal TRtCiclus of Nlrta termanalts (Table 31. an prcltmtnarya studies with samples from the bed na-

Ohlttatie and taaraine caticentration wsere signafi- cletas of the stria terminals. this fixation method wissint l\ anic ased an the bed tauIcleus of stria termanalas. fouand unsuitable for the present study. Instead. tt ssas

bai aitalainged lit the other nuLClet. Scrane andi 6ABA decided to tase rapid decapitation followed by freezingoitentrations siere uanchanged art afl regions wath a (0:-itl for the amino acid analysts. ThisITIable I I method. wshich allows preparation oif frwee-dried

66

1694 1, WALAAS and F. FoNNUM

In r4140 - oc-H +1 -H +1 +1 +

- e 0. 00

+1 +1 +1 -H - +1 +1 +1 +1 +19 =00 en 8

r4 .0 C4

o 0 0 0-

z V14 0' O

14 0. E0 -'

- 0' N - (N 0(N 0<a &

z +1 +1 +1 +1 +1 +41

+1 ~

-H + +1 +1 +1 +1 +1I m - c00 0 ='0o 3 5 1 + + t +

00 ~ 0 .. z - A 1 - - 4 -

E

- ~ --E

z .0 N .

67

(di lunate as a trasnsmlitter in hippocaripal ellcir, i-t

Meioi id thus1 lacilItaics the precise dissectioni of matumillars bod% tFi NNLt M & WASAAS. 197h. S1(io1

tie atrnplcs it wm i 51 it.. 1970; 1977). is not MAt OSL-N & OitSAI41. 1978. NiTS(J !i ill. 9

expcted to influence greal, tie concentration of glu- "A ALNAS & Fo*%Nt w 1979a. Z( /I K0 cif.. 19 91. andimate itc the concenitrationt of this amino acid somec aspects of these glutamate projections has e neen

.tppc,t, to be relatisel\ stahle in the rat brain post discussed presiously (WALAAS. I9NOri~h. MALtii -Tfltt citi ku\ 1At i NiiS& Nh *rR1I1ii . 1976). Host.- SA il ssN Omit N & VbALAAS. 19801. In contrast, thecs ci It is 11,1 possible to ascertain the GAHA concen- identification of glutamnate fibres to the anterior drug-na~tiont rciehi \kith this method. dute to the rapid onal band nittleu-. the bed nucleus of the striapo't-mlortenr) Increae III the les el of this amino acid terminalisoI :., triediobasal hypothalamus desersesB-i , ota Li \o\I & 1 Nh~I KilO1It. l)7S VAN Di iR Hi-s - some come., tit-iii %. )i Ki or I. KoRI & V1 itS111i . 119791. The GABA The iuca' 0.; Me thayonal hand. This contains aleseis reported in the present work are therefore prob- dense collection of acetylcholinesterase-posttiseibl\ higher than the anmount of 6ABA present in the neurons IJACOHiiWTI & PALKOwn?]. 19741. which

tissue mi iii,' iBAI(OM~ et alt.. 1975; FoNNUsi & project to the frontal cortex. septum and hippo-%\AxS. 1978: t\xtIIAS & FrINNINI. 1978: 1979a). campus (DiX-AC. 1975; - ALTHI-SR-NY,1\ N illci..

liossesecr. its all brains ssere treated similarls. com- 1980), The present studs indicates that this nucleuslmrisori heisseeni the (i-BA concentrations in corre- might receise glutamate fibres from the hippocampalspoirdtig1 reions1 froim uno1perated and lesioned ani- formation. Howeser. these fibres apparentl% terminate111.11s 1, 1t thath peisivible. Yhe lack of significant only Iin the rostral part of the diagonal band nucleus.Jiaiuius iii 0j-\B lesel agrees %srth the Lack oif effect as the glutamate uptake in samples from more caudal'I] 11ic it" Xi\ aplike parts of the nucleus was unchanged after fornix lesion

(results not shown). Thus these results are in agree-lh-nir 1I 0 1)),fl~ii In fit' foril 11fires mient with electirophysiological in sestigations in the

I lie nutclei trial ' sed in tire present study all seem to cat brain, which have shown that excitatory fornixreecise excitators' allocorticL. fibres through the for- fibres terminate in the anterior but not the posteriorIII\ atnd or for ni superior, ats judged from anatomical diagonal band nucleus JHHBRARD er uh.. 1979t. Theaiid lh~ siologreal studies iRsss ss. Ci iAs' & PowX- identity of the diagonal band neurons receis nv thisIll. 1166.; ~SkAS-s & ('is-ss 977, VI & glutamiate input is uncertain. HowNeser. destruction of

~i i I- ~i.-Nsii sistiat. 9~: i. i'N~'s& hippocampal and subreular neurons changes the

%IrII iI. 11974. f it i sIr ifci.. 11191. t-urthermore. aetylcholine lesel and choline uptake. but 110t thetie nuci drsplai\ a ser\ high concentration tif gluta- actis t>. of' choline aceisltransferase in the hippo-nraii. " hereas aspartate did not correspond to more campusi(HwAR(t.. ZAC/FK & COsaIJ. 1978: ZAC7iFK

thin 22 35. , of the lutamnate concentration. It is ei al.. 1979). This may be explained by an excitatorythrreitrc Tnipttani that the cinl\ :onrsistent and sig- input from the hippocampus regulating the actisits ofrirticant ierirocherical chainge in all theoe nuclei after cholinergic cell bodies projecting to the hippocampus.3\0101Miv Of tire: f~liri fibres %Na, a cnsiderable deC- These cells originate mainly in the nucleus of thecrecase In hothI the h1ih aiim rt\ upt ake of - marate diagonal band (SHrITr & Luss is. 1967; MAiirrn

a rid lin the ii1U seeticen1Iti at ion k i gIiitamlale In t he SORt.I NssiN 0i ci1.. 1980). It is therefore tempting to

dragttri cl bhirid nu1Ccus and in iht- maninullar\ bod~ a , suggest that some of the glutamate fibres reported insiriall sl'vccrei In ire lc% el tt a-spartate kkas Aso seen, the present stud,. may be responsible for regulation ofI tc tilier inluilo acids. ineludintg (iA BA. and the the eholinergic cell bodies.

i B kt Uptake ACtIs r did not decrease signilicanihi 7 he bed nucleuax o~j 1he vrio enaas The resultsdicl thi si,10in IItn1s. region. onl this nucleus expand the alreads considerable

I lisi. flutian r te i e ton i amri no icid t ransmi tier in umbher of transmitter candidates present in this nu-

candidatec coirsistentis% associated wkith these corticofri- cleus. Presiotis % trk has demonstrated high concen-

thee al"'rt~al ff.ren, rsethi amo aid s dcaro~vaseinthis nucleus )Bt s-ARt. Kk\.A/ 555 A

Itar~miiciMinr cntinent ofasprtae fbres & Z(;m Ni. 176.A A AAS & 1-0s'M Mi. 19701. and

niairiinillai\ bodxIlse r the sers small absolute arrise through the saria terminalis iLv6Ai LASALLI,

decrease III aspartate Irsels alter lesions compared to PAXIt'St. I-M'AN & Hi N-ARt. 19781. Due to the prox-trte dccise in geliaate lesels indicates that the imit of thle timrbria forii, and the stria terminalis. we11iitic Ahe tl ititis e aspartate fibres probabiN is, \er\ fotind it diicult ,clectis el\ to lesion one of theserall itliernalisel> the decrease in ;isparnate lesels bundles without affecting the oither IW~t kt's & Fo's-

riiieiit he a1ir itispecilic effect foiltiss g thre glutate Nt %i. 1974au) llo'~eer. tt evidence for a destro~ eddener sat ion GABA input. i.e. a lesioned stria rerminalis was

found in ant, oif the animals used in the I-p esent study.!'o'sihic fioh iionail imphic tnofls as measured by GABA content and 6ABA uptake. In

thlese ettircirisions estend presioris studies oin the contrast, the glutamate content and uptake decreasediriprio theIfi sepirini. thre nucleus acrimbens and the significantly in all animal%. Most oif the glutamate

1696 1 WAI AAS and F. 1-oNNI, m

fibres reaching the bed nucleus therefore appear to demonstrating selective deceases in hoth the gluta-traviel in the fornix. However, it cannot be excluded mate uptake and content after transection of the for-that some glutamate fibres also could arrive through nix fibres it) tbe h~pothalamus indicate that thesethe stria terminalis. fibres prov ide it major part of the glutamnate inner-

The hfvpothalamus. Most of the allocortical fibres %ation in the mediobasal h~pothalamus. Thcy mnasfrom the hippocampal formation to the hypothala- thus be important in the expression of endocrine funic-mus, which travel in the medial corticohypothalamic ttons initiated Ili the hippocampal formation (A/Mit I.Itracts, pass rostrocaudally through the hypothalamus & CONRAD. 1976. Kmta & HAiaS. 1963: Vuil AS(()

between the arcuate and ventromedial nuclei (SWAAN- & TATt ISNis'. 19691.SON & COWAN, 1977: MtFiBA(H & SiEFjAL. 1977). Cmk uinNeurons in this part of the mediobasal hypothalamus (~lsoi

are sensitive to glutamate-application IGFLLER. 1976) Thc present study indicates that the fornix bundleand newly synthesized glutamate is released from this in the rat hi am distributes massise projections ofregion in 1ico (MFiEKER & MYERS, 1979). The neurons putative glutamate-containing fibres to the septum.in the arcuate nucleus are very sensitive to the neuro- the basal telencephalon and the basal b~pothalamus

toxic effects of the parernteral administration of gluta- . , few aspartate-containing fibres mas also enter the

mate (OLNEY, 1979: WALAAS & FONNt M, 1978). and anterior diagonal band nucleus and the mammillar 'destruction of these neurons leads to gross endocrino- body. These results thus expand the already, consider-logical abnormalities (OLNMY 1979: RtuotIN;. able list of corticofugal fibres vwhich probabl use anSCHALLY. ARIML'RA & WAKOBOYASHi. 1971 * Nsosi- acidic amino acid ats neuroiransilitter I -iiN M xi al o..

SAWA. YANAt & KiKUYAMA. 1974). Also, application of 1979: .s 'N AS. 1980h).

non-toxic doses of glutamate or glutamate-analogs tothe mediobasal hypothalamus leads to rapid changesin the release pattern of the pituitarv gland lOt siN. 1, k Im,hqcme, We Thank liv lIi ss N and vIs i11R

1979). Thus, glutamate-neurotransmission mas be im- SI -, for technical assistance. and 11 J K \vs i -s for prep-portant in hypothalamic function. The present results irilion of fitiire,

R1tI t RI N(t[ S

ANDIRSIENP.. Bi 4,,t) B H & lt IOAR S DI) N1) trganreation ,(t the hjppo,.:oIipa l up1t 1,ip Biui R- t7. 17 1"AiMITIA E. C_ AR & (NRsAII 1, ( A 19761 tempotii effeli, "I lollis iraiivei1iion 0on bhi, puihiiisrss1.

activity and plasma corticosterone levels N euros'uidiv rinolsii 21, 1,1s 341)

BALCAR V. J & JOtiNST(N Gs (A R it1972) The st ructural pecitio iN of the high itti pi ike oI I -glIiii amri - and1-aspartate b3 rat brain slices J Neur, him 19, 265", 266h

BAiCOTM G. J. Li o\i R HI & Mi-iiluiott 1915 Regional -arnim u~ne acid Ice, li imia bai det il icd t

microwave fixation J Veuriit hei 24, 6091 613BAi~comG L. LF'six R It & Mi~iot i J101 1 097161 Regioinal yoila.itc levels Ini rat brain deter niiied afitr n1ir.uo ,uiC

fixation J. .. euroihen 26, 423 42581-N-Alu Y. KANA/AW A J & tilSIIIsni R F 119761 Regional iiiniriin if ' liniait el iisve iind G\\ % vvthin

the amygdaloid complex and stria termninalis svvtem of the rnit I %isrsiin 26. t1271) 1281CURTIS D. (19791 Problems in the evaluation of Olitarnaie I, a central linsr0 lillCn vlte tansmittert Ini glitai acid

.4di'nces in hiochi'mistr jndphtsi'oviiii.(1.d% VillS 1. 1 IR.G ( i ii NI S,- KkSM \1 K A RIkN ilsV & \I KiM i vR. J.t. pp. 161 176. Raven Press. New York]

Cu 'As D, R, & JonNs ro.N G. A. R. 1) 974i *Am01tl L)d tiianv~itiers In the imo nniili~in :ctit i ,iJ nersou ii'v tnt fw i1,n

Phnyviol. 69. 94 188.D1-IRANCF J. F. KiTAt S. T & Siwoisui 1 I 197'11 1 lctriiph~ vriilogicat ,inakil 0t lie hi~~.atplstilpiicIIIv I

Response and topographical characteristics IFsp Braiui R,- 17. 4-47 46_2DFFRANCI J. F & YOSHIHARA ti I10751 I:Imbria inipt I,, the nih leir avcimben'vepi IB,i R, 90, 1541 ]0DiVA- 1, It975) Magnocellular nuclei of the baval fiircbrain proiiect To neocionies. brin lerni inid iilfactiir\ bulb Rev ess of

soefunctionat correlates Brain Re, 93. 195 148DiSA( 1. FoNi SI F & STRoMA iAiIIsi 1 111)7t 9~i igh allinits uiiiaki if glrjiamie Oil tiITfliialv ot ci eotniIlit'i I ,fl

\urture. Londi 266. 377 378,

FoNNiVM F. Li il-KARI Mi% R . MA nu II - siiti %Ni . 1) Ski lii K K vk Vci vvv I I It)')I I i'Ldli/.liiuiii o 01il OttliS.11

mitters. particularly glutamate, in hippoucanlpki. wsptw. nii.ciiv asliiliiind upc -~ - ihoile, P-sj Hiam ,,i8Vol St.

1 0NNI M f'. STORM AlitI\ I & D 1)5 i I INKIli Hu iu.)tIii,.11 1'11 ,0 01 0 i m - , 1, 1,1 ill ..i 1C 1!1 ih

corticostriatail and corticothalamuc tibres in Iat braiin %, I, ,, n, , a,,illFiNNi'5 F . StoRM-qMAiiiiis J At WiA FIT", - iI19'I0 Gilaiioi AI~,. ,\ .u \ i it 0 111 % iirv~, '11vi, ot Irs',

rnlyrne in Purkinic cells .ind) boiion, in the LI B,m R. , 20. *' 51

FONiM F & WAL AAS 1 11978) The etfLls of liltiallu 1''On (m.11 1~i, 11, 1 '. '1 i-J I] ol. 11 1,rri "1' ic..?'

mit ters in hippocampuv and sept11 m .1 N r,, I. m It. I I -I I I X I

69

MiN N I. Vi \Ni 1-' 1% 1 Rsi k , 1 (1 1 1 alu.aiitn ti\H Icigi, hinrc damergic structure, in the

(ii IIk It M\L I'I ,I 1. feel of Nomxe pwi,e nciir,,l ai -.niiic',' 11) ciix AIts ol.(f txihercle h\ poilalamic neurons Ii rn,13a R, , 1(OM. 4231 4lii

[It i iiiii J. I. \Ili is R 0 & SiR! i I I 119x)1 Reipon~c' Ili thw diagoinal hand A~ Broca exoked b% itimulation of theoriii i tile cat. .1 Pi'l wxl liiin 292. 231 2491

i mi (Inm Ii I ) \I N. si \1 KI mI I / \1 119-.4) 1 opocrjphi: miI 't, c.itcholarri and acet% Icholinesteraxe-contadiningictirir, Ili tile iii hrtiii .1 Ppip \ 'l 157. 13 IS

Kim .1 S . 11 xxxi I R .s m P) & N'il, K S i l I 1-11c ol Ifitta ahlaion onl striatal glutlamic acid Joel in rat BrainRx \ 13. 1' -.4,

K I ii \1 & Poll II Ci I I '-19 I II ppicarn palI aftcrd isch arges d I ffereti thalI spr ea of act I it ho'A n b% t he [I "C]deox v-ghLic,'Cci liiiii S, i, , I1 , 1 2014. (,41 (,.4

K \,i ,i K \1 & JIi, \1 11) ,'I I x lcii I inihiiix c toeII II hppiicjma II IIn neural I-CLilatiin iof X.(TFI release

.I R & Ki IIm I R -\ 1 196"1 1 h, Ri:Ba 3'If ~ 1' 'ix Iti' of t, I-c, hrainm lLiO, P'art, ,I ilw Brainmi lI ilant & Vi ikAm,. Hilt iii i

I i0 \[ I \S\I I I ( .I'xix( I N'i N 11 & Hi1 \--'k \R1 i 'I N NcUIOI)~CcmIicl mnappingz of (; XB~ergi, xsteni, in the.11i\gIOL c1i1leriic i !cIl iIICI1CI "I the iria i ctiiili Bait R, , 155. N'- 40'3

,\iikR () 11I R, , ili,,i .it I I \xR \ I &%_ Rx\iixi I I IL045 11 protecin incaincment, ".ith the (-oliti phenol

II CexuI I ': i , 11 , " 191 'i

I i'-K ki i I, R IS I '\\i \I I 19l N, I .il I 'll eliltii.t1C Al ,, n)CIiroTIrinin1ilter In the caomicofligal fihres t,, thetaI ;itci~il Acicit ,l tld IItlillCII cxxp icljlikia III JAIN Bru01 R, 151. 45;' 46'

Mm ]III 1s"ki \-I\ 1 ) ()1,im \ I & \i \1 \\s I i l'i SclcctIic xlcxtiajin III kainic acid ,if neurons inimrated III

piiiixiiiitital t tx ,t"ICIttxII ',It :n int mi ika .f tie iiie,'nl hband Bunm R,' 182."\ xi1 5 iNx ) SkPin H K K I, Ii i'Nt %I I WiI~Ili (iliini-dcpendeni release oIf D-[ 'H]axpartate from the

iidxolitLcral x 'tei lict cci lxtiliiiII~r III lie tibim H? i- Si, ,ir,, i, . 1-'- It3

M[I IN\ \\ 11 & \111i Ik R I ; -4 I lie litil,.x-itIIIp,11 ,11Ii01 ,I n1Citr0iial Iicharge Ii thle xepliumi o the rat .1PIitlV-0i 231. '(1, W24

\It I ki HR & \IxiK R 1 ) i1-j Fil t''I imiii,' .ic,I 1ii01ilC, III dixcicte hi parhalaiic legioins diiring. pkiih-pIlt

\It aIxI if I? ( & 11 -I I,. \ I 10--) 1 !lct 'III ,Iince'i, Oil the liippoc.iinpp.0 firniaiiin in tile tat Bruit R, 124. 1tQ- 224N,, x Ix xI. N\, xI xt NX Ki1,ii)\ \ xx s I 1i, IcCrihcII nhtlxiii1Iii 01 fItmiat pialacting I, atid ith hormone

Aiid ItIIIIInI~II 21.1111

&Le lIietI1 't Mile I niono IM ,i,' dm i ltill.i1e AinIIIir'tered netidill 1, 1-

\xIx INi Ix .IK . '1i1\1 xii(\1 Iki xi Iltic Ii hippixinp110 CiT c Piiipitiii In tle I,]( on guilian' ie lcxci in

\\ ' I 'j4' I ix"\ ,t11,'' 1 ,1 c'tiillix'c jinl x'iieC tll,l u , ii,x ,in tile xlecloxping central netwu xo x xtenl IttItIl '1 0i. J, , i , I:1 , a; W~, led:c, \\li.iit VI, I I I,[ (Ifl Xl I Viilec & soti'. Nexx otak

11..' V, I I. I'' I -r .. \R' \ I I (x im iI s K x\i\I 1HR. Hixiix N\, i \ & Viikti\ K Ji ppl

i;' I kkIa i I I \Ii xI ii I R & I (,I liiik Ri H i \,:tetile eliiite-iniicd celxatii'nx Ili cerumn

P .\\ix i I,\x \% \1I, P. xI 1 P1 , i 1,l \ii xxpx"iliiixtiii i II til e etlereniI c0iniicctinx MI tie

It . 89. 11,1

H ii I N, I I. I xIi \ktt Hi I, & VI ikillixN \,Ili I 1 lil1 1 tI ic L!11,1iI Oniiisii eliaiacIi ~nIC

Si,~~ kHt\ F &lii I I I Ii Vi IQx \1 IIi~ ,;il 11011lIN ii,iti nJi hI WOe raxn.tet Iit IAtipIxcma I leet W i

ilx, i\ I 1 X I ix P iN 1\ 1 \f i 'th iIdw hioiiiilgiiii iit 'Iteiginigct la \te i niixitt ,i. th leeti aotnecd n' i

l ?, .i ~xi , .1118 - 'I l

xx~~ ~~ "\ iIi i 1\ \ II\ Kis \i I HQ- Kit Hi1 .I hi i Hiltile' I) II ii iix of -\ he eiiit , iret c rnncisf

hiiatot Il IIa'' ,i,\ o, 7.4

P\I~m/ 1 0'1 i~tihuton f ilutmmil dicaro\\~i~cin fisrel Na

70

1698 1. WALAAS and F. Fo)Nwum

WALAAS 1. & FoNNum F. (1978) The effect of parental glutamate treatment on the localization of neurotransmitters in themediobasal hypothalamus. Brain Res. 153. 549-562.

WALAAS I. & FoNNum F. (1979a) The effects of surgical and chemical lesions on neurotramamitter candidates in thenucleus accumbens of the rat. Neuroscience 4, 209-216.

WALAAS I. & FoNNUm F. (1979b) The distribution and origin of glutamate decarboxylase and choline acetyltransferase inventral pallidum and other basal forebrain regions. Brain Res. 177, 325-336.

WALAAS 1. & FONNUm F. (1980) Biochemical evidence for y-arninobutyric acid-containing fibres from the nucleus accum-bens, to the substantia nigra and ventral tegmental area in the rat. Neuroscience 5, 63-72.

YOUNG A. B.. OSTER-GRANITE M. L., HERNDON R. M. & SNYDER S. H4. (1974) Glutamic acid: selective depletion by viralinduced granule cell loss in hamster cerebellum. Brain Re. 73. 1-13.

ZACZEK R.. HEDREEN J. C. & COYLE J. T. (1979) Evidence for a hippocampal-septal glutamergic pathway in the rat. Exp.Neural. 65. 145-156.

(Accepted 5 May 1980)

T

71 PAPER VI

rain Reearch, 153 (1978) 549 -562 549i' Elsevier North-Holland Biomedical Press

THE EFFECT OF PARENTERAL GLUTAMATE TREATMENT ON THELOCALIZATION OF NEUROTRANSMITTERS IN THE MEDIOBASALHYPOTHALAMUS

IVAR WALAAS and FRODE FONNUM

Xorwegit, De/etce Rewarch Estabblshuem, Dhision for Toxkology. N-2007 Kjeller I Norway)

(Accepted January 18th, 1978)

SUMMARY

The localization of cholinergic, aminergic and amino acid-ergic neurones in the

mediobasal hypothalamus has been studied in normal rat brain and in brains whereneurones in nucleus arcuatus were destroyed by repeated administration of 2 mg.,gbody weight monosodium glutamate to newborn animals. In normal animals acetyl-cholinesterase staining, choline acetyltransferase and aromatic L-amino acid decar-boxylase were concentrated in the median eminence and the arcuate nucleus.Glutamate decarboxylase was concentrated at the boundary between the ventromedialand the arcuate nuclei, with lower activity in the arcuate nucleus and very low activityin the median eminence. Nucleus arcuatus contained an intermediate level of highaffinity glutamate uptake.

In the lesioned animals, there were significant decreases in choline acetyltransfer-ase. acetylcholinesterase staining and glutamate decarboxylase in the median emi-nence. whereas choline acetyltransferase activity and acetylcholinesterase staining, butnot glutamate decarboxylase activity. were decreased in nucleus arcuatus. Aromatic L-amino acid decarboxylase was unchanged in all regions studied. The high affinityuptakes of glutamate, dopamine and noradrenaline, and the endogenous amino acidlevels were also unchanged in the treated animals.

The results indicate the existence of acetylcholine- and GABA-containing ele-nients in the tuberoinfundibular tract. They further indicate that the dopamine cells inthe arcuate nucleus are less sensitive to the toxic effect of glutamate than other celltypes. possibly because they contain less glutamate receptors.

INTRODUCTION

The systemic administration of high doses of monosodium glutamate tonewborn animhls leads to neuronal necrosis in several brain regions, most consistently

72

550

in the nucleus arcuatus of the mediobasal hypothaiamus5,263-'. In this nucleus. whichthrough its connections to the median cminlence

17 constitutes ani important neuro-

endocrine center. 80 90", of the local neurones are destroyed after administration of 2mg/g body weight of glutamate (GLU);17. In contrast, glial cells and axons passingthrough the nucleus are left intact, and the neighbouring sentromedial nucleus is

completely unaffected'1 . The (LU-induced lesion is thus of' remarkable selectivity,and seems %kell suited for studies oft he anatomical and neurochcmical organization of'connections in mediobasal hypot halamus.

The aim of the present investigation was to utilize the toxic effect of GLU as atool for studying the localization of putative neurotransmnitters in mediobasal hypo-thalamus. Choline acetvlt ransferase (ChAT. E.C. 2.3.1.6). glutamnate decarboxylase(GAD. E.C. 4.1.1.15) and aromatic [.-amino acid decarboxylase (AAD, F.C.4.1.1.26) were studied in microdissected freeze-dried sections from rat hypo-thalamus. Acetyicholinesterase (A~ihE. E.C .3.1.1.7) w~as examined by histo-chemical staining. The relative abundance of noradrenaline (NA) and dopamine(DA) fibres in the arcuate nucleus and median eminence %% as csti mated by measuringthe high affinity uptake of these aiins. [mlls .N thle locaizationl oft putatis e amino1acid transmitters w\as investigated by measuring the high aflinity uptake of G LU andby amino acid analysis. Part of this ss ork has been presented in a preliminarsN form""0.

MATERIALS ANtD METHODS

[1-1 'Clacet) l-CoA, dans~ l-(dinicthyl-[;'H jamnino f-chloride, I -["Il I1glutamic acid.t-.[7,8- 1H I1noradrenaline and 3,4-[cth l-2-1H ]dopamine sserc obtai ned from New\England Nuclear, Boston. I-[I )tlglutamnic acid. mi-3,4-dih~droxy-[2-1iiC ]phe ny Ia a-nine and I 1C-labelled amnino acids ssere from thle Radiochemical Centre, Amershamn.Sodium t-glutamic acid anld Unlabelled da ns\ lated amnino acids were from SigmlaChemicals, St. Louis. Polyamide plates (V 1700) were fromi Schleicher and Schijll. D)-2433 Dassel. Nialamide h~drochloride ssas a gift from P1fizer. bei;tropine was a gitlfrom Merck. Sharp and l6ohnie. desmiethx -inlipranlinc ssas a gift from Ciba-(ieigx.

and ethopropahine "5as from Pharnia Rhodia. Niet hi lisoc~clopent Ifluorophosphate(sonlan) %%as synthesized in our laboratorN. White Wistar rats ssere obtained fromDyrlaege Mollergaard-flanisse ns AWskiboratoriuni. De)n mark, and allow ed to breedii our animials quarters.

Tiss'ue preparationMonosodium glutamnate was dissolved in distilled ssatcr and thle solution

neutralized to pH 7.2. Ness born rats of both sexes w\ere giveni daily subcutaneousinjections of' 2 mg 8 body sseight during their first sNeck of life. The animials were fedr normal lab chow. and water ad lib. and housed together s, ith uninjected littermates.which were used as controls.-The animials \Aere sacrificed at 6 9 weeks of age. Forenzyme studies, the animals were decapitated. the brain rapidly but carefully dissectedout, placed onl a microtome chuck and frozen with a CO 2 jet. Serial transverse sections(40 pm) from hypothalamus %%ere cut in a cryostat at 18 C. freeze-dried and stored

73

551

I1ig. 1.- A: cell -stained 40) tim sect ion from medI basI h p I I mul~s of normal rat. Dissection offree/e-diied sample,; is indicated on the left par 0 fi 1, mein eminence. 11. nucleus arcuatu,:III, 'lateral arteCIC I V. n uc li %ent romedia, \l, n ucleus dor sonled.ialjs BI: section from G I.. -treated rat, demonstrating hea\\ cell losin the arctiate nucIls. tDissected samples asin A. 25. Ma%

(I n nsald (;iemnii stain.

-pre\viously described2- For histological staining. 20 40piscon%\rcletdon mijcros;copic slides and allowe Io dry at room) temperature. For amnino acidanalysis, the animals Nsere sacrificed b\ Cxpositig their heads to a bcarn of focussedmicro%%a~e radiation in a Lit ton 70 50 Oven for 3 cec to rapidly inact i'ate all en/yrnes

andl aw id postmortem artifactsi11i. The head, '.\ crc then rapidix\ cooled oin ice and thehrain carefullN taken oul. Byv this procedure. it proxed impossible to preserve themediatn eminernce. but microscopic inspect ion of transws ei slices, demonstrated thearcutate amid %etitroniedial nuclei to be intact.

Iisw 'Ciol0

The freeze-dried sect ions \%ecre compared to corresponding cell-stained sectionstinder ;t stercomicroscope. and the different repions \\ecre identified after the at lasof K~inic and KIi ppe[21 " .the regions I tken 01mm for anl\ sis iocluded thle miedian

etn inence. the t11IL1 ucles acLKa us. lie boundarv bet scceethe arcuate and the \entro-miedial niuclei (named 'lateral arcuate'). the \entromnedial niucleus and the dorsornedial

FIJCCA (ig. ). The samples (0. 1 2 ,og dry sseight) Nr \%cighed on afs-ocqatIi her balanCe2s and assayed directl\ for emiyrne activ ities.

For uptake studies. the brains \ere placed \\ilb the \entral surface up. and themedian eminence \-.,s dissected vsith raior blades mtd smiall forceps tinder a stereo-microscope inr the cold room according to ( nello et at.,;. int addition 1o median

- . . .. . . .- -- --

74

552

Fig. 2. Cell-stained sections from normal and GLU-treated animal, demonstrating the selective neu-ronal loss and shrinkage of the arcuate nucleus (ar). Me, median eminence. 75, May GrunwaldGiemsa stain.

eminence this sample contained most of the arcuate nucleus and the proximal stalkeminence. Samples were pooled and kept on ice until homogenization.

For amino acid analysis, the hypothalami of the cooled brains were frontallysliced with two razor blades, one at the level of the entrance of the optic tract into thebrain and the other at the level of the stalk median eminence. From this slice, atriangular sample (Vig. I) containing the arcuate nucleus and the medial part of theL ventromedial nucleus was taken.

Hi.stological preparationThe air-dried sections were stained with a routine May GrUnwald/Giemsa

method for cell staining, or for AChE by the thiocholine method as previouslydescribed5'-. AChE sections were incubated for 6-24 h. routinely using ethopropazine(10 M)i as pseudocholinesterase inhibitor'. The staining was compared with sectionsfrom animals which had been pretreated with a lethal dose (0.2 pg/g body weight) ofsoman and killed 15 min after injection, thus inhibiting all cholinesterase.

Biochemical nettodChAT was assayed by a previously described micromethodt2 . GAD was assayed

by a CO,.-trapping method as described by Henke and Fonnum'7 with a finalconcentration of 20 mf GLU, using Triton X-l00 to ensure penetration of substrateand to minimize non-GAD-released CO,2r. AAD was assayed as previously de-scribedt. with final concentration of 0.3 mA .-DOPA. Protein was determined asdescribed by Lowry et al. '-' . with bovine serum albumin as standard.

75

553

For high affinity uptake, the pooled samples were homogenized in 40 ,l 0.32 Msucrose at 800 rev/min in a glass-Teflon microhomogenizer. For catecholamineuptake, homogenate containing 30-40 pg protein was added to 0.5 ml Tris-bufferedmedium which contained 10-4 M ascorbic acid and 12.5 pM nialamide hydrochloride.Benztropine (10-6 M) followed by [3H-]NA (10- 7 M), or desmethylimipramine (10-6

M) followed by [3H]DA (10- 7 M) were added, the uptake performed at 37 'C for 5min, and tcrminated by Millipore filtration as previously described 3 . Benztropineand desmethylimipramine inhibited DA or NA uptake, respectively. The GLUuptake was performed with 10- 7 M GLU at 25 'C for 3 min in the same manner.

The amino acids were analysed by the dansylchloride method with 14C-labelledamino acids as internal standards, as described in detail by Lund-Karlsen and Fon-num 29. The dansylated amino acids were dissolved in acetone-acetic acid (3:2) beforescintillation counting. Comparison between biochemical results was performedwith the Wilcoxon two-sample test.

RESULTS

Previous histological examinations of the arcuate nucleus after GLU administra-tion have shown that the lesion is characterized by neurones undergoing necrosis andphagocytosis, followed by gliosis within the first 48 h5 ,.6

,35 . Our cell-stained sections

from 7-week-old animals seem to indicate that most neuronal perikarya havedisappeared irreversibly from the arcuate nucleus, while the remaining cells seem to becharacterized by small nuclei and lack of Nissl substance. They thus probablyrepresent glial cells (Fig. 2). The study of the cell-stained sections led us to make fulluse of our dissection technique which, in contrast to the 'punching' technique42 , allowscomplete sampling of regions with irregular outline. The distribution studies wereperformed on the median eminence, the triangular nucleus arcuatus, the almostperikarya-free region separating the arcuate and the ventromedial nuclei, and samplesfrom the ventromedial and the dorsomedial nuclei (Fig. 1).

In normal animals, both ChAT and AAD showed low activities in the ventro-medial nucleus, with higher levels in the arcuate samples and the median eminence(Table 1). The highest activity of GAD was found in the region between the arcuateand the ventromedial nuclei and in the dorsomedial nucleus. The activity in thearcuate nucleus was intermediate, and the median eminence displayed low activity.The dorsomedial nucleus exhibited intermediate ChAT and AAD activity (Table 1).

In the lesioned animals the levels of enzyme activities in the dorsomedial andventromedial nuclei were unchanged. AAD was also found to be unchanged in thelateral arcuate, the arcuate nucleus proper and the median eminence. GAD wasunchanged in the lateral arcuate and the arcuate nucleus, but a 50% decrease inactivity was found in the median eminence. ChAT was significantly decreased in all 3areas, but most pronounced in the arcuate nucleus and the median eminence (Table I).

AChE staining of sections from normal animals demonstrated that the vertro-medial nucleus was almost devoid of staining product. The arcuate nucleus, however,did exhibit intermediate staining, in agreement with the distribution of ChAT. The

O M NI=2 . .. . ... . . .. . . . .. . .. . .. ..... ... -

76

554

000

+h +v +- -H +

It_

0rq

0 0

1<c -

tj II - -

-~~~ a'd 'Ir

77

.~C

.0

u0

1 C

E L.

.aV.M

u

rJ

E0 E

00

ca E

E~ E

E 2~

78

556

TABLE II

rhe effect of GL U-administration on the high affinity uptaA e ottransiutter candidw.tes in median elypinence-nucleus arcuotus

Values presented as mean ± S.E. (number of experiments).

Normal brain GLU-treated brain€nmotelmin/g prorein) (per cent of norntal)

GLU 26.2 = 2.8 (6) 93 ± 10(7)DA 0.72 ± 0.09 (8) 113 ± 24 (6)NA 0.29 ±0.02 (8) 109± 6 (6)

TABLE III

The effect of GL U adninistration on endogenous amino acids in nucleus arcuotus-ventrornedialis

Values presented as mean ± S.E.

Normal brain (n = 6) GL U-treated brain (in 9)(Qtnole/g protein) (per cent olnormal)

ASP 35 ± 3 91± 9GLU 105 ±8 103± 8GLY 26 ± 2 86:13ALA 12 ± I 116 ±17GABA 41 -f 3 87± 6GLN 87 ± 7 80 6TAU 57 ± 5 123 15

median eminence was found to exhibit staining in the lateral, external regions,particularly in the caudal part (Fig. 3). This staining was most clearly demonstratedafter long incubation time, eg. 24 h, whereas the staining of perikarya in the arcuatenucleus was optimal after shorter incubation. After preparation of sections fromanimals treated with a lethal dose of soman, the staining was absent in both regions. Inthe GLU-injected animals the AChE staining intensity was reduced in the arcuatenucleus and median eminence, but unchanged in the other parts of the hypothalamus.

The other biochemical parameters were investigated on homogenates obtainedby dissection of wet tissue samples. The high affinity uptake of GLU in the medianeminence-arcuate sample was found to be of intermediate activity, being approxi-mately 30" of that in neostriatum, which is one of the regions with the highest GLUuptake, but twice as high as the uptake in more lateral samples, and 4 times as high asthe uptake in globus pallidus (data not shown). The GLU uptake was found to beunchanged in samples from lesioned animals (Table II). The high affinity uptake ofcatecholamines demonstrated that DA at l0-7 M was the most active ,ubstrate inagreement with a previous report8.The uptake activities of both DA and NA were alsounchanged in the median eminence-arcuate sample from lesioned animals (Table 11).

Analysis of endogenous amino acids in an arcuate-ventromedial sample demon-

79

557

strated cons;iderable concentration of GLU. aspartic acid (ASP) and ;'-aminobutvricacid (GA HA) (Table 11l). In the lesioned animials no signilicarft change,. %%ere detected(Table Ill).

t)I Sf tiSS ION

Several approaches, hase'been used for the studN of hypothalamic neuro-ransmIlitters. both histochemical (for review see ref. i7), and microdissection comn-

binied wsith microchernical assay s (for reviess see ref. 2). Inl the present insestiation wschase taken adsantae of the selective lesion of local neurones in the nucleus arcuatusatter GLU administration to ness born animials to study the origin and distribution ofneulrotransmitters in this region. As Gi LU administration to newborn or adult animalshas been reported to have interesting endocrinological consequenceS"i 3 . 1.45~. suc.studies ss ould presumnabi also t hro%% ;ome light on the neurochemical basis of thesephenomena.

The small dimensions ot the mediobasal hvpot halanmus necessitate "ery precisedissect ion procedures for such studies. I nitiall\ 'se analyzed the enzyme actisities inthe median eminence in samples from fresh tisstue dissected as described by Cuello etal.';. Ho\\es.er. in these studies (results not shown) sse did not find the same relativev alues and the same decreases in actis ity as we later found using microdissection onfreeme-dried section,,. This is probably due to contamination oif tle median eminencess it I neiahhourin iiqtruct tires also containing the cnz7 lies in question, and indicatesthat Such samples are not Suitable for detailed studies on the median eminence proper(s;ee also ref. 7). The microdissection on freez-dried samples illossecd us to asoid suchcoiltam inat ion. In addition. it enabled us to dissect samples of ss idely different outlinesan sseio-hts from thin s;ections compared to sections used in pnhtg.Teutk

studies. howsever. had to be performed on a combined median eminence arcuateI sample. and t he amino acid anaissis on anl arcuate lsent romedial sample.The involvement of acetylcholine in neuroendocrine hypothalamic function has

been suspected on pharmacological and ph\siological groundst ' 211. Our results

demons;trate that the specific cholinergic marker ChAT I displays anl intermediateactis itv in thle dorsomedial nucleus, a lo\\ activity' in the \entromedial nucleus, andhigh activi ty in the arctuate nucleus and median eminence. Compared to pres iouLy,,\published results;'. \%e find a relatively higher activity in the arcuate nlucleus;. which is

probably due to a more suitable dissection procedure as discussed abose. In thelesiotied animals. the arcuate nucleus and the median emitnence displayed a 70%>decrease in ChAT activity-. while the actilit in the other regions %%as unchanged. Asthe rat median eminence is almost devoid of nerve cells2i1. this indicates the esistence ofaI cholinergic fiber tract originating in the arctiate nucleus and terminating in themedia n eminice. in aiereemient \s ith t he itidepentdent j nsstigation of (arsoti et al.'.

* The \%eight of our free-tirk'd 1":dj'1lItIC1 "ninc mples iflkll.Iic' ai iuiaI \~i%C egt o -0 qO QO cof tile lidin emlinecnce, in areement wth ii (iocchtio et idi.

80

558

The remaining ChAT activity in the arcuate nucleus could represent minor projectionsfrom other areas, as suggested from knife lesions by Brownstein et al. 4.

Our results with AChE staining would also support the concept of a cholinergictuberoinfundibular tract. In sections from normal animals we found staining in theexternal lateral part of the median eminence, i.e. the same region which displays mostintense staining for DA and for hypothalamic releasing factors18. Both the medianeminence and nucleus arcuatus displayed considerably less intense staining in thelesioned animals, indicating that this staining represents a visualization of tubero-infundibular fibres. In contrast, other workers19,43 have failed to find AChE stainingin the median eminence. Our conditions, by using thick sections from frozen tissue,avoiding prefixation, and using long incubation periods, would seem to favourstaining of AChE. The conditions which lead to staining of median eminence are,however, at the expence of high resolution. From the effect of the cholinesteraseinhibitors ethopropazine and soman we conclude that the staining obsecved doesindicate the presence of 'true' AChE in the median eminence of the rat.

The distribution of the GABA-marker GAD" reported here is an extension ofthe results recently reported by others55 .We here demonstrate that the highest activityof GAD in the mediobasal hypothalamus is located not in any of the nuclei, but in theregion between the arcuate nucleus and the medial part of the ventromedial nucleus,with less activity in the arcuate nucleus proper. The median eminence displays muchlower activity. Previous investigators have proposed that GABA may not be impor-tant in median eminence function22 ,55. However, our results demonstrate that in thelesioned animals a decrease of 50% in GAD activity occurred in the median eminence,while the rest of the hypothalamus had unchanged activities. This could be due eitherto destruction of a GABAergic fiber tract terminating in the median eminence or thedestruction of intrinsic GABAergic neurones in the median eminence. However, suchintrinsic cells are reported to be very scarce23, and we therefore think it is mostlikely that our results indicate the existence of a GABAergic tuberoinfundibular tract,which originates in the arcuate nucleus. This tract could possibly represent theanatomical basis for the endocrinological effects of GABA 20.39.40. The small andinsignificant changes in GAD activity found in the arcuate nuclei of the lesionedanimals probably mean that most of the GAD activity in this nucleus is concentratedin afferent terminals, and only a small proportion is located in local nerve cells. This isin agreement with the results of other studies, which demonstrate that a major part ofGAD disappears in the arcuate nucleus but not in the median eminence afterdeafferentiation of the mediobasal hypothalamus4 .

The well-known DA-containing tuberoinfundibular tract, originating in the A12cell group in the arcuate nucleus56, was also studied. AAD is found in both DA-, NA-and indoleamine-containing cells, in addition to being localized in capillaries andpossibly glial cells21 .25 .However, the relative values of AAD reported in this study arein good agreement with the corresponding values of tyrosine hydroxylase 48,49, whichin these regions is acknowledged as being a selective marker for DA elements'. Wealso analyzed the high affinity uptake of DA and NA in median eminence-arcuatesamples in order to detect any loss of DA structures 4. However. our results indicate

N1

81

559

that the DA innervation of the arcuate nucleus and median eminence has not beensignilicantly destroyed in the lesioned animals. Amine fluorescence in the arcuatenucleus has been shown to disappear after acute injection of the GLU analogue kainicacid3 to adult animals32 or after repeated injections of very high doses (4 mg g body\%t) of GLU to newborn animals3 l. Our results indicate th:t these DA cells, despitethis apparent GLU sensitivity, are more resistant to the neurotoxic effect ofGLU thanother types of cells in the arcuate nucleus. This is in agreement with results from otherregions in brain and in retina' 9'.

Except for GABA, the amino acid transmitter candidates have received rela-tiselv scant attention in the hypothalamus. The preferential release of GABA, GLUand ASP from hypothalamic synaptosomes has been demonstrated 10, and it has alsobeen sho%% n that both systemic and intraventricular application of GLU hasendocrineeffectsl s 4. This endocrine action of GLU is probably located in the arcuate nucleus.a this is the only hypothalamic region where GLU is accumulated following systemicadministration 11. Our results on the high affinity uptake of GLU seem to indicate thatthis amino acid could be a transmitter in the median eminence-arcuate complex 9 '5.The result. do howkever indicate that the 'glutamatergic" terminals have survived theaccumulation of GLU in the arcuate nucleus, and argue against a major population ofintrinsic glutamatergic cell bodies in this region. Alternatively, the glutamatergic cellbodies could be resistant to GLU accumulation. The levels of endogenous amino acidsin the arcuate nucleus and medial part of the ventromedial nucleus did not show anydecrease in the lesioned animals, also arguing against a major population of 'aminoacid-ergic' cells being destroyed by the GLU lesion.

In conclusion, this study extends and contirms previous findings on theIocdi/ation of neurotransmitters in mediobasal hypothalamus based on knife lesionsand punching technique: 1.49.5 1 5. We suggest that the tuberoinfundibular tract, inaddition to the preiously described peptidergic and aminergic elements, containscholinergic and GABAergic fibres. These are destroyed by GLU administration, andtherefore probably originate in nucleus arcuatus. This destruction could be a part ofthe neurochemical basis for the endocrine dysfunction in the GLU-treated animals.Also. \\c find that the DA cells in this region are resistant to 2 mgg body weight ofGLU. They therefore probably do not contain a high number of GLU receptors.

Finally. a relatively high concentration of "gluiamatergic' terminals was found in thearcuate region, which is probably not due to intrinsic neurones.

A(KNOWL[GXFMENTS

We thank Dr, I. Grofovdi for photographical help, and Dr. M. Tappaz forhelpful disctIssions.

I. Walaas is a fello\N of the Norwegian Research Council for Science and theHumanities.

82

560

REFERENC ES

I llroch, 0. ).,Jr. and Foninum, I-., The regionralI and stitvellular distrihouIion of calechol.-rrieths I-transferase in the rat brain. J. \'urrwhem.. 19 (1972 2049 2055.

2 Bro%%nstein ., N eurotiansmittets and I-I pothalamnic hormnones ti the central niervous %sfsem.FI'd. Proc., 36 (197711960 1963.

3 Bro'.%nstein, Ni.. Koba~asbi, R., Palkosir'.. NI. andi Saa'.edra. J. Wi. Choline acet~ltransferaselevels in diencephalic nuclei of the rat,.J. 'Fiumoiiemn.. 24 1 975(13 .

4 Bro\%nstein, M., Valko'.its. M.,Tappa, NI. L.. Saasedra. J. M. and Ki/er. J. S.. Effect of murgicalisolation (if the h. pot ha lamnuson itis neu rot ra nsnu t ter content. rain Re'.iohl. 11I7 (1976o 28 7 -295.

5 Burde. R. Ni.. Schainker. B. and Kas es. J.. Acute effect of oral and subcutaneous admtinistrationOf mon0OSOdItiM glutamate1 On the arlcuate nucleus oft hei h.\ potha1lirS in mice and rat. eawue4Loid. ,.233 (1971) 58 60.

6 Carson, K. A.. Nerneroff. C. B.. Rone. NI. S_. N oung-Blood. W. W., Prange. A. 3.. Jr.. Hanker. J. S.and Ki.:er. 3. S.. Hiociernical and histochemoical eiidence for the esistence ofa tuberoinfundibularcholinergic pathv.a in the rat. Brain Rerar, h. f29 (1977) 169 173.

7 ('hiocc'iioS. R..Cannata. NI.A. and Tramneiani.J. Il. .Ticsize. %k eighIt andt catIecholamine contentof the median eminence of the rat. Brain Re iear, 110 11976) 612 618.

8 Cuello. A. C., Horn. A. S.. NiacK a\ .A. %. 1P. anid Is ersert L. L. ( atecholamnines in the moedianteminence: ne'. es idertee fIm a iraiiir norad renergic input. N alie'Lond. . 243 (1973) 465 467.

9 Di ac. I., I [,nn um. I - andi Storm-MNathisen. J.. High allini iv uptake of glutamate in terminalIs oifcorticrisi a awn~is. Num intone , d 266 119771377 378.

MI Fd'.'ardson. J. A.. Bennet, 0i. \k. and Bradford. H. I . . Release iif amino-acids and neurosecretorvsuhbstanees after st inmulat ion of nmr.e-endinrgs Is'.narPtosoies isolated fronit fle IiNpitha Iani s.Niature (Loud.1, 2401(1972f,554 55;6.

11 Fornnti F.. The localiaiton oft gIlutamate idecairbo\ lase. choline acet'. transfecrase. and alroniaticaiino acid decarbos\ lasein mnraimalian and insertebrate nert %ou t iss-ue. In S. Bei. I). I). Clark eand 1). Sch neider (Ed. . %ti'ribsljitf n~itriro i arnd N 'ntamiiii'in Plenum tire,.Ne' York. 1975. pp. 99 122.

12 _-Ionnum. t .. A rapid radiocheiicail method for, the detcimination of chiiline acet'. transferase. J.Neu~w. Iim. .24 11Io) .1117 41).

13 1-onnum. I-.. Walaas. 1. and Iseisen. I .. I .ocaalawn of (i-N) -crgic. cliolinergic and airinergics.tructutre, in the rite,oIimii ic s'.stem../. N ,iri, im. . 29 11977 F '2 1 23;1).

14 GO Wsei n, NI.. .n tagnosre. B., Hat . . ., .. h1k' n. S, kk -amtis \atani .S.. Studies of amine . inthe striaturi in niorke~ %i inivrah lesions 3The dispoi I iiorr hiss~k1 rilesis. atid Frleta ho~lites of(I (I I-do 'amrine, and Il serirtorir ii tnIhe IF riarur UIII. News ~hr i.. 110909) (45 ,5,1.

IS (it andismn. L.. Cielaro. Ni. iitd Niertes .Inhiiion of' pirlactii scriction h.% clrolinergic tlrUgS.Pi 5 S,(. '.xp. Rim?. 1 145)11974) 1 23fi 1239.

16i (ijiidotti.A.herre'. I I krbici N. ~ruh a.N V ng..An a'.irsU.A.Iotsdtricr,,.'.a. e radthaion: a technfit)uei to iinmi/e postmoirtemri changes of ct dio nurcleot ide,. D)1)lA

andchoineandto resr t 'rin mirhioirg . Nemq'hmi naii/igi.13( 19741 111'; 1122.17 IHenke. 1f. and f inrium. I .. I iipog aph ical andi sihilil dlistrI Ihuiiort ofeholitie acet\ IIIratrstrasc

nmd gltit arate deca rbos\ hiae in pigeoni optic rectui. i.1. Nerohro. . 27 (1976) 3S7 391,18 Hokfelt.. Johansson. 0., l' tis. K_.. Ithe. R_. Coltlstein. Ni.. Park. 1) , IEfenthic. S.. -Lft. R.,

[riser. If end Jeffkcoate. S , H'. porhalantic doprmrrre neturorns anti h\ pothalarrie pepi-des. In I.( osia and (K i, . (,ess,i N il, i, N ,'r ioa t(ipjiiiijrit I, Nip ii 4

11i11hi , IHim hemi P,.i, boiif/ri i

maiu, of/ . t 111. 16. Ra. en 1 e,.'.Ne\ No 'If . pp- Il 99 Il0819 JalcohoIt/. I) and, l'.rlkmi Is. \I . I opm :iphic: irlis of carechiilari anid acet'. lclrslirresterase-

conftaining rneuron, tin thett r im I I Jrht~mt I mmi,\ Niiio/ . 15' (I,)~-- I I I'201 Jones. %i. 1 , 110hihi00se. I \\ .i) Ml130hcmmI. .I fiIO %f '.rrrIiiS ptit,iI'.e OFFr~r~nsrrteu Thre

,ecretwnfli ~tcirm~ri-ejea'.inig It, Iurine Ifrirri Irle ra t\ '. I or hiafliu], in ill 1i, .1 irroilel M fihenieiiiirr~insiier, rr'.mhm .1 1.. .(,,) (Yr I it).

I2 Kelloi gg. (i .. Lund hi rg. 1'. art) R,iirtedr . t , nl.s' o. tlsillat a ind pa rench\ ira) awni mi Irrirrini acid dee,r ho' l~ase d\~i it miiegimirtl firain areas d uringo mungenerti,: sheeluprtert in ihr ir.Ilrmi Ruiiapi h. 50)(19' 4) 16r9 ) '

72Kite .. S Palk... it,. I. ap,.N Kehahimn_. 3 .ui Hrstn'tirr. \I1.Irtihiiinr eeirfactirs hrigenr. imine... andi retired ci,, in [fie' I'm ite rieitivritr"rre. I id, mvo/tw 9''

83

561

23 Kobayashi. H.. Matsui, T. and Ishii, S., Functional electron microscopy of the hypothalamic me-dian eminence. Int. Rer. Crial.. 29 (1970) 281 381.

24 Kibnig. J. F. R. and Klippel, R. A.. The Rat Brai,. A Ster,'otactic At/a.% of the Forebrajiand LotierParts ofthe Bra/in Stemn, Williams and Wilkins. Baltimore. 1963.

25 Kuntzman, R., Shore. P. A.. Bogdanski. D. and Brodie, B. B., Microanalytical procedures forfluorometric assay of brain DOPA-5-HTP de:a rboxy lase, norepinephrine and serotonin. and adetailed mapping of decarboxylase activity in brain. J. Netorochem, .6 (1961) 226- 232.

26 Lemnkey-Johnston, N. and Reynolds. W. A.. Nature and extent of brain lesions in mice related toingestion of monosodium glutamate. J. Neuropath. exp. Vettrol.. 33 (1974) 74 .97.

28 Lossry. 0. H.. Rosebrough. N. J., Farr, A. L. and Randall, R. J.. Protein measurement sith theFolin phenol reagent. J. hiof. Chem,., 193 (1951) 265-275.

29 Lund-Karlien, R. and Foninum. F.. The tosic effect of sodium glutamate on rat retina: changes inputative transmitters and their corresponding enzymes. J. Neur,,dhen., 27 (1976) 1437 1441.

30 Lund-Karlsen, R., Fonnum. F. and Walaas, I., The toxic effect Of L-glutamate on neurotransmitter,in rat retina and hypothalamic nuclei, Ahbstr. 6th, Int. kfeet. ISN.. 1977, 200 pp.

SI NMaDonnell, P. and Greengard, 0.. The distribution of glutamate decarboxylase :n rat tissues.iotopic %s. luoromectric a'sa) s. J. .Vewroc/u', 24 (1975) 615 618.

32 M i/ukassa. K., Shimizu, K., Matsuura,.T.. I bata. Y. and Sano. Y., The influence of kainic acid onthe tILI l,oinfundibular dopaminergic tract of the rat: fluorescence histochermistrs, and electronmicroscopic insestigation. Ate, Hi!.to(hemi. Cwtonheui..9(19761 315-322.

3 3 Nagasa%%sa. H.. N anai. R. and Kik)yamia. S.. I rreversible inhilition of pituitary prolactin and gro%% thhot mone scretion and of mammary gland deselopmcnt in mice by monosodium glutamate ad-ministered neonatalls, Att .4c ,socr. ,Khh. .75)(1974) 249 259.

34 NemrerotY. C. B.. Grant. L. D).. Bissette. (i.. Eroin. Gs. N.. Harrell. L. E. and Prange, A. J.. Jr.,Gross th. endocrinological and behas oral deficit, after mionosodim b-glutamate in the neonatalra t: possible ins olement of arcuaite dopamiiine neuron damiage. Ps.i , Iii,',ws-4etids,riiiolgi., 2 119771179 196.

35 ()lne~ ,. J. V_. Toxic effects of glutamnate and related amino acids on the deseloping central nersouss\ stem. In W. 1.. NN han ( Ed.)I. Heritaib/e Dsiwido e, oI .4,iiis 4, ii t Ictahodmi. Wj ile) and Sons,Ness N ork. f974, pp. i0l 512.

36 OlneN. J. kV., Rhee. \r. and Ho. 0. L.. Kainic acid: a posserful neuretoxic analogue of glutamate.Brajin Rceri hi, 77 (1974) 507 5I 2.

37 0lneo. J. W., Schainker. B. and Rhee. V. Chemical lesioning of the hypothalamus as a means ofStUds ineg neuroendlocrine function. In E. Sachar ( Ed.) Ioroicnes. Behavior and P,,iciopatfiicog.Rasen Press. Ness NYork. 1975. pp. 15.3 158.

38 ()lne . J. W., Cicero. T. J.. Me'er. E. R. and de Gubarctr. T.. Acute glutamiate-induced elevationsIn "'rum testosierone and ltcini.'ing hormone. Brain Re~c.,ii 1121(1976) 420-424.

39 Ondo. J. G.. Giantn'a-aminobuity rc acid effects on pituitary gonadlotropin secretion. Science. 18611l914) 738 739.

401 (ndo. J1 G. and Pass, K. A.. The effects of neurally actise acids on prolactin secretion. bndot~iniil,-i! i ,98 (19 76)1 12 48.

41 Ondo. J. 6-. [lass, K. A. and Baldss in. R., The effects of neuralls actise amnino acids on pituitarNgonadol ropin secretion. S inoeiiadairiiiiif ,-i 21 11976179. 87.

42 Palkos its. \I_ .Isolated remosal of K\ pothalamic or other bran nuclei of the rat. Tai,, R,.sccnrli. 591973)449 4511o.

41 P'arent. -V. and Butcher. L. L.. COrganization and morphologies of acetyleholinesterase-containingneui ons in the thalamuILs and h\ lothalamus of the rat. J. icillp . .\euriiI., 170 (1977) 205 -226.

44 P'ere/. V.3. and ()lnces.J. VW.. Accumulation of glutamic acid in the arcuate nucleus of the hN pothala-itis of the infant Mouse folloO ng subcutaneous administration of monosodium glutamaie. J.\etcicwot.. 19 (1972) 1777 1782.

45 l'i., .. Batrnhart. 3. E. and F-ansloss. D. J.. Monosodium gl utamate admi nist ration to the ness -hoi n reduces repr-oductise ahilits in female and miale mice. Science, 196 (1977) 452 454.

46, Redding. T. W. .Schalle\ . A. V. Arimura. A. and Wakobo\ashi. I., Effect of mionosoidium iglutamiateon some endoscrine fuinet ions. Neitioenlt,fcrincifcg, 8 (19711 245 2501.

47 Renaud, L. 1P.. Tuberoinfundibular neurons in the basomiedial hypothalamus of the rat: electro-ph siological es idcnce for awon collaterals to h pothalamic and estrah~ pothalamiic areas. BroiniRewain 1h. 10)5 (1976) 59- 72.

.48 Sai.,cdiw. I NI.. Bi os nstei, M., Palkos its. NI.,- Kiter. J1 S. and Aselrod. J.. T\ rosine hidro\' lase

84

562

and dopamine-fi-hydroxylase: distribution in the individual rat hypothalamic nuclei, . Neurochem.,23(1974)869-871.

49 Saito, M., Hirano, M., Uchimura, H., Nakahara, T. and Ito, M., Tyrosine hydroxylase activity in

thecatecholamine nerve terminals and cell bodies of the rat brain, J. Neurochem., 29 (1977) 161-165.50 Schwarcz, R. and Coyle, J. T., Neurochemical sequelae of kainate injections in corpus striatumn and

substantia nigra of the rat, Life Sci., 20 (1977) 431-436.51 Silver, A., In A. Neuberger and E. L. Tatum (Eds.), The Biology of Cholinesterases, Vol. 136, North-

Holland, Amsterdam, 1974.52 Storm-Mathisen, J., Quantitative histochemnistry of acetylcholinesterase in rat hippocampal region

correlated to histochemical staining, J. Neurochern., 17 (1970) 739- 750.53 Storm-Mathisen, J., Glutamic acid and excitatory nerve endings: reduction of glutamic acid uptake

after axotomy, Brain Research, 120 (1977) 379-386.54 Tappaz, M. L. and Brownstein, M. J., Origin of glutamate decarboxylase (GAD) containing cells

in discrete hypothalamic nuclei, Brain Research, in press.55 Tappaz, M. L., Brownstein, M. J. and Kopin, 1. J., Glutamate decarboxylase (GAD) and J-amino- I

butylic acid (GABA) in discrete nuclei of the hypothalamus and substantia nigra, Brain Research,125(1977) 109-121.

56 Ungerstedt, U., Stereotaxic mapping of the monoamine pathways in the rat brain, Adta physiol.scand., 82, Suppl. 367 (1971) 1-48.

57 Zimmermann, E. A.. Localization of hormone secreting pathways in the brain by immunohistoche-mistry and light microscopy: a review, Fed. Proc., 36 (1977) 1964 -1967.

85 PAPER V11

.' .... . i" (. - io m ': w

The Effects of Kainic Acid Injections on GuanylateCyclase Activity in the Rat Caudatoputamen,

Nucleus Accumnbens and Septum

Ivar Walaas

Abs.tract: the ofti ol u dile and paisticiilate guin ate Cs clase (EIC4 6 1 21 has been .minipaiid %k ith ltly (lixlibulioil 0ititicuimatiniter candi-dale, i IIIlitee tat loicl'i inl ii lei and tht: vtc .I i ol kaiL~i acid Inj~ectionsinto thesec numci hisc ben tesied Soiiile giiai~ ate ck claise %kas hijzhlyi'iic:cnttated III both the cai~t'ii.inand the nutcleu, actinobens. Aith

It o A etiJ 11% It % fouind tin the wpIturn. this dish butuuion coincided u ith niarlers1,1i .tClcholinc and motwnomnes. hut not %k.ith rmikeis for y antnohuts rate((AA Aft i liaate neuons. InI conti ast. paiticulaite lguans late c'.clase "sas

cqu'dk aIt ein al ,egions, L ocal in -ect ions of kainic acid. sm hich destros edholiii~fgi, and (003BA neurons in the cauidatoptitamnen and int the nucleuisiiiihkcn' - ciiid i apid 70 90' Jdecrease in the soluble insaeccas

anid I 1,,,Aci '(1 60', fill in the particulate guian~ late c~clase in these nuclei.

I lie q-ritin. "inhere kainate destroyed (jABA cells but not cholinergicIiciions. the giuan~ late cls actis it'. "as unChanged after the lesion. 1hus.

oli sihV iibl1-c and parltiulatie gulan% late ec elases appear to he concentratedIn 10,J1 ral111 neIinin the CAiidato1Putamnen and nuieleus aeeumhens. In the septum.hi'" :' i, miost ol I he giians ate c),clase actis it% is located out side kainate-sensit Sc icions Ke,. %%ords: ('cltc (,MP -Neurons -Glia -1-orebrainnucleci ' alaas 1. the effects of kainic acid injection% (in guan~ ate c~elaseact us iti iii the tat caiidatoputlamen. nutcleus accuirnbens and sepitum. J

Ait, mn 3h. 231 241 11981 I

(ruan~ale c~clasc (EC 4.6.1.2). the cn;. me interest to compare in detail the topographical ands> nthewiing guanosine 3 .5' -nonophosphate (cy- cellular distribution (of the soluble and particulateclic (IMl11 I liardirran and Sutherland, 1969). is pres- guanylate cyclases in the catidalopulamen. nucleusent in brain in both soluble and particulate forms accumbens and the septlern, brain regions %%ith rela-Nak&u,;uka and Sano, 1974-; Kimura and Niurad. tively %%ell-defined glutamate and ACh neuronal1975. IDeguchi el at.. 1976-. Troyer and Ferrendelli. systems (Fonnumn and Walaas. 1978. 1979; Walaas1976: 'rros~er el al.. 1978). The physiological func- and Fonnum, 1979). For this purpose. the kainictions, these eni~mes posibly have in the brain are acid lesioning technique has been used to distin-lirgecly unkno%&n tFerrendelli. 1978: Greengard. guish between local nerve cells and glial cells (Hat-1978). but a 'second messenger' role has been sug- tori and McGeer. 1977. Schwarcz and Coyle. 1977.gested for c'.clic (;MP in mediating neurotransmis- Coyle et al., 1978. Fonnum and Walaas. 1978:sion due 1o ;iCetylcholine (ACh) (Lee et al.. 1972; Malthe-Sorenssen et al., 1980). and biochemicalStone and la~lor. 1977; Hanley and Iversen. 1978; markers for putative ACht. y-aminobutyrate% oodv et al.. 1978) and gluttamate (Mao et al.. 1974, tGABA). monoamine and glutamate nerve termi-Iiiggio el aL.. 1978). It was therefore deemed of nals and glial membranes have been analysed to-

IRLSckicdi Juir% S. ittuti. cptird Juil% t. 1980 A hhrri .uiu.wn.% au[,cd A AD. Aromatic -amino acidAcddics ktiespondennv anti ieptifli rewicsts ii' liar NWalaas. decatbox~lase, ACh. Acet~ lchotrnc. ChAT. Choline aertyl-

1k paritnenii o' Pf'iniat~,oog . lat t'nitet~ . School. of tiiinsretase. GAIIA. ), Amiruohuti nc acid. 6Al). t GlutamaltNiedim ine Nev fflssnn. ( uonfi twill 0010t. 11 S A eih5ls

233

86

L!Lteih "x ith the soilublle andlk pVr tIIIl~t. gunsI iAlt 11idl.n Iirl '.1 l iu ILI hill .. 'Itt It,~ fill

:c Ila se acI , I tie. S. Ma It , it i 1,wi .iI id. I ti utike~ stidik-. the .Ili

Initial iesults titli the c-imiittpiixipitaii mtlit tled I' \\I ci Iiii~f Il f COld V I' %I IIIII '(i 1tic ,\

it decrease in gian)[late e~s id ler kaitiate lv- I ii' III I, pit 7. -3 2 %i ) % sltl 10 stijikts 11) V2l.1" I cth'

sions. and thus a neuronal tocalijation tit the en Ioiiiieeis i S O at ' rp.m.. kept on ice and ari.il1 sed fll

iim.toss es r. a simultaneous report pentd uptake acti\ it\ \A tlhin 30) nin F-or .iiiilk is (it cr/ nie,~ me. prsented ,i'.ities. the %amiple, %%c homogenised in 20) ol (if told

Cx idence suggest ing that kiiinatc md ucd neuronal 10m' int-hs- i( 1 I p1 7. .r1 ;', an d on cc I he ii.kcineci osis incireased the guaillilati: tx tasc ajctIi itvi in eivns te xxie m-d\,l ced .ifici aditliiun ill II 2 I wonllrat eaudatoptiinen boiigciates I Wiuhaitmar et x 10' 1 %%\al.. 1979). The present stiid hall theiclore also (Itnx li cx 1,,c ka ixdictd .iii ,ditiion tit dithio-

insestigated some charakcteristjis oi' lte cni\ mie and ihieiltol (final tili cit~tiii fit, fil te I t hiittcied

confhrmed that the I ikIO oh nC 1ItI I pt tiiict ft' if ed in irfoenis.SiiKCliilIIA tiion1f.iiiiii (It AlIqltIs I Ik II

he pie sent as'.a met hod is c clic (,%it'. filhe lilicil sal pits xx 1 elIci dllen liii t'la

Some prelmntrs results has eC been pi)I"II 1sei 1 in lit W) iiiii at 4StO00 X. %\ fill It his, h-cii liiiind tol toiritl\

ibstiact lorIm ( Walaas. 19791. xcp1.ii,1it Ie CSollible Amd pA ItIIt iLmIC Me C % Lif Ls IxcI iti1 oxcin et id- 1978i, the uipeinaiont iix 'ohihlc

%,itioix xc tollicicd on ice. "t itle the pellet, xxecMAt I FRIAILS AND) %I[ I 1101)5 \ktashed III Ilitol ,,itmmwit bufft.irexentrihiied ind

xIIIILlcilt~ 11 11111 t"iit Q11-1 \0111i11C siie utculd 3)0 nist- t isMatlc %k ixtl fit, HNO~l 200( g hod\x %k c101. L c I c oh tl I II% 'II lrbl!c IIIC11. pli 7 " . I' 11win' Of 1:1i"'

iiricil 1iori %I.liiD.teiii.iik, j" I ( *l I P \k.1s lb IVI ti- 11" lii11 itjmix oi ip.II.0Ioii ts Vce theCir L

taIiiied Ililiti Nsex t iiglaiiit N'.clemi Btosin. \isI i.,tic 1 I' I Itt 11 1 11,1 loll Ix 1i .1 4 ( ti'l fi1t 1t11

sithiietts. ind the Ritliuiciiical (enuic. Aniihirn. Dt'iue .n.il xi 111iltil61iIK i- S t~iI (1 lcI'. G%1 1) 581 (YNit' and I i IA' - HAIII Oi i itii xcitI 1011 t hie t{.itwlrthe nit i ( , tie ew I m ttO'xiu1Cush,1ni. I K K.~ - It I lic (1\uiti. Is., Ht 10 1 11.Iii no h /i

M~ P4S A I. i 2 I I eiiw.,mic IIdIp 2-' dihx- Ilii xx\ s il s h\ tillt mct'liilo l 01 1 xi 0%%j l.diO\ I'helx1 1Idlinifitc intlk I I 'I( A I \i I ( 0A .~ CxIC iii I 1011 1114; 1 xx fllt hox fie set lifi .1hill'ni ., xtmdddd l1I,1ilksNet'. I 11):in111 NIl cil I III.ifiiiie Andi .11L r1u1e n)iu clikles tiifttiuing dithiothicutol xxere TI citilcth to 10 u f CIt Illixodiii siAs xobit\I I iich\l.iinci kaiinu it li iniiilerence h% the thuol icugetrii Lk icue [It,slif I tIc-1i 1, 11 11-( 1144 11 i I .~ lix I uII, Iu IIII'I IcirlC phll Irh11di]urul decukibed mtcilurds\ t xxei c i fd lit utaaxw ldi-itf it ( ', 1 .4 17 11i liti bcc ht.oi xxieC fholn SignIa maic 'I aniunr acLid dcicaih-osaw Li .% At). I-C 4 t 1 I

htri)Ciicl ('I t SI luls \1trxxiiiu tt''cx "I)" - 4 lit 03wlch andJ ,3ijunt J972). -choline teu inlrslini -'11) 400) IIrucxhi mid tDox'e\ t - (l WI litni (ChA'I. I-U 2. .1 06 itoutum. 1971. Illu arniatc die2))00 400) Illh Cxu 111ton ICIIkIIl I liuka AlG. itiuchs. neictal -iilruisae KIAI). I-U 4.1 t1.15 0~ Ilinuim ei At.. 19701.

.iiitiin11 gI itmd C Ii Ix 'A t Iot i ici1. I- xIt l1uxkCee. 11 k ulie iliiOcllrduliliall fifne ciuninne itistinse CpoIleths Icitirilu i cellulose I IDC I vutigpl-iiutt.enf I Au I iI stei Ii .Iad I onnium. 1980). and the menibiane

xx-l Actiom Illt cickl D~mii uisidi. Vi ext (siian\ cix inc, I li elileiIU1.3 ,I.%xlhich x'.ix.nils sedOtIher ,litnital, cxi c c iuiiciu ll\ iibtiied ( liomouiii " lit) (10' mst i 11, AMP as, siubstiiti (A iuich aiid \%ilriaphic xiiltuiins ill t il) 'AtIt eie l him " H Ad 1. 1i. I 4' 1 i l 1h .,Itutritt Uptake ol 1 2.3- '4 eiitaim,

I I libiti~ i i it ihtrul tIilnxclldtiiln t0 : -,I s t ldied %k ith \1u11i

pie fuht.itiiin ii'ed tot) imiilte the nfl uibation as de

the l.ll1l-i I i 1 iiil.ited on the liltegs xix detet-I till muinials x'eiv i.ies1thtetl xx ilih It r.iitx IAifit I' l tntd 11 I II Atu- illtlwi ti ii uilucu- olnit % ititillititn liuiid

I Hoich.0(7 tri p I theti. 'ifub rll .. I lxslcd lit iib 'lulI tsl fined "filh Ii ablllie ac~tit\ itani I vA 141) itii xotiiiti pholi rspi (t filci, [,pt 4 %k. 1,%s I t I ti li It i i~ix s

wiiIxd itili file tii lt liilii tiiliiit Nx -IAI'u . I ( I iii l kIt1t else iisc At anal\ scd 6%-it moiidilatioir (il

2IN, .itd % 700 ?tl Kiitig and Klhi-ptel 110.; Iii tinit. filie niilui'd dIl ibed b\. Krixhni itid Kisifian t'J7liIrriiil kl.iitma it) I) " ;11 hiiffei 'Ax gl, tist trAIC i l4 the I icliis intl im tbaiiitei efi at. 11979) Hi li nkliairon mitute

atii,,ibens ait A 4tM, I 12)110 and DV' I1)W1 as dcesItied cItititne11d (final ttliicltiiiont: 10) min I its HIt. PH 7 'bs Vi laas aiitd I ilnnilin ( 14-91 and 10) niol kititte int 0 xxi ith 0I '1 mst-(i I P containing 1 2 -t(1 d p m. lipi t-tiferix tintlwwdt intoltIre sep~lfti .1 A 72001. t1 611) and 'PG II P ind 4 nm~ MnCI. axl xubstraie. I mik-Lch, 6\i1MI

DV 101) for 2 min I Ni lilie Sx't enxxen el at . 19801 Jhe intl t m xi- iblit s met h~l Isnthine tol pies em hiceakdosx ntcvItet %%ia' tell in push fun for 5 F11ii beftire %k it hdu dxlal ito ififhe cxc lie (1 MP formed. approlxumatel\ 31.10( k p ni -pies tnt leakagtie. t he ainnfas xx ec decapitated at the in- I'lI jclue (IP toi mnitor recos rs. and It pl en,\rmedixtaied ismsal tiniex. and 01 6 0.8 fitrm frontal hrain pueparat ion or homnogenisation buffer blanks.. in a to(iatsticex xleie prepared xl ith a tIsslue chopper at 4 U. 'ihe %olline of 50 tpl. When indicated. NaN, (final concen-atidaiiipirlinen xl ax dissected bet xlecn appt tuimate ti a:ion I mwi or NaF: (iunal concentr ation 20 mmi %ki x

lex Clx A 46W)1 and A OWlM. the ileleis tcch benx \k.ll, adided uirthe test tithes at the sitart oit the pitteincubaiiontiken .tnteftor tot A S400l. and the septu iin las taken be After ", nun pietin iAlmt on at 37 U. the reacitin mxas mt-litileen A 86(lit and A 6911il Konug and Ktippet. 19631 tinder ;tied bs additfitnil Gr I'llP. NlnCIt. and triated cslce (,NlP.microslcopic guidance. I the septiat %ample contained both and t er miniatedi alter m in ife ubat ion at 37 U b\ add itio n

J Nu..,' hem Vo .i 6i N, 1, MVR/

rd! 1\ )/ 1/1 C IiI \ 1% /Ril ! UA'lIB,1\ \1CII

- I 'T ~ (I 11,liric ci ~IIL. 19-4. (Iilkilicip t:I aL. 147S . arid thc1ii , It, i I In I j I -Al),li , I1 * l \k acic iht'rctotrr c sd in thle testI of' this Ntiid% I tic

p hI l , iii I' , . ' I., 'i id o ) . Iiiii 'It i.II silubic crr/%nme ltus pieptied fionm the t-Itidalo.nfl oi.i-I K 1 .. 11, k111illn. riultnricti el hhiied %Aivsical MichaclisN Menten ki-

Hot!i. i 1 A wIIiirC Sit A[(if Ot il' aIririrIrI nctics. \%il it Sill, lor G(11) of* 47 )ui StInmean of taskoiI".0 lilt a ~ddcd IS, 6 [111 DIIrtrlixexrrticrii dcitrnirratirrns. similar to pie, ioirslik reported -.al-

loiiII .d rd P rnd Hi l.drI0.,:Ir itkl~ IN. ric tics, fhorn iicni Iltajn CI io\ er and I crrendelli.'I'Ah I i-lc quA llitI.il S"1iir 19'76: rter et al.. 197S) .Addlition of' I miki NaN

kIia'd .d \tii9lta. _41 In ., ia.lkd i tir. t) tile different preparations increased the fnran\ late\Stici 0111 ,, I'.iMur plt S\k'liii I 1o~cd 1 t1 .1 1 11i\ cack.lase actiit in homogenates, and particulate

I .1 511 ( rMP II I .1liC %IAN IllifllItII iii 't ) I i!l fioctions 3.5- and IS-fold. respcctiael\ . but left thelIc I '. r IkII 1.l I il'It rid u 101 1 1 \ S.l,%It , It IIII' t I Int in Ii srlible acti% it ] amost uncha[ itnged Il abIle 11. Sii-

Ii Icr., Ii I om I.utcraiiid~ i1 lleiik I 1,. Ikl , larl\. pri cutbation \A.ith 11riton V IMX rut reased the

I - I iik 01 1' n ti " 11 11 'I I' lrlu lgtnatc anid particuilate actia ites, Nelectiikelsti~~~~~~~ ak'i'iI riltld.ble I I In t outrast, incuibation likith '10 rnit-Nal-

I i i IIa r lit' P I Strill tirn 1i1Q I ll 11 1 I ~. . 11 fid ntrl increase citvelte (I M P s~ uthesis in anx of thck 'l li'. 111\d Ii.i I I Siril .11111 'hil icdIL r0 IIill tr .littns(dalta nomt sho1As n. NaN, v%.v, al-L eflfectiac

ii'ii~'ii I)i I ),tsa "0 r il 1h,riI1iiit- c' .1 l1- 111 rcrGtnNetr ~IP xxuttiesis in honiogenatesI I, listSc I 2 (I turnl (t \ . i I I l 1,II I1 11-4 1 0 tld i .rid pirlitruirEi fi.ittons ( I Ifould) preparedilti or %k 11t'I ll I'l Ill, 1.-.lii .0 ii0 ( till (I ' I: iilit I ioin korirai lcsinet S.rmdatluprrtamcn (data from

p1 tISt V i 0 ~1 1 i. ) 1kIdII 11111 1 tlit lIt. I INC In I i 2 I. I hexec it-riliN all ,onf'kiii toI the propertiesl,. Ii- It( , I .i I il' NIL!( I IWi 1 ",(HiIll,-, prca- iotisl% des~iihed for Viianikkle criclase in

''ic .111d (It i i I .1.1 \int' a iP \Irl M I'. (nmmhi bta Inm and ttad. 1975. Kim-

.ill-Ii 1, d r,-1%, Ii 1 i-ri 5) 2 l 1.1 l Ii rita ci al.. 1975. I'toser et al.. 1978)

ri' i.r ipoli.i~,.1 1.:iIic, ti"iY p]SC (traractet nation of the neuIititmamrsrter com-K iii i ~'~i I' I P14 It T-, in ' I I I J1c 1ir position of the tecgions tinder stud\ ,hit\&ed that

stt I:; ri-i It Ir I" 1,,: i' 1-' isA 11 itirIg the nionolaniuc mnarker cnirne AA) aind the AChiii u t~.-'l' 'r . 'I rid il rcIll "j, marker cnixme ('hAT \kere concentrated in the nu-

'~ dI'.I d.' Ii - '.rirecleus accumbens and the cauidattopuitatnen. vi hulekSi It'I i, .1 .iii 'Il I lic i 11 ,11, 1 IlI "ti II ic the GABA marker eni~ me GAD) \Siits most aclia e in

5%ii 'iiI tii pi itiI i.'~' ntI l~iiu1 r 1

0641 the septim and nucleus accuinbensi (Tia ble 21. Thet, i' Ic I lk I " pi'd oll I t rip'll'-d ill xcliru gIlit amate uptake Capacity % as greatest in the

ikcri lini iip, iicd twiiiiuu k ill ri sirirpic Irui theiao ua e . h ciiy o 5-ice td~

10' 1 ii :1101upics turnl I"Iil 'liop'l.e and iiilu'rw:aaoncentrated in the septum and the nucleus ac-Iiii 11it Willi I(t it-i 1.dC ul t I 1' Il .ii1ii1 11.11 1 11t 11 UITIllens. as hie the general mitochondrial marker

siiihic Shi'tl I'.0i1" t-iic ateix Itransferase ilkas eisenly distributedp~t "ic nu~ulhisl x ii ii Ilsi is I hic 21 (;riailslatecc clase %lia most con,. ent rated

Ililn hIrilogt.ttesW from the catidatopuitlamen a-id nu-

HI- St I Sens i.cciiitnbens. mud this as as appat entlx due Iiltie actt atw~ itN oluble cnix nrc piesent in thesl nu-

,iiui\ tue ~ .Ir as~s Ins t I~rrtct ed i tis I I able 11 In dnntraAt. the patiitiltic guianil

n.1u0rsnxhu it)( i..tlltiltqritairtcti hithtbtion ' %k ~Li~t 55 as esenl distributSed in all tiegrons.0it1 liiiiit5 n~t's %k ith it I(I I P hulld hN~ thin as tit minus sed in utireai'ted pieparatrons or after

as i iiitiutigriplxol the put lied piodul tcteigeti estiactiron (lable HI.shised I' udrmicia txIn thle .\tilc ()Nt NI put lilhe eff-ects of karnit allid %scire tested in some

onkl ( not shoisn) li ither. all ~P in the put ufied detail i the catidatoupit.tmen taso da% s after injec-Ill olmi Ieru flhi d like iii licrntic k Nc c ( (1 .11 P on both lion of 1(0 nmol karnate . homiogenates from this nu.

lOoasc 1 0( and lDo\Aci I coluns.- \Ahilc more than cleuis had lost 60 70"; of the GAD and (ThAT ac-96', ol rte P 1 label co miriated as ith Ill (MP on tit'ities present. ashile guanylate cyclase %sas de-Itois el ') ateri phophiudesteraise treatment (data creased by 50 60'7 both in untreated (Fig. 1).riot hoiAniu. IitsN. all P libethl lit the pitfied prod- detergent-extracted or azide-aciivated homollgenatles1it1t . 1 p1C.OCILd to repiexIil ec'SI 11,0I ". 1-uther (H-g. 2) illhen compared with similarly treated con-sntdies shoited that gritnx ate cticlase actia-ity %as trol preparations. Carnitine acetyltransferase wasslighil) lorver in caudatoptitamen homogenates, pre- decreased by 30% and 5'-nucleotidase was de-pared in dithioth-icitol contltining buffer compared creased by 15%. w hile AAD activity was unchangedlioith preparations iiiithout the thiol reagent (40.5 * (Fig. 1). Separate analysis confirmed that most of1 1 acisusN 66.5 7.9 rol tmin mg protein, mean , the soluble guanylate cyclase had disappeared after% I.Al.. n 8). Blhot-corntaiing picparattons asoid this lesion (Fig. 2). In contrast. the untreated par-a possiblet oikidatiaei tLctianaton of gitanylrite cxelase ticulate preparation displayed significantly higher

i eAm . VolI 16. N.' 1. I981

88

I , IF I., ti t %fill, if I nd C C' 'i,/, CIi'~~~h Clr d ~ 141"~ CICC

III I JI b ' 1 i1 illi pil, 1 /C

( ji, laic cCi,CL ~ ?1 81 1 elin1

Ill o flC' atie 401 3 4(o0 11 3 ' 1 9

I rCiCn X I W~ 149(0 6 A 171 7 16 7 1 2 7'

NaN, 296 -171)

Stiluhibi 240.8 12 6 224 7-246 t,-

I Frilon X - I(1W 241 2 19.3

Na , 248Al 2 1 6

PailiCUlaile it, I 1, 6IS 2 - ) S 14 ?4- 1 to

1I ilt~lX I1XW 123.0) 13 4 144 9-47 101l 0 217

NaN225.6 141;

ResultS ate1 expic-%%d as% proitil in rng pi~iein aind are presentfled as flCican

% is S.irnple% ere prepaieJ in hule tac fiilfg 1; nisi da hoit lre iiol. 1 fltin

X 100l i NaN ,CIC.Ied j, deccihed in IC si

I ARI+I 2. R, gCCiCCJ thip~illmliJ1C) 4I A Al). I -I (1r4 ), It/CII)JC

off Si Inl 61 ill

AA) 0 91 1 1 0I YS - to 14 to 1' 0 01

ChAl 1f 11124 2'~-i 1 1 oII IW0 1IA

GAD 2 Ii I0 1 (2;120 0 28 4 11?1 I I I

" Nii'e~itias 5,41I 0 62 S 4l ) 0 14 9 W2 - (1 1I

( arnitine 34 A 0I 1 A7 4 4 4 - I

ao~CI5 Iran~lcrii~c(,iCCi.,rnale Ctiilke 0I069 0 (MIS (0044 - (I I 0 l44" ( 0 W4

Hesiilt% aire e~pfc%eid CS flflI iin ii) portiein and are presented a, mean

]5 2 DAYS AFTER LE90ON (N=7) 0

S7 DAYS AFTER ILESION (Nr6)

0

z

50 50

GAYAE AAD -~ChAT AO -5N LE -CRNITINE GLUTAMATECCAETIDASE ACETYL-T UPTAKE

FIG. 1. Biochemical activities in rat caudatopulamen homogenates following local inhection of 10 nmol kainic acid Rtesultspresented as mean Si8E m percent of control (cf Table l and 21 Significantly different Irom control. single asterisk. p < 0 05.double asterisk. p < 0.01.

J Ne~iwrai. Vol M. No. 1. 1981

89

I\)/ 4 11-- ( )(I I.S1. /V RA /l 0R1'LBRAIN\ AI( f(1 2.?7

riji.ii lite L: '%clas a, n\ it\. tha~n ,imilair Control pier- trast. gtian~ latec~ ls in the septum AaJ% un-it ions at I hvs iiinc. '.k lilc the NiN.,-activatd Changed after %cptail Injection of 10 nmol kainat' e

pal tic iile icti' % its a lightIl and the dcii gnt- xs hcn tested both in homogenates (Fig. 4) and inc tacted painic iite ac t 'kas significant I. de- soluble and particulate fractions (not shos n i. Thecit-cased oipai cd \ it h si nilar'. tiwaled cot ntrol kainaic lesion of the nucleus accumbens destroyedpr cp itions [1ig 2). 75 K5'( (if the (JAI) and ChAT acti,.ities present.

Sc' cii dai' ' atici the lesion. (6Af aind ('hA V ac- s% hie scptal injection% deslro~ed 70'7 of the GADt i' tics, in the "adiipraun~Ci e leduced bN acti'.it) but left the ('hAT activity in the septum80) 851'. c:iIII11 n\in ,c\It infrae.a rcdticed unchanged. 5'-Nuclcotidase had increased consid-b\ 40 . ic5 icctdi'cn'x'~as gnificantly etably. 'A hue carnitine acetyltransferase \ks~ de-

i.ICased to about 150 (' of the initial acti'. it) (Fig. creased in both legions. (Figs.. 3 and 4).1). A Al acti'.iti, and glultmate uptake \ke re un-chanpt.d. At this timec point. puan) latecy~clase in the DISCUSSIONdillicic n homnogcnate plc paratlon. wkas dcreasedbN 70 We'. \while the soluble acti'.ity \%as de- The aim of the present insestigation \kas to iden-cmca'.ed b) about 80';. The increase in the untreated tify s.onic properties of the guanylate cyclase ac-paittiuidltc activity had disappeared. and at consid- tivitie'. in the rat caudatoputamen. nucleus ac-ciaibic decrease wkas found in this fraction also irre- cumbens and s.eptum. In the first part of the study.spe1(cti'.c of treatmrent t(Fig. 2). subcellular fractional ion demonstrated guanylate

[he otheri nuclei w crc .inal) sed 10 day'. after c'.clasc aicti'.it% in both soluble and particulatekariiat1C nli ictlon'.. and detergent estraction wAas piepau ations firm all three regions.. %kith the solubleloutincl\ used oin homogenates and particulate acti'.it) most concentrated in the caudatoputamenfimctlons to cspres'. maximal particulate guany'late and nuLcleus accumbens. The dis.tribution of thisL S c lasev act,'.ity . InJection of 5 nmol kainate into the soluble guanylate c~clase coincided with the activ-nuicleus% accumbhens decreased guanylate cyclase ity of the cholinergic marker ChAT (Fonnum et al.,aicti'.it) in suich I, ton iwated homogenates from 1977). but not \kith the high-affinity uptake ofthis nucleus by 70';. this decicase representing an glutamate. a ptitti'.e marker for glutamnate nerveappi .' imnae 90"; fall in the soluible actis it) and a terminal'. W onnum and Walaas. 1978; Walaas andW"t till Ii the particulaite acti'.iti, W ig. 3). In con- Fonnum. 1979). Interestingly. acetylcholine has

150r - 2 DAY7S AFTER tt- SION (Nz6)10

S7 DAYS AFTER LESION (N=6)

Rioo - -- 100P.-z0

z

so- C,.50

1.1. TRITON 1mM NaN3HOMOGE NATE SOLUBLE PARTICUL.ATE FRACTION

FRACTION

FIG. 2. Guanytate cycilase in homogenates and subcettutar fractions from rat caudlatoputamen after locat injection of 10 nmolkainic acid Samptes prepared in 30 mm-Tris-HCt. pH 7 5 ptus 5 mm-dilhiothreitot Results iepresefll percent of similarly treatedcontrol preparations (Table 1), and are given as mean -S3E M Singte asterisk, p , 0 05. double asterisk, p < 0.01.

NeKu',.hem . Val 30. No. I. I9WI

90

2381. ti AL.A.A%

?s-505

200-20

_J FIG. 3. Biochemial activities in rat nlu-t0- -15 cleus accumnbens '10 days after local

Z.15 injection of 5 nmol kainic acid0Q Ij Guanylate cyclase was analysed in

U. homogenates (HOM). soluble fractions0 (SOL) and particulate fractions (PAR).

Z the other activities were studied inW ~homnogenates Results presented as

I in Figs 1 and 2 5 -NIJC. 5 -nucleo--J --10tds AAT. carnitine acetyltrans-

I ferase Significance single asterisk.p 0 05. double asierisk. p 0 01.n 6

50- -.. 50

IIMMSt.PR GAD ChAt 5-NUC CAR-ATGUANYLATE CYCLASE

been suggested to increase cyclic GMP levels in the ports. In the caudatoptamen. the local ACh andcaudatoputamen through an activation of the solu- GABA neurons %%ere destroyed and the afferentble guanylate cyclase only (Hanle) and Iversen. monoamnine and glutamate fibres %%ere intact I week1978). In contrast, both the untreated and the more after kainate injection. In addition. 5'-nucleotidase.active detergent -extracted paticulate guanylate cy- which in the brain appears to be confined to glialclase activity was evenly distributed between the membranes (Kreutiberg et al., 1978). increased innuclei, unlike the neuronal markers. This difference activity in parallel %kith the impressive reactivein distribution pattern between the two guanylate a~trocytosis that follows kainate lesions (Coyle etcyclases is thus in general agreement w~ith previous al.. 1978). It is therefore suggested that an increasestudies from other brain regions (Nakaiawa and in the activity of this enzyme might reflect prolif-Sano. 1974; Greenberg et al.. 1978. Troyer et al., eration of glial cells. Indeed. injury-induced pro-1978). liferation of 5'-nuc leot idase -staining glial cells has

The cellular localization of guanylate cyclase in been reported in other regions (Kreutzberg et al..these regions was then studied with the kainic acid 1978).lesioning technique. Previous studies have de- The destruction of local neurons and accom-scribed the local neuronal necrosis and glial cell panying gliosis led in the caudatoputamen to a con-proliferation induced by this agent in the caudato- sistent decrease in the homogenate guanylate cy-putlamen (Hattori and McGeer. 1977: Schwarcz and clase activity, irrespective of pretreatment, and atCoyle, 1977. Coyle et al., 1978). the nucleus accum- both 2 and 7 days after the lesion. Thus, most of (hebens (Walaas and Fonnum. 1979). and the septum guanylate cyclase activity in the caudatoputamen(Malthe-Sorenssen et at.. 1980). The present appeared to be concentrated in local neurons. Inbiochemical results are in agreement with these re- agreement, both the soluble and the detergent-

J Newrorhrm. Vol X6 No 1. 1981

91

0

FIG. 4. Biochemical activities in hornog'-,natns0 from rat septum 10 days after local injection of 10

Z nmol kainic acid Results presented as in Fig IUi ~Single asterisk. p 0 05. double asterisk. p 0 01.

505050 50

GUANYLATE GAD ChAT -i 5NuC LE0- CARNITINECYCLASE TIDASE ACETYL-T

extracted particulate a~ti~j ies decreased consid- ofdlegeneration. A complete characterization of thiserabl) after kainate lesion%, oin ltle ,%:itles sintilar it) phenomenon %k ill require fkiriheir -. uds.the rapid degeneration of neurtonal c) tosol aind the In conclusion, both the soluble and pArticulatemore sloly esolsing degezneration seen in thc par- gian)latc cyclas-es appear to be concentrated inric tiltic nerke cell component% after kainatie lesions local neuctrons in the rat catidatoputamen. A prefer-44 ftle catdatoputamen I Hauttori and McGeer. 1977. critial neuional h'cali/ation for guan~ late cyclaseCov~letc al.. 1978). The non-detergent-treated par- ha% also been suggested from ptckious studies ontuculate guan~late cyclase prescnied a mote compli- soluble gtian~klate c~clase fiom %khole rat braincated rtesponse. ht)%eser. In most tissues%. this f(Gidis and Moigan. 1973) and from studies usingp itictilautc eniymc is present in a partly occluded cell culttures from chick brain (Goridis et al.. 1974).form, and the full actisity is only expressed after Ho%%eser. a glial localization tias recently -.ug-detergent treatment (Kimura and Murad. 1975; gested. based oin an increase in apparent guanylateDurham. 1976; Goldberg and Haddox. 1977). In cyclase actis ity in kainate-lesioned rat cau4a-,*ic, of the consistent decrease in such detergent- toputamen (TjOrnhammar et al., 1979). The differ-extracted particulate actisity. it was not expected ent guanylate cyclase methods %cre therefore corn-that a significant increase in the non-detergent- pared during a collaboration with the latter authors.treated parliculate guan% late cyclase activity would Despite using aliquots from the same homogenates.be found 2 days after lesion Thus, despite an ap- the radiotimmunoiassay analysis performed by Tjbrn-parent decrease in total particulate guanylate cy- hammar et al. %till sho%%ed a pronounced increaseclase. the kainate lesion also appeared to induce a in apparent guanylate cyclase activity. %hile thereversible, detergent -sensit ive activation of the re- present radiochemnical method showed a W7 de-maining particulate eni% me during the initial phase crease in activity 2 days after lesion (results not

J Aru.,,.remi. Vol J.6 No. 1. /of/

92

240 1. WALAAS

shown). The cause of this discrepancy has, then, Bohme E., Jung R., and Mechler 1. (1974) Guanylate cyclase innot been identified. However, the apparent reliabil- human platelets. Methods Enzymol. 38C, 199-202.

ity f te prsen radochmica mehod Krihna Broch 0. J., Jr. and Fonnum F. (1972) The regional and subcel-ity f te prsen radochmica mehod Krihnalular distribution of catechol--methyltransferase in the rat

and Krishnan, 1975; Craven and De Rubertis, 1976; brain. J. Neurochem. 19, 2049-2055.Birnbaumer et al., 1979; this paper) suggests that Coyle J. T., Molliver MI. E., and Kuhar M. J. (1978) In situ in-

radjoimmunoassay of cyclic G MP formed in jection ofkainic acid: A new method for selectively lesioning

kainte-esioed audtopuame homgentes neurona) cell bodies while sparing axons of passage. J.kainte-esi~ed audtopuame homgentesComp. Neurol. 180, 301-324.

might be subject to interfering factors. Craven P. A. and De Rubertis F. R. (1976) ChromatographicThe results from the nucleus accumbens extend separation of cyclic guanoslne 3'.5'-monophosphate from

our previous studies (Fonnum et al., 1977; Walaas guanylate cyclase reaction mixtures. Anal. Thor/tern 72.

and Fonnum, 1979) and show that the neuro- 4549Deguchi T., Amano E.. and Nakane MI. (1976) Subcellular distn-

chemical characteristics of this nucleus are virtually bution and activation by non-ionic detergents of guanylateidentical to those of the caudatoputamen. Similar cyclase in cerebral cortex of rat. J. Neurochern. 27.guanylate activities were found in all types of 1027-1l034.

preparations, and all neurochemical markers re- Durham J. P. (1976) Guanylate cyclase: Assay and properties ofspnddsiilry okainate lesions. Thus, both the particulate and supernatant enzymes in mouse parotid,

spondd siilary toElr. J. Biochem. 61. 535-544.guanylate cyclases are predominantly located in Ferrendelli J. A. (1978) Distribution and regulation of cycliclocal neurons in the nucleus accumbens, as in the GMP in the central nervous system, in Advances in Cyclic

caudatoputamen. In the septum, however, kainate N5.ucleotide Research. Vol. 9 (George W. 3. and Ignarro

selectively destroyed the GABA neurons but left L. J.. eds). pp. 453-464. Raven Press. News York.Fonnum F. (197S) A rapid radiochemnical method for the deter-

the guanylate cyclase activities intact. Thus, most mination of choline acet yIt ransfe rase. J. Neurochern. 24,of the cyclic GMP synthesis in this region appears 407-409.to take place in cells not destroyed by kainate, such Fonnum F. and Waiaas 1. (1978) The effect of intrahippocampal

asgilcells or the ACh neurons in the media) sep- ksainic acid injections and surgical lesions on neurotrans-as glialmitters in hippocampus and septum. J. Neijrochem 31.

tum (Zaczek et al., 1979-. Malthe-Sorenssen et al., 1173 - 1181.1980). Kainate sensitivity may be restricted to Fonnum F. and Walaas 1. (1979) Localization of neurotrans-neurons receiving glutamate fibres (Herndon and mitier candidates in neostnatum. in The Neostriatum (DivacCoyle. 1977: McGeer and McGeer. 1978; Biziere I and Oberg G E., eds). pp 53 -69 Pergamon Pre%-s. O0.

and Coyle, 1978; Kohler et al.. 1978; Malthe- frFonnum F . Storin-Mathisen i J. and walberg F. 19701 Gluts-

Sorenssen et al.. 1980). In the septum, such mate decarboxylase in inhibitory neurons. A stud) of the

cnanvrlo guryaeccaeatvtacnRe2029glutamate-innervated neurons, therefore, appear to enz) me in Purkinje cell axons and broutons in the cat Brain

clusion which argues against a general 'second mes- Fonnum F . Walaa% I and Iversen E. (1977) Localization ofsne'role for cyclic GMP in glutamate neuro- GABAergic. cholinergic and aminergic structures in the

senger~mesolimbic system J Neurochern. 29. 221-230transmission (Mao et al., 1974. Biggio et al.. 1978) (ioldherg N D and Haddox MI. K (1977) Cyclic GMPor kainate neurotox icily (Honlegger and Richelson. metabolism and insolveinent in biological regulation. Anna.

1977). Re% Bimchem. 46. 823 - 896.Goldberg N. D.. Graff G.. Haddox NI. K.. Stephenson J. H..

Glass D, B.. and Moser NI. E. (19781 Redox modulation ofACKNOWLEDGMENT splenic cell soluble guanylate cyclase activity: Activation by

hy~drophilic and hydrophobic oxidants represented by as-I thank Liv Eliassen for valuable technical assis- corbic and dehydroascorbic acids, fatty acid hydro-

tance, Dr. 0. Walaas for help with the thin-layer peroxides. and prostaglandin endloperoxidecs. in Advances inchromatography and E. Odden for help with the (St (ii Nucleiitide Research. Vol. 9 (George W. J. and lg.

septal esions.narro I.- J .. eds). pp. 10 1 - 130. Raven Press. New York.septa lesins. (ondis C. and Morgan 1. G. (1973) Guanyl cyclase in rat brain

suhcellular fractions. FEBS Lett. 34. 71-73.REFERENCES Gondis C . Massarelli R.. Sensenbrentser M.. and Mandel P.

(1974) Guanyl cyclase in chick embryo brain cell cultures:Avruch J. and Wallach D. F. H. (1971) Preparation and prop- E'.idence of neuronal localization. J. Neurochem. 23.

erties of plasma membrane and endoplasmic reticulum 135 118fragments from isolated rat fat cells. Bioc him Biphv'.. Ai Ia Gireenberg L. H.. Troyer E., Ferrendelli J. A.. and Weiss B.233. 334.- 347. (1978) Enzymatic regulation of the concentration of cyclic

Bigo G., Cords M. G.. Casu M.. Salis M.. and Gessa 0. L (MP in mouse brain. Neuropharif ncology 17. 737-745.(1979) Disappearance 'if cerebellar cyclic ONIP induced by Greengard P (1978t (Cvclii Nucleotides. Phosphorylated Pro-kainic acid. Brain Re.. 154. 203 - 208. teinsi and Vironal Function. Raven Press. New York.

Birnbaumer L.. Torres H. N.. Flawit NI. NI.. and Fricke R. F. Hanley MI. R. and Iversen L. L. (1979) Muscarinic cholinergic(1979) Improved methods for determination of guanylyl cy- receptors in rat corpus striatumn and regulation of guanosineclase activity and synthesis of (a'PIOTP. Anal. Thor/ter cyclic 3 .5 -monophosphate. Mol. Pharmacol. 14.246-255.93. 124- 133. Hardman J. G. and Sutherland E. W. (1969) Guanyl cyclase, an

fliziere K. and Coyle 3. T. (1978) Influence of corticostriatal af- enzyme catalyzing the formation of guanosine 3.5-.mono-ferents on striatai kainic acid neurotoxicity. Nei~rosc,. L~eff. phosphate from guanosine triphosphate. J. Biol. C/tem. 244,8. 303-310, 6363 -6370.

f. Plewnoehem.. Val. A6. No f. 19I8

93

GUANYLATE CYCLASE IN RAT FOREBRAIN NUCLEI 24!

Hattori T. and McGeer E. G. 11977) Fine structural changes in cyclic guanosine monophosphate in rat cerebellum: Possible

the rat striatumn after local injections of kainic acid. Brain involvement of putative aminlo acid neurotransmtters. Brams

Re.%. 129. 174 -ISO. Resc. 79. 510 -514.Herndon R. M. and Coyle J. T. (19771 Selective destruction of McGeer E. G. and McGeer P. L. t 19781 Some factors influencing

neurons by a transmitter agonist. St tenie 198. 71 72. the neurotoxicity of intrastriatal injections of kainic acid.Hodges J L. and Lehmann E. L. 11964i Bai( C(epts oft Neciro hent. Resc. 3. 501 - 517.

Probabilit ' and Statistics. Holden-Day. San Francisco. Nakazawa K. and Sano M. 11974) Studies on guanylate cyclase.

Honegger P. and Richelson E. t1977i Kainic acid alters neuro- A new assay method for guanylate cyclase and properties of

chemical developmeni in fetal rat brain aggregating cell cul- the cyclase from rat brain. J. Biol. Chem. 249. 4207 - 4211.tures. Brain Res. 138. 580 - 584. Randerath E. and Randerath K. (1964) Resolution of complex

Kimura H.- and Murad F. (1975) Two forms of guany late cyclase nucleotide mixtures by two-dimensional anion-exchange

in mammalian tissues and possible mechanisms for their thin-layer chromatography. J. Chromatogr. 16, 126 -129.

regulation. Msetabolisi 24. 439-445. Salomon Y.. Londos C.. and Rodbell M. (19741 A highly sensi-

Kimura H.. Mittal C. K.. and Murad F. 11975) Increases in cyclic tive adenylate cyclase assay. Anal. Biochem. 58. 541 -548.

LiMP levels in brain and liver with sodium azide. an activa- Schwarcz R. and Coyle J. T. (1977) Striatal lesions with kainic

tor of guanylate cyclase. N.aure 257. 700-702. acid: Neurochemical characteristics. Brain Res. 127.

Kobayashi Y. and Maudsley D. V. 11974) Biological Applica- 235-249.

lions of' Liquid Scintillation Ccicntinxv. Academic Press. Sterni S. H. and Fonnum F. (1980) Acetyl-CoA synthesizing en-

London. zymes in cholinergic nerve terminals. J. Nemrochem. 33.

Kohler C.. Schwarcz R.. and Fuxe K. (19781 Perforant path Iran- 249-254.sections protect hippocampal granule cells from kainate le- Stone T. W. and Taylor D. A. 11977) Microiontophoretic studies

sion. .Xurcsci. Left. 10. 241-- 246. or the effects of cyclic nucleotides on the excitability of

Konig J. F. R. and Klippel R, A. 11963) Thr Rat Brain. A neurones in the rat cerebral cortex. J. Phvsiol. tLond.) 266.

Sterecctaxic Atlas of the Forebrain ancd Lower Parts of the 523-543.

Brain Stemi. Williams & Wilkins. Baltimore. Tjornhammar M -L.. Schwarcz R.. Bartfai T.. and Fuxe K.

Kreutzberg G. W.. Barron K. 0.. and Schubert P. 11978) (1979) Guanylate cyclase activity increases after kainic acid

Cytochemical localization of 5 -nucleotidase in glial plasma lesion of rat striatum. Brain Res. 171, 567 -572.

membranes. Brainc Res. 158. 247 -257. Troyer E. and Ferrendelli J. A. (1976) Two forms of guanylate

Krishna G. and Krishnan N. (1975) A rapid method for the assay cyclase in cerebellum. Fed. Proc. 35. 456.

of guanylate cyclase. J. Cvcl 14( clec'tide Ri's. 1. 293 -302. Troyer E. W.. Hall 1. A.. and Ferrenkdelli J.- A. 1j1978) Guanylate

Lee T. P.. Kuo J1. F.. and G3reengard P. (1972) Role of muscarinic cyclases in CNS: Enzymatic characteristics of soluble and

cholinergic receptors in regulation of cGMP content in particulate enzymes from mouse cerebellum and retina, J.

mammalian brain, heart muscle, and intestinal smooth mus- Neuirachem. 31. 825-833.

cle Prot Nall. Arad. Srci. LI'SA 69. 32147- 3291. Walaas 1. 11979) Topographical and subcellular distribution of

Lowry 0. H.. Rosebrough N. V. Farr A. I... and Randall R. J. guanylate cyclase in glutamate-innervated brain nuclei. Acta

4 195 1) protein measurement with the Folin phenol reagent. Lnicc-rsitatis Tamperepecis trampere., Finland). Sen. B.

J Blio ('hem. 193. 265 -275. 10. 37.

MaltheSerensseti D.. Odden E_. and Walaas 1. 119801 Selective Walaas 1. and Fonnum F. (19791 The effects of surgical and

destruction by kainic acid of neurons innervated by putative chemical lesions on neurotransmitter candidates in the nu-

glutamergic afferents in septum and nucleus of the diagonal cleus accumbenis of the rat. Neuroscience 4. 209-2 16.

band. Brain Resc 182. 461 465 wc-ody C. D.. Swatz B. E.. and Gruen E. (1978) Effects of

Mao C. C. and Guidotti A. 119741 Simultaneous isolation of acetylcholine and cyclic GMP on input resistance of cortical

adenosine 3 .5 -cyclic monophosphate tcAMP) and guano- neurons in awake cats Brain Res. 158. 373-395.

sine 3 .5 -cyclic monophosphate icGMP) in small tissue Zacek R.. Hedreen J. C.. and Coyle J. T. 419791 Evidence for a

samples. Anal Biccihem. 59. 63 68. hippocampal-septal glutamergic pathway in the rat. Exp.

Mao C.- C. Guidoiti A . and Costa F 11974) The regulation of A euriil. 65. 145 - 156.

95 PAPER VIII

Brtlw Re .serch, 177 (1979) 325-336 325i' Elsevier.North-Holland Biomedical Press

THE DISTRIBUTION AND ORIGIN OF GLUTAMATE DECARBOXYLASEAND CHOLINE ACETYLTRANSFERASE IN VENTRAL PALLIDUM ANDOTHER BASAL FOREBRAIN REGIONS

IVAR WALAAS and FRODE FONNUM

Norwegian Defence Research Esrablishmtent, ivision For To.ricolog)', N-2007 Kjeller (NorwayI)

(Accepted January 4th, 1979)

SUMMARY

The distribution of choline acetyltransferase (ChAT) and glutamate decarboxy-lase (GAD). and the histochemical reaction for acetylcholinesterase have been studiedin the basal forebrain and globus pallidus of unoperated rats and in rats with anelectrolytic lesion of the nucleus accumbens. ChAT was highly concentrated in thesubstriatal region, the neostriatum and the lateral part of the rostral substantiainnominata. The strongest intensity of staining for acetylcholinesterase was found inthe substriatal grey and the neostriatum. Very high GAD activity was found in thesubstantia innominata, being even slightly higher than that in the pars reticulata ofthe substantia nigra. The lateral preoptic area, the bed nucleus of the stria terminalisand the globus pallidus also showed high activity of GAD. After lesions of the nucleusaccumbens the activity of GAD decreased significantly in the substantia innominataand in a restricted part of the rostroventral globus pallidus, but not in the otherregions studied. ChAT activity and acetylcholinesterase staining were unaffected in allregions.

The results indicate that a dense GABAergic projection originates in the nucleusaccumbens and terminates in the rostral substantia innominata and rostroventralpart of the globus pallidus. The study gives neurochemical support to the suggestionthat nucleus accumbeis may be regarded as a ventral part of the neostriatum and thatthe rostral substantia innominata may be regarded as a ventral part of the globuspallidus.

INTRODUCTION

The basal forebrain of the rat is composed of several indistinctly outlined cellgroups. among them the preoptic nuclei. substantia innominata and the bed nucleusof stria terminalis. However, despite the close proximity of these nuclei and similarities

~LL~A.G i.igmagr1jS

96

326

in cytoarchitectonics, they receive afferent projections from different parts of thebrain. The bed nucleus of stria terminalis is innervated predominantly by projectionsfrom the amygdala and subiculum,3.'k2 while the preoptic areas receive their majorinnervation from the septum and from the bed nucleus of stria terminalist 11. Thesubstantia innominata is especially difficult to distinguish from the lateral preopticarea3. ' .1i 6i ,3 but recent work has shown that the rostral part of the substantiainnominata as defined by Swanson31 receives a dense projection from the nucleusaccumbens septi' .1 6 and from the olfactory tubercle15 . The accumbens projection,which seems to continue into a restricted part of the rostroventral globus pallidus :3 1.3

is not shared by the other basal forebrain nuclei, and Heimer and co-workers havetherefore introduced the terms 'ventral striatopallidal pathway' and *ventral pallidum"to describe these fibres and their termination I,.

We have previously found a high activity of the ,,-aminobutyric acid tGABA)marker enzyme, glutamate decarboxylase (GAD. EC 4.1.115), and an intermediateactivity of the acetylcholine (ACh) marker enzyme. choline acetyltransferase (ChAT,EC 2.3.1.6), in what was tentatively described as substantia innominata and whichshould correspond to the "\erltral palliduin' . In the present communication theseobservations ha%e been extended. The detailed distribution of ChAT and GAD andthe acetylcholinesterase (AChE, EC 3. 1. 1.7) staining pattern in the cell groups of thebasal forebrain have been studied, special attention being paid to precise dissection.In particular, we have investigated whether the concept of 'ventral pallidum' with a'ventral striatopallidal' projection is supported by neurochemical data. The tran,-mitter in the '\entral striatopallidal' libre. has also been identified. A preliminaryreport has been presented : ,.

EXPERIMENTAL PROCEDURESI Surgical operations

Male Wistar rats of 170-200 g body weight. were anaesthetized with a mixture ofdiazepam and Hypnorm Vet 'Mekos', and the animals \%ere placed in a Kopfstereotactic apparatus according to the atlas of K,nig & Klippel2'2. A tungstenelectrode (0.5 mm diameter), insulated to the tip, was lowered through a burr hole intothe nucleus accumbens at coordinates A: 9.5, L: 1.0. V: -- 1.0, another electrode wasclipped to the edge of the wound, and lesions were made with an alternating current(0,2 Al for 3-10 sec. Three sham-operated animals were subjected to the sameprocedure, but no current was applied, (All animals were operated on and, or sacrificedbetween 15 April and 15 May).

Tiosiw pr(parationFive to seven days after operation, both lesioned and sham-operated rats,

together with nine unoperated animals were decapitated between 09.00 and 11.00 h.The brains were rapidly removed and frozen with a CO., jet. and frontal 40pm sectionscut in a cryostat as previously described". Sections from the whole rostrocaudalextent of the nucleus accumbens were thawed on microscope slides, air dried and

97

0 7 CC327

-~ Ac -~ 10 050 Mm

TOL

1-i. (onpo.it. rasagssh~sag heesen o th eecroytc esinsofnule cc ubninrt

RI R Ac~nuclus acumbns:CA. ommisuraanteior C.corps calosu: .41 dtemtamenoa nttr~i ltator~nuceus OT.olfcto~ tucrce: Y. prirrniCore L aea etm

SM.mcdii epam T. dagoal an: TL. ateal lfator trct VL laera vetrile.Numer

indicate ~ ~ ~ ~ ~ ~ Ac ditac infotoCnAarl ie oiidfo KngadKirl

stated iththininformapingof he esios (ig.I).Begnnig a leel 75022sectons ncluing he aterl propti are. gobuspalldus r sustatnir r

eihrclecefezedidad trda reiul esrbdt o hwdomirocpeslds ndsaie fr ~E 2 , rwih otiecel stin Ltinno a

Griinw8.6d mims)

DiS.ccAioTheregonsto e sudid eredisectd udera inoula mirosopewCh

allowed~~T corsodn tie n reedid etost eFee iutnosy.Base onthe orpoloicalchaacteistcs n ou stinedsecion (Fi. 2 th

folowig rgins eretakn ut:a smpe fom he eotritumimmditel dosa

98

328

VL~

CA

L A.

abrvain as in Fi.1.0

AD-AI07 453 NORWEGIAN DEFENCE RESEARCH ESTABLISHMENT KJELLER F/6 6/lbNEURD RANSMITTER MECHANISMS IN THE NUCLEUS ACCUMRENS SEPTI AND -ETC(U)JUN Al 1 WALAAS

UNCLASSIFIED NDRE/PU8L-Al1501 NL.2*fffffffffff

''1111 111112.2111012.5 flI~ ~JIB1 1 .6

MJCROCOPY RE SOlUJTICN I 1 t, CHARTNAT1 NAIkI- ii. i -. AND AH

99

329

IBST

LPO

Fig. 3. AChE-stained frontal section (40 1m) of basal forebrain regions in the rat, taken from levelA 6900 .GP, globus pallidus. Other abbreviations as in Figs. I and 2. - 37.

to the commissura anterior and extending from the bed nucleus of the stria terminalismedially to the medial border of the substriatal grey rcgion" laterally, the bed nucleusof the stria terminalis both ventral and dorsal to the commissure30, the lateral preopticarea immediately ventral to the bed nucleus, the substriatal grey region, and the rostralsubstantia innominata I)ing between the last two regions (Fig. 2). in unoperatedanimals the last sample was divided into lateral and medial parts. At more caudallevels, the rostroventral tip of the globus pallidus adjacent to the bed nucleus wasdissected instead of the neostriatum. In addition, more caudal samples from thecentral part of glohus pallidus and from the rostromedial pars reticulata of substantianigra were taken for comparison.

Biochenical analysisThe samples (I -5 pg dry weight) were weighed on a fishpole quartz fibre

balance 3, and assayed for ChAT or GAD activity with previously reported radio-chemical methods 9. t.I2. Statistical significance between results was assessed with thenon-parametric Wilcoxon two-sample test17.

RESULTS

Histological staining of the basal forebrain (Fig. 2) revealed several neuronal

100

330

TABLE I

Distribution of choline acetyltransferase and glutamate decarhoxylase in the basal Jfrebrain and relatedregions of the rat

Results from 3-9 animals, presented as mean ± S.E.M. (number of samples).

Region Choline acety! transJerase Glutamate decarhok lase(/imnol/'h/g dry wt.)

Neostriatum 84.3 ! 9.7 (4) 200 - 22 (6)Globus pallidus,

rostroventral part 20.9 1 3.2 (4) 700 1 29 (6)central part - 546 -1 II (5)

Bed nucleus of stria terminalis,ventral part 27.8 ± 1.3 (5) 619 , 63 (7)dorsal part 21.9 ± 2.2 (5) 446 -f 50 (7)

Lateral preoptic area 32.2 ± 3.3 (4) 755 A 75 (5)Substantia innominata,

medial part 34.8 ± 4.3 (4) 1075 j 71 (6)lateral part 79.7 ± 8.5 (5) 957 ± 106 (8)

Substriatal grey region 150.0 f 7.9 (5) 291 A 43 (5)Substantianigra, parsreticulata - 892 ± 70 (23)

populations showing different characteristics under optical microscopy. The ventraland dorsal parts of the bed nucleus of the stria terminalis both consisted of rathersmall, densely packed neurones. Ventral to this, the lateral preoptic area containedlarger, more scattered cells, while the substantia innominata displayed irregular, largeand unevenly scattered neurones (Fig. 2). Lateral to the substantia innominata andventral to the temporal limb of the anterior commissure, the substriatal grey regioncontained small neurones, which resembled those of the neostriatum. The rostro-ventral part of globus pallidus, dissected at a slightly more caudal level, had similarcell morphology as substantia innominata. AChE stained sections (Fig. 3) demon-strated a dense staining in the neostriatum, the substriatal grey and the horizontalnucleus of the diagonal band. In contrast, the bed nucleus of the stria terminalis wasalmost devoid of staining product, while the globus pallidus and the lateral preopticarea had a few AChE positive ne,. 3nes.

The substantia innominata had both some neuropil staining and AChE-positiveneurons (Fig. 3). The latter are probably a dorsal extension of the dense aggregationsof such cells in the horizontal diagonal band nucleusS.ts. The distribution of ChAT inthese regions (Table 1) paralleled, with few exceptions, the AChE pattern. Highactivity was found in the substriatal grey and the caudate-putamen (neostriatum),while the bed nucleus and rostral globus pallidus displayed low activities. The lateralpreoptic area and medial part of the substantia innominata had slightly higher ChATactivity than the bed nucleus, while the lateral substantia innominata had approxima-tely the same activity as the caudate-putamen.

The GAD activity was distributed differently. Both parts of the substantiainnominata showed very high activity. These were clearly higher than any pat ts of the

101

331

TABLE 11

Glutainate decarboxylase in substantia inmornifata and glebma pallidus following electrocoogulationoftnucleus occmbe,,s

Results presented as mean ± S.E.M. (number of samples). Lesioned side significantly lower thanunlesioned side: *** P -_ 0.001; 00 P <. 0.01; 00 P <~ 0.02; $ P . 0.05 (Wilcoxon two-sample test).

Ani- Substiwaja jlmnottina: a Rostroventrol globus pallidus Central gobus pallidus

nuti Unesioned side Lesioned side (Unles joned Lesioned side Unlesioned Lesio ued side(1111o0/ImIg '~of unlesioned side (11moll 0 of un- side (smol/ % of un-drY wt.) hig dry wt.) lesioned h/g dry wt.) /esioned

RI 1 845 ±24(6) 52±46)** 582 ±7 (3) 85 ±18(3)R2 1192 ±73 (8) 29±6(6)** - -

R3 841 ± 82 (3) 40±~7(3)* 540±47(3) 60±9 MO) - -

R4 1099 120(4) 49±6(5)** 694±44(5) 62±5 (5)* 579±14(3) 94±4 (3)R5 1002 ±40 (6) 56±4(6)** 596± 51(4) 65±4 (4)** 490± 16(3) "99+10(3)R6 1361-r5S (11) 63 ±5(l 1)0* * 859 ±56(3) 82±4 OP) 983±82(3) 82±8 (3)R7 It 37 108(9) 73--5(I0)* 683 ±37(2) 87±7 (4)* -RS 946.: 54 (7) 99±4(7) 654± 78(3) 100±15(3) 658±9 (3) 108±8 (3)R9 930 147(5) 96 i 9(7) - -- -

globus pallidus and even slightly higher than the activity in pars reticulata ofsubstantia nigra (Table 1). The rostroventral part of globus pallidus, the ventral part ofthe bed nucleus and the lateral preoptic area also showed high GAD activity, while inthe dorsal part of the bed nucleus it was slightly less. Compared with these regions, thesubstriatal grey and the caudate-putamen showed rather low GAD activity.

Electrocoagulations of the nucleus accumbens (Fig. I) caused different types oflesions depending on the duration of the lesioning current. In animals RI-R4 almostall accumbens tissue was destroyed, in 115-117 the rostrodorsal part of the nucleus wasdestroyed but the ventrocaudal part was intact, and in R8 and R9 only a small partsituated rostrodorsally was affected. In addition, the anterior olfactory nucleus, partsof the septum, and the medial caudate-putamen were affected to a limited degree in allanimals (Fig. 1).

In lesioned animals the GAD activity on the contralateral, unlesioned side wassimilar to the activity in samples from unlesioned animals (compare Table I and 11).The effect of lesion is therefore expressed as the percentage of the activity in thecontralateral side of the lesioned animals (Table 11).

Following lesions in rats RI-R7 the GAD activity in both substantia innomi-nata and the rostroventral part of the globus pallidus decreased significantly. Themaximum decrease was found after the largest accumbens lesions (Table 1I, RI-R4),while no significant decrease was found after the small lesions in R8 and R9. Incontrast, there was no correlation between the extent of the septa( or striatalinvolvement in the lesions and the decrease in GAD activity (compare the septallesions in R I vs R8 and the striatal lesions in R5 vs R9). GAD in the central part of theglobus pallidus was unaffected by all types of lesions, and this was also true in the bed

102

332

TABLE III

Choline acevyItransferase in substantia innominata, rosral globus pallidus and ventral neostriatunrfollowint electocoagulation of nucleus aceumbens

Results from 6-15 samples from animals R3, RS, R6 and R7. presented as mean ± S.E.M.

Unlesioned side Lesioned side(Imol/h/g dry wt.) °o of unlesioned

Substantia innominata 54.9 ± 5.6 106 i I IRostroventral globus pallidus 21.0 ± 4.2 112 ± 8Ventral neostriatum 83.3 ± 12.0 97 - II

nucleus of stria terminalis (94 ± 8 %), lateral preoptic area (120 ± 14%) and in theneostriatum adjacent to the substantia innominata (105 ± 13%) (mean ±S.E.M.,data from It 1, R3, R4 and R5).

In contrast to GAD, ChAT activity in animals with large, intermediate or small

accumbens lesions was unchanged in both the substantia innominata and the

rostroventral globus pallidus (Table III). In addition, ChAT in samples from theneostriatum adjacent to the lesion was not changed (Table 1i1), a finding whicheliminates the possibility that an unspecific heat effect was responsible for the decreasein GAD activity in substantia innominata and globus pallidus. The AChE stainingpattern was not visibly affected by any of the lesions.

DISCUSSION

Two points deserve special consideration in the present study: the identity of the

transmitter in a major efferent projection of the nucleus accumbens, and theclassification of the region in the basal forebrain where these fibres terminate.

The indistinct morphology of the basal forebrain in the rat has been an obstacleto a clear functional classification of the nuclei present s . 5. ,1-SO. This is also reflected in

neurochemical studies, which have shown high concentrations of several putativetransmitters in this region, but which - except for the bed nucleus of the stria

terminalis - have failed to provide for a detailed neurochemical characterization ofthe different neuronal populations present .2',2 1 ,2 6. In particular, the precise distribu-tion of the very high GAD activity in the region has not been unequivocallydefined 2 ,14, 3 3 . In an effort to clarify these issues, we have in the present workcharacterized the morphology of the basal forebrain in the rat with thionin and AChEstaining. The different nuclei were identified following the description of Swanson30 ,and the activity of the specific ACh and GABA markers, ChAT and GAD 10, wereanalysed.

The results indicate that the neurones in the basal forebrain adjacent to the

anterior commissure can be divided into 4 principal groups, e.g. the bed nucleus of thestria terminalis, the lateral preoptic area, the substantia innominata together with the

103

333

rostroventral globus pallidus, and the substtiatal grey region together with theneostriatum. Morphologically, 3 regions are well discernible. The neurones in thelateral preoptic region and rostral substantia innominata are rather similar, and noclear border separates these nuclei. In contrast, the neurones in the bed nucleus of thestria terminalis are clearly different from those in the lateral preoptic area and thesubstantia innominata, and these regions are well separated by a cell-free zone.Similarly, the neurones in the substriatal grey region is markedly different from thosein the substantia innominata, and a sharp border also separates these regions (Fig. 2).The rostroventral globus pallidus demonstrates morphological similarities with thesubstantia innominata and lateral preoptic area, while the neostriatum is similar to thesubstriatal grey region.

This classification is also, at least partly, reflected in the transmitter compositionof the regions. The bed nucleus of the stria terminalis has a dense population ofGABAergic fibres, but an unimportant cholinergic innervation, in agreement withother reports .2. In contrast, the substriatal grey region is densely innervated bycholinergic fibres, and has only a sparse population of GABAergic terminals, like theneostriatum. The lateral preoptic area, the rostral substantia innominata and therostroventral globus pallidus seem to be intermediate. They all have a population ofAChE-positive cells, which are most numerous in the substantia innominata. Thecholinergic innervation of these nuclei as indicated by the ChAT activity is quan-titatively unimportant in the rostroventral globus pallidus. more dense in the lateralpreoptic area, and quite considerable in the lateral substantia innominata. In contrast.a very dense GABAergic innervation is found in all these nuclei. These GABAterminals are most highly concentrated in the rostral substantia innominata. which haseven higher GAD activity than the pars reticulata of the substantia nigra.

Thus, only small morphological and neurochemical differences are foundbetween the lateral preoptic area and the rostral substantia innominata and rostro-ventral globus pallidus. The 3 regions differ, however, in their fibre connections. Thelateral preoptic area receives fibres from the septum and sends efferents to thehabenula5 .24.30, while the rostral substantia innominata and the rostroventral globuspallidus have a major input from the nucleus accumbens and do not project fibres tothe habenula'5. Also, the origin of the GABAergic fibres in the regions seems to bedifferent. In the present work, the lesion experiments indicate that most of the GABAterminals in the rostral substantia innominata and the rostroventral part of globuspallidus originate in the nucleus accumbens while those in the lateral preoptic area donot. Thus, the decrease in GAD in the substantia innominata and rostroventralpallidum correlated well with the extent of the accumbens lesion, it was found to beindependent of the involvement of the septum or the neostriatum, and it was restrictedto those parts of the basal forebrain and globus pallidus which are known to receiveaccumbens fibres. These results make it unlikely that the decrease in GAD activity wasthe result of damage to fibres passing through the nucleus accumbens but originatingelsewhere. Judged on the amount of GAD activity destroyed by this lesion, this fibretract must be one of the most dense GABA projections in the rat brain.

We thus suggest that most of the GABA terminals in the rostroventral globus

33 104

334

pallidus and the rostral substantia innominata belong to an accumbens-relatedfunctional system28, which is clearly different from that of the hypothalamus-relatedpreoptic nuclei5' 30. The nucleus accumbens has traditionally been regarded as part ofthe 'limbic system' 26.27. However, recent anatomical and biochemical data indicatethat the nucleus in many ways is similar to the neostriatum. both regarding afferentconnections and intrinsic neurones'2 . 16 ,34. It is therefore interesting that the presentwork also indicates similarities in efferent connections. Thus. the neostriatum has beenfound to send a dense GABAergic projection to the main body of globus pallidusl3.3"whereas nucleus accumbens, as shown in the present work. probably sends aGABAergic projection to a ventral extension of this nucleus. This finding consequentlygives additional credence to the proposal that the nucleus accumbens might beregarded as a *ventral striatum', that the rostral substantia innominata might beregarded as a 'ventral pallidum', and that these two regions are connected by a *ventralstriatopallidal' projection' 5, 6. The remainder of the GAD activity in the *ventralpallidum' may possibly derive from fibres originating in the olfactory tubercle.Although these fibres, which probably would be unharmed by our lesions (Fig. I),terminate in a slightly more ventral and lateral part of the substantia innominata thanthe fibres from the nucleus accumbens'5, they would still be included in our samples.

The origin of cholinergic fibres in the basal forebrain remains unanswered bythis study. However, we have previously shown that the basal forebrain does notreceive ascending cholinergic fibres' 2, and it has also been shown that the ChATactivity in samples from the lateral preoptic area which probably included thesubstantia innominata as defined in this paper, is decreased by local kainic acidapplications . It is theref'ore possible that most of these cholinergic fibres are derivedfrom local neurones, some of which probably belong to the predominantly cholinergicsystem of "magnocellular nuclei of the basal forebrain', which project widely to thecortex and brain stem 7.20.2., .

ACKNOWLEDGEMENTS

We thank Professor L. Heimer for helpful advice and Dr. I. Grofova forphotographical help.

I. Walaas is a fellow of the Norwegian Research Council for Science and theHumanities.

REFERENCES

I Ben-Ar. Y.. Kanazawa, I. and Zigmond, R. E., Regional distribution of glutamate decarboxylaseand GABA within the amygdaloid complex and stria terminalis system of the rat, J. Neurocrom.,26 (1976) 1279- 1283.

2 Ben.Ari, Y . Zigmond, R. E., Shute, C. C. D. and Lewis, P. R., Regional distribution of cholineacetyltransferase and acetylcholinesterase within the amygdaloid complex and stria lerminalissystem, Brain Research. 120 (1977) 435-445.

3 Blackstad, T. W., Notes sur l'hodologie comparative de structures rhinenciphaliques (limbiques)- quelques relations mutuelles et ndocorticales. In J. de Ajuriaguerra et R. Tissot (Eds.). Rhine.ciphale. Neurotransinetteurs et Psrchoses. Georg Masson, Geneve. Paris. 1977. pp. 17-60.

T ________________________________________

105

335

4 Conrad. 1. C. A. and Pfaff. ID. W., Autoradiographic tiracing of nucleus accumbens efferents in therat. Brain Rese~arch. 113 (19761 589 596.

5 Conrad. L. C. A. and Pfaff. 1). W.. Efferents front medial basal forebrain and hypothalamus inthe rat. 1. An autoradiographic study of the miedial preoptic area, J. com~p. Neural.,. 1 97(1)185 220.

6 13L Olmos. J. and Ingram. W. R., The projection field of the stria tcrminalis in the rat brain. Anexperimental study. J. ronrp. Nettrol.. 146 (1972) 3033134.

7 Divac, L.. Magnocellular nuclei of thc basal forebrain project to neocortex, brain stem and olfactorybulb. Review of some functional correlates. Brci,, Re.%etrch, 93 (1975) 385--398.

8 Ernson. P. C., Paxinos, G.. LecCal La Salle, G., Ben-Ari. Y. and Silver. A.. Choline acetyltransfer-ase and acetyleholinestcrise containing projections from the basal forebrain to the amygdaloidcomplex of the rat. Bra/i, Research. 165 (19791 27t1-281.

9 Foninum. FK. A rapid radiochemical methcd for the determination of choline acetyltransferase,J. Nc'uromlw,,.. 24 (1975) 407 409.

10 Fonnum. F.. The localization of glutamnate dcarboxylase. choline acet) Itransferase and aromaticamino acid dccarboxsslase in manmalian and invertebrate nervous tissue, In S. Berl, D. 0. Clarkeand D. Schneider (Eds.). Metabolic Compeuina'ntsion cod ':oouoi/t/n NATO AdvancedStudy Institutes series, Plenum Press. Nesw York. 1975. pp. 99-122.

11 Fornn, F.. Storni-Mathisen. i. and Walberg. F.. Glutamate decarboxylase in inhibitory neurons.A stud\ of the enesmc in Purkinje cell axons and boutons in the cat. rcin Research. 20 (1970)259 275.

12 Foninum. 1'.. Walaa,t 1. and lscrscn, E.. Locali,ation of GABAergic, cholinergic and aminergicstructurcs in the mesolimbic ssstcm., J. Aeur-ihem, 29 t 1977) 221 2,30.

1.3 I-onnum. F., (lottesfeld. Z and Grofosa. I.. Distributicn of glutamiate decarboxylase. cholineactN It ra nsferascaitnd araiat ic anino aci d deCarboXy lase i n the l'asalI ga nglia of normal and operatedrats. Es idence for striatopallidal. st,-iatvcntol'eduncular and siriatonigral GABAergic fibre'.Br-ai, Reveiir-h. 143 fI 1918 125 1 3S.

14 (iottc~feld. Z. and Jacobovsit7. 1). M_. Neurochemical and anatomical studies of GABAergicneurons. In S. Garattini, J. F. Pujol and R. Samanin tFd.,.t. hinemimns Betw-een Pitrative Neurtp-it amm5illet.it thei: Brain, Raven Pre-,%, Nev% York, in press.

15 Heimer, L.. The olfactory cortex and the vemtral striatum. In K. Livingston and 0. H'ornykiewici(Eds. I. The Coniii,miig Erohtijio, iii uhe Limbhic SYste,, Con~cepti Plenum Press, New York. 1978.pp. 95 18S7.

16 Hetmer. L. and Wilson, R., The subcortical projection-, of the allocortex: similarities in the neuralassoctattons of the hippocampus. the piriforin cortex, and the neocortex. In M. Santini (Ld.),P %P(' fpireC ji *erobio/ogi-. Gu'/gi Ceirrntidid S1 inposviin. Rav en Press-. New York. 1975. pp.177 193.

17 Hodges. J . L. and Lehmann. E. L.. Bash- Concept.% tit/Proah/'a/,i and Statistics, Holden-Day, SanFrancisco. 1964. 375 pp.

18 Jacobov iti. D. M.and Palkos its. M .. Topographic atlas of catecholamine and acetyleholinesterase-containing neurons in the rat brain. 1. Forebrain tielenceplsalon. diencephalon). . camp. Nlettra.,157 (1974) 13 28.

19 Johnson. T. N. and RossotO. ". E., Topographic projections on the globus pallidui and the sub-stantia nigrit of selectively placed lesions in the pirecommissural caudate nucleus and putamen inthe monkey, .%/)-. Netid.. 33 (19711584 596.

201 Jones, E. G., Burton, H.. Saper. C. B. and Swanson. L. W.. Midbrain. diencephalic and corticalrclatonships oif the basal nucleus of Meynert and associated structures in primates. J. ,-ol,,p.

\:'w-ol. 107 11976) 385 420.21 Kataoka. K.. Sorimachi, M.. Okuno. S. and Mi~uno. N.. Enzymnatic evidence fot a meso-limbic

slopaminergic innervation in the olfactory tuberclc of the rabbit. Brain Researc-h. 88 (1975) 513 517.22 Khnig. J. F. R. and Klippel. R. A.. The Rat 8,-ala. .4 Ster,'oni.%i Ala-s of the [forhrttipt and Lower

PortsA o h fil Bain Stem,,. Williams and Wilkins. Baltimore. 1963.23 Lowry. 0. H4.. The quantitative histochemistry oif the brain. J. 1fistoc/,en. Ci,heip.. 1 (19531

4!0 428.24 Meibach. R. C. and Siegel. A.. Etlerent connections of the septa) area: an analysis utiliz~ing retro-

grade and anterograde transport methods. Brait, Research. 119 (19771 1.20.25 Mesulan M.-M. and Van Hoesen. C. W., Acetylcholinesterase-rieh projections from the basal

forebrain of the rhesus monkey to neocortcx. Brain Research. 109 (1976) 152- 157.26 Palkovits, M.. Saavedra. J. M., Kobayashi, R. M. and Brossnstein. M.. Choline acetyltransterase

content of limbic nuclei of the rat. 8Brain Re.sar-h, 79 (1974) 443-450.

106

336

27 Powell, E. W. and Leman. R. B., Connections of the nucleus accumbens, Brain Research, 105(1976) 389-403.

28 Pycock, C. and Horton, R., Evidence for an accumbens-pallidal pathway in the rat and its possibleGABA-minergic control, Brain Research, 110 (1976) 629-634.

29 Storm-Mathisen, .,. Quantitative histochemnistry of acetylcholinesterase in the rat hippocampalregion correlated to histochemnical staining. J. Ne,,roclhem., 17 (1970) 739-750.

30 Swanson, L. W., An autoradiographic study of the efferent connections of the preoptic regions inthe rat, J. comnp. Neurol., 167 (1976) 227-256.

31 Swanson, L. W. and Cowan, W. M., A note on the connections and development of the nucleusaccumbens. Brain, Researc, 92 (1975) 324-330.

32 Swanson, L. W. and Cowan, W. M., An autoradiographic study of the organization of the efferentconnections of the hippocampal formation in the rat, J. com~p. Neural., 172 (1977) 49-84.

33 Tappaz, M. L., Brownstein, M. J. and Palkovits, M., Distribution of glutamate decarboxylase indiscrete brain nuclei. Brain Research, 108 (1976) 371-379.

34 Walaas, 1. and Fonnum, F., The effects of surgical and chemical lesions on neurotransmitter can-didates in nucleus accumbens, Neuroscience, 4 (1979) 209-216.

35 Walaas, 1. and Fonnum, F., The distribution of putative monoamine, GABA, acetylcholine andglutamate fibres in the mesolimbic system. In H. Kofod, P. Krogsgaard-Larsen and J. Scheel-Kruger(Eds.), GABA-Neuro-transinitters, Alfred Benazo,, S ,posiw,,i 12, Munksgaard, Copenhagen, 1978,pp. 60-73.

36 Williams, D. J., Crossman. A. R. and Slater, P.. The efferent projections of the nucleus accumbensin the rat, Brain, Research, 130 (1977) 217-227.

107 PAPER IX

BIOCHEMICAL EVIDENCE FOR y-AMINOBUTYRATECONTAINING FIBRES FROM THE NUCLEUS

ACCUMBENS TO THE SUBSTANTIA NIGRA ANDVENTRAL TEGMENTAL AREA IN THE RAT

1. WALAAS and F. FoNNumNorwegian Defence Research Establishment. Division for Toxicology. N-2007 Kjeller. Norway

Abstract Glutamate decarboxylase activity, a specific marker for ;,-aminobutyrate-containing neurons.has been analysed in microdissected samples from rat mesencephalon following unilateral electrocoagu-lations of the nucleus accumbens. This lesion resulted in a consistent decrease of 50",, in the enzymeactivity in the rostromedial substantia nigra. and a slight, but insignificant decrease I 01in themedial parts of the caudal pars compacta of the substantia nigra. No change was round in the lateralPars COMPaCta Or the central pars reticulata. In the ventral tegmental area, the highest activity was foundin the rostromedial part. adjacent to the mammillary body. At this level, a significant decrease of 20".was found in the ventral tegmental area on the lesioned side. In contrast, the actisities in the medialaccessory optic nuicleus and the caudal ventral tegmental area adjacent ito the interpenduncular nucleuswere unchanged

The results indicate that the nucleus accumbens sends 7--aminnobutyrate-containinig fibres to the rostro-medial substantia nigra and to the rostral ventral tegmental area. The caudal %entral tegmental area. thelateral pars compacta and the central pars reticulata do not receive measurable amounts of such fibres.

THE NL'ct.E's accumbens septi distributes nerve fibres effort to resolve this problem. taking advantage of theto a considerable number of brain regions, and seems precise and restricted distribution of the fibres from

to be able to exert a widespread influence on brain the nucleus accumbens (NAUTA et ill.. 1978). Follow-

functions (POWAELL & LEMAN. 1976: CONRAD & PFAFF. tng unilateral electrocoiagulations of the nucleus1976. WILLIAMS. CROSSMAN & SLATE.R. 1977: NAL'TA. accumbens we have dissected parts of the substantia

SMITH. FAILL & DOMESICK. 1978). One of the quanti- nigra and ventral tegmental area from freeze-driedtatively dominating projections terminate% in the sections, and analysed the activity of the specificmesencephalon. where the fibres distribute densely to GABA marker glutamate decarboxylase (GAD. EC

rostral parts of the ventral tegmental area of Tsai and 4.1.1.1 5). on both the lesioned and unlesioned aide.

the dlorsomedial substantia nigra (SWANSON & Adjacent regions have been analysed as control. SomeCOWAN. 1975: CONRAD & PFAFF. 1976 NAITA eq al.. preliminary results have been presented (WALAAS &1978. PHItLipsoN. 1978). These fibres may be impor- FONNum, 1979a).

tant both as a regulator of the 'mesolimbic' dopamineneurons in the ventral tegmental area, which in turn EXPERIMENTAL PROCEDURES

project badt to the nucleus accumbens (UNGERSTEDT. a1Otei~iIs1971 ; SiMoN. LFMOAL. GALEY & CAROo. 1976; White Wistar male rats of 1810 2101z body weight wereNAVTA et il.. 1979k. and as an input to the nigral from Mollergaard-Hanssens Alslat Iratorium. Denmark.complex. The identification of the neurotransmitter(s) i.-[l-'4 Cglutamic acid was from The Radiochemical

in these fibres would therefore be of considerable Centre. Amersham. U,.

interest. -,--Aminobutyric acid (GABA) has potent Iiii

actions on the dopamine-containing cells in the ven- The animals were anaesthetized with fentanyl citrate andtral tegmental area (FIXF. HOKFEILT. L" tNGDAHI.. fluanisone (Hypnorm vet.. Mekos. 0.1 ml se.c and %aliumA6NATI. JOHIANSO)N & PtRUz DE L.A MORA. 1975: PAl.- (Roche, 0.75mig tip.l. and placed in a stereotactic frameFRFYMAN. IHOT. LIPPERT & SCHECITIER. 1978) and after Kbisic & Ktipit (19631, The skull was opened withhas been proposed as a transmitter candidate on a denial drill, and a tungsten clectrode insulated to the tip

physiological and pharmacological grounds (WoLi . 10.5 mm diameter) was lowered into the nucleus accumbens

OLPE. AVRiTH & HAAS, 1978; STEENs. 1979). How- 9.5 mm in front of the interaural line. 1.2 mm lateral to the

ever, previous neurochemical studies have given con- midline and 5.5 mm below the dura. Another electrode was

flicting results (FtNNTM. WAI.AAS & IVERSYN. 1977; clipped to the wound edge. and lesions were made with an

MCGEER. M-GFER & HATTORI. 1977; WADDINU TON alternating current (0.2 Al for l0os. The lesions were placed

& C7Ros 1978). The present study was initiated in a .n randomly on the right or the left side in different rats.

Tissue' pre'paratrion'

Abhrriafliti: GABA. y--aminohutyrate; GAD. t.-gluta- Five six days after operation the rats were decapitated.

mate dlecarboxylase the brains rapidly removed and frozen on a microtome

63

108

64 I. WALAAS and F. FONNM

RI R2 R3 R4 R5 R6

" A

FK,. I. Schematic outhne of lesions destroying the nucleus accumbens in animals RI Rt. A. nucleusaccumbens; CP. caudatoputamen. S. septum: OT. olfactory tubercle; TD. diagonal band nucleus.

chuck with a C-O2 Jet. A knife mark ,*as made in the cortex animal in a stereomicroscope. The lesioned side ,*as identi-on the lesioned side. and frontal 40 pm sections were cut at fled by means of the knife cut in the cortes. and the- 16 C. Sections from the rostral telencephalon were samples were dissected wilth broken ra,'or-blade splints~thaw*ed on microscopic slides, air-dried and stained w*ith Rostrally. the medial accessory optic nucleus, the fasca-

thionin for mapping of he lesions Fig. I and Table I. culus retrofleus and the mammillar) pendune Were1.Between levels A 280.0 and A 1300 (IKi iI & KiipptiL. easily sisible and used as reference points. The sentral19631 each 6th section was taken for cell-staining, while the tegmental area ,*as dissected bet,*een the mammillar.,intermediate sections w*ere collected, freeze-dried and peduncle. the medial lemniscus and the medial accessorystored as previously described IFO) NI'I STORM-MATHIS1-\ optic nucleus. The rostromedial substantia nigra w*as dis-& WALRFR(;. 1970). sected adiacent to the medial accessory optic nucleus, and

ll~sd .i (dprtp Irthmthe central parts reticulata of the substantia nigra was dis-I~iM~Ivih~I pr'pacara~nsected dorsal to the crus cerebri IFig. 21l. Some samples1: noperated animals were perfused through the ascend- w*ere also taken from the medial Iccessory optic nucleus.

ing aorta w*ith .5Om1 Isotonic saline followed by Bouin's More caudally the sentral tegmental area was taken adja-fixatise for 30 mm. The brains w*ere fixed osernight in cent to the easily sisible interpeduncular nucleus, whbile theBouin's fixative, dehydrated, embedded in paraffin and substantia nigra was di,,ided into medial and lateral parssectioned in series at 5pJm. Every tenth section from the compacta and central pars retaculata I Fig. 21.mesencephalon w*as mounted and stained with toluidine Some of the dissected sections were mounted on eggblue albumin-glyeerol coated slides, fixed in I0",, formalin and

Dl,,,e rl.1stained for acetylcholinesterase ISToRsq-M THISi '. 1970t,in order to vizualize the sentral tegmental area and parsThe freeze-dried sections w*ere inspected simultaneously compacta of the substantia nigra IFoNN, sq et cal.. 1977t.w*ith adjacent cell-stained 40 jm sections from the same Inspection of these slides revealed a precise dissection of

TARI I I. ExrITNT 01 tI.F(CTRO<'SActI.ATK\S 0*- THI NI ('t 11, Ac ('I MSIlN-

Nucleus Olfactory. Diagonal band Bed nucleus ofAnimal accumbens Caudatoputamen tubercle nucleus stria terminals Septum

RI +++-~+ +-lt -R2 .- + + + -- +R3 +++ - -r *r -R4 ++. - 4-.-- ---R -- .... . -5---

R6 -+*r

*

r. Roxttral: d. dorsal The number of plus signs indicates the estent of destruction

CID~5 CPCD PCI I

109

;-Aminohutyratc in nucleus accumbons efferents 65

medial accessory optic nucleus was found between thecrus cerebri and the mammillary peduncle. Itextended from the ventral surface of the brain,between the medial substantia nigra and the ventral

tegmental area, and the medium-sized cells found

seemed to be preferentially oriented in a ventro-medial dorsolateral direction iFig. 3). This nucleus

was easily visible in the freeze-dried sections. having a

more dense appearance than adjacent regions whenilluminated from below. The rostromedial nigralsample comprised a region extending from the dorsal

border of the crus cerebri to the dorsal parts of the

'B medial pars compacta. Medially. this sample borderedupon the accessory optic nucleus, and it extended

laterally 0. 1-0.2 mm into the substantia nigra (Fig. 21.Ventrally, this sample contained tissue from the pars

reticulata. with a few medium-sized neurons, whileSAN sdorsally a dense population of both fusiform and pyr-

amidal pars compacta neurons (FALLON. RILEY &MxRE. 1978: JURASKA. WILSON & GRovEs. 19771 was

P included. The ventral tegmental area comprised morescattered neurons. most of them being identical in

Flo. 2. Drawings based on toluidine-stained sections from appearance to the pars compacta cells (Fig. 3).the mesencephalon. demonstrating approximate outline of In more caudal regions. the medial pars compactathe different nuclei and localization of dissected samples. was pierced by fascicles from the oculomotor nerve(A) rostral mesencephalon Dissected samples: I. central and by perforating blood vessels. Laterally it blendedpars reticulata: II. rostromedial substantia nigra: Ill. with adjacent regions without any distinct border.medial accessory optic nucleus; IV. ventral tegmental area. Also at this level, the pars compacta neuronsIB Caudal mesencephalm. I. central pars reticulata: II.lateral pars compacta: Ill. medial pars compacta: IV. extended without interruption into the caudal ventralventral tegmemal area. tegmental area, where they became more sparse in the

Ahhreriariois: AON. medial accessory optic nucleus: region dorsolateral to the interpeduncular nucleus

FR. fasciculus retroflexus: IP. interpeduncular nucleus: (Fig. 2).CC. crus cerebri: NR. nucleus ruber: ML. medial lem-niscus; MB. mammillary body: MP. mammillary peduncle: Et.,cts of the lesion on glutamate decarhoxyluse

SNC. substantia nigra. pars compacta: SNR. substantia lctiiity

nigra. pars reticulata: OC. oculomotor nerve Anatomical studies have demonstrated a strict uni-lateral distribution of both the striatonigral (BtNNEY

the accessory optic nucleus, the ventral tegmental area and & AGHAJANIAN. 1976: TULL H. ARRBTHNOTT &

the rostromedial substantia nigra. while some samples WRIGHT, 1978) and the accumbens mesencephalic

from the caudal pars compacta extended slightly into the fibres (NAUTA et al.. 19781. and we have previously

dorsal pars reticulata. or into the tissue between the sub- shown that a hemitransection of the brain is withoutstantia ntgra and the medial lemniscus. Thus. some varl- effect on GAD in the contralateral mesencephalonation would be expected in the results from the caudal pars (FoNNU1M et al.. 19771. In the present study. we havecompacta. especially in the lateral part. therefore used the unlesioned side as 'normal', and

Bochemial inal vsits compared the biochemical results found here with

Under microscopic guidance the samples (0.2 1.5 tig dry those on the lesioned side.

weight) were collected, moved to a fishpole quartz fibre Distrihution ot glutamate decarhoxrlase in the

balance (LOWRY. 1953 weighed and transferred to the mesencephalrm. On the unlesioned side. the highest

assay tubes. Glutamate decarboxylase (GAD) activity was GAD activity was found in the central pars reticulata

analysed by a C0 2-trapping method as described (FONNUM and in samples from the rostromedial substantia nigraet al.. 1970; 1977). with a final concentration of 20mM (Table 2). The latter sample comprised both the parst-glutamic acid. The Wilcoxon two sample test was used to reticulata and the pars ompacta. and separate analy-evaluate the results statistically. sis of these regions revealed equal activities in both

(results not shown). Lower activity was found in bothRESULTS medial and lateral parts of the caudal pars compacta

Histological studies (Table 3). In the ventral tegmental area. the highest

Histological staining of sections from the mesen- activity was found rostrally (Tables 2 and 3?. The dis-

cephalon was performed in order to characterize the tribution of GAD in the caudal ventral tegmental

neuronal population present in the regions investi- area and caudal pars compacta is in general agree-

gated in this study (Fig. 2). In the rostral samples. the ment with our previous results (FoNNtum et al., 1977).

110

66 1. WALAAS and F. FONNUM

TABLE 3. GLUTAMATE DECARBOXYLASE ACTIVITY IN MUSENCEPHAIi' NUCIEi Fit LIiN(6 IJl( TR(XOA(UI I ATIO\ (iF THF

NILEUS A(('t'MB4ENS

Contralateral side to lesion Ipsilaleral side to lesionRegion (umol h g dry ssW ,, of unlesionedi

Caudal ventral tegmental area 178 + 15(91 97 + 9(151Caudal substanti nigra. medial pars compacta 354 + 34114) S5 + 111!Caudal substantia nigra. lateral pars compacta 396 + 67(111 9x + 22 10()

Medial accessory optic nucleus 190 + 120 )9 92 5(8 1

Results from five animals, presented as mean ±S.E.M. (number of samples).

The medial accessory optic nucleus had distinctly nucleus, regions which receive fes. or no fibres fromlower activity than the adjacent ventral tegmental the nucleus accumbens (NAt TA ct al.. 9781 all dis-samples, and very much lower than the rostromedial played unchanged activity after the lesions (Table 21.nigral region immediately lateral to it (Tables 2and 3). DISUt SSION

Extent of the lesions. Inspection of the telencephalicsections demonstrated that the nucleus accumbens as 7-Aminohutyrate in the mesencephaui

usually defined (K6NtG & KLIPPEL, 1963) has been The GABAergic fibres reaching the mesencephalondestroyed in all animals, and that the globus pallidus. probably have an important influence (in the localthe caudal caudatoputamen and the anterior hypo- dopaminergic cells (Ci-mERAMY. Nli t Io'. & Gt.ow-thalamus were unharmed (Fig. 2). However. the iNSKI, 1978: CATTARnI. BL ,ATTI. GROPIt r i. MAG(,i.lesions were not restricted to the nucleus accumbens. PARENTI & RAAGNI. 1978L but recer -,. d,, haveIn animals RI. R2 and R6, a part of the rostromedial shown that they also may be insolsri in behaviourcaudatoputamen was also involved, and in RI and R6 not mediated by monoamines (S( HFt= -KRI otR. ARTthe lesion also encroached slightly upon the lateral & MAGIFL.ND. 1977: Di CHiAR,. POR(t It)t . MOR-septum and rostral parts of the dorsal bed nucleus of ELLI. MULAS & GUSSA. 1978: WAtU)I'(.TO%. 1978. seethe stria terminalis (Table I). The nucleus of the dia- also review by DRAY. 19791. Thus. the functionalgonal band was involved in animal R6. In animals R3. organization of these fibres may be more complicatedR4 and R5 the lesion left the caudatoputamen and the than previously belieed. The major GABAergic inputbed nucleus of the stria terminalis intact. However, to these regions has been well defined: it arises in thethese lesions involved most of the ventral ipsilateral caudatoputamen and possibly in the globus pallidusseptum and the nucleus of the diagonal band. and terminates in the substantia nigra (KIM. BAK.

Animals R4 and R5 also had small lesions in the HASSLER & OKADA. 1971: HATTORI. MCGFER. FIRIGER

mediorostral olfactory tubercle (Table I). & MCGEFR. 1973: FONNUM. GRoFoV,4. R%,niK.Glutamate decarhoxvlase ipsilateral to the lesion. STORM-MATHISFN & WALER;. 1974: BROWNSTEIN.

Biochemical analysis on the lesioned side showed that MRoz. TAPPAz & LEEMAN. 1977: FoNs1,. GoT'Tms-the activity of GAD was markedly reduced in the FEL.D & GROFOVA. 1978: JFEsts. EMso,. PAXNOS &rostromedial substantia nigra but not in the adjacent CUtLO. 1978). However. other minor GABAergicmedial accessory optic nucleus. This decrease was projections may also terminate parts of this nucleus.found both in ventral and dorsal parts of this sample. especially in the pars compacta (FoNNtM et al.. 1977:and was independent of the involvement of the BRoWNsrItIN et al.. 19771. In addition the GABAergicseptum, the caudatoputamen. the nucleus of the dia- innervation to the medialmost mesencephalic nuclei.gonal band or the bed nucleus of the stria terminalis which also might be functionally important (SrvEs.(Table 2). Therefore it seems to be a specific result of 19'791 is as yet undefined: the present study shows thatthe accumbens lesion only. In the ventral tegmental these regions have a highly specific pattern ofarea rostraL to the interpenduncular nucleus, smaller GABAergic innervation. The highest GAD activitiesdecreases in GAD (10-31",) were found in five in the pars compacta of the substantia nigra and theanimals. These decreases were most substantial in ventral tegmental area were found in the rostralanimals R3 and R5. both with considerable septal in- mesencephalon. At this level the rostromedial parsvolvement, but they were also statistically significant compacta and pars reticulata exhibited the samein R2 and R6 (P < 0.05 and P < 0.01. respectively), density of GABAergic terminals as the central parsanimals where the septum was barely touched by the reticulata, while the caudal pars compacta had onlylesion. 60', of this GAD activity. Considerably lower activi-

A slight but insignificant decrease in GAD activity ties were found in the medial accessory optic nucleus.was also found in the medial part of the caudal pars the caudal ventral tegmental area and the interpedun-compacta of the substantia nigra. The lateral pars cular nucleus (FoNNUM et IL.. 1977. and this paper).compacta, the central pars reticulata. the caudal yen- These data lead us to suggest that, at least in the rat.tral tegmental area and the medial accessory optic both the medial part of the dopaminergic cells group

W.f11c

VT .-

e .' ~*~ * * d

* ,i,, ~ *

AONa. *

cc..

Fi(;.~ ~ ~ 3. Toudr-5ie seto hwn9h yooia etrso h oto eilsbtni ir

an ajcet eios.Cmpr wthFg . breiainsa in Fi. 2 a- 04

113

-Aminobutyrate in nucleus accumbens efferent% 69

- = .r- 4 x

-.-

7 0

-.. - -

-' -

I

+

/ ... ., , .,.*,+,

V

E7 E

C 7

E

S +1 +1+1 ---- +

id

114

70 1. WAi.AAS and F. Fossum

A9 in the pars compacta. and the dopamninergic cell 1DM1usicK 1977: NAUIA t'l I., 1979). Lastly. ourgroup Alt) in the ventral tegmental area (DAHLSTRONI quantitative results show that the majority of the& Fuxin. 1965) might preferentially be innervated by GABAergic fibres from the nucleus accumbens to theGABAergic fibres rostrally. mesencephalon terminates in the medial substantia

.4fi'repirs from rhe nuclus ccunsbens. Our study nigra and not in the ventral tegmental area. Theyalso defines the origin of one of the GABAergic inputs therefore probably innervate cells which are situatedto these regions: it significant population of GABA- laterally to those dopaminergic cells fin the ventralergic terminals disppears from both the rostromedial tegmental area which project back to the nucleussubstantia nigra. the rostra] ventral tegmental area accumbens. This projection pattern has also beenand possibly a part of the caudomedial pars compacta found in the neostriatonigral pathway 1DoM.St K.after destruction of the nucleus accumbens. This effect 1977: FALLON & MWRE. 19781): restricted parts of theis not dependent on the involvement of nuclei adja- neostriatum project most densely to nigral regionscent ito the nucleus accumbens. and the distribution of immediately lateral to those nigral cells which inner-these (iABAcrgic terminals coincides well with the vate that particular striatal locus. Taken together.termination oif the efferent fibres from the nucleus these data give additional credence to the proposedaccumbens as visualized by anatomical methods neostriatal nature of the nucleus accumbens ( Hi imi it(t'Oti,I & P Atf. 1976: NALTA Vf t 1.. 1978). Thus. & Wit.sws. 1975: SWANSON() & CoWAN. 1975: NAt'TAthese (iARAergic fibres probably originate in the et i).. 1979; N AUTA. 1979: W ALAAS & FiNI't W. 1979h.

nucleus aceumbens proper and do not represent fibreseiT poaat. doestrowed hv the lesion. Some fibres fromthe nutcleus .i:,umhens also, extend into the lateral I'ii~mhle fulotioa iinph (111011

pars c,'mpa.tai IN %t i s, I. 1/ t7XI. and these might The nigral neurons influenced h% the (iAHAergicAmo p -ssfl.nitiin smaitl ctintingents of (iABAcrgic fibres from the nucleus accumbens hase: not beentibles It ,, h ... Acer thes ire too fe% ito be detected defined fNAI TA 0 11.. 197S), It has been shosso thath% the le~ iniities eniplosed in the present study In (iABAergic terminals in the rostral substantia nigra.>"I'It ist I these rekiii'n %c told thait the caudal %lt) influence nigrostriatal dopamninergic cells (JAMIs &

crawp I, Ott icnu 0 ii emcnial area I% not innersated SiARItt. 197S) and the accumbens fibres could %%ell he111CisA~ir INC im'iiits .t, \4ACrgI, fibre,, from the insolsed lin this system. However, also nigrotectal.i, t iix 1 hi, , In ittrcincit % i tl our nigrothalamic or n igroret icu tar neuronts IC \Rm Pti NlIII.

"Itnir.g~e e him.ins It"" Mii~i NAI,4 it & Kim, 1976:' Rl'%%IK. (jiOWi\SA & Oi11 It-* ' ti SH isci I, the medilaccs st %, 1976: 6114AYBUI I. 1978: Bi IKIM At). DOMIs S1( K &

* ' 'Pt . i - xriiticdi h,. the lesion, and NAt FA. 19791 in the pars reticulata ma% he insolsed.ti lii C (11...*''I agticintri %ith the It is well knowni that locomotor acit, mat, he ini-

o I -1 1, t, -1 - mi itii miser it fit\. of the tiAted bN dopamninergic mechanisms Ill the nucleusi..i.'It" , ilii,i i 111lAw11 se. reli I si I \ acumbens (I i t R6 & VA-, R issi xi. 19731, and

ih)w I,,,, h,,\'%s that the (.5HA- time Lack of this doparninc effect hats indeed been sug-i, I1. 1)It,, 1-i. ci') 1a.. ,umtss -111.a oni, be gested ito be responsible for the akinesia found in the

!.iincl 'th Pre is , i. 'fse.ti, ekthniques this Parkinsonian syndrome tA~ili N & Pift t N 19771.

.. ivi..c that' i h.....t tindings ont (,AD) in The (i'BAergic projection from the nucleus accum-'..mpls ' i~hi sihsioita igia itier large lesioins hens to the substantia igra. or alternatiselit the serl,

II thc mii, :ntc-'ir iatii i smi *sR (sti R & dense (iAHAergic projection to thle ninral' pallidumtIIA 1114.x 'i ft% r.iot lIe 0i-nteli, Lorrect (WAI AAS & FVo-Ct. 1979h) could possibli represent

the piesent stuths further demontrates that the output pathways mediating the-se motor r esponses. asnuinjeus av..iimhoi in manN% va~s hhi.es like an both these nuclei seemn to hlae connections to theiteygrl pairt of the rosItoentral neostriatum Fi-rstly. central motor sy'stem %it the thalamus or the reticular'he biochemnical idtentification o(it GAHA as a transmit- substance t~ Mi Rw. 197X: (AR11TI'i Ni i l.176.ter in the accumbens nigril pathway is comnsistent Ill ( STt At1) illa.. )1979.vitidi the findings for the %triatonigral projection Il- Lastly, our results demonstrate that some GABA-si M ,r fl 11k74, Ul-(% %I W (, 11. 1978). Secondly. the ergic fibres pass from the nucleus accumbens to thlerostromedial termination of these (iARAergic fibres rostromedial part of the xentral tegmental areal.in the substainuta nigra is comparable to the distribu- Retrograde transport techniques and elect ro phv sto-toon of the r4ostral neostiatal fibres which are known logical studies indicate that this region contains onlyto terminatec preferentially in medial parts of the sub- few cells projecting to the nucleus accumbens (NAtTAsuantli fligra. svhil the caudal neoistritum sends i ill.. 1978: Woi er itl.. 1978). and our results there-fibre,, more laterally lI-iNN & A(IIAJANIAN. 1976. fore do not seem ito gisre strong esidence inl support ofDr i if ei ci t.. 1978 . The dorsal termination of the a mesolimbic inhibitorv 'feed-back* system, i.e. that

a.CriMbenS fibres also agrees with anatomical studies. dopaminergic Alt) cells tre itnfluenced by af GAHA-which haive shot that the sentral neostriatum sends ergic projection from thle nucleus accumbens. Thefibres preferentially to dorial nigral regions, while the majority of the mesolimbic dopaminergic cells pro-dorsal necostriatumn sends fibres to the ventral nigra jecting to the nucleus aecumbens., as demonstrated

115

-Aminobuisyrate ti nucleus accumbens clerents,7

by retrog.raide transport techniqueis. is located more At kninit ldqenivin s We thank Li%. Ut 1AS~i % (or' technicalcaudallv N'. liA Clill.. 1979). and these cells are not assistance. Dr A. SOR1t ii for help %kith the histOlog'. andinnersatled bv (AIRAergic libres from the roslral P. J KAkl SJ " for preparation of the drawtingsforebrain (F-ti\N t till. 1977: and this paper.

RFFERENCFS

Asax % N 4I. & Jii.i I s H, 1977) [ffect of local application of apomorphine it) the corpus striatumn and to the nucleusaccumbens till the reserpine-induced rigiditN in rats. Brini Re's. 133. 1X6 3x 9

Bit Kil Al) R. M.. lDo'.i siii V. H. & NAki A W, J H. 09791 Efferent connections of the substanti nigra and sentralte~gmen tat area tin the rat )4,,i Res. 1715. )')1 21 7

Biuiw Sit I's NI. J.. Iki/ F A . TAPPiA/ M. L. & Li I MAN S. E. (19771 On the origin of' substance P and glutamic aciddecarhti.sase (6A01 tin the subsiantia ntgra. Brain Re,% 135. 315 323.

Hi \ii 8,H S. & At(iiAJAsi% Gi K. 1197M1 The precise localization of nigral afierents in the rat as determined by aretrograde tracing techniqlue. HraIn Re%. 117. 423 435.

CAtti'st~ii M H. NAKA%ii K. & KiM R. 419761 Nigrothalamioc projections in the monkey demonstrated by autoradio-graphic technics J. tnip \tiir,4. 165. 4141 41t6

('AFlA11ii1 F. Hi (,-%Il A. (iittit''i ITt A.. MA(.at A.. PARfNI-' M. & RAtAt,-0 G. (197h4 GABA and doparmne; theirmutual regulation tin the nigrostriatal system In All4,'tu~rn~ratr~..ired Btiton Simposium 12 teds KRtK,%-(,A'0tD-IAitsi \ P. St iii it-Kiti (dlt J & Kiiiiio H.i. pp. 10)7 117. Nlunksgaard. Copenhagen.

Ciii AMN A . it ii ti o A & 0vii iw~ J. tl97NI 6AHA-ergic proticswes iniolved in the control ot dopainine releasefrom nigrtistriaital dopaiminergic neurons in the cat. Eiir. J. Pharnkaiol. 49. 281 295.

Cli~kAil L C. A & Pt st 11) W 119764 Autoradiographic tracing of nucleus accumbens efferents in the rat. Bruin Res.11.4. 589 596

DA1tisriisti A & Ft mi K 019651 Fsidence for the evistence of monoiamine-containing neurons in the central nervoussystem 1. Demonistration of monoamnine'. in the cell bodies of brain stem neurons. ia phi-siol.arni. 62. Suppl. 232.1 55

131 CHIARA 6i. Pink(t I oot M..Misi i.1 M., MVLAS M. L & GZsSA G. L. (19784 Strio-nigral and nigro-thalamic6ABA-urgic neuirons as output pathways for striatal responses. In GAB.4-'urotraisirt'rs. AlfJred Be'n:ein St-mplosillm12 teds KRoiHsi,AAitt-LAHit'.- P.. S( i t i~-Kkt (a R J & Kt-ot HA.) pp. 465 481. Munksgaard. Copenhagen.

I~oli sit K V. H. 1)19771 The topographical organization of the striatonigral connection in the rat. A4nal. Rec. 187, 567.DRA i A 11919)1 he striatumn and substantia nigra a commentary on their relationships. Ntoirost it'ne 4. 1407 1440.FAE 1 0% J. H & MixmJ- R N 119)7h4 Catecholamine innervation of the basal forebrain IV. Topography of the dopamnineI projection Iii the basal forebrain and neostriatum. J. tomp. .Vt'uro/. 180. 545 580).FAt to-, . H- RutN J N & Moin ti Y 19781 Substanta nigra dopamine neurons: separate populations project to

neostriatun aind ,illiicorte. %:,urost iemeC Liitvr s 7. 157 162.Foi'%t 54 F. STtiRM-MAliiiSI % J & WAi PI.Rt. F. 1197()) Glutamate decarboxylase in inhibitory neurons. A study of the

enzyme in Purkinie cell avons and boutons in the cat. Braiin Ries. 20. 259 275.Foi%tl M F. (Gol-i Sit 1 ) & (ibouiiA I141978) Disribution of glutamate decarboxylase. choline acetyltransferase and

aromatic amino acid decarboxylase in the basal ganglia of normal and operated rats. Evidence for striatopallidal.striatoentopenduncular and striatonigral GAHAergic fibers. Brain Res.' 143. 125 138.

Fisitm F. Gtovoi. A I , Ri%%i.i F. STOass-MArtisis % & W~t.ntjt(, F. 119741 Origin and distribution of glutamnatedecarhoxylase tin substatntia nigra oif the cat. Braini Rr'~s. 71, 77 92.

Fti%Nt M F.. WAL AAS 1. & lvi Rsii E 119774 Localization of GABAergic. cholinergic and aminergic structures in themesolimbic svstenr. J.tiiur~ilit. 29. 221 2301.

Fiui K.. HoKiit Y T.. Lit %tit)At J%. AI.%ATi L.. JiiA%SSO)% 0. & PRI~r t I.A MORA M. (19751 Evidence for aninhibitory gabergic control of the mesoliinbic dopamine neurons: possibility of improving treatment of schizophreniaby combined treatment with neuroleptics and gabergic drugs. Me Bill. 53. 177 183.

014AY011 t A M. 1197S) Organization of the nigrotectal connection:, an experimental tracer study in the cat. Brain Res.. 143.339 148.

HATTORI T.. Mi (i P L.. IFIRnstit H C & MtGil-lR E.G6.049731 On the source of GABA-containing terminals in thesubstantia nigra Flectron mictroscoipic ,iutoradtographic and biochemical studies. Brain Re.. -04 103 114.

Hei e L. 01971 The olfactory cortex. and the ventral striatum. In Limhit Mtechtanisms teds Livi-wusrs K, E- &HOR%YKii sw U . 04. pp. 95 187. Plenum Press. New York.

lHlitmi R L. & Wit ois R 1). 119751 The %ubeortical projections of the allocortex. similarities in the neural association$ ofthe hippocampus. the priform cortex. and the neocortex. In Gilqi Centennial Simposiutn Pu'rspet rile'. in Neurobiollwi.ted. SA'Tri's MA. pp 177 193. Raven Press. New York

JAmUS T. A. & STARR M. S. 1197m) The role of (jABA in the substantia nigra. Nture, 1,111d. 2754, 229 230.Jtxsxt.i T. M. Fswt P C. PA'.i'os Gi & Ci i i iit A. C. 119781 Topographic projections of substances P and GABA

pathways in the striatti- and pallidsnigral %sistm, a biochemical and immunohistochemical stud). Brain Res. 152.487 498

RtRASKA J M-. Wit %i0% C J & (i1tvtS P M 0t)774 The subsiantia nigra of the rat. a Golgi study. J. iiimp ,Neiuri'I 172.585 64(1.

KiM J.S.. OAK I J. 11ASSlit R & OKADSA V 119711 Role of -,aminobutyric actd I(iAHAl in the extrapyrainidal motorsystem 2 Some evidence (or the esistenice tsf it type of CIABA-rich strionigral neurons. Expi Brit Rues. K4 95 104.

116

72 1. WAI AAS and F. FtONt-m

Kio'sii J. F. R. & K LIPPEL Rt. A. J1963) The Rat Braipt. .4 Stereta,w thitcs of the Fitrehriom tindi Loiter Pair, ol the BrinStem. Williams & Wilkins. Baltimore.

Lowity 0. H. (1953) The quantitative histochermstry of the brain. J. Histotheiii. 1uohm . 420 42K.MCGFHRJ P. L.. MCGFrR E. 6, & HATTORit T. (1977) Dopamine-acetylcholine-iABA necuronal linkages in the extrapyra-

midal and limbic systems. In N'oni.%riattil Doptiminerytu .%euroii. .4dr. Bignheni. Ps i-hophiirmiof.. Vol. 16 (eds C1A F.& GiSSA Gi. L.). pp. 397 402. Raven Press. New York.

MoiFNSO' Ci. 3.. WI M. & MANCHANDA S. K. 11979) Locomotor activity initiated by microinfusions of picrotosin intothe ventral tegmental area. Brain Res. 161. 311 319,

NAGY J. L. CARTER D. A. & FunIGIit H. C. 09K) Anterior striatal projections to the glohus pallidus. entopenduncularnucleus and substantia nigra in the rat: the GABA connection. Brajin Re's. 144. 15 29.

NAt TA H. J. W. t)979) A proposed conceptual reorganiz.ation of the basal ganglia and eclencephalot.Veroien.te. 4.1875 IX8I.

NAUTA W. J. H. SMITH (j. P.. FAt ILt R. L. M. & DOMFSICK V. B. (1197K Efferent conneciins and nigral afferents of thenucleus accumbens septi in the rat. Neurou ien&' 3. 3K5 40),

PALFREYSIAN M Gi. HUim S.. LIPP'ERT B. & Su'HFiHTV.R P. J. 119781 The effect of ;--accylenic (;ABA. an cn~smc-aetuvaicdirreversible inhibitor of (ABA-transniinase. on dopamnine Pathways of the extrapyramidal and limbic systems. Ea4r J.Pharmwvo/. M0. 325 336.

P11IiIuPON 0. T. (1979) Afferent projections to AIO dopaminergic neurons in the rat as show~n by the retrograde transportof horseradish peroxidase. Neurosciepice Letters 9. 353 359.

PUNUNRURKG A. . 3. & VAN Rossum J. MI. f (973) Stimulation of locomotor activitY fosllowing intecc ion of( dopamnine intothe nucleus accumbens. J. Piirni. Pharrmai. 25. 13 I(K).

Powti L. E. W. & Li-MAN Rt. B. 11976) Connections oif the nucleus accumbens Brain Rev. 1105. 3X9 4W0RtiviiK E., (iROiii S I. & OTTuRs.% 0. P. (1976) Demonstration of nigrotectal and nigroreticular projections in the Cal hsv

axonal transport of proteins. Brain Res. 112. 388 394.SCHFE-L-KRC 6FR J.. ART Y. & MAGtI IN I') . 11477) Behav ioural stimulation induced by muscimol and oither GABA

antagonists injected into the substantia nigra. eiirisient-r Let ters 4. 351 356.SimoN H.. LMOAL M.. GALFY D. & CARo B. (1976) Silver impregnation of dopaniinergic systems after radiofreq~ucnc

and 6-OHDA lesions of the rat %entral tegmentum. Brain Re's. 115. 215 23).STEVENS J. R. 11979) Schizophrenia and dopamine regulation in the mesolimbic system. lreflds .Veiriw,. 2. 1t)2 1015STORM-MATHISL5I 3. (1970) Quantitative histochemnistry of acetvlcholinesterase in t'he rat hippocampal region correlated to

histochemnical staining. J . Neunrwhem. 17, 739 750),SWANsoN L. W. & COWAN W. M. (1975) A note on the connuections and development of the nucleus accumbens Braini

Res. 92. 324 330.TtLjLLOCH I. F. ARRtLTHNOTT Ci. W. & WRtioTr A. K. 119781 Topographical organiztion of the striatonigrall path15av

revealed by anterograde and retrograde neuroanatomical tracing techniqlues. J. Aiiat. 127, 425 44).UNUERSTFUT U. (1971) Stereotaxic mapping of the monoamine pathways in the rat brain. .4eta phti-sl. wcand. 82. Suppl.

WADDINGTON] L. (1978) Rotational behaviour and striatal dopamine metabolism following unilateral activation of nigralGAAmechanisms: GA BA-like modulation of dopaminergic and non-dopaminergic neurons in rat substantia nigraBr.J. liirru.63. 384P.

WADDINGTON J. L. & Citoss A. J. (1978) Neurochemical changes following kainic acid lesions of the nutcleus accumbens.implications for a GABAergic accumbal-ventral tegmental pathway. Life Sii1. 22. [i Ml (14

WALAAS 1. & Fo%-.Lm F. 11979a) Amino acid transmitters linking nucleus accumbens to the batsal ganglia % eiirfosieis.vLetters Suppl. 22().

WALAAS 1. & FoNNum F. (1979h) [Distribution and origin of gtamate decarboxylase and chohinc acetytansferast: inventral pallidum and other basal forebrain regions Brain Ri's.. in press%.

WILLIAMS D. J.. CROSSMAN A. R. & SLATER P. (1977) The efferent protections of the nucleus accunibens in (he rat. BrinRes. 130. 217 227.

WOIut. P., Ot Pt H.-R.. AV'RIt 1). & HAAS H. L. (1979) CiABAergic inhibition of neurons in the ventral tegmentail areaExpertentia 314. 73 74.

14 epted 30 411ipot It Y79)