Substance P synaptic interactions with GABAergic and dopaminergic neurons in rat substantia nigra: An ultrastructural doublelabeling immunocytochemical study

Substance P Synaptic Interactions With GABAergic and Dopaminergic Neurons in Rat Substantia Nigra: An Ultrastructural DoubleLabeling Immunoc~ochemical Study I. MENDEZ.

K. ELISEVICH

AND B. FLLJMERFELT’

Received 19 September

1991

rrillf G4~.4cr,~ic ~!zd dopminw~ic~ BRAIN RES BULL 28(4) 557563, 1902.--The distribution of substance P (SP), tyrosine hydroxylase (TH). and glutamic acid decarboxylase (GAD) immunoreactivity in the substantin nigra of the rat was studied by means of an ultrastructural double-labeling immunocytochemical method. Direct synaptic contact between SP-immunoreactjve t~rnlinals and GAD-positive nigral neurons was more often observed in the pars lateralis than the pars reticularis and was rarely observed in the pars compacta. Substance P-positive terminals also formed synapses with cell bodies and dendrites of TH-positive, dopaminergic neurons in the pars compacta and pars reticulata. Multiple SP-immunoreactive terminals were often observed with symmetrical and, less frequently. asymmetrical synapses on individual TH-containing dendrites. Evidence of SP-containing termtnals contacting both GABAergic and dopaminergic neurons tn the substantia nigra suggests a direct excitatory action upon nigral projection neurons. MENDEZ.

I.. K. ELISEVICH.

~~mr~v~c cn INI ~~r/~.s~un~icr rri,yw

Neostriatum

Striate-nigrai

AND 9. FLUMERFELT. .-In dhuc/rrcc~/wd

pathway

S’~~/J.WIIWP s,rr~upric irrtc~ciions

~~orrhlc-lahclir7~

‘Tyrosine hq’drowylase

substantia nigra (SN) contains the highest density of substance P (SP)-immunoreactiv,e terminals in the rat brain (6,27,32.35) and the majority ofthem arise from neurons in the ipsilateral neostriatum and glohus pallidus (6,X,32.47.48). The SP projection to the SN is widespread (I 5.46) and has been shown to influence the activity of nigro-tectal GABAergic ncurons (37). In addition. evidence for an SP-mediated excitatory action on SN GABAergic neurons has been provided by studies in which injections of SP into the SN resulted in ipsiversive rather than contraversive circling behavior in rats (2X). The locomotor behavior was explained by a subsequent inhibition of the nigrost~atal dopaminergic projection. The SP projection has also been shown electrophysiologically to exert an excitatory action on nigral neurons ( 12.4 I). Such an action appeared to induce an increase in dopamine turnover in the striatum (28). Morphological studies have. in fact, suggested the presence ofsynapses between SP-positive terminals and nigral dopaminergi~ neurons (7.3 1.36).The mirror technique employed by Kawai and co-workers provided indirect evidence of such synaptic interaction between nigral neurons (3 1). Chang (7) employed a double labeling technique. but demonstrated only apTHE

’ Requests for reprints should be addressed to Dr. B. A. Flumerfelt. Ontario,

London.

Ontario.

Canada

N6A 5C

1.

irn,nr~~70~:1.1r~(~h~~~771(.~11strrdy

Glutamic

acid decarboxylase

Nigral interneurons

position between SP-positive terminals and tyrosine hydroxylase (TH)-containing neurons. An immunogold double-labeling technique has also been used with some success (36). However. the poor tissue penetration characteristic of gold-labeled antibodies may have caused the number and variety of immunoreactive synapses to be underestimated. The purpose of this study, therefore. was to employ a sensitive double labeling ultrastructural immunocytochemical technique to address the relationship between SP-containing terminals and GABAergic neurons in the SN, and further characterize synaptic interactions between SP-containing terminals and nigral dopaminergic neurons. MATERIALS AND METHOt>

Adult female Wistar rats (200-300 g) were given an overdose of sodium pentobarbital and were perfused trans~rdially with a fixative consisting of 4% paraformaldehyde. 0.34> t-lysine monohydrochlo~de, and 0.05% sodium periodate in 0. I M phosphate buffer (pH 7.3). The fixative was followed by an infusion

Department

of Anatomy.

Health

Sciences

Center.

University

of Western

MENDEZ.

558

of 10% sucrose in buffer, and the brain was then dissected out and stored overnight in 30% sucrose in buffer. Coronal sections (40 wrn) were cut with a freezing microtome and collected serially in buffer. The aforementioned fixative was used for single- and double-labeling immunocytochemical procedures for the detection of SP and TH. A zinc-aldehyde perfusate was employed for the immunocytochemical detection of GAD (39). The primary fixative contained 4% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.3), followed by a solution containing 4%’paraformaldehyde and 0.2% zinc salicylate in 0.1 M cacodylate buffer (pH 6.8). After immunocytochemistry (see next), sections were washed in phosphate buffer, mounted on chrome-alum gelatincoated slides, dehydrated, and cover-slipped with DPX. Electron Microscop), For ultrastructural study, animals were perfused transcardially with a solution of 4% paraformaldehyde and 0.08% glutaraldehyde in 0.1 A4 phosphate buffer (pH 7.3). The brains were removed and stored overnight at 4°C in a solution of 4% buffered parafonnaldehyde. For detection of glutamic acid decarboxylase (GAD), two successive solutions were used. The first contained 4% paraformaldehyde and 0.08% glutaraldehyde in 0. I M cacodylate buffer (pH 7.3); and the second solution contained 4% paraformaldehyde and 0.2% zinc salicylate in 0.1 M cacodylate buffer (pH 6.8). Sections were cut on a vibratome at a thickness of 40 pm, collected in phosphate buffer, and processed for singleand double-labeling immunocytochemistry. After the peroxidase reaction, the sections were postfixed in 1% osmium tetroxide for 1 h at room temperature, then rinsed in buffer, dehydrated. cleared in propylene oxide, and embedded in a mixture of Poly/ Bed and Araldite. Ultrathin sections were collected on copper grids and stained with uranyl acetate and lead citrate for viewing on a JEOL 100 CXII transmission electron microscope.

Single-labeling. Substance P-, TH-, and GAD-immunoreactive structures were demonstrated with the peroxidase-antiperoxidase technique. All incubations were performed at room temperature unless otherwise stated. Sections through the SN were incubated for I h in 0.1 M phosphate buffer (pH 7.3) containing 10% normal rabbit serum (NRS) and then placed in a solution containing a monoelonal antibody (1: 1000) raised against SP (Sera Lab) or rabbit antiserum raised against TH ( I : 1000) (Eugene Tech International). Following this incubation, sections treated with SP antiserum were rinsed in buffer for 30 min and incubated in rabbit antirat IgG (1: 100) for 1 h, rinsed again, and incubated in rabbit PAP complex (1: 100) for 1 h. Sections treated with TH antiserum were incubated in goat antirabbit IgG (1: 100) for 1 h, and then incubated in rabbit PAP complex ( 1:100) for 1 h. Detection of GABAergic structures was performed by incubating tissue sections in GAD antiserum ( I : 500) (obtained from Dr. W. H. Oertel) for 48 h at 4°C. The sections were then washed in 0.1 M phosphate buffer, incubated in rabbit antisheep IgG (1: 100) for 1 h, and finally incubated in goat PAP complex (1: 1000) for 1 h. The tissues treated in each case were then rinsed again and reacted with 0.05% 3,3’-diaminobenzidine tetrahydrochloride (DAB) and O_Ol% hydrogen peroxide in 0.1 Mphosphate buffer (pH 7.3). Sections were then washed and processed for either light or electron microscopy. Double-labeling. The dual antigen localization technique described by Levey et al. (34) was employed. The procedure initially followed the immunocytochemical protocol for single-labeling of SP using DAB as the chromogen. Following the peroxidase reaction, the sections were rinsed in 0.01 M phosphate buffer

ELISEVKH

AND t-LlJMERFfil~

i

(pH 6.8) for 20 min and processed immunocytochemicaliy forthe second antigen, using an antibody raised against TH or GAD. incubations were carried out sequentially in rabbit antisera directed against bovine TH (1: 1000) for 48 h at 4”C, goat antirabbit IgG (l:lOO), and then rabbit PAP complex II: 100). both fur 1 h. For detection of GAD immunoreactivit). the sections were sequentially treated in sheep anti-GAD immunoglobulin ( 1: 5000) for 48 h at 4”C, rabbit antisheep IgG (I : IOO), and goat PAP complex (l:lOOO), both for 1 h. The sections were rinsed in 0.01 ‘$4 phosphate buffer (pH 6.7) for 20 min and then incubated in a solution of 0.01%~ benzidine dihydrochloride (BDHC) and 0.025% sodium nitroferricyanide containing 0.002% hydrogen peroxide in phosphate butTer for j-10 min. The reaction was stopped by rinsing the sections in 0.01 .t/ phosphate buffer (pH 6.7). Double-labeling immunocytochemistry was performed for SP/TH and for SP/GAD. C’0ntrol.s /iv irnrnunocyto~/~~~~~7i.stt~~~. Several controls were employed for all immunocytochemical procedures. Sections were incubated in preimmune serum or normal serum without the primary immunoglobulin or solutions of primary antibody that had been preadsorbed with TH or SP. The results of singlelabeling were compared with those of double-labeling procedures to ensure consistent results with the same antibody in both procedures. The primary antiserum of either the first or second immunocytochemical procedure was omitted to rule out the possibility of inappropriate immunostaining due to cross reaction. Finally, sites of antigenicity were visualized with DAB and BDHC. thereby ensuring comparable reactivity with both chromogenic

substrates.

KESUL-IS

Light A4icroscop~~ In single-immunolabeled tissue sections of the SN, large numbers of TH-immunoreactive somas were found in the SNC, and a dense plexus of immunoreactive fibers extended throughout the SNC and entered the SNR where TH-positive somas were distributed in clusters (Fig. 1A). Cell bodies that were immunoreactive for GAD were found in greater numbers in the SNL than the SNR. A dense GAD-immunostained fiber network and numerous punctate immunoreactive structures covered the entire area of the SNR, SNL, ventral tegmental area (VTA), and, to a lesser extent, the ventral SNC (Fig. 1B). Some GADpositive neurons were covered by GAD-positive punctate structures outlining both the soma and proximal dendrites. In the ventral mesencephalon, SP-immunoreactivity was confined to the SN and interpeduncular nucleus [Fig. l(C)]. A dense network of SP-immunoreactive fibers was observed in the SNR and SNL, whereas the SNC contained a less dense SPimmunostained fiber plexus, mainly confined to its ventral aspect. For those tissue sections examined at the light microscope level for double-labeling the DAB and BDHC reaction products were easily differentiated. The DAB product was brown, diffuse, and homogeneous, whereas that for BDHC was blue and granular. Substance P-glutamic acid decarboxylase. GAD-immunoreactive structures were identified with BDHC, and were easily differentiated from SP immunoreactive structures labeled with DAB (Fig. 2B). Substance P-immunoreactive fibers were Seen among GAD-positive somas and fibers in the SNR, SNL, and, to a lesser extent, in the VTA (Fig. 2A and B). Substance Ppositive varicosities were also seen outlining the profiles of GADcontaining neurons, particularly in the SNL. No presynaptic contact of GAD-immunoreactive terminals with axon terminals

SP PROJECTIONS

TO GABA-

AND

DA-POSITIVE

NIGRAL

TH

FIG. I, Coronal sections of the left ventral mesencephalon showing the substantia nigra and ventral tegmental area. (A) TH-immunoreactive neurons in large numbers in the pars compacta (SNC) and the ventral tegmental area (VTA). and in smaller numbers in the pars reticulata (SNR) and pars lateralis (SNL) (X47). (B) GAD-immunoreactivity in the substantia nigra. which is particularly prominent in the pars reticulata (X47). (C) SP-immunoreactivity in the substantia nigra pars compacta and reticulata (~47).

of GAD- or SP-immunoreactive neurons was observed. The double-labeling method also did not demonstrate any dual immunoreactivity of SP and GAD in individual fibers or terminals in the SN. S~thsfanc~c~I’-/jvnsirw h~drvo~.\-!km~. Substance P- and THimmunoreactivity were also easily distinguished (Fig. 2C and

NEURONS

559

D). Substance P-immunoreactive structures possessed the characteristic diffuse reaction product of DAB. whereas THimmunoreactivity was demonstrated with BDHC. A few immunoreactive fibers were seen to pass through the VTA which contained numerous TH-immunoreactive neurons. At higher magnification, SP-immunoreactive tibers were observed to course among immunoreactive somas and hbers both in the SNC and SNR. Occasionally. SP-positive fibers and their varicosities appeared to run along the surface of TH-positive cell bodies and dendrites. outlining the DA neuron. This observation was especially prominent in the SNC.

Sh~uncc~ I’-~hurnic~ ucid (k~~(,~ll.ho\-!./u.\(‘. GAD-immunoreactivity was identified at the ultrastructural level by the tilamentous, crystalline reaction product of BDHC (Fig. 3). Somas. dendrites, axons, and terminals containing GAD were abundant in the SNR and SNL. In contrast. the SNC contained GADimmunoreactive terminals and tibers, but no G.i\D-immunoreactive somas were observed. The GAD-immunoreactive terminals varied in size and contained pleomorphic or round, electron-lucent vesicles. Substance P-positive terminals. characterized hy the diffuse immunoprecipitate of DAB. were observed to make symmetrical synaptic contacts with GAD-positive dendrites and somas in the SNR (Fig. 3A) and. in particular. in the SNL (Fig. 3B and C) where the density ofGAD-positive neurons was highest. Very rarely were SP-positive terminals seen contacting CiAD-positive dendrites in the SNC. Occasionally. GADpositive dendrites received synaptic contacts from unlabeled terminals in addition to SP-positive contacts, S/ths/u~c~ ~-?-~IwYI‘IK~ h?,tlro\-~,/~l.\t. At the ultrastructural level. SP-immunoreactivity was observed within axon terminals as indicated by the presence of the diffuse electron dense DAB reaction product (Fig. 4). Tyrosinc hydroxylasc-immunoreactivc structures were characterized by the reaction product of BDHC. Typically, the BDHC precipitate was found in proximal and distal dendrites and in cell bodies which varied in shape considerably and contained round nuclei with an occasional indcntation. The BDHC-reaction product was easily distinguished from that of DAB. which was present in the SP-immunoreactive terminals. Two types of SP-positive terminals were identified (Fig. 4B). The most frequently encountered type contained pleomorphic electron-lucent vesicles with occasional dense core vesicles, The second type contained spherical vesicles. Substance P-immunoreactive terminals of the first type usually formed symmetrical synaptic contacts with TH-immunoreactive dendrites both in the SNC and SNR (Fig. 4B and C). Less frequently, terminals of the second type formed asymmetrical synaptic contacts (Fig. 4B). The frequency of SP-TH axo-dendritic synaptic interactions was greater in the SNR than the SNC. Axo-somatic interactions were rare and appeared to be confined to the SW. which contained large numbers of TH-immunoreactive somas. At times two or more SP-positive terminals contacted a single TH-positive dendrite, which also received synaptic input from a numhcr of unlabeled terminals (Fig. 4C).

The SN contains a large number of dopaminergic neurons that are highly concentrated in the SNC (3. I I .25). Nondopaminergic neurons are found mainly in the SNR and a large number of them contain gamma-aminobutyric acid (GABA) (8,9. I3,43.44). Two neurochemically distinct striatal projections to the SN have been described ( 17.19.2 I). The GABAergic inhibitory

MENDEZ.

ELISEVICH

ANI)

FLIIMERFEI.

i

FIG. 2. Double-labeled coronal sections of the substantia nigra. (A) Simultaneous demonstration of SP and GAD immunoreactivity (X47). (B) An enlargement of the rectangle shown in (A). Note the characteristic granular BDHC reaction product which is present in the GAD-immunoreactive neurons. SP-immunoreactive structures in the background contain diffuse DAB reaction product (X385). (C’)Simultaneous demonstration of SP and TH immunoreactivity. TH-immunoreactive neurons are clearly seen against the background of SP-immunoreactivity (X47). (D) An enlargement of the rectangle shown in (C). TH-immunoreactive neurons contain granular BDHC reaction product, while SP-immunoreactive fibers and neuropil are labeled with DAB (X385).

pathway arises from the posterior striatum (6). The SP facilitatory pathway arises from both the anterior and posterior striatum and is topographically organized such that the posterior group projects to the SNL and the anterior group projects to both the SNC and SNR (32). The striatal influence on the SNR is sufficiently widespread that very few neurons remain unaffected by changes in striatal activity (2). Electrophysiologic studies revealed that more SNR neurons appear to be under a facilitatory influence than an inhibitory one (1). This effect may be due to either an excitatory striato-nigral pathway mediated by SP or to the presence of an inhibitory GABAergic interneuron receiving an inhibitory striato-nigral pathway containing GABA. Although quantitative morphological studies indicated that only 20% of striato-nigral fibers contain GABA (4), Albe-Fessard et al. (2) observed a facilitatory influence on more than 50% of SNR neurons, suggesting that SP might indeed play a significant role in mediating nigral output via either dopaminergic or GABAergic intermediaries. SP und GAD Synaptic Interactions

The principal finding of this study identifies a direct synaptic contact between SP-immunoreactive axon terminals and GAD-

immunoreactive neurons in the SN. Such contacts are particularly numerous in the SNL where the largest proportion of GAD-immunoreactive somas is observed. Injection of SP into the SNL resulted in a distinctive ipsiversive circling behavior (28), suggesting that excitatory SP-mediated synaptic input onto GABAergic-inhibitory nigral interneurons may, in turn, influence the dopaminergic projection neurons of the SNC. Interneurons containing GABA in the SN were identified by neuroablative methods (42). and Golgi studies identified these neurons lying immediately ventral to projection neurons in the SNC and synapsing extensively with them (16,22,30.45). Axo-dendritic contacts between SP- and GAD-immunoreactive neurons were frequently observed in the SNR, but were rarely found in the SNC. These results are consistent with electrophysiologic evidence that SP exerts a tonic excitatory action on nigro-tectal GABAergic projections from the SNR (37). Furthermore, the excitatory effect has been shown not to be mediated via nigral GABAergic receptors, indicating that GABAergic interneurons are not part of this pathway. Inhibitory recurrent axon collaterals of nigro-thalamic and nigro-tectal projections from the SNR have also been demonstrated ( 14,2 1). These ostensibly GABAergic collaterals appear likely to exert their effect postsynaptically upon GABAergic or

SP PROJECTIONS

TO GABA-

AND

DA-POSITIVE

NIGRAL

NEURONS

561

The present study provides convincing evidence that SP-immunoreactive terminals contact TH-containing neurons in the substantia nigra. and suggests strongly that SP influences nigral DA neurons by direct synaptic action. The double-labeling technique employed in this study (34) was of greater sensitivity than those methods used previously (7.31.36) and permits greater definition of ultrastructural detail concerning synaptic structure. The proportion of TH-immunoreactive dendrites receiving synaptic input from SP-immunoreactive terminals was found to be much larger than that suggested by an earlier investigation (36). Because that study employed gold-iabelcd antibodies. which were subject to poor tissue penetration, this disparity was not surprising. This finding was consistent with electrophysiologic cvidence for a significant facilitatory influence mediated by SP on the neurons of the SNR (2). In agreement with other reports (31.36). both symmetrical and asymmetrical synaptic contacts between SP-immunoreactive

FIG. 3. SP-GAD interactions in the substantia nigra. (A) A GAD-immunoreactrve soma containing BDHC reaction product (arrows) and receiving synaptic contacts from two SP-immunoreactivc terminals (*) in pars reticulata (X5.200). (B) An SP-immunoreactive terminal (*) in pars lateralis synapsing (arrow) on a GAD-immunoreactive dendrite (D) that contains BDHC reaction product (open arrow) (X 17,420). (C) Two SP-immunorea~tive terminals (*) in pars lateralis forming synaptic contacts (arrows) with a GAD-immun~~reacti~~e dendrite (D) containing BDHf reaction product (open arrow) (X19.500).

dopaminergic neurons, because no evidence of presynaptic contact with axon terminals of GAD- or SP-immunoreactive fibers was seen in the present study. No evidence was found in the present study to support the notion that single striato-n&al Wimmunoreactive neurons also contained GAD-immunoreactivity within their terminals. It was reported that small numbers of presumptive striato-nigral neurons possess such dual immunoreactivity (IO). Using a doublelabeling technique at the light microscopic level, Penny et al. (40) reported the coexistence of GAD- and SP-immunor~activity in individual neostriatal neurons and their terminals in the globus pallidus. The possibility exists that a similar coexistence of neurotransmitters may yet be identified in the terminals of striatonigral neurons.

FIG. 4. Electron micrographs showing double-labeling for SP and TH in the pars compacta. (A) A TH-immunoreactive dendrite containing BDHC reaction product (arrowhead) and contacted by an SP-immunoreactive terminal (*) (~31.000). (B) A ‘TH-immunoreactive dendrite (arrowhead) contacted by two SP-immunoreactive axon terminals. Note the presence of pleomo~hic and dense-core vesicles in one terminal (*f. The other terminal contains spherical vesicles and possesses an asymmetrical contact (x24.880). (C) SP-positive terminal (*) making synaptic contact (arrow) with a GAD-positive dendrite containing BDHC reaction product (open arrow). Note the presence of an adjacent unstained terminal making contact (arrowhead) with the same dendrite (x 17.400).

567

MENDEZ.

E1.ISEVKt-I

AND

FLUMERFEI

multiple synapses with the soma and proximal dendrites

FIG. 5. Diagram showing the SP input and related connections in the substantia nigra: A massive SP projection from the ipsilateral striatum contacts dopaminergic (DA) neurons in the pars compacta and pars reticulata. SP-GABA and GABA-GABA (not shown) interactions were observed in the pars reticulata and pars lateralis. The pars reticulata and lateralis GABAergic neurons are likely of two types: (1) GABAergic intemeurons that, in turn, interact with nigro-striatal DA neurons in both the pars compacta and reticulata; (2) GABAergic projection neurons to the thalamus, superior colliculus, and reticular formation.

terminals and TH-immunoreactive neurons were identified, although the majority in both the SNC and SNR were ofthe former type. The ultrastructural characteristics of two distinct types of SP terminals were also identified. The more commonly observed SP terminals contain pleomorphic vesicles in keeping with the predominance of symmetrical synaptic contacts on TH-immunoreactive neurons. The second type of SP terminal contains spherical vesicles and makes asymmetrical contacts. This input is presumed to be of nonstriatal origin, because lesions of the striatum have not resulted in degeneration of these terminals in the ipsilateral substantia nigra (19,46). A number of TH-immunoreactive dendrites in the SNC and SNR are contacted by two or more SP-immunoreactive terminals forming symmetrical synapses with their target. These contacts are found among a number of unlabeled terminals, particularly in the SNR. Single SP-immunoreactive fibers in the SNC make

I

ofsmglc

nigral neurons in the rat (46). Substance P-containing terminals ofstriatal neurons ending upon TH-immunoreactive nigral neurons appear to form one limb of a reciprocating striato-nigro-striatal loop (5.18). Tyrosine hydroxylase-immunoreactive axonal boutons have been demonstrated in synaptic contact with SP-immunoreactive neostriatal neurons, suggesting that nigro-striatal dopaminergic neurons monosynaptically influence striato-nigral SP neurons (33). Moreover, pharmacological studies indicated that this nigrostriatal input was required to maintain striatal SP levels (23.5 I). Pharmacological and behavioral studies suggested an excitatory role for SP on nigral DA neurons. and injection of SP into the substantia nigra has been shown to increase the release and turnover of DA in the ipsilateral striatum (23,38.50). Furthermore. selective injections of SP in the SNR alone elicited dose-dependent contraversive circling (28). Interestingly, SP applied unilaterally to the SNC or SNL induced ipsiversive circling, suggesting an inhibitory interneuron as the target of SP terminals and highlighting the complexity with which nigro-striatal dopaminergic activity was regulated. A final question concerns the ultimate regulatory effect of GABA- and SP-containing neurons on nigral DA neurons that form the other limb of the striato-nigro-striatal loop. The substantia nigra receives a substantial input of both neurotransmitters from the striatum (6,17,24,26,42,47,48). Local control of neurotransmitter release in the substantia nigra has been suggested by in vitro studies showing that release of SP from nigral slices may be inhibited by GABA. whereas picrotoxin increases its release (29). However, there is little evidence of a similar presynaptic modulatory function of SP at GABAergic terminals (49). This study found no evidence of either type of axo-axonic contact, suggesting that GABAergic presynaptic modulation of SP release may be minor within the nigra. However, examination of SP-containing terminals and GABA receptors will be necessary before this issue is resolved. In summary, this study has demonstrated direct SP input onto GABAergic nigral neurons and has further characterized SP input onto nigro-striatal dopaminergic neurons (Fig. 5). This SP input may modulate striatal output via nigral GABAergic projection neurons. In addition, SP-mediated excitatory action may directly influence nigro-striatal dopaminergic activity. or may do so via inhibitory GABAergic neurons. ACKNOWL.EDGEMEIU”l-S

The excellent technical assistance of Mrs. J. Sholdice is gratefully acknowledged. We thank Dr. W. H. Oertel and the Laboratory ofclinical Sciences, NIMH. who provided the GAD antiserum. This investigation was supported by the Medical Research Council of Canada and the Upjohn London Neurosciences Program.

REFERENCES 1. Albe-Fessard, D.; Sanderson, P. A demonstration of tonic inhibitory and facilitatory striatal actions on substantia nigra neurons. In: Carpenter, M.; Jaramian, A., eds. The basal ganglia II. New York: Plenum Press; 1987:32 I. 2. Albe-Fessard, D.; Sanderson, P.; Mavoungou, R. The influence of striatum on the substantia nigra: A study using the spreading depression technique. Brain Res. Bull. 24:2 13-2 19; 1990. 3. Anden, N. E.; Magnusson, T.; Waldeck, B. Correlation between noradrenaline uptake and adrenergic nerve function after reset-pine treatment. Life Sci. 3: 19-25; 1964. 4. Araki, M.; McGeer, P. L.; McGeer, E. G. Striatonigral and pallidonigral pathways studied by a combination of retrograde horseradish peroxidase tracing and a pharmacohistochemical

method for gamma-aminobutyric acid transaminase. Brain Res. 331:17-24; 1985. 5. Beckstead, R. M. Striatal substance P cell clusters coincide with the high density terminal zones of the discontinuous nigrostriatal dopaminergic projection system in the cat: A study by combined immunohistochemistry and autoradiographic axon-tracing. Neuroscience 20:557-576; 1987. 6. Brownstein, M. J.; Mroz, E. A.; Tappaz, M. L.; Leeman, S. E. On the origin of substance P and glutamic acid decarboxylase (GAD) in the substantia nigra. Brain Res. I35:3 15-323; 1977. 7. Chang, H. T. Substance Pdopamine relationship in the rat substantia nigra: a light and electron microscopy study of double immunocytochemically labeled materials. Brain Res. 448:391-396; 1988.

SP PROJECTIONS

TO GABA- AND DA-POSITIVE

NIGRAL

8. Chevalier. G.: Vacher, S.; Deniau, J. M.: Desban. M. Disinhibition as a basic process in the expression of striatal functions. I. The striatonigral influence on tecto-spinal/tecto-diencephalic neurons. Brain Res. 334:215-226; 1985. 9. Childs, J. A.: Gale. K. Evidence that the nigrotegmental GABAergic projection mediated stereotype induced by apomorphine and intranigral muscimol. Life Sci. 33: 1007-1010; 1983. IO. Christensson-Nylander. 1.: Herrera-Marschitz. M.: Staines, W.: Hokfelt. T.: ‘Terenius, L.: Ungerstedt. U.; Cuello. C.: Oertel. W. H.: Goldstein. M. Striate-nigral dynorphin and substance P pathways in the rat: I. Biochemical and rmmunohistochemical studies. Exp. Brain Res. 64: 169-l 92: 1986. II Dahlstrom. A.: Fuxe. K.: Olson. L.: Ungerstedt. U. Ascending systems of catecholamine neurons from the lower brain stem. Acta Physiol. Stand. 624X5-486: 1962. I3 Davies. J.: Dray, A. Substance P in the substantia nigra. Brain Rcs. 10762%6’7: 1976. 13 Deniau. J. M.: Chevalier, G. Disinhibition as a basic process in the expression of striatal functions. II. The striatonigral influence of thalamocorticul cells of the ventromedial thalamic nucleus. Brain Res. 334:277-233: 1985. 14 Deniau. J. M.: Kitai. S. T.: Donoghue. J. P.: Grofova. I. Neuronal interactions in the substantia nigra pars reticulata through axon collaterals of the projection neurons. Exp. Brain Res. 47: 105-l 13: 19X2. 15 Dihglia. M.; Aronin. N.: Leeman. S. E. Immunoreactive substance P m the substantia nigra of the monkey: light and electron microscopic localization. Brain Res. 233:3X1-388: 1981. 16 Francois. C‘.: Percheron. G.: Yelnik. J.: Heyner. S. Demonstration of the existence of small local circuit neurons in the Golgi-stained primate substantia nigra. Brain Res. 172: 160-164: 1979. 17. Gale. K.: Hong. J. S.: Guidotti. A. Presence of substance P and GABA in separate striatonigral neurons. Brain Res. 136:37 l-375; 1977. IX Ciraybiel. ,A. M.; Ragsdale. C. W.: Moon-Edley. S. Compartments in the striatum of the cat ohservcd by retrograde cell labelling. Exp. Brain Res. 34: I X9-195: 1979. I9 Grofova. I.: Rinvik. E. An experimental electron microscopic study on the striate-nigral projection in the cat. Exp. Brain Res. Il:249262; 1970. 20 Grofova. 1.; Deniau. J. M.: Kitai. S. 7‘. Morphology ofthe substantia nigra pars rcticulata projection neurons intracellularly labeled with HRP. J. Camp. Neurol. 20X:352-368: 1982. 21 Groves. P. M. A theory of the functional organization of the neostriatum and the neostriatal control of voluntary movement, Brain Res. 2X6: IO9-- 132: 19X3. 22 Ciulle). R. L.: Smithberg. M. Synapses in the rat substantia nigra. Tissue Cell 3:69 I-700: 197 I. 23 Hanson, G.: Alphs. L.: Pradhan, S.: Lovenberg, W. Response 01 striatonigral suhstancc P systems to a dopamine receptor agonist and antagonist. Neuropharmacology 20:541-548; 1981. 24 Hattori. T.; McGcer. P. L.: Fibiger. H. C.: McGeer. E. G. On the source of GABA-containing terminals in the substantia nigra. Electron microscopic autoradiographic and biochemical studies. Brain Res. 54:103-l 14: 1973. 25. Hokfclt. 7 .; Ungerstedt, Ii. Electron and fluorescence microscopical studies on the nucleus caudatus putamen of the rat after unilateral lesions of ascending nigro-neostriatal dopamine neurons. Acta Physiol. Stand. 76:415-426: 1969. 26. Hong, J. S.: Y’ang. H. Y.: Racagni. G.; Costa, E. Projections of substance P containing neurons from neostriatum to substantia nigra. Brain Res. 122:541-544; 1977. 27. Inagaki. S.: Sakanaka. M.: Shiosaka. S.; Senba, E.; Takatsuki. K.: Takagi. H.: Kawai, Y.; Minagawa, H.: Tohyama, M. Ontogeny of substance P-containing neuron system of the rat: Immunohistochemical analysis-I. Forebrain and upper brain stem. Neuroscience 7125 I-271: I’)87. 28. James. T. A., Starr, M. A. Effects of substance P injected into the substantia nigra. Brit. J. Pharmacol. 65:423-429; 1979. 29. Jessell. T. M.: Emson. P. C.; Paxinos, G.: Cuello, A. C. Topographic projections of substance P and GABA pathways in the striato- and pallido-nigral system: A biochemical and immunohistochemical study. Brain Res. 152:487-498: 1978. 30. Juraska. J.: Wilson. C. J.: Groves, P. M. The substantia nigra of the rat: a golgi study. J. Comp. Neural. 172:585-600; 1977.

NEURONS

563

31. Kawai. Y.: Takagi. H.: Kumoi. Y.; Shiosaka, S.: Tohyama, M. Nigrostriatal dopamine neurons receive substance P-ergic inputs in the substantia nigra: Application of the immunoelectron microscopic mirror technique to fluorescent double-staining for transmitter-specific projections, Brain Res. 401:371-376: 1987. 32. Kohno. J.: Shiosaka. S.: Inagaki. S.: Tohyama. M. Two distinct strionigral substance P pathways in the rat: An experimental immunohistochemical study. Brain Res. 308:309-317: 1984. P neurons 33. Kubota. Y.: Inagaki, S.; Kito. S. Innervation ofsubstance by catecholaminergic terminals in the neostriatum. Brain Rcs. 375: 163-167: 1986. 34. Levey. A. 1.: Bolam. J. P.: Rye. D. B.: Hallanger. A. E.: Demuth. R. M.: Mesulam. M. -M.: Wainer. B. H. A hght and electron microscopic procedure for sequential double antigen localization using diaminobenzidene and benzidine dihydrochloride. J. Histochem. Cytochem. 34:1449-1457: 1986. 35. LJungdahl. A.: Hokfelt. T.: Nilsson. G.: Goldstein. M. Distribution of substance P-like immunoreactivity in the central nervous system of the rat-II. Light microscopic localization in relation to catecholamine-containing neurons, Neuroscience 3:945-976: 197X. of interactions between sub36. Mahalik. T. J. Direct demonstration stance P immunoreactive terminals and tyrosine hydroxylase immunoreactive neurons in the suhstantia nigra of the rat: An ultrastructural study. Synapse 2:50X-5 15: 1988. 37. Melis. M. R.: Gale, K. Evidence that nigral substance P controls the activity of the nigrotectal GABAcrgic pathway. Brain Res. 295:3X9393; 19x4. M. F.: Cheram). A.; 38. Michelot. R.: Leviel. L.: GiorguieIf-Chesselet. Glowinski. J. EIfects of the unilateral nigral modulation ofsubstance P transmission on the activity of the two nigro-striatal dopamincrgic pathways. Life Sci. 24:715-724: 1979. 39. Mugnaini. E.: Oertel, W. H. An atlas ofthe distribution ofGABAergic neurons and terminals in the rat CNS as revealed hy GAD immunohistochemistry. In: Bjorklund, 4.: Hokfelt. T. eds. Handbook of chemical neuroanatomy. Amsterdam: Elsevier: 1985:436-622. 40. Penny. G. R.: Afsharpour. S.: Kitai. S. T. The glutamate decarboxylase-, lcucine enkephalin-. methionine cnkephalin- and substance P-immunoreactive neurons in the neostriatum of the rat and cat: Evidence for partial population overlap. Neuroscience 17: IO I I-1045: 1986. 41. Pinnock. R. D.: Woodruff G. N.: Turnbull. M. J. Actions of substance P. MIF, TRH and related peptides in substantia nigra. caudate nucleus and nucleus accumbens. Neuropharmacology 22:687-696: 1983. 42. Rihak. C. E.: Vaughn. J. E.: Roberts, E. GABAergic nerve terminals decrease in the substantia nigra following hemitranxctions of the striatonigral and pallidonigral pathways. Brain Res. 924 13-420: 1980. 43. Ribak. c‘. E.: Vaughn. J. E.: Saito. K.: Barber. R. P.; Robert\. E. Immunocytochemical localization of glutamate decarboxylase in rat substantia nigra. Brain Res. 116:2X7-298: 1976. 44. Scheel-Kruger, J. Dopamine-GABA interactions: evidence that GABA transmits. modulates and mediates dopaminergic functions in the basal ganglia and the limhic system. Acta Neurol. Stand. 73fSuppl. 107): l-49; 1986. 45. Schwyn. R. C.: Fox. C. A. The primate suhstantia nigra; A golgt and electron microscopic study. J. Himforsch. I5:95- 126: 1974. 46. Somogyi, P.: Priestley, J. V.: Cuello. A. C.: Smith. A. D.: Bolam. J. P. Synaptic connections of substance P-immunoreactive nerve terminals in the substantia nigra of the rat. Cell Tissue Res. 223: 469-486: 1982. 47. Sperk. G.; Smger. E. A. In vivjo synthesis of substance P in the corpus striatum of the rat and its transport to the substantia nigra. Brain Res. 23X:127-135: 1982. 48. Staines. W. A.: Nagy. J. I.; Vincent. S. R.: Fibiger. H. c‘. Neurotransmitters contained in the e&rents of the striatum. Brain Res. 194:391-402: 1980. 49. Starr. M. S. Investigation of possible interactions between substance P and transmitter mechanisms in the substantia nigra and corpus striatum of the rat. J. Pharm. Pharmacol. 30:359-363: 1978. 50. Waldmeier. P. C.: Kam, R.: Stocklin. K. Increased dopamine metabolism in the rat striatum after infusion of substance P into the substantia nigra. Brain Res. I59:223-227; 1978. 51. Zivkovic. B.: Oblin. A.; Bartholini. G. Pogabide reverses the nigral substance P reduction induced by chronic impairment of dopaminergic transmission. Eur. J. Pharmacol. Il2:253-255; 1985.