Brain Research, 117 (1976)423-435 © Elsevier/North-Holland Biomedical Press, Amsterdam - Printed in The Netherlands
423
THE PRECISE LOCALIZATION OF NIGRAL AFFERENTS IN THE RAT AS DETERMINED BY A RETROGRADE TRACING TECHNIQUE
B. S. BUNNEY and G. K. AGHAJANIAN
Departments of Psychiatry and Pharmacology, Yale University School of Medicine, New Haven, Conn. 06508 (U.S.A.) (Accepted April 13th, 1976)
SUMMARY
Afferent innervation of the rat substantia nigra (SN) was studied by the retrograde horseradish peroxidase (HRP) method. High concentrations of HRP were deposited in discrete subregions of the SN by means of a microiontophoretic delivery system. Using this technique it was possible to demonstrate that the caudatonigral projection system is arranged topographically. All portions of the caudate-putamen except for a central medial core were found to contain HRP positive cells, indicative of retrograde transport. In the positive areas a much larger percentage ofcells (30-50 %) were found to particin~.te in this projection than has previously been reported. Only medium size cells (12-20/~m) were found to contain the HRP reaction product. Other areas found to heavily innervate the SN were the globus pallidus, central nucleus of the amygdala and dorsal raphe nucleus. Areas containing fewer reactive cells but which also appear to innervate the SN included the prefrontal cortex and lateral habenula. These results emphasize the, importance of striatonigral projections which recent studies have suggested contain a G ABAergic link.
INTRODUCTION
The substantia nigra (SN), which has been studied intensively in recent years because of its demonstrated involvement in the pathogenesis of Parkinson's disease 1~, 24, consists of two divisions" the zona compacta (ZC) and the zona reticulata (ZR). The ZC and the ZR have been differentiated histochemically, pharmacologically and electrophysiologicailylt,17,4~. The ZC has been shown to consist almost exclusively of dopamine-containing cells that project to the caudate-putamen nucleus u,47. The ZR appears to possess both interneurons and neurons which project to the striatum and thalamus ~,15,33. The aff~.rent connections to the SN are less well known. The first evidence for a projection from the basal ganglia to the SN was reported in 191113.
424 Since then numerous degeneration studies using both light and electron microscopic techniques have der,lonstrated striatonigral and pallidonigral pathways 21,4a,40. in addition, light and electron microscopic autoradiographic studies have also demonstrated both striatal and pallidal projections to the SN 22. These latter studies strongly suggest that the caudatonigrai pathway predominantly innervates the ZR whereas the pallidonigrai projection is mainly to the ZC. None of these studies, however, specifically identifies the location of the cells of origin for these pathways within the caudate nucleus (CN) and globus pailidus (GP). Recently, retrograde axonal transport of horseradish peroxidase (HRP) has been added as a new technique for studying the afferent connections of the SN z°. Using this technique the existence of both a striatal and a pallidal input to the SN has been confirmed in the cat e0. The experiments reported here, which use the HRf" technique to study rat SN afferent connections, reveal the ~pecific location of cells projecting to the SN and demonstrate a topographical arrangement for the striatal innervation of the SN. METHODS Male Charles River albino rats weighing 240-300 g were used (N == 26). Two series of experiments were performed. In both, the rats were deeply anesthetized with chloral hydrate during the deposition of HRP. In the first series (N = 10) HRP (Sigma VI) was injected by means of a 5/~! Hamilton syringe with a fixed 3!-gauge needle. -The needle was inserted stereotaxically into the SN using two approaches: a vertical approach (coordinates A 1950 ~tm, L 1800 ttm; according to K/Snig and Klippe126) and an oblique approach (same anterior coordinates but from the side at an angle of 49°). In both cases care was taken not to damage the crus cerebri. HRP was dissolved in distilled water to a concentration of 25 % and 0.05/~1 was illjected over a 0.5 h period. In another series of experiments (N -- 16) a more precise localization within the SN was obtained by the use of a microiontophoretic delivery system for depositing the HRP. Again both a vertical and an oblique approach were used. Using a modification of the technique described by Graybiel and Devor TM, it consisted of the following seeps. Glass tubing (2 mm) filled with fiber glass was pulled into micropipettes, the tips of which were then broken back to a diameter of approximately 40/~m. The pipettes were filled with 6/~1 of a 25 % solution of HRP in 0.01 M NaCI by direct injection into the hub of the mic~'opipette with a Hamilton syringe. After a pipette was stereotaxically lowered into the SN, a 4/~A positive current was applied to the H R P solutien for 5 min by means ofa DC constant current source thereby ejecting HRP molecules into the SN. Control ejections (N = 6) were made into the dorsal raphe nucleus, the reticular formation dorsal to the SN, and the crus cerebri (causing deliberate damage to fibers with the injection cannulae)just posterior to the SN. The HRP histochemical technique used is a modification of that described by Kristensson et al. 2s and LaVail and LaVail al. Twenty-four hours after the placement of HRP within the brain the animals were anesthetized and the brain fixed by intracardiac perfusion for 0,5 h with a solution consisting of 1% paraformaldehyde and
425
.
-."
~'...:
.
.
•
~ . , _ _
•
.
o
,i.
' ~
. ~
-
. . . . .
"
' -
%,:-
•
-"
.-
It. : ~"
\
,.
" ,
"
. . . .
-;,
..
-
~'
, o
~
~
." : ....
•
,,
"
".
"
x;-
.
"
' ',
;.,
"%"
"i
,-:
.
'~
.-
~'
.
,
'i , - ,
.:,.. '. •
,t
.
¢ "~,,"~i'~i~..
-
.
•
"
-
b
•a
•
- . " . . . . t.
'4
"
,
t-'.:.-';~:r "" .. '
.'• -.~
"
e,t
:.~"ll..
"
~ ,.....,. ,:
I," t
,, -
,-~,'.
"~ " '" : . . .
..'
.
,, •
•
• • S
-
-%
. "
;.. hl;~/!"--: ":/,lllt:'~'.
.".
i
.
I'.
".
...
,: : ,
/
."
/.." "" i i
,f/
'-,:,..,,:-.-"..liinll;~,-. "
/
.
e , ~ t ~ . ' ! l . l ~ l ~ ' '~
.,
y',,"
i.I
Fig. I. Bright field photomicrograph illustrating the distribution of HRP reaction product after microiontophoretic ejection into the right substantia nigra. Section is unstained.
1% glutaraldehyde in 0.05 M phosphate buffer (pH 7.3). After storage overnight in 0.05 M Tris buffer with 5 ~o sucrose, frozen serial transverse sections (50 fire) of the brain were cut and collected in 0.05 M Tris buffer (pH 7.6). The sections were immediately transferred to trays containing 3 ml of a 1.4 mM solution of 3,3-diaminobenzidine (DAB) (free base, Sigma) in Tris buffer which had been filtered through Whatman glass fiber paper GF/B under vacuum. The sections were incubated for 30 rain at 35 °C. After cooling to room temperature 0.4 ml of a 0.03 o~ solution of H202 was added to the DAB solution which was then gently agitated for I h. Sections were then rinsed twice in Tris buffer (0.05 M, pH 7.6) and once in distilled water. Every fourth section was mounted from gelatin alcohol (0.5% gelatin, 80~'0 alcohol), dehydrated, coverslipped and examined in a light microscope using both bright and dark fie!d illumination. RESULTS
In the experiments using a Hamilton syringe, both the vertical and oblique approaches resulted in the appearance of diffuse HRP reaction product within portions of the ZC and the ZR. With the oblique approach, since the needle tract followed the natural contour of the SN, the reaction product was limited to the confines of the nucleus itself. In this case no damage occurred to fibers in the crus cerebri. Nevertheless, in these experiments fibers in the crus cerebri contained dense accumulations of HRP reaction product and could be traced to the internal capsule and into the
426
Fig. 2. Dark field photomicrogl~tphs of the caudate nucleus (anterior 78 %, according to Kt~nig and Kiippel"% The lateral portion of the caudate nucleus (A) is seen to contain numerous H R P reactive cells whereas the medial portion (B) is totally devoid of labeled cells. A ~breviations: CC, corpus callosum; LV, lateral ventricle. Scale bar: 20 ,ml.
427 TABLE
I
Topography o f caudatonigral projections Abbreviations: ZC, zona compacta ; ZR, zona reticulata; V, v e n t r a l ; VL, ventral lateral; L, lateral; D L , dorsal lateral; D, dorsal; T, tail. Abbreviations correspond to areas m a r k e d V, L, D a n d T in Fig. 2. .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Location o f HRP ejection sites hi substantia nigra*
Location of reactive cells" at 5 representative frontal planes withht the caudate Ilttc[etls
Frontal plane
Anterior 9410" *
Laterality
Zone
Anterior 7890
Anterior 6790 .
.
.
Anterh~r 5660 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Anterior 41 !0 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Anterior A2420 A2180 A2180 A2180
Intermediate Lateral Intermediate Medial
ZC and Z R ZC and ZR ZR ZC
V -D, L, V V
VL -DL VL
L L L DL
VL L D D
T T T T
( , ***) ( - ~ ) (--) (~
Lateral Intermediate Intermediate Medial
ZR ZC ZR ZC
V, VL -D, DL, L ~
V,VL V D, DL. L L
V, VL V D, DL DL
VL,L V D D
T ( : ! ) T (~) T ( . ) T( : )
Lateral Intermediate Intermediate
ZC Z C and Z R ZR
----
V VL, L, L
V, VL VL L
V V VL
T ~i ~ ) -- (--) T (.)
Middle AI950 A!950 AI950 Ai950
Posterior AI760 AI760 AI610
* Represent area of densest H R P deposition. ** Frontal planes according to K6nig and Klippel"'k *** -!, 5 - 1 0 H R P reactive cells per section; ! - , , 20 H R P reactive cells per section.
tail of the caudate nucleus to cells with granular (particulate) HRP reaction product. Under bright field microscopy these fibers did not show the uniform brown color which, according to other investigators 37, is characteristic for an axonal reaction secondary to injury. Rather they had a granular appearance with multiple varicosities visible under dark field illumination. In the experiments using microiontophoresis HRP was deposited in small (500 #m diameter) well localized, but highly concentrated amounts within various subregions of the SN (Fig. I). The use of both a vertical and oblique approach al!owed us to control for damage to fibers of passage everywhere except within the SN itself. However, the small tip diameter of our micropipettes (40-50/~m) coupled with the fact that H RP is ejected iontophoretically rather than hydrolically ensured a minimum of tissue damage at each injection site and eliminated the problem of passage of HRP up the tract made by the micropipette. Using this technique it was possible to limit the deposition of HRP to the lateral, intermediate, or medial one-third of the SN as well as to the ZR or to the ZC. The latter was made pn~,ible by the natural dorsal barrier to dorsal spread made by the fibers of the medial lemniscus. The Hamilton syringe injection studies were u ~ d as a general guide to areas of the brain which might innervate the SN. Only results which were common to both sets of
42g
\"
A 7890
DD
•
•
io.;.ibt, . //-st -Z !v v v/:
c'~A,/;.",'}..ttt.]/ •
A 6790
A5660
•
I
]
$R
__.,: .-",....i-:-?
A4110
)
AI950
A550
Fig. 3. Schematic tlra~ings of representative frontal sections o1" rat brain demonstrating location of cells containing H RP reaction l')oduct (black do!,, and bold face lettersl following an ejection of HRP into the substantia nigra. The letters (I), L. V, Tj ~ithin the caudate nucleus (Cp) designate lopographical projection :lrcas to the SN described in Table I. Drawings modified from K~.Snig and Klippel'-'~L Abbreviations: ac, nucleus amygdaloideus centralis: CA, conanaissura anterior; CAI, capsula intcrna: CL, claustrum ; C'p, nucleus caudatus putamen ; FM P, fascictdus medialis prosencephall; GP, globus pallidus: HI, hippo~'ampus: ip, nucleus interpeduncularis; LM, iemniscus n)edialis; mr, nucleus medianus raphes: PCS. pedunculus cerebe!laris superior; SNR, substantia nigra, zona reticulata; SR, sulcus rhinalis, I), di_~rsal: L. lateral; V, ventral; T, tail. experiments are reported here. Every grour) of cells labeled in these experiments occurred on the side of the brain ipsilaterai to the injection site.
.4rea.~ with a high de,,siti" o.1"H R P retr'tive cells ( ! ) Bu.salganglia. The head of the CN contained numerous reactive cells which ranged fi'om 12 to 20 l~m in size (Fig. 2). No "giant" cells were found with a positive H RP reactione:'. Examination of section~ of the C N which were stained with cresyi violet revealed that between 30 and 50",, of neurons in a given H RP labeled area of the C N project to the SN. The combined result,, o f our SN H RP studies showed that all parts o f the CN contain cells which project to tiae SN except for a medial cure (,Figs. 2, 3). Only a few scattered cells were found to be labeled in the latter area
429
Fig. 4. l)ark field photomicrograph of the cerebral cortex tA}, just dorsal to tl~e rhinal sulcus, and globus p:~llidus (B) sho~ing cells containing HRP reaction product follo~ing an ejcction of HRP into the substantia nigra. Scale bar: 20 ,,,m.
430 regardless of the SN ejection site. The distribution of labeled cells varied according to thesite of HRP deposition within the SN and demonstrated a topographical projection from the CN to the SN in both a medial-lateral and an anterior-posterior arrangement (Table I). Thus HRP ejected into the medial or lateral one-third of thc SN resulted in reactive cells in the dorsal and lateral areas of the CN respectively. Anteriorally placed ejections in the SN resulted in labeled cells in anterior but not posterior areas of the CN. However, an ejection into the ZC often resulted in different caudate cells being labeled from those labeled when HRP was ejected into the ZR (even though both ejections were in the same frontal and sagittal plane), thereby greatly adding to the complexity of the topographical arrangement (see Table I). Reactive cells were also seen in the tail of the CN (Fig. 3). The greatest number of cells in this area were seen labeled after lateral ejections, but ejections into most other parts of the SN also resulted in labeled cells in the tail of the CN (Table I). In the GP (Figs. 3 and 4) reactive cells were most numerous in the posterior-lateral portion and were progressively less frequent toward the anterior pole. Ejections into the lateral one-third of the SN, no matter what the anterior-posterior location, resulted in the labeling of raany cells within the central nucleus of the amygdaloid complex (Fig. 3). No other nuclei within that complex were labeled. (2) Accumbens nucleus. The deposition of HRP into the medial one-third of the SN and adjacent ventral tegmental area resulted in labeled cells within the accumbens nucleus. (3) Dorsal raphe nucleus. Every HRP ejection into the SN resulted ill reactive cells in the ipsilateral half of the dorsal raphe nucleus (Fig. 3).
Areas with a low density of HRP reactive cells (1) Cortex. Every HRP ejection into the SN resulted in occasional reactive cells scattered throughout the parietal, insular and temporal cortex (according to Zeeman and Innes 5°) (Figs. 3 and 4). No labeled cells were ever seen ventral te the rhinal sulcus. The prefrontal cortex consistently contained a few heavily labeled cells. Regardless of the cortical area labeled, only layer 5 appeared to contain reactive cells. (2) Habenula. Microiontophoretic ejection of HRP into the lateral one-third of the SN resulted in an occasional reactive cell in the lateral habenula. (3) Hypothalamus. Every ejection into the ZC of the SN yielded a scattering of la~beled cells in the hypothalamus although the location within the hypothalamus varied from animal to animal. No preoptic nuclei contained reactive cells. HRP reactive cells after control injections When ejections were made into the crus cerebri posterior to the SN, reaction product was found in fibers and cells immediately surrounding the ejection sites, but never resulted in labeling of either the internal capsule or the cell groups described above. Contro| ejections into the reticular formation dorsal to the SN resulted in the labeling of cells in the parietal and prefrontal cortex, hypothalamus (ventral part of the dorsal medial nucleus), locus coeruleus and contralateral dentate and
431 emboliformis nuclei. Deposition of HRP into the dorsal raphe nucleus failed to label any cells in the CN, GP, or insular, temporal, or parietal cortex. DISCUSSION Nigral afferents from the basal ganglia have been demonstrated by several investigators ~0,22,45,49. However, due to the limitation of the techniques used in these studies (degeneration and autoradiography) it has not been possible to identify exactly which cells project to the SN. Previously it has been suggested that only 5 O//oofcaudate cells leave the CN to innervate distal structureslS,~L Using the retrograde H RP technique it has been possible to identify precisely which cells form the caudatonigral projection. Our study suggests that 30-50 O//oofcaadate cells, except for a medial core, project to the SN. Since most of the CN is involved in this projection, CN cells projecting outside the CN are clearly much more numerous than previously expected. Other investigators, using degeneration techniques in monkeys36,4~, 49 and cats as, 4s,49, have found a topographical distribution of caudatonigral fibers although they have differed as to the specific arrangement. Our study in the rat demonstrated a mixed anterior-posterior medial-lateral topographical distribution of the striatal efferents to the SN as well as v difference between ZC and ZR caudatonigral afferents. Because of this complex topographical arrangement an enormous number of highly localized ejections would be necessary to obtain a complete topographical map. The 14 microiontophoretic ejections performed in this study, therefore, do not allow us to designate fully the striatonigral topographical projections. Although we found that tbe tail of the CN projected to all parts of the SN, the lateral one-third received the heaviest innervation from this area. These findings, in the rat, are in marked contrast to those tbund in the monkey where the tail of the caudate has been reported to project to the caudal SN posterior to the emergence of NIII fibers from the midbrain 4'~. The finding of HRP reaction product in cells of both the CN and the GP adds further support to the concept of two pathways from the basal ganglia to the SN2°,~L Autoradiographic studies in the rat by Hattori et al. 2~ suggest that the GP cells mainly project to the ZC. We found more HRP reactive cells in the GP after discrete ejections into the ZR than after well localized ZC ejections. A recent behavioral study has suggested a feedback pathway to the SN involving cell: bodies located in the claustrum TM. However, no HRP reactive cells were found in this structure. Of equal importance to demonstrating the anatomical connections to the SN is determining the neurotransmitter used by each afferent system. The two highest ranking candidates for mediators of chemical transmission in the caudatonigral and/or pallidonigral pathways are gamma-aminobutyric acid (GABA) and acetylcholine (ACh). Substantial evidence now exists in favor of GABA and against ACh 17'aa. Thus lesion of the globus pallidus or transection of the brain between the basal ganglia and SN results in decreased levels of GABA and glutamic acid decarboxylase (GABA synthesizing enzyme) activity in the SN. The same lesions, however have no effect on the concentrations of ACh in the SN ~7,a3. Labeled GABA injected into the GP has been shown both chemically and by electron microscopic radio-
432 autography to be transported to the SN whereas, when a similar injection was made into the CN, no radioactivity could be detected in the SN 17. These latter experiments, therefore, suggest that the pallidonigral pathway is GABAergic, whereas the caudatonigrai pathway is not. Electrophysiological studies also have provided evidence for a GABAergic pathway from the basal ganglia to the SN. Stimulation of the head of the caudate nucleus has been shown to cause depression of neuronal activity in the SN. This depression was prevented by intravenous administration of picrotoxin (a presumed GABA antagonist) 4°. In addition, microiontophoretically applied GABA has been found to depress SN neurons and this depressant effect can in turn be blocked by microiontophoreticaily applied picrotoxin 1°. Paralleling studies of reciprocal connections between the SN and the CN, biochemical and electrophysiological studies on the mechanism of action of certain drugs also have provided indirect evidence for a feedback pathway from the accumbens nucleus to dopaminergic neurons in the midbrain ventral tegmental area (A I0). A recent autoradiographic study carried out in 2-weekold rats has reported a p:ojection from the accumbens nucleus to the mediaimost pole of the SN and to the A l0 area 44. in our studies in adult rats, microiontophoretic placement of HRP into the extreme medial edge of the SN adjacent to the A 10 area resulted in reactive cells within the accumbens nucleus, providing further anatomical evidence for the existence of such a pathway. A dopaminergic input to the central nucleus of the amygdala (AC) has been known fo~ some time 47. Recent biochemical studies have shown that the AC contains *,he highest concentration of dopamine of any nucleus within the amygdaloid complex .',~. Injection of HRP into the midbrain tegmental area has been reported to result in heavy labeling of cells in the AC 2a. However, as degeneration studies confined to the AC have resulted in marked fiber degeneration in the median forebrain bundle (M FB) 1~ it is possible that the labeling of cells in the AC resulting from the injection of HRP into the tegmentum may have been due to damage of the MFB which passes through this area. On the other hand, in the studies reported here, tissue damage was minimal and discrete injections into the lateral one-third of the SN, all of which were far from the MFB, consistently resulted in heavy labeling of cells in the AC. Such results suggest the possibility of a neuronal feedback system from the amygdala to the SN, similar to that known to exist for the nigrostriatal and mesolimbic dopaminergic systems. As several investigators using a variety of techniques have demonstrated a serotonergie (5-HT) projection to the SN 7,11,29'39, the finding of cells containing HRP reaction product in the dorsal raphe nucleus was predictable. Our study does not permit us to determine w,t,,ich part ofthe SN, ZC or ZR receives the 5-HT innervation. However, when 5-HT is applied directly to cells in the SN by means of microiontophoresis it has little effect on ZR neurons but totally blocks glutamate excitation of ZC cells 1. Such evidence suggests that the ZC dopaminergic neurons may be the ones primarily influenced by the serotonergic input. The findings of HRP reactive cells in parts of the cortex is consistent with the long standing suggestion of the existence of a corticonigral pathway 6,aS,4a. However, in other studies few if any degenerating terminals were observed in the SN after large
433 lesions of the ipsilateral cortex27, 41. In addition, a recent autoradiographic study of projections of the precentral motor cortex fails to mention any labeling in the SN '~0. The prefrontal cortex was unique in that it consistently demonstrated the highest density of labeled cells and these were also the most heavily labeled. These results may be explained by the finding that a fiber contingent from the prefrontal cortex passes laterally over the dorsal surface of the SN 3z. However, it has also been suggested that because these fibers become fewer in number as they pass caudad, some of them are terminating in the SN 32. Although not answered by our study the question of the existence of a corticonigral pathway is of some importance (especially if coming from the parietal cortex) as it would provide a second feedback pathway system for sensory motor areas to the basal ganglia and/or thalamus (via the SN). Further autoradiographic studies such as those recently reported by Kiinzle 30 may provide a definitive answer to this question. Two other areas of the brain were found to contain reactive cells - - the habenula and the hypothalamus. Never more than one or two cells were seen labeled on any one section. The existence of projections from these structures to the midbrain has previously been reported s,34. However, innervation of the SN by these areas must await confirmation by the use of other techniques. In conclusion, the importance of the existence of a pathway from the basal ganglia to the SN has been underscored in recent years by increasing evidence that certain drug effects in the CNS may be induced by an action on such a neuronal system. For example, a feedback pathway has been suggested to mediate the effects of amphetamine 9 and the antipsychotic drugs 4 on dopamine turnover in the CN by inducing changes in dopaminergic cell firing rate. We report here the localization of the cells of origin for two such pathways in the rat. The rat pallidonigral pathway was found to be similar to that recently described for the cat 2°. However, the cells making up the rat caudatonigral pathway appear to have a somewhat different distribution. Thus, in the rat, all parts of the CN appear to project topographically to the SN except for a medial core. in addition there is a relativley large projection from the tail of the CN to the SN. These findings suggest that physiological and pharmacological studies designed to characterize further these pathways are feasible. Selective stimulation and lesioning of these areas should enable one to determine the role each of these pathways plays in modulating the spontaneous activity of cells in the ZC and ZR and in mediating the effect of drugs on the activity of dopaminergic neurons. For example, recent electrophysiological studies in our laboratory 2 have shown that the depressant effects of D-amphetamine on dopaminergic cell firing rate 3, which have been presumed to be mediated via a neuronal feedback mechanism, can be blocked by lesioning the tail of the caudate nucleus. ACKNOWLEDGEMENTS This research was supported by NIMH Grants (MH-17871; MH-14459; MH-25642), The Benevolent Foundation of Scottish Rite Freemasonry, Northern Jurisdiction, U.S.A., and the State of Connecticut.
434 REFERENCES 1 Aghajanian, G. K. and Bunney, B. S., Dopaminergic and non-dopaminergic neurons of the substantia hi,ca: differential responses to putative transmitters. :in J. R. Boissier, H. Hippius and P. Pichot (Eds.), Proc. 9th Int. Congr. of the Collegium Interna, ionale Neuropsychopharmacologicure, Excerpta Medica Foundation, Amsterdam, 1974. 2 Bunney, B. S. and Aghajanian, G. K., D-Amphetamine-induced inhibition of central dopaminergic neurons: mediation by a striatonigral feedback pathway, Sctence, 192 (1976) 391-393. 3 Bunney, B. S., Waiters, J. R., Roth, R. H. and Aghajanian, ¢2. X., Dopaminergic neurons: effect of antipsychotic drugs and amphetamine on single cell ac, i','it?', J. Pharmacol. exp. Ther., 185 (1973) 560-571. 4 Carlsson, A. and Lindqvist, M., Effect of chlorpromazine a~.~d haloperidol on formation of 3methoxytyramine and normetanephrine in mouse brain, Acta pharmacol. (Kbh.), 20 (1963) 140-144. 5 Cole, M., Nauta, W. J. H. and Mehler, W. R., The ascending efferent projections of the substantia nigra, Trans. Amer. neuroL Ass., 89 (1964) 74-78. 6 Combs, C. M., The distribution and temporal course of fiber degeneration after experimental lesions in the rat brain, J. comp. NeuroL, 94 (1951) 123-176. 7 Conrad, L. C. A., Leonard, C. M. and Pfaff, D. W., Connections of the median and dorsal raphe nuclei in the rat: an autoradiographic and degeneration study, Y. comp. Neurol., 156 (1974) 179-206. 8 Conrad, L. C. A. and Pfaff, D. W., Axonal projections of medial preoptic and anterior hypothalamic neurons, Science, 190 (1975) 1112-1114. 9 Corrodi, H., Fuxe, K. and H6kfelt, T., The effect of some psychoactive drugs on central monoamine neurons, Europ. J. Pharmacol., 1 (1967) 363-368. 10 Cro~sman, A. R., Walker, R. J. and Woodruff, G. N., Picrotoxin antagonism of ~,-aminobutyric acid inhibitory responses and synaptic inhibition in the rat substar.~ia nigra, Brit. J. Pharmacol., 49 ~1973) 696-698. 11 Dahlstr/~m, A. and Fuxe, K., Evidence for the existence of monoamine containing neurons in the central nervous system, Acta physiol, scand., 62, Suppl. 232 0964) I. 12 De Olmos,J. S., The amygdaloid projection field in the rat as studied with the cupric-silver method. In B. E. Eleflheriou (Ed.), The Neurobiology of the Amygdala, Plenum Press, New York, 1972, pp. 145-204. 13 Edinger, L., Vorlesungen fiber den Bau der nervosen Zentralorgane, VoL 1, Vogel, Leipzig, 8th ed., 1911. 14 Ehringer, H. and Hornykiewicz, O., Verteilung yon Noradrenalin und Dopamin (3-Hydroxytyramine), im Gehirn des Menschen und ihr Verhalten bei Erkrankungen des extrapyramidalen Systems, Klin. Wschr., 38 (1960) 1236-1239. 15 Fauli, R. L. M. and Carman, J. B., Ascending projections of the substantia nigra in the rat, J. comp. Neurol., 132 (1968) 73~92. 16 Fleisher, L; N. and Giick, S. D., A telencephalic lesion site for D-amphetamine-induced contralateral rotation in rats, Brain Research, 96 (1975) 413-417. 17 Fonnum, F., Grofov~i, I., Rinvik, E., Storm-Mathisen, J. and Walberg, F., Origin and distribution of glutamate decarboxylase in substantia nigra of the cat, Brain Research, 71 (1974) 77-92. 18 Fox, C. A., Andrade, A. N., Schwyn, R. C. and Rafols, J. A., The aspiny neurons and P,a glia in the primate striatum: a Golgi and electron microscopic study, J. Hirnforsch., 13 (1971) 341-362. 19 Graybiel, A. M. and Devor, M., A microelectrophoretic delivery technique for use with horseradish peroxidase, Brain Research, 68 (1974) 167-173. 20 Grofov~i, I., The identification of striatal and pailidal neurons projecting to substantia nigra. An experimental study by means of retrograde axonal transport of horseradish peroxidase, Brain Research, 91 (1975) 286-291. 21 Grofovfi, I. and Rinvik, E., An experimental electron microscopic study on the striatonigral projection in the cat, Exp. Brain Res., 11 (1970) 249-262. 22 Hattori, T., Fibiger, H. C. and McGeer, P. L., Demonstration ofa pallidonigral projection innervating dopaminergic neurons, J. comp. Neurol., 162 (1975) 487-504. 23 Hopkins, D. A., Amygdalotegmental projections in the rat, cat and rhesus monkey, Neurosci. Lett., 1 (1975) 263-270. 24 Hornykiewicz, O., Dopamine (3-hydroxytyramine) and brain function, Pharmacol. Rev., 18 (! 966) 925-964.
435 25 Kemp, .l.M. and Powell, T. P. S., The structure of the caudate nucleus of the cat: light and electron microscopy, Phil. Trans. B, 262 (1971) 383-401. 26 K0nig, J. F. R. and Klippel, R. A., The Rat Brain: a Stereotaxic Atlas, Krieger, New York, 1970. 27 Knook, H. L., The Fibre Connecti~ns of t/'.e Forebrain, Royal Van Gorcum, Assen, 1965, p. 81. 28 Kristensson, K., Olsson, Y. and Sj0strand, J., Axonal uptake and retrograde transport of exogenous proteins in the hypoglossal nerve, Brain Research, 32 (1971) 399-406. 29 Kuhar, M. J., Aghajanian, G. K. and Roth, R.H., Tryptophan hydroxylase activity and synaptosomal uptake of serotonin in discrete brain regions after midbrain raphe lesions: correlations with serotonin levels and histochemical fluorescence, Brain Research, 44 (1972) 165-176. 30 Kiinzle, H., Bilateral projections from precentral motor cortex to the putamen and other parts of the basal ganglia. An autoradiographic study in Macaca foscicularis, Brain Research, 88 (1975) 195-209. 31 LaVail, .1. H. and LaVail, M. M., Retrograde axonal transport in the central nervous system, Science, 196 (1972) 1416-1417. 32 l.eonard, C. M., The prefrontal cortex of the rat. I. Cortical projection of the mediodorsal nucleus. II. Efferent connections, Brain Research, 12 (1969) 321-343. 33 McGeer, P. L., Fibiger, H. C., Hattori, T., Singh, V. K., McGeer, E. G. and Maler, L., Biochemical ne~ roanatomy of the basal ganglia. In R. D. Myers and R. R. Drucker-Colin (Eds.), Advances in Behavioral Biology, Vol. !0, Plenum Press, New York, 1974, pp. 27-47. 34 McGeer, P. L., Hattori, T. and McGeer, E. G., Axonal transport in the habenular interpeduncular cholinergic tract, biochemical and autoradiographic studies, Neurosci. Abstr., 1 (1975) 851. 35 Minkowski, M., Etude sur les connections anatomiques des circonvolutions rolandiques, pari6tales et frontales, Schweiz. Arch. Neurol. Psychiat., 12 (1923) 71-104; 227-268. 36 Nauta, W. J. H. and Mehler, W. R., Projections of the lentiform nucleus in the monkey, Brahz Research, 1 (1966) 3-42. 37 Nauta, H. J. W., Pritz, M. B. and Lasek, R. J., Afferents to the rat studied with horseradish peroxidase. An evaluation of a retrograde neuroanatomical research method, Brain Research, 67 (1974) 219-238. 38 Niimi, K., Ikeda, T., Kawamura, S. and Inoshita, H., Efferent projections of the head of the caudat~ nucleus in the cat, Brain Research, 21 (1970) 327-343. 39 Parizeg, J., Hassler, R. and Bak, I. J., Light and electron microscopic autoradiography of substantia nigra of rat after intraventricular administration of tritium labeled norepinephrine, dopamine, serotonin and the precursors, Z. Zellforsch., 115 (1971) 137-148. 40 Precht, W. and Yoshida, M., Blockage of caudate-evoked inhibition of neurons in the substantia nigra by picrotoxin, Brain Research, 32 (197 I) 229-233. 41 Rinvik, E. and Walberg, E , Is mere a coi'tlco-nigra| tract.' A comment based on experimental electron microscopic observations in the cat, Brain Research, 14 (1974) 742-744. 42 Schwyn, R. C. and Fox, C. A., The primate substantia nigra: a Golgi and electronmicroscopic study, J. Hirnjorsch., 15 (1974) 95-126. 43 Sherrington, C. S., Further note on degenerations following lesions of the cerebral cortex, J. PhysioL (Lond.), 11 (1890)399-400. 44 Swanson, L. W. and Cowan, W. M., A note on the connections and development of the nucleus accumbens, Brain Research, 92 (1975) 324-330. 45 Szabo, J., Topical distribution of the striata! efferents in the monkey, Exp. Neurol., 5 (I 962) 21-36. 46 Szabo, J., The course and distribution of ~fferents from the tail of the caudate nucleus in the monkey, Exp. NeuroL, 37 (1972) 562-572. 47 Ungerstedt, U., Stereotaxic mapping of the monoamine pathways in the rat brain, Acta physiol. ~cand.o Suppl. 367 (197!)" 1. 48 Usunoff, K. G., Hassler, R., Wagner, A. ar~.dBak, I. J., The efferent connections of the head of the caudate nucleus in the cat: an experimental morphological study with special reference to a projection to the raphe nuclei, Brain Research, 74 (1974) 143-148. 49 Voneida, T. J., An experimental study of the course and destination of fibers arising in the head of the caudate nucleus in the cat and monkey, J. comp. Neurol., 115 (I 960) 75-87. 50 Yehezkel, B., Zigmond, ~.. E. and Moore, K. E., Regional distribution of tyrosine hydroxylase, norepinephrine and dopamine within the amygdaloid complex of the rat, Brain Research, 86 (1975) 96-101. 51 Zeemz~n, W. and Innes, J. R. M., Craigie's Neuroanatomy of the Rat, Academic Press, New York, 1963.
423
THE PRECISE LOCALIZATION OF NIGRAL AFFERENTS IN THE RAT AS DETERMINED BY A RETROGRADE TRACING TECHNIQUE
B. S. BUNNEY and G. K. AGHAJANIAN
Departments of Psychiatry and Pharmacology, Yale University School of Medicine, New Haven, Conn. 06508 (U.S.A.) (Accepted April 13th, 1976)
SUMMARY
Afferent innervation of the rat substantia nigra (SN) was studied by the retrograde horseradish peroxidase (HRP) method. High concentrations of HRP were deposited in discrete subregions of the SN by means of a microiontophoretic delivery system. Using this technique it was possible to demonstrate that the caudatonigral projection system is arranged topographically. All portions of the caudate-putamen except for a central medial core were found to contain HRP positive cells, indicative of retrograde transport. In the positive areas a much larger percentage ofcells (30-50 %) were found to particin~.te in this projection than has previously been reported. Only medium size cells (12-20/~m) were found to contain the HRP reaction product. Other areas found to heavily innervate the SN were the globus pallidus, central nucleus of the amygdala and dorsal raphe nucleus. Areas containing fewer reactive cells but which also appear to innervate the SN included the prefrontal cortex and lateral habenula. These results emphasize the, importance of striatonigral projections which recent studies have suggested contain a G ABAergic link.
INTRODUCTION
The substantia nigra (SN), which has been studied intensively in recent years because of its demonstrated involvement in the pathogenesis of Parkinson's disease 1~, 24, consists of two divisions" the zona compacta (ZC) and the zona reticulata (ZR). The ZC and the ZR have been differentiated histochemically, pharmacologically and electrophysiologicailylt,17,4~. The ZC has been shown to consist almost exclusively of dopamine-containing cells that project to the caudate-putamen nucleus u,47. The ZR appears to possess both interneurons and neurons which project to the striatum and thalamus ~,15,33. The aff~.rent connections to the SN are less well known. The first evidence for a projection from the basal ganglia to the SN was reported in 191113.
424 Since then numerous degeneration studies using both light and electron microscopic techniques have der,lonstrated striatonigral and pallidonigral pathways 21,4a,40. in addition, light and electron microscopic autoradiographic studies have also demonstrated both striatal and pallidal projections to the SN 22. These latter studies strongly suggest that the caudatonigrai pathway predominantly innervates the ZR whereas the pallidonigrai projection is mainly to the ZC. None of these studies, however, specifically identifies the location of the cells of origin for these pathways within the caudate nucleus (CN) and globus pailidus (GP). Recently, retrograde axonal transport of horseradish peroxidase (HRP) has been added as a new technique for studying the afferent connections of the SN z°. Using this technique the existence of both a striatal and a pallidal input to the SN has been confirmed in the cat e0. The experiments reported here, which use the HRf" technique to study rat SN afferent connections, reveal the ~pecific location of cells projecting to the SN and demonstrate a topographical arrangement for the striatal innervation of the SN. METHODS Male Charles River albino rats weighing 240-300 g were used (N == 26). Two series of experiments were performed. In both, the rats were deeply anesthetized with chloral hydrate during the deposition of HRP. In the first series (N = 10) HRP (Sigma VI) was injected by means of a 5/~! Hamilton syringe with a fixed 3!-gauge needle. -The needle was inserted stereotaxically into the SN using two approaches: a vertical approach (coordinates A 1950 ~tm, L 1800 ttm; according to K/Snig and Klippe126) and an oblique approach (same anterior coordinates but from the side at an angle of 49°). In both cases care was taken not to damage the crus cerebri. HRP was dissolved in distilled water to a concentration of 25 % and 0.05/~1 was illjected over a 0.5 h period. In another series of experiments (N -- 16) a more precise localization within the SN was obtained by the use of a microiontophoretic delivery system for depositing the HRP. Again both a vertical and an oblique approach were used. Using a modification of the technique described by Graybiel and Devor TM, it consisted of the following seeps. Glass tubing (2 mm) filled with fiber glass was pulled into micropipettes, the tips of which were then broken back to a diameter of approximately 40/~m. The pipettes were filled with 6/~1 of a 25 % solution of HRP in 0.01 M NaCI by direct injection into the hub of the mic~'opipette with a Hamilton syringe. After a pipette was stereotaxically lowered into the SN, a 4/~A positive current was applied to the H R P solutien for 5 min by means ofa DC constant current source thereby ejecting HRP molecules into the SN. Control ejections (N = 6) were made into the dorsal raphe nucleus, the reticular formation dorsal to the SN, and the crus cerebri (causing deliberate damage to fibers with the injection cannulae)just posterior to the SN. The HRP histochemical technique used is a modification of that described by Kristensson et al. 2s and LaVail and LaVail al. Twenty-four hours after the placement of HRP within the brain the animals were anesthetized and the brain fixed by intracardiac perfusion for 0,5 h with a solution consisting of 1% paraformaldehyde and
425
.
-."
~'...:
.
.
•
~ . , _ _
•
.
o
,i.
' ~
. ~
-
. . . . .
"
' -
%,:-
•
-"
.-
It. : ~"
\
,.
" ,
"
. . . .
-;,
..
-
~'
, o
~
~
." : ....
•
,,
"
".
"
x;-
.
"
' ',
;.,
"%"
"i
,-:
.
'~
.-
~'
.
,
'i , - ,
.:,.. '. •
,t
.
¢ "~,,"~i'~i~..
-
.
•
"
-
b
•a
•
- . " . . . . t.
'4
"
,
t-'.:.-';~:r "" .. '
.'• -.~
"
e,t
:.~"ll..
"
~ ,.....,. ,:
I," t
,, -
,-~,'.
"~ " '" : . . .
..'
.
,, •
•
• • S
-
-%
. "
;.. hl;~/!"--: ":/,lllt:'~'.
.".
i
.
I'.
".
...
,: : ,
/
."
/.." "" i i
,f/
'-,:,..,,:-.-"..liinll;~,-. "
/
.
e , ~ t ~ . ' ! l . l ~ l ~ ' '~
.,
y',,"
i.I
Fig. I. Bright field photomicrograph illustrating the distribution of HRP reaction product after microiontophoretic ejection into the right substantia nigra. Section is unstained.
1% glutaraldehyde in 0.05 M phosphate buffer (pH 7.3). After storage overnight in 0.05 M Tris buffer with 5 ~o sucrose, frozen serial transverse sections (50 fire) of the brain were cut and collected in 0.05 M Tris buffer (pH 7.6). The sections were immediately transferred to trays containing 3 ml of a 1.4 mM solution of 3,3-diaminobenzidine (DAB) (free base, Sigma) in Tris buffer which had been filtered through Whatman glass fiber paper GF/B under vacuum. The sections were incubated for 30 rain at 35 °C. After cooling to room temperature 0.4 ml of a 0.03 o~ solution of H202 was added to the DAB solution which was then gently agitated for I h. Sections were then rinsed twice in Tris buffer (0.05 M, pH 7.6) and once in distilled water. Every fourth section was mounted from gelatin alcohol (0.5% gelatin, 80~'0 alcohol), dehydrated, coverslipped and examined in a light microscope using both bright and dark fie!d illumination. RESULTS
In the experiments using a Hamilton syringe, both the vertical and oblique approaches resulted in the appearance of diffuse HRP reaction product within portions of the ZC and the ZR. With the oblique approach, since the needle tract followed the natural contour of the SN, the reaction product was limited to the confines of the nucleus itself. In this case no damage occurred to fibers in the crus cerebri. Nevertheless, in these experiments fibers in the crus cerebri contained dense accumulations of HRP reaction product and could be traced to the internal capsule and into the
426
Fig. 2. Dark field photomicrogl~tphs of the caudate nucleus (anterior 78 %, according to Kt~nig and Kiippel"% The lateral portion of the caudate nucleus (A) is seen to contain numerous H R P reactive cells whereas the medial portion (B) is totally devoid of labeled cells. A ~breviations: CC, corpus callosum; LV, lateral ventricle. Scale bar: 20 ,ml.
427 TABLE
I
Topography o f caudatonigral projections Abbreviations: ZC, zona compacta ; ZR, zona reticulata; V, v e n t r a l ; VL, ventral lateral; L, lateral; D L , dorsal lateral; D, dorsal; T, tail. Abbreviations correspond to areas m a r k e d V, L, D a n d T in Fig. 2. .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Location o f HRP ejection sites hi substantia nigra*
Location of reactive cells" at 5 representative frontal planes withht the caudate Ilttc[etls
Frontal plane
Anterior 9410" *
Laterality
Zone
Anterior 7890
Anterior 6790 .
.
.
Anterh~r 5660 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Anterior 41 !0 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Anterior A2420 A2180 A2180 A2180
Intermediate Lateral Intermediate Medial
ZC and Z R ZC and ZR ZR ZC
V -D, L, V V
VL -DL VL
L L L DL
VL L D D
T T T T
( , ***) ( - ~ ) (--) (~
Lateral Intermediate Intermediate Medial
ZR ZC ZR ZC
V, VL -D, DL, L ~
V,VL V D, DL. L L
V, VL V D, DL DL
VL,L V D D
T ( : ! ) T (~) T ( . ) T( : )
Lateral Intermediate Intermediate
ZC Z C and Z R ZR
----
V VL, L, L
V, VL VL L
V V VL
T ~i ~ ) -- (--) T (.)
Middle AI950 A!950 AI950 Ai950
Posterior AI760 AI760 AI610
* Represent area of densest H R P deposition. ** Frontal planes according to K6nig and Klippel"'k *** -!, 5 - 1 0 H R P reactive cells per section; ! - , , 20 H R P reactive cells per section.
tail of the caudate nucleus to cells with granular (particulate) HRP reaction product. Under bright field microscopy these fibers did not show the uniform brown color which, according to other investigators 37, is characteristic for an axonal reaction secondary to injury. Rather they had a granular appearance with multiple varicosities visible under dark field illumination. In the experiments using microiontophoresis HRP was deposited in small (500 #m diameter) well localized, but highly concentrated amounts within various subregions of the SN (Fig. I). The use of both a vertical and oblique approach al!owed us to control for damage to fibers of passage everywhere except within the SN itself. However, the small tip diameter of our micropipettes (40-50/~m) coupled with the fact that H RP is ejected iontophoretically rather than hydrolically ensured a minimum of tissue damage at each injection site and eliminated the problem of passage of HRP up the tract made by the micropipette. Using this technique it was possible to limit the deposition of HRP to the lateral, intermediate, or medial one-third of the SN as well as to the ZR or to the ZC. The latter was made pn~,ible by the natural dorsal barrier to dorsal spread made by the fibers of the medial lemniscus. The Hamilton syringe injection studies were u ~ d as a general guide to areas of the brain which might innervate the SN. Only results which were common to both sets of
42g
\"
A 7890
DD
•
•
io.;.ibt, . //-st -Z !v v v/:
c'~A,/;.",'}..ttt.]/ •
A 6790
A5660
•
I
]
$R
__.,: .-",....i-:-?
A4110
)
AI950
A550
Fig. 3. Schematic tlra~ings of representative frontal sections o1" rat brain demonstrating location of cells containing H RP reaction l')oduct (black do!,, and bold face lettersl following an ejection of HRP into the substantia nigra. The letters (I), L. V, Tj ~ithin the caudate nucleus (Cp) designate lopographical projection :lrcas to the SN described in Table I. Drawings modified from K~.Snig and Klippel'-'~L Abbreviations: ac, nucleus amygdaloideus centralis: CA, conanaissura anterior; CAI, capsula intcrna: CL, claustrum ; C'p, nucleus caudatus putamen ; FM P, fascictdus medialis prosencephall; GP, globus pallidus: HI, hippo~'ampus: ip, nucleus interpeduncularis; LM, iemniscus n)edialis; mr, nucleus medianus raphes: PCS. pedunculus cerebe!laris superior; SNR, substantia nigra, zona reticulata; SR, sulcus rhinalis, I), di_~rsal: L. lateral; V, ventral; T, tail. experiments are reported here. Every grour) of cells labeled in these experiments occurred on the side of the brain ipsilaterai to the injection site.
.4rea.~ with a high de,,siti" o.1"H R P retr'tive cells ( ! ) Bu.salganglia. The head of the CN contained numerous reactive cells which ranged fi'om 12 to 20 l~m in size (Fig. 2). No "giant" cells were found with a positive H RP reactione:'. Examination of section~ of the C N which were stained with cresyi violet revealed that between 30 and 50",, of neurons in a given H RP labeled area of the C N project to the SN. The combined result,, o f our SN H RP studies showed that all parts o f the CN contain cells which project to tiae SN except for a medial cure (,Figs. 2, 3). Only a few scattered cells were found to be labeled in the latter area
429
Fig. 4. l)ark field photomicrograph of the cerebral cortex tA}, just dorsal to tl~e rhinal sulcus, and globus p:~llidus (B) sho~ing cells containing HRP reaction product follo~ing an ejcction of HRP into the substantia nigra. Scale bar: 20 ,,,m.
430 regardless of the SN ejection site. The distribution of labeled cells varied according to thesite of HRP deposition within the SN and demonstrated a topographical projection from the CN to the SN in both a medial-lateral and an anterior-posterior arrangement (Table I). Thus HRP ejected into the medial or lateral one-third of thc SN resulted in reactive cells in the dorsal and lateral areas of the CN respectively. Anteriorally placed ejections in the SN resulted in labeled cells in anterior but not posterior areas of the CN. However, an ejection into the ZC often resulted in different caudate cells being labeled from those labeled when HRP was ejected into the ZR (even though both ejections were in the same frontal and sagittal plane), thereby greatly adding to the complexity of the topographical arrangement (see Table I). Reactive cells were also seen in the tail of the CN (Fig. 3). The greatest number of cells in this area were seen labeled after lateral ejections, but ejections into most other parts of the SN also resulted in labeled cells in the tail of the CN (Table I). In the GP (Figs. 3 and 4) reactive cells were most numerous in the posterior-lateral portion and were progressively less frequent toward the anterior pole. Ejections into the lateral one-third of the SN, no matter what the anterior-posterior location, resulted in the labeling of raany cells within the central nucleus of the amygdaloid complex (Fig. 3). No other nuclei within that complex were labeled. (2) Accumbens nucleus. The deposition of HRP into the medial one-third of the SN and adjacent ventral tegmental area resulted in labeled cells within the accumbens nucleus. (3) Dorsal raphe nucleus. Every HRP ejection into the SN resulted ill reactive cells in the ipsilateral half of the dorsal raphe nucleus (Fig. 3).
Areas with a low density of HRP reactive cells (1) Cortex. Every HRP ejection into the SN resulted in occasional reactive cells scattered throughout the parietal, insular and temporal cortex (according to Zeeman and Innes 5°) (Figs. 3 and 4). No labeled cells were ever seen ventral te the rhinal sulcus. The prefrontal cortex consistently contained a few heavily labeled cells. Regardless of the cortical area labeled, only layer 5 appeared to contain reactive cells. (2) Habenula. Microiontophoretic ejection of HRP into the lateral one-third of the SN resulted in an occasional reactive cell in the lateral habenula. (3) Hypothalamus. Every ejection into the ZC of the SN yielded a scattering of la~beled cells in the hypothalamus although the location within the hypothalamus varied from animal to animal. No preoptic nuclei contained reactive cells. HRP reactive cells after control injections When ejections were made into the crus cerebri posterior to the SN, reaction product was found in fibers and cells immediately surrounding the ejection sites, but never resulted in labeling of either the internal capsule or the cell groups described above. Contro| ejections into the reticular formation dorsal to the SN resulted in the labeling of cells in the parietal and prefrontal cortex, hypothalamus (ventral part of the dorsal medial nucleus), locus coeruleus and contralateral dentate and
431 emboliformis nuclei. Deposition of HRP into the dorsal raphe nucleus failed to label any cells in the CN, GP, or insular, temporal, or parietal cortex. DISCUSSION Nigral afferents from the basal ganglia have been demonstrated by several investigators ~0,22,45,49. However, due to the limitation of the techniques used in these studies (degeneration and autoradiography) it has not been possible to identify exactly which cells project to the SN. Previously it has been suggested that only 5 O//oofcaudate cells leave the CN to innervate distal structureslS,~L Using the retrograde H RP technique it has been possible to identify precisely which cells form the caudatonigral projection. Our study suggests that 30-50 O//oofcaadate cells, except for a medial core, project to the SN. Since most of the CN is involved in this projection, CN cells projecting outside the CN are clearly much more numerous than previously expected. Other investigators, using degeneration techniques in monkeys36,4~, 49 and cats as, 4s,49, have found a topographical distribution of caudatonigral fibers although they have differed as to the specific arrangement. Our study in the rat demonstrated a mixed anterior-posterior medial-lateral topographical distribution of the striatal efferents to the SN as well as v difference between ZC and ZR caudatonigral afferents. Because of this complex topographical arrangement an enormous number of highly localized ejections would be necessary to obtain a complete topographical map. The 14 microiontophoretic ejections performed in this study, therefore, do not allow us to designate fully the striatonigral topographical projections. Although we found that tbe tail of the CN projected to all parts of the SN, the lateral one-third received the heaviest innervation from this area. These findings, in the rat, are in marked contrast to those tbund in the monkey where the tail of the caudate has been reported to project to the caudal SN posterior to the emergence of NIII fibers from the midbrain 4'~. The finding of HRP reaction product in cells of both the CN and the GP adds further support to the concept of two pathways from the basal ganglia to the SN2°,~L Autoradiographic studies in the rat by Hattori et al. 2~ suggest that the GP cells mainly project to the ZC. We found more HRP reactive cells in the GP after discrete ejections into the ZR than after well localized ZC ejections. A recent behavioral study has suggested a feedback pathway to the SN involving cell: bodies located in the claustrum TM. However, no HRP reactive cells were found in this structure. Of equal importance to demonstrating the anatomical connections to the SN is determining the neurotransmitter used by each afferent system. The two highest ranking candidates for mediators of chemical transmission in the caudatonigral and/or pallidonigral pathways are gamma-aminobutyric acid (GABA) and acetylcholine (ACh). Substantial evidence now exists in favor of GABA and against ACh 17'aa. Thus lesion of the globus pallidus or transection of the brain between the basal ganglia and SN results in decreased levels of GABA and glutamic acid decarboxylase (GABA synthesizing enzyme) activity in the SN. The same lesions, however have no effect on the concentrations of ACh in the SN ~7,a3. Labeled GABA injected into the GP has been shown both chemically and by electron microscopic radio-
432 autography to be transported to the SN whereas, when a similar injection was made into the CN, no radioactivity could be detected in the SN 17. These latter experiments, therefore, suggest that the pallidonigral pathway is GABAergic, whereas the caudatonigrai pathway is not. Electrophysiological studies also have provided evidence for a GABAergic pathway from the basal ganglia to the SN. Stimulation of the head of the caudate nucleus has been shown to cause depression of neuronal activity in the SN. This depression was prevented by intravenous administration of picrotoxin (a presumed GABA antagonist) 4°. In addition, microiontophoretically applied GABA has been found to depress SN neurons and this depressant effect can in turn be blocked by microiontophoreticaily applied picrotoxin 1°. Paralleling studies of reciprocal connections between the SN and the CN, biochemical and electrophysiological studies on the mechanism of action of certain drugs also have provided indirect evidence for a feedback pathway from the accumbens nucleus to dopaminergic neurons in the midbrain ventral tegmental area (A I0). A recent autoradiographic study carried out in 2-weekold rats has reported a p:ojection from the accumbens nucleus to the mediaimost pole of the SN and to the A l0 area 44. in our studies in adult rats, microiontophoretic placement of HRP into the extreme medial edge of the SN adjacent to the A 10 area resulted in reactive cells within the accumbens nucleus, providing further anatomical evidence for the existence of such a pathway. A dopaminergic input to the central nucleus of the amygdala (AC) has been known fo~ some time 47. Recent biochemical studies have shown that the AC contains *,he highest concentration of dopamine of any nucleus within the amygdaloid complex .',~. Injection of HRP into the midbrain tegmental area has been reported to result in heavy labeling of cells in the AC 2a. However, as degeneration studies confined to the AC have resulted in marked fiber degeneration in the median forebrain bundle (M FB) 1~ it is possible that the labeling of cells in the AC resulting from the injection of HRP into the tegmentum may have been due to damage of the MFB which passes through this area. On the other hand, in the studies reported here, tissue damage was minimal and discrete injections into the lateral one-third of the SN, all of which were far from the MFB, consistently resulted in heavy labeling of cells in the AC. Such results suggest the possibility of a neuronal feedback system from the amygdala to the SN, similar to that known to exist for the nigrostriatal and mesolimbic dopaminergic systems. As several investigators using a variety of techniques have demonstrated a serotonergie (5-HT) projection to the SN 7,11,29'39, the finding of cells containing HRP reaction product in the dorsal raphe nucleus was predictable. Our study does not permit us to determine w,t,,ich part ofthe SN, ZC or ZR receives the 5-HT innervation. However, when 5-HT is applied directly to cells in the SN by means of microiontophoresis it has little effect on ZR neurons but totally blocks glutamate excitation of ZC cells 1. Such evidence suggests that the ZC dopaminergic neurons may be the ones primarily influenced by the serotonergic input. The findings of HRP reactive cells in parts of the cortex is consistent with the long standing suggestion of the existence of a corticonigral pathway 6,aS,4a. However, in other studies few if any degenerating terminals were observed in the SN after large
433 lesions of the ipsilateral cortex27, 41. In addition, a recent autoradiographic study of projections of the precentral motor cortex fails to mention any labeling in the SN '~0. The prefrontal cortex was unique in that it consistently demonstrated the highest density of labeled cells and these were also the most heavily labeled. These results may be explained by the finding that a fiber contingent from the prefrontal cortex passes laterally over the dorsal surface of the SN 3z. However, it has also been suggested that because these fibers become fewer in number as they pass caudad, some of them are terminating in the SN 32. Although not answered by our study the question of the existence of a corticonigral pathway is of some importance (especially if coming from the parietal cortex) as it would provide a second feedback pathway system for sensory motor areas to the basal ganglia and/or thalamus (via the SN). Further autoradiographic studies such as those recently reported by Kiinzle 30 may provide a definitive answer to this question. Two other areas of the brain were found to contain reactive cells - - the habenula and the hypothalamus. Never more than one or two cells were seen labeled on any one section. The existence of projections from these structures to the midbrain has previously been reported s,34. However, innervation of the SN by these areas must await confirmation by the use of other techniques. In conclusion, the importance of the existence of a pathway from the basal ganglia to the SN has been underscored in recent years by increasing evidence that certain drug effects in the CNS may be induced by an action on such a neuronal system. For example, a feedback pathway has been suggested to mediate the effects of amphetamine 9 and the antipsychotic drugs 4 on dopamine turnover in the CN by inducing changes in dopaminergic cell firing rate. We report here the localization of the cells of origin for two such pathways in the rat. The rat pallidonigral pathway was found to be similar to that recently described for the cat 2°. However, the cells making up the rat caudatonigral pathway appear to have a somewhat different distribution. Thus, in the rat, all parts of the CN appear to project topographically to the SN except for a medial core. in addition there is a relativley large projection from the tail of the CN to the SN. These findings suggest that physiological and pharmacological studies designed to characterize further these pathways are feasible. Selective stimulation and lesioning of these areas should enable one to determine the role each of these pathways plays in modulating the spontaneous activity of cells in the ZC and ZR and in mediating the effect of drugs on the activity of dopaminergic neurons. For example, recent electrophysiological studies in our laboratory 2 have shown that the depressant effects of D-amphetamine on dopaminergic cell firing rate 3, which have been presumed to be mediated via a neuronal feedback mechanism, can be blocked by lesioning the tail of the caudate nucleus. ACKNOWLEDGEMENTS This research was supported by NIMH Grants (MH-17871; MH-14459; MH-25642), The Benevolent Foundation of Scottish Rite Freemasonry, Northern Jurisdiction, U.S.A., and the State of Connecticut.
434 REFERENCES 1 Aghajanian, G. K. and Bunney, B. S., Dopaminergic and non-dopaminergic neurons of the substantia hi,ca: differential responses to putative transmitters. :in J. R. Boissier, H. Hippius and P. Pichot (Eds.), Proc. 9th Int. Congr. of the Collegium Interna, ionale Neuropsychopharmacologicure, Excerpta Medica Foundation, Amsterdam, 1974. 2 Bunney, B. S. and Aghajanian, G. K., D-Amphetamine-induced inhibition of central dopaminergic neurons: mediation by a striatonigral feedback pathway, Sctence, 192 (1976) 391-393. 3 Bunney, B. S., Waiters, J. R., Roth, R. H. and Aghajanian, ¢2. X., Dopaminergic neurons: effect of antipsychotic drugs and amphetamine on single cell ac, i','it?', J. Pharmacol. exp. Ther., 185 (1973) 560-571. 4 Carlsson, A. and Lindqvist, M., Effect of chlorpromazine a~.~d haloperidol on formation of 3methoxytyramine and normetanephrine in mouse brain, Acta pharmacol. (Kbh.), 20 (1963) 140-144. 5 Cole, M., Nauta, W. J. H. and Mehler, W. R., The ascending efferent projections of the substantia nigra, Trans. Amer. neuroL Ass., 89 (1964) 74-78. 6 Combs, C. M., The distribution and temporal course of fiber degeneration after experimental lesions in the rat brain, J. comp. NeuroL, 94 (1951) 123-176. 7 Conrad, L. C. A., Leonard, C. M. and Pfaff, D. W., Connections of the median and dorsal raphe nuclei in the rat: an autoradiographic and degeneration study, Y. comp. Neurol., 156 (1974) 179-206. 8 Conrad, L. C. A. and Pfaff, D. W., Axonal projections of medial preoptic and anterior hypothalamic neurons, Science, 190 (1975) 1112-1114. 9 Corrodi, H., Fuxe, K. and H6kfelt, T., The effect of some psychoactive drugs on central monoamine neurons, Europ. J. Pharmacol., 1 (1967) 363-368. 10 Cro~sman, A. R., Walker, R. J. and Woodruff, G. N., Picrotoxin antagonism of ~,-aminobutyric acid inhibitory responses and synaptic inhibition in the rat substar.~ia nigra, Brit. J. Pharmacol., 49 ~1973) 696-698. 11 Dahlstr/~m, A. and Fuxe, K., Evidence for the existence of monoamine containing neurons in the central nervous system, Acta physiol, scand., 62, Suppl. 232 0964) I. 12 De Olmos,J. S., The amygdaloid projection field in the rat as studied with the cupric-silver method. In B. E. Eleflheriou (Ed.), The Neurobiology of the Amygdala, Plenum Press, New York, 1972, pp. 145-204. 13 Edinger, L., Vorlesungen fiber den Bau der nervosen Zentralorgane, VoL 1, Vogel, Leipzig, 8th ed., 1911. 14 Ehringer, H. and Hornykiewicz, O., Verteilung yon Noradrenalin und Dopamin (3-Hydroxytyramine), im Gehirn des Menschen und ihr Verhalten bei Erkrankungen des extrapyramidalen Systems, Klin. Wschr., 38 (1960) 1236-1239. 15 Fauli, R. L. M. and Carman, J. B., Ascending projections of the substantia nigra in the rat, J. comp. Neurol., 132 (1968) 73~92. 16 Fleisher, L; N. and Giick, S. D., A telencephalic lesion site for D-amphetamine-induced contralateral rotation in rats, Brain Research, 96 (1975) 413-417. 17 Fonnum, F., Grofov~i, I., Rinvik, E., Storm-Mathisen, J. and Walberg, F., Origin and distribution of glutamate decarboxylase in substantia nigra of the cat, Brain Research, 71 (1974) 77-92. 18 Fox, C. A., Andrade, A. N., Schwyn, R. C. and Rafols, J. A., The aspiny neurons and P,a glia in the primate striatum: a Golgi and electron microscopic study, J. Hirnforsch., 13 (1971) 341-362. 19 Graybiel, A. M. and Devor, M., A microelectrophoretic delivery technique for use with horseradish peroxidase, Brain Research, 68 (1974) 167-173. 20 Grofov~i, I., The identification of striatal and pailidal neurons projecting to substantia nigra. An experimental study by means of retrograde axonal transport of horseradish peroxidase, Brain Research, 91 (1975) 286-291. 21 Grofovfi, I. and Rinvik, E., An experimental electron microscopic study on the striatonigral projection in the cat, Exp. Brain Res., 11 (1970) 249-262. 22 Hattori, T., Fibiger, H. C. and McGeer, P. L., Demonstration ofa pallidonigral projection innervating dopaminergic neurons, J. comp. Neurol., 162 (1975) 487-504. 23 Hopkins, D. A., Amygdalotegmental projections in the rat, cat and rhesus monkey, Neurosci. Lett., 1 (1975) 263-270. 24 Hornykiewicz, O., Dopamine (3-hydroxytyramine) and brain function, Pharmacol. Rev., 18 (! 966) 925-964.
435 25 Kemp, .l.M. and Powell, T. P. S., The structure of the caudate nucleus of the cat: light and electron microscopy, Phil. Trans. B, 262 (1971) 383-401. 26 K0nig, J. F. R. and Klippel, R. A., The Rat Brain: a Stereotaxic Atlas, Krieger, New York, 1970. 27 Knook, H. L., The Fibre Connecti~ns of t/'.e Forebrain, Royal Van Gorcum, Assen, 1965, p. 81. 28 Kristensson, K., Olsson, Y. and Sj0strand, J., Axonal uptake and retrograde transport of exogenous proteins in the hypoglossal nerve, Brain Research, 32 (1971) 399-406. 29 Kuhar, M. J., Aghajanian, G. K. and Roth, R.H., Tryptophan hydroxylase activity and synaptosomal uptake of serotonin in discrete brain regions after midbrain raphe lesions: correlations with serotonin levels and histochemical fluorescence, Brain Research, 44 (1972) 165-176. 30 Kiinzle, H., Bilateral projections from precentral motor cortex to the putamen and other parts of the basal ganglia. An autoradiographic study in Macaca foscicularis, Brain Research, 88 (1975) 195-209. 31 LaVail, .1. H. and LaVail, M. M., Retrograde axonal transport in the central nervous system, Science, 196 (1972) 1416-1417. 32 l.eonard, C. M., The prefrontal cortex of the rat. I. Cortical projection of the mediodorsal nucleus. II. Efferent connections, Brain Research, 12 (1969) 321-343. 33 McGeer, P. L., Fibiger, H. C., Hattori, T., Singh, V. K., McGeer, E. G. and Maler, L., Biochemical ne~ roanatomy of the basal ganglia. In R. D. Myers and R. R. Drucker-Colin (Eds.), Advances in Behavioral Biology, Vol. !0, Plenum Press, New York, 1974, pp. 27-47. 34 McGeer, P. L., Hattori, T. and McGeer, E. G., Axonal transport in the habenular interpeduncular cholinergic tract, biochemical and autoradiographic studies, Neurosci. Abstr., 1 (1975) 851. 35 Minkowski, M., Etude sur les connections anatomiques des circonvolutions rolandiques, pari6tales et frontales, Schweiz. Arch. Neurol. Psychiat., 12 (1923) 71-104; 227-268. 36 Nauta, W. J. H. and Mehler, W. R., Projections of the lentiform nucleus in the monkey, Brahz Research, 1 (1966) 3-42. 37 Nauta, H. J. W., Pritz, M. B. and Lasek, R. J., Afferents to the rat studied with horseradish peroxidase. An evaluation of a retrograde neuroanatomical research method, Brain Research, 67 (1974) 219-238. 38 Niimi, K., Ikeda, T., Kawamura, S. and Inoshita, H., Efferent projections of the head of the caudat~ nucleus in the cat, Brain Research, 21 (1970) 327-343. 39 Parizeg, J., Hassler, R. and Bak, I. J., Light and electron microscopic autoradiography of substantia nigra of rat after intraventricular administration of tritium labeled norepinephrine, dopamine, serotonin and the precursors, Z. Zellforsch., 115 (1971) 137-148. 40 Precht, W. and Yoshida, M., Blockage of caudate-evoked inhibition of neurons in the substantia nigra by picrotoxin, Brain Research, 32 (197 I) 229-233. 41 Rinvik, E. and Walberg, E , Is mere a coi'tlco-nigra| tract.' A comment based on experimental electron microscopic observations in the cat, Brain Research, 14 (1974) 742-744. 42 Schwyn, R. C. and Fox, C. A., The primate substantia nigra: a Golgi and electronmicroscopic study, J. Hirnjorsch., 15 (1974) 95-126. 43 Sherrington, C. S., Further note on degenerations following lesions of the cerebral cortex, J. PhysioL (Lond.), 11 (1890)399-400. 44 Swanson, L. W. and Cowan, W. M., A note on the connections and development of the nucleus accumbens, Brain Research, 92 (1975) 324-330. 45 Szabo, J., Topical distribution of the striata! efferents in the monkey, Exp. Neurol., 5 (I 962) 21-36. 46 Szabo, J., The course and distribution of ~fferents from the tail of the caudate nucleus in the monkey, Exp. NeuroL, 37 (1972) 562-572. 47 Ungerstedt, U., Stereotaxic mapping of the monoamine pathways in the rat brain, Acta physiol. ~cand.o Suppl. 367 (197!)" 1. 48 Usunoff, K. G., Hassler, R., Wagner, A. ar~.dBak, I. J., The efferent connections of the head of the caudate nucleus in the cat: an experimental morphological study with special reference to a projection to the raphe nuclei, Brain Research, 74 (1974) 143-148. 49 Voneida, T. J., An experimental study of the course and destination of fibers arising in the head of the caudate nucleus in the cat and monkey, J. comp. Neurol., 115 (I 960) 75-87. 50 Yehezkel, B., Zigmond, ~.. E. and Moore, K. E., Regional distribution of tyrosine hydroxylase, norepinephrine and dopamine within the amygdaloid complex of the rat, Brain Research, 86 (1975) 96-101. 51 Zeemz~n, W. and Innes, J. R. M., Craigie's Neuroanatomy of the Rat, Academic Press, New York, 1963.
