- Email: [email protected]
The inhibitory input from the substantia nigra to the mediodorsal nucleus neurons projecting to the prefrontal cortex in the cat
BRAIN RESEARCH Brain Research 649 (1994) 313-318
ELSEVIER
Short Communication
The inhibitory input from the substantia nigra to the mediodorsal nucleus neurons projecting to the prefrontal cortex in the cat Yoshihisa Miyamoto a, Kohnosuke Jinnai b,, a Department of Neurosurgery and b Department of Physiology, Shiga University of Medical Science, Ohtsu, Shiga 520-21, Japan
(Accepted l0 March 1994)
Abstract
The effects of stimulation of the substantia nigra pars reticulata (SNr) on the neurons of the mediodorsal nucleus (MD) projecting to the prefrontal cortex (PF) were studied in cats. The MD neurons projecting to the ventral part of the PF tended to be located in the ventral part of the MD, while those projecting to the dorsal part of the PF in the dorsal part. Spontaneous discharges of 30/57 tested MD neurons were suppressed by SNr stimulation at a latency ranging from 2 to 15 ms. The latency of the suppression corresponded well to that of antidromic responses of SNr neurons elicited by MD stimulation (from 1.4 to 14.0 ms). Intracellular recordings in a few MD neurons showed IPSP by SNr stimulation. The SNr is considered to exert an inhibitory effect on the MD neurons projecting to the PF. Key words: Nigrothalamic projection; Substantia nigra; Mediodorsal nucleus; Prefrontal cortex; Electrophysiology; Cat
The nigro-thalamo-cortical pathway is one of the main output systems of the basal ganglia, but its functional role remains obscure partly because of paucity of anatomical and electrophysiological data concerning this pathway. Previous anatomical studies revealed that the substantia nigra pars reticulata (SNr) projects to the medial devision of the ventral lateral nucleus (VLm), magnocellular part of the ventral anterior nucleus (VAmc) and the mediodorsal nucleus (MD) in monkeys [7,14] and rats [4,12]. Velayos and ReinosoSuftrez [18], and Kemel et al. [8] found SNr-MD connection in cats by retrograde horseradish peroxidase ( H R P ) or autoradiographical studies, though Rinvic [13], Kultas-Ilinsky et al. [10] and Tokuno et al. [16] did not demonstrate the connection. Nigro-thalamic output is suggested to be sent to the prefrontal cortex (PF) because all of these thalamic nuclei (VLm, VAmc, and MD) were reported to project mainly to the PF [6,7,11,12,14,15]. Although the cortical projection neurons in the V A m c have been shown to receive inhibitory synapses from the SNr in monkeys [9], the direct-relay of SNr output to the PF through the other thalamic nuclei remains to be examined. The nigro-
* Corresponding author. Fax: (81) 775-48-2146. 0006-8993/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0006-8993(94)00295-N
thalamic projection is considered to be inhibitory because monosynaptic inhibitory post-synaptic potential (IPSP) by stimulation of the SNr was recorded in ventral medial nucleus (VM) neurons in cats [17], and because degenerated synapses in the V A m c after SNr lesioning were symmetrical ones with pleomorphic vesicles [9]. Influence of nigro-thalamic input on MD neurons has not yet been studied. In this study, the effect of SNr stimulation on M D neurons projecting to the PF were studied electrophysiologically. The aim of this study is to test: (1) whether SNr input is relayed to the PF by MD neurons, and (2) whether nigral input to the MD is inhibitory or excitatory. The experiments were performed on adult cats under general anesthesia, which was induced with an intramuscular injection of ketamine hydrochloride (40 m g / k g ) , and maintained with sodium pentobarbital (30 m g / k g ) . Cats were intubated, and maintained on spontaneous respiration. The skull was fixed in a stereotaxic frame and the frontoparietal cortex was exposed bilateraly by craniectomy. Four pairs of bipolar stimulating electrodes were inserted into the right PF, and 4 concentric stimulating electrodes were stereotaxically inserted into the right substantia nigra (SN). After the left parietal cortex was aspirated to expose the dorsal surface of the left thalamus, a glass microelectrode
314
}: ..tliyamom. K. .limmi /" Brain Rcwarch 04Y (I9~)4) 713 -.'~/~,'
1(I-30 MS/) were used. R e c o r d i n g sites of M D n e u r o n s were m a r k e d by passing negativc c u r r c n t s at 1 0 / z A for 10 min t h r o u g h the glass microelectrodc. To study SNr n e u r o n s projecting to the MD, three concentric stimulating electrodes wcrc inserted into the M D in the same m a n n e r as the i n t r o d u c t i o n ol recording microelectrode into the MD. T h e activities of SN n e u r o n s were recorded extracellularly with a glass-coated elgiroy microelectrode, and the latencies of a n t i d r o m i c response to M D stimulation were measured. R e c o r d i n g sites of SN n e u r o n s wcrc m a r k e d by passing positive D C c u r r e n t s (10 ~ A for 30 s) through the elgiroy electrode. At the e n d of each experiment, the brain was perfused with 10% formalin ( c o n t a i n i n g 2% potassium
filled with 2% P o n t a m i n e Sky Blue in {).5 M sodium acetate solution, was inserted across the midline into the right M D from the dorsal surface of the contralateral t h a l a m u s at an angle of 45 ° to the vertical axis. T h e cxtracel[ular unitary recordings of M D n e u r o n s which r e s p o n d e d antidromically to stimulation of the ipsilateral PF were carried out. T h e effects of stimulation of the SNr (less than 0.5 mA, d u r a t i o n 0.3 ms, r e p e t i t i o n rate 1.0-0.8 Hz) on M D n e u r o n s were observed with an aid of peri-stimulus time histogram (PSTH). Postsynaptic p o t e n t i a l s evoked by SNr stimulation were recorded intracellularly from M D n e u r o n s in several experiments. For this purpose, glass microelectrodes filled with 2% P o n t a m i n e Sky Blue dissolved in 0.5 M potassium acetate ( D C resistance;
b
PFstimulation i
c
I ,f
"'"~--k
-.
/
"X
a
"//~
.
"
SNr stimulation _scLj~-'..
~B
scr(
I
! i
,
;,
,ii
!
.
,
A o
mesial view
;
,:
0
J s'
\
~"
lateral view
II
25ms [[i
d /
A 10.2 !I
/' s . . . . . ~ .7)" 'X ...... ,. . . . t' ML) } •
A 9.1
/~,
- >E 5ms d
e ventral PF
A 7.9
A 8.3
/ (\ } CLN )
~"I
' /" PC/"
"~, ,~
-,,/
-
•
)
~.._.,~ pc
/
. _ t'_
.:'~/
!
-!
-
'\
i Rf ~ '
J
\
",
"
,
dorsal PF
/
/' C M
'
I
Fig. 1. a: PSTH showing suppression of discharges of a MD neuron after SNr stimulation. Bin width 1 ms, 128 sweeps addition, b: Antidromic response of the MD neuron in 'a' elicited by PF stimulation and its collision with a spontaneous discharge, c: the location of the tip of the stimulating electrodes in dorsal (square), ventral (circle), caudo-medial (triangle) part of the PF. The filled symbols indicate the stimulating sites which evoked antidromic response in the MD neurons inhibited by SNr stimulation. SCr, curuciate sulcus; SCot, coronal sulcus; SPs, presylvian sulcus, d: location of MD neurons projecting to PF. Square, circle, and triangle demonstrate the neurons projecting to the dorsal, ventral, and caudomedial parts of PF, respectively. Filled symbols indicate the MD neurons which were inhibited by SNr stimulation. The number above each coronal section (A 7.9-10.2) indicates the distance from the interaural plane. CLN, central lateral nucleus; CM, centre median nucleus; HL, lateral habenular nucleus; LP, lateral posterior complex; PC, paracentral nucleus; RF, reticular fiber, e: Latency histograms of antidromic response of MD neurons following stimulation of ventral part of PF and dorsal part of PF.
Y. Miyamoto, K. Jinnai / Brain Research 649 (1994) 313-318
ferrocyanide in the case of SN recording). All recording sites in the MD and the SNr were reconstructed on the basis of location of the marked spots. In the extracellular studies, 97 MD neurons were activated antidromically by stimulation of the PF at a threshold lower than 0.5 mA. All of these neurons showed the collision interaction with spontaneous discharges (Fig. lb). Fig. lc shows the location of tips of the stimulating electrodes in the PF, which elicited antidromic responces of MD neurons as shown in Fig. ld. The MD neurons antidromically activated by stimulation of the dorsal part of the PF, i.e., the rostral part of the frontal gyrus, dorsal part of the proreus gyrus and dorsal part of the medial bank of the presylvian sulcus, were located in the dorsal part of the MD (squares in Fig. lc). The MD neurons activated by
a
,:i 1
5
10
b /)
stimulation of the ventral part of the PF, i.e., rostral and ventral parts of the rectus gyrus, and ventral two-thirds of the medial bank of the presylvian gyrus were located in the ventral part of the MD (circles in Fig. lc). The MD neurons projecting to the ventral part of the PF showed shorter antidromic latencies (mean + S.D. = 4.2 + 2.2 ms) than those projecting to the dorsal part (mean + S.D. = 6.7 + 4.1 ms) as shown in Fig. le. The stimulation of caudal part of the medial PF may have activated the fibers projecting to the more rostral PF by current spread to the underlying white mater. Therefore, the MD neurons antidromically activated by those stimulating sites were classified into another category (triangle in Fig. lc). Among the 57 MD neurons projecting to the PF, the spontaneous discharges of 30 MD neurons were sup-
C
N=19 M ~ SD: 5.2 :t: 3.6 ms
N=30
15 ms
0
A 7.2
315
d
5
MDstimulation
10
15 ms
e
!
LP
I -~
"
~
-
-
~
~
O0 o
A45 /
'
i A 3.7
Fig. 2. a: latency histogram of inhibitory responses of M D neurons following SNr stimulation, b: distribution of the tips of stimulating electrodes in and around SN. The n u m b e r above the each frontal section (A 2.9-7.2) indicates the distance from the interaural plane. Closed circles indicate the tips of electrodes which elicited the inhibitory response of M D neurons. Open circles indicate the tips of stimulating electrodes which did not show the inhibitory effect. The bar over, across, or below the circles, indicates the classification of stimulating sites into three groups, i.e., dorso-marginal, central, and ventro-marginal parts of SNr. PP, pes pedunculi; RN, red nucleus; RR, retro-rubral nucleus; SNc, pars compacta of SN; VTA, ventral tegmental area of Tsai. c: Latency histogram of antidromic response of SNr neurons following M D stimulation d: Antidromic response of a SNr neuron by M D stimulation, and its collision with spontaneous discharge, e: location of stimulating electrodes in M D which induced the antidromic activation of SNr neurons in 'd'. f: location of SNr neurons showing the antidromic response to M D stimulation with threshold less than 0.45 mA. Most of them were located in the central or medial part of SNr.
Y. Mivamoto K. .linnai / Brain Research 049 (1994) 313-.~1S
316
pressed after SNr stimulation at intensities ranging from 0.15 to 0.5 m A (0.3 ms duration). Fig. la shows an example of inhibitory response to SNr stimulation. The latencies of inhibitory responses in MD neurons ranged from 2.0 ms to 15.0 ms, and most of them (19/30) were 2.0 to 6.0 ms (Fig. 2a). MD neurons inhibited by SNr stimulation were located in the dorsal and central portions of the MD, but not found in the ventral part, although the ventromedial parts of the MD were scarcely explored in this study (Fig. ld). The tips of the nigral stimulating electrodes eliciting the inhibition located mostly in the medial two-thirds of the SNr, and none in the lateral part of the SN, i.e., the SNL according to Berman [1], was effective. As for the dorsoventral direction, the electrodes in the central part of the SNr induced the inhibition more effectively than those in the dorsal and ventral margins of this nuclei as follows. Twenty-three MD neurons out of 48 tested were suppressed by stimulation of the central part of the SNr, whereas 10 and 6 out of the 43 and 33 tested MD neurons, respectively, were inhibited by stimulation of the dorsal and ventral border regions of the SNr. (Fig. 2b; One MD neuron was sometimes counted twice when it was tested by two stimulating electrodes.) Six MD neurons showed an increase in discharge rates after stimulation around the SNr at a latency longer than 6.0 ms. The effective stimulating
b
a
SNr"~\stimulal /ti'Y~ °nl
sites for the excitatory responses were not confined to the SNr. The location of SNr neurons projecting to the MD (Fig. 2f) and the latencies of antidromic responses to MD stimulation (Fig. 2c) were investigated to examine whether the SNr-induced inhibitory responses of MD neurons were mediated directly through the nigrothalamic pathway. Nineteen SNr neurons were activated antidromically by MD stimulation with thresholds less than 0.45 mA (Fig. 2d). Most of them were located in the medial two-thirds of the SNr (Fig. 2f). The latencies of antidromic activation ranged from 1.4 to 14.0 ms but were mostly less than 6.(1 ms (Fig. 2c) and this distribution of latency was coincident with that of the inhibitory response in MD neurons following SNr stimulation (Fig. 2a). These antidromic responses cannot be attributed to the activation of thc nigrothalamic fibers projecting to the other thalamic nuclei than the MD, e.g., the VM, because antidromic responses activated by electrodes in the rostro-ventral part of the MD did not show a lower threshold. Furthermore, the antidromic latency was much longer than expected from the latency of the monosynaptic IPSPs of VM neurons induced by SNr stimulation [6]. Five MD neurons were studied intracellulary with resting m e m b r a n e potentials lower than - 4 0 inV. Fig. 3 shows two examples of changes in membrane poten-
c //,,
SNrstimulation 1
d
,I i[
'I!i
J
A 4.8
\,%/~ !i!
....
-
./
/
__J
2
rnV
20 ms Fig. 3. a: hyperpolarization induced by SNr stimulation recorded intra-cellularly from a MD neuron, b: intracellular recordings from another MD neuron, showing hyperpolarization with a fixed latency. The hyperpolarizing response followed triple shocks (100 Hz) with the same latency, c: the recording sites of neurons 'a' and 'b'. d: the stimulating sites eliciting the responses in neurons 'a' and 'b' shown in the same manner as in Fig. 2b.
Y. Miyamoto, K. Jinnai / Brain Research 649 (1994) 313-318
tials in MD neurons after SNr stimulation. In Fig. 3a, the hyper-polarization was elicited at the fixed latency of 1.9 ms. In Fig. 3b, hyperpolarization with longer latency (5.9 ms) was seen when the SNr was stimulated by single pulse, and this response could follow triple shock at 100 Hz at a constant latency. The fixed latency of these hyperpolarizing responses suggested that they were induced monosynaptically. Antidromic responses of these neurons to PF stimulation could not be obtained due to technical difficulties, such as rapid deterioration of membrane potential. The suppression of spontaneous discharges of MD neurons was induced exclusively by stimulation of the medial two-thirds of the SNr where antidromic responses of SNr neurons to MD stimulation were recorded, and the suppression was elicited more effectively by electrodes in the central rather than the dorsal or ventral marginal regions of the SNr. Latency of MD-induced antidromic responses (1.4-14.0 ms) coincided well with that of suppression of MD neurons induced by SNr stimulation (2.0-15.0 ms). These findings strongly suggest that the inhibitory effect is not due to a current spread to the surrounding structures such as the cerebral peduncle or the medial lemniscus but is mediated by SNr neurons projecting to the MD. It might be possible that the suppression could be due to disfacilitation of MD neurons following the inhibition of VM neurons, because SNr neurons projecting to the VM were reported to be located in the similar part [17]. However, this is most unlikely because none of the previous histological studies injecting H R P into the MD reported a retrograde labelling of VM neurons. Therefore, the postulated 'disfacilitation' should be mediated through more than two synapses. In this study, hyperpolarizing responses with fixed latency to SNr stimulation, were recorded in MD neurons intracellulary. Such fixed latency response cannot be attributed to the polysynaptic disfacilitation. Disfacilitation through the nigro-tectal inhibitory pathway is also unlikely for the same reason. Furthermore, stimulation of the lateral part of the SNr (SNL; according to Berman [1]), which includes nigro-tectal neurons, did not suppress MD neurons at all. Terminal distribution of nigro-thalamic fibers within the MD was studied in details in rats [4,5] and monkeys [7,10], but not so precisely in cats. In monkeys, Carpenter [2] reported that it was confined to the paralamellar part of the MD (MDpl), and other authors [7] found the projection to the parvicellular part of the MD (MDpc) and the magnocellular part of the MD (MDmc) in addition to that to the MDpl. In this study in cats, MD neurons with SNr input were found in the dorsal and central parts of the nucleus, and were not confined to the lateral part of the MD. The difference between the two species may be related to absence of subnucleus corresponding to the MDpl in cats.
317
The topographical projection from the MDpc and the MDpl to the dorsolateral convexity of the PF [6], and that from the MDmc to the orbitofrontal cortex [6,7], were reported in monkeys. Although such a subdivision is not applicable to the MD of cats [15], topographical projections from the dorsal MD to the dorsal PF, and from the ventral MD to the ventral PF was observed in this study. It is possible that the dorsal MD-PF projection in cats corresponds to the MDpcdorsolateral PF projection in monkeys for the following reasons. Namely, Cavada [3] investigated the afferents to the PF from other cortical areas in cats and compared them with those in monkeys. The dorsal PF in cats received projections from the neocortical sensory association and premotor areas [11,19], in the same manner as the dorsolateral convexity of the PF in monkeys. On the other hand, the ventral PF in cats received fibers from the olfactory cortices [3], in the same manner as the orbitofrontal cortex in monkeys. Therefore, the dorsal PF in cats may well correspond to the dorsolateral convexity of the monkey's PF, which is projected by the MDpc. The longer latency of antidromic responses of the dorsal MD neurons to stimulation of the dorsal PF suggested the slower conduction velocity and the smaller cell sizes of those neurons. Most of MD neurons receiving inputs from the SNr, which projected to the dorsal and central parts of the PF as shown in the filled symbol in Fig. 1, belong to this dorsal projection system. It is interesting that the nigral output through the MD reaches to the dorsal part of the PF and interacts with cortical inputs from the neocortical sensory and motor association areas. Both the nigro-thalamic and pallido-thalamic projections were regarded as two main output systems from the basal ganglia. Pallido-thalamic output has been known to be inhibitory and directly relayed to the motor and premotor areas of the cerebral cortex. This study confirms that the nigro-thalamic output to the MD is also inhibitory, and it is relayed at least to the dorsal prefrontal area in cats. Consequently, this nigro-thalamo-cortical pathway is considered as an equivalent system to the pallido-thalamo-cortical pathway. The authors wish to thank Prof. T. Yokota and Prof. J. Handa for critical reading of the manuscript and constant encouragement.
[1] Berman, A.L., The Brainstem of the Cat. A Cytoarchitectonic Atlas with Stereotaxie Coordinates, The University of Wisconsin Press, Madison, MI, 1968. [2] Carpenter, M.B.,Nakano, K. and Kim, R., Nigrothalamic projections in the monkey demonstrated by autoradiographic technics, J. Comp. Neurol., 165 (1976) 401-416. [3] Cavada, C. and Reinoso-Sufirez, F., Topographical organization
318
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
~: Mo'amoto. K. .linnai / Bruin Reseanh 649 q 1994) ,71,7-,7IS of the cortical afferent connections of the prefrontal cortex in the cat, J. ('omp. Neurol., 242 (1985) 293-324. Cornwall, J. and Phillipson, O.T., Afferent projections to the dorsal thalamus of the rat as shown by retrograde lectin transport - 1. The mediodorsal nucleus, Neuroscience, 24 (1988) 1035-1049. Deniau, J.M. and Chevalier, G., The lamellar organization of the rat substantia nigra pars reticulata: distribution of projection neurons, Neuroscience, 46 (1992) 361-377. Goldman-Rakic,P.S. and Porrino, L.J., The primate mediodorsal (MD) nucleus and its projection to the frontal lobe, Z Comp. Neurol., 242 (1985) 535-560. llinski, 1.A., Jouandet, M.L. and Goldman-Rakic, P.S., Organization of the nigrotbalamocortical system in the rhesus monkey, J. Cornp. Neurol., 236 (1985) 315-336. Kernel, M.L., Desban, M., Gauchy, C., Glowinski, J. and Besson, MJ.. Topographical organization of efferent projections from the cat substantia nigra pars reticulata, Brain Res.,, 455 (1988) 307-323. Kultas-llinski, K. and Ilinski, 1.A., Fine structure of the magnocellular subdivision of the ventral anterior thalamic nucleus (VAmc) of Macaca mulatta: II. Organization of nigrothalamic afferents as revealed with EM autoradiography, J. Comp. Neurol., 294 (1990) 479-489. Kultas-Ilinski, K., llinski, 1.A., Massopust, L.C., Young, P.A. and Smith, K.R., Nigrothalamic pathway in the cat demonstrated by autoradiography and electron microscopy, Exp. Brain Res., 33 (1978) 481 492. Markowitsch, H.J., Pritzel, M. and Divac, 1., The prefrontal
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
cortex of the cat: anatomical subdivisions hdscd on retrograde labeling of cells in the mediodorsal thalamic nucleus, l'Jxp. BraiH Res., 32 (1978) 335-344. Ray, J.P. and Price, J.L., ]'he organization ol the thalamocortical connections of the mediodorsal thalamic nucleus in the rat. related to the ventral fl)rebrain - - prefrontal cortex topography, J. Comp. Neurol., 323 (1992) 167-197. Rinvic, E., Demonstration of nigrothatamic connections in the cat by retrograde axonal transport of horseradish peroxidasc, Brain Res., 90 (1975) 313-318. Russehen, F.T., Amaral, D.G. and Price, J.L., The afferent input to the magnocellular division of the mediodorsal thalamic nucleus in the monkey, Macaca f~lscicularis, .I. Comp. Neurol., 256 (1987) 175-210. Tanibuchi, 1., Electrophysiological and anatomical studies on thalamic mediodorsal nucleus projections onto the prefrontal cortex in the cat, Brain Res., 580 (1992) 137-146. Tokuno, H., Nakamura, Y., Kudo, M. and Kitao, Y., Laminar organization of the substantia nigra pars reticulata in the cat, Neurosciencc, 38 (199(/) 255 270. Ueki, A., The mode of nigro-thalamic transmission investigated with intracellular recording in the cat, Exp. Brain Res., 49 (1983) 116 124. Velayos, J.L. and Reinoso-Sufirez, F., Topographic organization of the brainstem afferents to the mediodorsal thalamic nucleus, J. Comp. Neurol., 206 (1982) 17-27. Velayos, J.L. and Reinoso-Suarez, F.. Prosencephalic afferents to the mediodorsal thalamic nucleus, J. Comp. NeuroL, 242 (1985) 161-181.
ELSEVIER
Short Communication
The inhibitory input from the substantia nigra to the mediodorsal nucleus neurons projecting to the prefrontal cortex in the cat Yoshihisa Miyamoto a, Kohnosuke Jinnai b,, a Department of Neurosurgery and b Department of Physiology, Shiga University of Medical Science, Ohtsu, Shiga 520-21, Japan
(Accepted l0 March 1994)
Abstract
The effects of stimulation of the substantia nigra pars reticulata (SNr) on the neurons of the mediodorsal nucleus (MD) projecting to the prefrontal cortex (PF) were studied in cats. The MD neurons projecting to the ventral part of the PF tended to be located in the ventral part of the MD, while those projecting to the dorsal part of the PF in the dorsal part. Spontaneous discharges of 30/57 tested MD neurons were suppressed by SNr stimulation at a latency ranging from 2 to 15 ms. The latency of the suppression corresponded well to that of antidromic responses of SNr neurons elicited by MD stimulation (from 1.4 to 14.0 ms). Intracellular recordings in a few MD neurons showed IPSP by SNr stimulation. The SNr is considered to exert an inhibitory effect on the MD neurons projecting to the PF. Key words: Nigrothalamic projection; Substantia nigra; Mediodorsal nucleus; Prefrontal cortex; Electrophysiology; Cat
The nigro-thalamo-cortical pathway is one of the main output systems of the basal ganglia, but its functional role remains obscure partly because of paucity of anatomical and electrophysiological data concerning this pathway. Previous anatomical studies revealed that the substantia nigra pars reticulata (SNr) projects to the medial devision of the ventral lateral nucleus (VLm), magnocellular part of the ventral anterior nucleus (VAmc) and the mediodorsal nucleus (MD) in monkeys [7,14] and rats [4,12]. Velayos and ReinosoSuftrez [18], and Kemel et al. [8] found SNr-MD connection in cats by retrograde horseradish peroxidase ( H R P ) or autoradiographical studies, though Rinvic [13], Kultas-Ilinsky et al. [10] and Tokuno et al. [16] did not demonstrate the connection. Nigro-thalamic output is suggested to be sent to the prefrontal cortex (PF) because all of these thalamic nuclei (VLm, VAmc, and MD) were reported to project mainly to the PF [6,7,11,12,14,15]. Although the cortical projection neurons in the V A m c have been shown to receive inhibitory synapses from the SNr in monkeys [9], the direct-relay of SNr output to the PF through the other thalamic nuclei remains to be examined. The nigro-
* Corresponding author. Fax: (81) 775-48-2146. 0006-8993/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0006-8993(94)00295-N
thalamic projection is considered to be inhibitory because monosynaptic inhibitory post-synaptic potential (IPSP) by stimulation of the SNr was recorded in ventral medial nucleus (VM) neurons in cats [17], and because degenerated synapses in the V A m c after SNr lesioning were symmetrical ones with pleomorphic vesicles [9]. Influence of nigro-thalamic input on MD neurons has not yet been studied. In this study, the effect of SNr stimulation on M D neurons projecting to the PF were studied electrophysiologically. The aim of this study is to test: (1) whether SNr input is relayed to the PF by MD neurons, and (2) whether nigral input to the MD is inhibitory or excitatory. The experiments were performed on adult cats under general anesthesia, which was induced with an intramuscular injection of ketamine hydrochloride (40 m g / k g ) , and maintained with sodium pentobarbital (30 m g / k g ) . Cats were intubated, and maintained on spontaneous respiration. The skull was fixed in a stereotaxic frame and the frontoparietal cortex was exposed bilateraly by craniectomy. Four pairs of bipolar stimulating electrodes were inserted into the right PF, and 4 concentric stimulating electrodes were stereotaxically inserted into the right substantia nigra (SN). After the left parietal cortex was aspirated to expose the dorsal surface of the left thalamus, a glass microelectrode
314
}: ..tliyamom. K. .limmi /" Brain Rcwarch 04Y (I9~)4) 713 -.'~/~,'
1(I-30 MS/) were used. R e c o r d i n g sites of M D n e u r o n s were m a r k e d by passing negativc c u r r c n t s at 1 0 / z A for 10 min t h r o u g h the glass microelectrodc. To study SNr n e u r o n s projecting to the MD, three concentric stimulating electrodes wcrc inserted into the M D in the same m a n n e r as the i n t r o d u c t i o n ol recording microelectrode into the MD. T h e activities of SN n e u r o n s were recorded extracellularly with a glass-coated elgiroy microelectrode, and the latencies of a n t i d r o m i c response to M D stimulation were measured. R e c o r d i n g sites of SN n e u r o n s wcrc m a r k e d by passing positive D C c u r r e n t s (10 ~ A for 30 s) through the elgiroy electrode. At the e n d of each experiment, the brain was perfused with 10% formalin ( c o n t a i n i n g 2% potassium
filled with 2% P o n t a m i n e Sky Blue in {).5 M sodium acetate solution, was inserted across the midline into the right M D from the dorsal surface of the contralateral t h a l a m u s at an angle of 45 ° to the vertical axis. T h e cxtracel[ular unitary recordings of M D n e u r o n s which r e s p o n d e d antidromically to stimulation of the ipsilateral PF were carried out. T h e effects of stimulation of the SNr (less than 0.5 mA, d u r a t i o n 0.3 ms, r e p e t i t i o n rate 1.0-0.8 Hz) on M D n e u r o n s were observed with an aid of peri-stimulus time histogram (PSTH). Postsynaptic p o t e n t i a l s evoked by SNr stimulation were recorded intracellularly from M D n e u r o n s in several experiments. For this purpose, glass microelectrodes filled with 2% P o n t a m i n e Sky Blue dissolved in 0.5 M potassium acetate ( D C resistance;
b
PFstimulation i
c
I ,f
"'"~--k
-.
/
"X
a
"//~
.
"
SNr stimulation _scLj~-'..
~B
scr(
I
! i
,
;,
,ii
!
.
,
A o
mesial view
;
,:
0
J s'
\
~"
lateral view
II
25ms [[i
d /
A 10.2 !I
/' s . . . . . ~ .7)" 'X ...... ,. . . . t' ML) } •
A 9.1
/~,
- >E 5ms d
e ventral PF
A 7.9
A 8.3
/ (\ } CLN )
~"I
' /" PC/"
"~, ,~
-,,/
-
•
)
~.._.,~ pc
/
. _ t'_
.:'~/
!
-!
-
'\
i Rf ~ '
J
\
",
"
,
dorsal PF
/
/' C M
'
I
Fig. 1. a: PSTH showing suppression of discharges of a MD neuron after SNr stimulation. Bin width 1 ms, 128 sweeps addition, b: Antidromic response of the MD neuron in 'a' elicited by PF stimulation and its collision with a spontaneous discharge, c: the location of the tip of the stimulating electrodes in dorsal (square), ventral (circle), caudo-medial (triangle) part of the PF. The filled symbols indicate the stimulating sites which evoked antidromic response in the MD neurons inhibited by SNr stimulation. SCr, curuciate sulcus; SCot, coronal sulcus; SPs, presylvian sulcus, d: location of MD neurons projecting to PF. Square, circle, and triangle demonstrate the neurons projecting to the dorsal, ventral, and caudomedial parts of PF, respectively. Filled symbols indicate the MD neurons which were inhibited by SNr stimulation. The number above each coronal section (A 7.9-10.2) indicates the distance from the interaural plane. CLN, central lateral nucleus; CM, centre median nucleus; HL, lateral habenular nucleus; LP, lateral posterior complex; PC, paracentral nucleus; RF, reticular fiber, e: Latency histograms of antidromic response of MD neurons following stimulation of ventral part of PF and dorsal part of PF.
Y. Miyamoto, K. Jinnai / Brain Research 649 (1994) 313-318
ferrocyanide in the case of SN recording). All recording sites in the MD and the SNr were reconstructed on the basis of location of the marked spots. In the extracellular studies, 97 MD neurons were activated antidromically by stimulation of the PF at a threshold lower than 0.5 mA. All of these neurons showed the collision interaction with spontaneous discharges (Fig. lb). Fig. lc shows the location of tips of the stimulating electrodes in the PF, which elicited antidromic responces of MD neurons as shown in Fig. ld. The MD neurons antidromically activated by stimulation of the dorsal part of the PF, i.e., the rostral part of the frontal gyrus, dorsal part of the proreus gyrus and dorsal part of the medial bank of the presylvian sulcus, were located in the dorsal part of the MD (squares in Fig. lc). The MD neurons activated by
a
,:i 1
5
10
b /)
stimulation of the ventral part of the PF, i.e., rostral and ventral parts of the rectus gyrus, and ventral two-thirds of the medial bank of the presylvian gyrus were located in the ventral part of the MD (circles in Fig. lc). The MD neurons projecting to the ventral part of the PF showed shorter antidromic latencies (mean + S.D. = 4.2 + 2.2 ms) than those projecting to the dorsal part (mean + S.D. = 6.7 + 4.1 ms) as shown in Fig. le. The stimulation of caudal part of the medial PF may have activated the fibers projecting to the more rostral PF by current spread to the underlying white mater. Therefore, the MD neurons antidromically activated by those stimulating sites were classified into another category (triangle in Fig. lc). Among the 57 MD neurons projecting to the PF, the spontaneous discharges of 30 MD neurons were sup-
C
N=19 M ~ SD: 5.2 :t: 3.6 ms
N=30
15 ms
0
A 7.2
315
d
5
MDstimulation
10
15 ms
e
!
LP
I -~
"
~
-
-
~
~
O0 o
A45 /
'
i A 3.7
Fig. 2. a: latency histogram of inhibitory responses of M D neurons following SNr stimulation, b: distribution of the tips of stimulating electrodes in and around SN. The n u m b e r above the each frontal section (A 2.9-7.2) indicates the distance from the interaural plane. Closed circles indicate the tips of electrodes which elicited the inhibitory response of M D neurons. Open circles indicate the tips of stimulating electrodes which did not show the inhibitory effect. The bar over, across, or below the circles, indicates the classification of stimulating sites into three groups, i.e., dorso-marginal, central, and ventro-marginal parts of SNr. PP, pes pedunculi; RN, red nucleus; RR, retro-rubral nucleus; SNc, pars compacta of SN; VTA, ventral tegmental area of Tsai. c: Latency histogram of antidromic response of SNr neurons following M D stimulation d: Antidromic response of a SNr neuron by M D stimulation, and its collision with spontaneous discharge, e: location of stimulating electrodes in M D which induced the antidromic activation of SNr neurons in 'd'. f: location of SNr neurons showing the antidromic response to M D stimulation with threshold less than 0.45 mA. Most of them were located in the central or medial part of SNr.
Y. Mivamoto K. .linnai / Brain Research 049 (1994) 313-.~1S
316
pressed after SNr stimulation at intensities ranging from 0.15 to 0.5 m A (0.3 ms duration). Fig. la shows an example of inhibitory response to SNr stimulation. The latencies of inhibitory responses in MD neurons ranged from 2.0 ms to 15.0 ms, and most of them (19/30) were 2.0 to 6.0 ms (Fig. 2a). MD neurons inhibited by SNr stimulation were located in the dorsal and central portions of the MD, but not found in the ventral part, although the ventromedial parts of the MD were scarcely explored in this study (Fig. ld). The tips of the nigral stimulating electrodes eliciting the inhibition located mostly in the medial two-thirds of the SNr, and none in the lateral part of the SN, i.e., the SNL according to Berman [1], was effective. As for the dorsoventral direction, the electrodes in the central part of the SNr induced the inhibition more effectively than those in the dorsal and ventral margins of this nuclei as follows. Twenty-three MD neurons out of 48 tested were suppressed by stimulation of the central part of the SNr, whereas 10 and 6 out of the 43 and 33 tested MD neurons, respectively, were inhibited by stimulation of the dorsal and ventral border regions of the SNr. (Fig. 2b; One MD neuron was sometimes counted twice when it was tested by two stimulating electrodes.) Six MD neurons showed an increase in discharge rates after stimulation around the SNr at a latency longer than 6.0 ms. The effective stimulating
b
a
SNr"~\stimulal /ti'Y~ °nl
sites for the excitatory responses were not confined to the SNr. The location of SNr neurons projecting to the MD (Fig. 2f) and the latencies of antidromic responses to MD stimulation (Fig. 2c) were investigated to examine whether the SNr-induced inhibitory responses of MD neurons were mediated directly through the nigrothalamic pathway. Nineteen SNr neurons were activated antidromically by MD stimulation with thresholds less than 0.45 mA (Fig. 2d). Most of them were located in the medial two-thirds of the SNr (Fig. 2f). The latencies of antidromic activation ranged from 1.4 to 14.0 ms but were mostly less than 6.(1 ms (Fig. 2c) and this distribution of latency was coincident with that of the inhibitory response in MD neurons following SNr stimulation (Fig. 2a). These antidromic responses cannot be attributed to the activation of thc nigrothalamic fibers projecting to the other thalamic nuclei than the MD, e.g., the VM, because antidromic responses activated by electrodes in the rostro-ventral part of the MD did not show a lower threshold. Furthermore, the antidromic latency was much longer than expected from the latency of the monosynaptic IPSPs of VM neurons induced by SNr stimulation [6]. Five MD neurons were studied intracellulary with resting m e m b r a n e potentials lower than - 4 0 inV. Fig. 3 shows two examples of changes in membrane poten-
c //,,
SNrstimulation 1
d
,I i[
'I!i
J
A 4.8
\,%/~ !i!
....
-
./
/
__J
2
rnV
20 ms Fig. 3. a: hyperpolarization induced by SNr stimulation recorded intra-cellularly from a MD neuron, b: intracellular recordings from another MD neuron, showing hyperpolarization with a fixed latency. The hyperpolarizing response followed triple shocks (100 Hz) with the same latency, c: the recording sites of neurons 'a' and 'b'. d: the stimulating sites eliciting the responses in neurons 'a' and 'b' shown in the same manner as in Fig. 2b.
Y. Miyamoto, K. Jinnai / Brain Research 649 (1994) 313-318
tials in MD neurons after SNr stimulation. In Fig. 3a, the hyper-polarization was elicited at the fixed latency of 1.9 ms. In Fig. 3b, hyperpolarization with longer latency (5.9 ms) was seen when the SNr was stimulated by single pulse, and this response could follow triple shock at 100 Hz at a constant latency. The fixed latency of these hyperpolarizing responses suggested that they were induced monosynaptically. Antidromic responses of these neurons to PF stimulation could not be obtained due to technical difficulties, such as rapid deterioration of membrane potential. The suppression of spontaneous discharges of MD neurons was induced exclusively by stimulation of the medial two-thirds of the SNr where antidromic responses of SNr neurons to MD stimulation were recorded, and the suppression was elicited more effectively by electrodes in the central rather than the dorsal or ventral marginal regions of the SNr. Latency of MD-induced antidromic responses (1.4-14.0 ms) coincided well with that of suppression of MD neurons induced by SNr stimulation (2.0-15.0 ms). These findings strongly suggest that the inhibitory effect is not due to a current spread to the surrounding structures such as the cerebral peduncle or the medial lemniscus but is mediated by SNr neurons projecting to the MD. It might be possible that the suppression could be due to disfacilitation of MD neurons following the inhibition of VM neurons, because SNr neurons projecting to the VM were reported to be located in the similar part [17]. However, this is most unlikely because none of the previous histological studies injecting H R P into the MD reported a retrograde labelling of VM neurons. Therefore, the postulated 'disfacilitation' should be mediated through more than two synapses. In this study, hyperpolarizing responses with fixed latency to SNr stimulation, were recorded in MD neurons intracellulary. Such fixed latency response cannot be attributed to the polysynaptic disfacilitation. Disfacilitation through the nigro-tectal inhibitory pathway is also unlikely for the same reason. Furthermore, stimulation of the lateral part of the SNr (SNL; according to Berman [1]), which includes nigro-tectal neurons, did not suppress MD neurons at all. Terminal distribution of nigro-thalamic fibers within the MD was studied in details in rats [4,5] and monkeys [7,10], but not so precisely in cats. In monkeys, Carpenter [2] reported that it was confined to the paralamellar part of the MD (MDpl), and other authors [7] found the projection to the parvicellular part of the MD (MDpc) and the magnocellular part of the MD (MDmc) in addition to that to the MDpl. In this study in cats, MD neurons with SNr input were found in the dorsal and central parts of the nucleus, and were not confined to the lateral part of the MD. The difference between the two species may be related to absence of subnucleus corresponding to the MDpl in cats.
317
The topographical projection from the MDpc and the MDpl to the dorsolateral convexity of the PF [6], and that from the MDmc to the orbitofrontal cortex [6,7], were reported in monkeys. Although such a subdivision is not applicable to the MD of cats [15], topographical projections from the dorsal MD to the dorsal PF, and from the ventral MD to the ventral PF was observed in this study. It is possible that the dorsal MD-PF projection in cats corresponds to the MDpcdorsolateral PF projection in monkeys for the following reasons. Namely, Cavada [3] investigated the afferents to the PF from other cortical areas in cats and compared them with those in monkeys. The dorsal PF in cats received projections from the neocortical sensory association and premotor areas [11,19], in the same manner as the dorsolateral convexity of the PF in monkeys. On the other hand, the ventral PF in cats received fibers from the olfactory cortices [3], in the same manner as the orbitofrontal cortex in monkeys. Therefore, the dorsal PF in cats may well correspond to the dorsolateral convexity of the monkey's PF, which is projected by the MDpc. The longer latency of antidromic responses of the dorsal MD neurons to stimulation of the dorsal PF suggested the slower conduction velocity and the smaller cell sizes of those neurons. Most of MD neurons receiving inputs from the SNr, which projected to the dorsal and central parts of the PF as shown in the filled symbol in Fig. 1, belong to this dorsal projection system. It is interesting that the nigral output through the MD reaches to the dorsal part of the PF and interacts with cortical inputs from the neocortical sensory and motor association areas. Both the nigro-thalamic and pallido-thalamic projections were regarded as two main output systems from the basal ganglia. Pallido-thalamic output has been known to be inhibitory and directly relayed to the motor and premotor areas of the cerebral cortex. This study confirms that the nigro-thalamic output to the MD is also inhibitory, and it is relayed at least to the dorsal prefrontal area in cats. Consequently, this nigro-thalamo-cortical pathway is considered as an equivalent system to the pallido-thalamo-cortical pathway. The authors wish to thank Prof. T. Yokota and Prof. J. Handa for critical reading of the manuscript and constant encouragement.
[1] Berman, A.L., The Brainstem of the Cat. A Cytoarchitectonic Atlas with Stereotaxie Coordinates, The University of Wisconsin Press, Madison, MI, 1968. [2] Carpenter, M.B.,Nakano, K. and Kim, R., Nigrothalamic projections in the monkey demonstrated by autoradiographic technics, J. Comp. Neurol., 165 (1976) 401-416. [3] Cavada, C. and Reinoso-Sufirez, F., Topographical organization
318
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
~: Mo'amoto. K. .linnai / Bruin Reseanh 649 q 1994) ,71,7-,7IS of the cortical afferent connections of the prefrontal cortex in the cat, J. ('omp. Neurol., 242 (1985) 293-324. Cornwall, J. and Phillipson, O.T., Afferent projections to the dorsal thalamus of the rat as shown by retrograde lectin transport - 1. The mediodorsal nucleus, Neuroscience, 24 (1988) 1035-1049. Deniau, J.M. and Chevalier, G., The lamellar organization of the rat substantia nigra pars reticulata: distribution of projection neurons, Neuroscience, 46 (1992) 361-377. Goldman-Rakic,P.S. and Porrino, L.J., The primate mediodorsal (MD) nucleus and its projection to the frontal lobe, Z Comp. Neurol., 242 (1985) 535-560. llinski, 1.A., Jouandet, M.L. and Goldman-Rakic, P.S., Organization of the nigrotbalamocortical system in the rhesus monkey, J. Cornp. Neurol., 236 (1985) 315-336. Kernel, M.L., Desban, M., Gauchy, C., Glowinski, J. and Besson, MJ.. Topographical organization of efferent projections from the cat substantia nigra pars reticulata, Brain Res.,, 455 (1988) 307-323. Kultas-llinski, K. and Ilinski, 1.A., Fine structure of the magnocellular subdivision of the ventral anterior thalamic nucleus (VAmc) of Macaca mulatta: II. Organization of nigrothalamic afferents as revealed with EM autoradiography, J. Comp. Neurol., 294 (1990) 479-489. Kultas-Ilinski, K., llinski, 1.A., Massopust, L.C., Young, P.A. and Smith, K.R., Nigrothalamic pathway in the cat demonstrated by autoradiography and electron microscopy, Exp. Brain Res., 33 (1978) 481 492. Markowitsch, H.J., Pritzel, M. and Divac, 1., The prefrontal
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
cortex of the cat: anatomical subdivisions hdscd on retrograde labeling of cells in the mediodorsal thalamic nucleus, l'Jxp. BraiH Res., 32 (1978) 335-344. Ray, J.P. and Price, J.L., ]'he organization ol the thalamocortical connections of the mediodorsal thalamic nucleus in the rat. related to the ventral fl)rebrain - - prefrontal cortex topography, J. Comp. Neurol., 323 (1992) 167-197. Rinvic, E., Demonstration of nigrothatamic connections in the cat by retrograde axonal transport of horseradish peroxidasc, Brain Res., 90 (1975) 313-318. Russehen, F.T., Amaral, D.G. and Price, J.L., The afferent input to the magnocellular division of the mediodorsal thalamic nucleus in the monkey, Macaca f~lscicularis, .I. Comp. Neurol., 256 (1987) 175-210. Tanibuchi, 1., Electrophysiological and anatomical studies on thalamic mediodorsal nucleus projections onto the prefrontal cortex in the cat, Brain Res., 580 (1992) 137-146. Tokuno, H., Nakamura, Y., Kudo, M. and Kitao, Y., Laminar organization of the substantia nigra pars reticulata in the cat, Neurosciencc, 38 (199(/) 255 270. Ueki, A., The mode of nigro-thalamic transmission investigated with intracellular recording in the cat, Exp. Brain Res., 49 (1983) 116 124. Velayos, J.L. and Reinoso-Sufirez, F., Topographic organization of the brainstem afferents to the mediodorsal thalamic nucleus, J. Comp. Neurol., 206 (1982) 17-27. Velayos, J.L. and Reinoso-Suarez, F.. Prosencephalic afferents to the mediodorsal thalamic nucleus, J. Comp. NeuroL, 242 (1985) 161-181.