Antidromic identification of dopaminergic and other output neurons of the rat substantia nigra

Brain Research, 150 (1978) 69-84 © Elsevier/North-HollandBiomedicalPress

ANTIDROMIC IDENTIFICATION OF DOPAMINERGIC AND OUTPUT NEURONS OF THE RAT SUBSTANTIA NIGRA

69

OTHER

P. G. GUYENET and G. K. AGHAJANIAN Departments o f Psychiatry and Pharmacology, Yale University School of Medicine and the Connecticut Mental Health Center, New Haven, Conn. 06508 (U.S.A.)

(Accepted November 3rd, 1977)

SUMMARY In the present study dopamine (DA)-containing and other output neurons of the substantia nigra (SN) were identified by antidromic stimulation from postulated target nuclei, the caudate-putamen, the thalamus, the cortex and the pontine reticular formation. To guide electrode placements, the topography of the nigrostriatal projection system was determined by retrograde tracing methods. Spontaneously active cells present in the SN were then classified in two groups according to the shape of their action potentials and their firing rate. Type I cells were located mainly in the pars compacta and could be antidromically-activated (AD-activated) from various locations along the course of the nigrostriatal pathway (caudate-putamen, globus pallidus, MFB) but not from other brain areas (ventromedial thalamus, motor cortex, pontine reticular formation). These neurons had a slow bursting pattern of firing, a very slow conduction velocity (0.58 m/sec), and a wide action potential. Their firing rate was dramatically reduced following the intravenous administration of apomorphine (ID 50:9.3 #g/kg), or the iontophoretic application of DA and GABA. Type II cells were located predominantly in the pars reticulata; most of them could be AD-activated from the ventromedial thalamus and the MFB but not from the motor cortex. A few of these cells could be AD-activated from the pontine reticular formation and the thalamus. A minority of type II cells, located in or near the pars compacta could be AD-activated from the caudate-putamen. In addition, their conduction velocity was much higher (2.8 m/sec) and their firing rate far in excess of that exhibited by type I neurons. These neurons were inhibited by the iontophoretic application of G A B A but not of DA. The microinjection of 6-hydroxydopamine (a neurotoxin relatively specific against catecholamine-containing neurons) in the vicinity of the MFB blocked selectively the propagation of antidromic spikes in type I but not type II cells.

70 It is concluded that type I cells are the DA neurons of the nigrostriatal pathway. Type 1I cells are mainly output neurons that project to the ventromedial thalamus, the pons and the forebrain. This telencephalic projection most likely constitutes a second, non-DA, fast-conducting nigrostriatal pathway.

INTRODUCTION The single unit activity of neurons in the substantia nigra (SN) has been recorded in many speciesr,12A3.14,21: some of the cells studied may have been nigro-striatal dopaminergic (DA) neurons, but others could have been non-DA output neurons, interneurons or even cells in the reticular formation. Some attempts have been made to identify physiologically the nigro-striatal DA as opposed to other types of SN neurons. Bunney et al. 6 described in the rat SN the existence of a distinctive population of slow cells, localized mainly in the pars compacta (PC) where DA neurons are concentrated. These cells were found to be extremely sensitive to a number of drugs known to alter the metabolism or the release of dopamine in the striatum (caudate-putamen). The pharmacological properties of these neurons together with the fact that they could not be found following the administration of 6-hydroxydopamine (a neurotoxin relatively specific against catecholamine-containing neurons) led Bunney et al. 8 to postulate that they were the DA neurons of the SN. In another study, antidromically identified nigrothalamic neurons in the rat and the cat have been found to lie mainly in the pars reticulata but there was no investigation of the electrophysiological differences between the nigrothalamic and nigrostriatal projections 12. In addition to the nigrostriatal and nigrothalamic pathways ~,18,a°, efferent projections to the tectum and the pontine reticular formation have been described recently~4,34. The existence of a rtigrocortical and a nigroamygdaloid pathway has also been suggested 4,25,31. Adding to the complexity of this organization, it is possible that more than one type of neuroncontributes to some of these projections. For example, the fact that locally applied dopamine exerts mainly a depressant action on neurons in the caudate-putamen ~ while the nigrostriatal pathway seems to be mainly excitatory'6, 32 has been tentatively explained by postulating the existence of a second nonDA pathway 16. This hypothesis has also received some indirect support through biochemical and histological experiments 17,29. The purpose of the present study was to identify the DA and other output neurons of the SN by antidromic stimulation from the main postulated target nuclei, the caudate-putamen, the thalamus, the cortex and the pontine reticular formation. The caudate-putamen is a large structure and the nigro-striatal projection known to be topographically organizeda,15; thus the chances of recording an antidromic response in randomly encountered nigral neurons when the caudate-putamen is stimulated, are very small. To increase this probability, the topography of the projection was first determined using the technique of the retrograde transport of horseradish peroxidase 28. This paper describes the electrophysiological characteristics of dopa-

71 minergic cells, provides some additional evidence in favor of the existence of a second non-DA nigrostriatal pathway and presents a preliminary account of the properties of nigrothalamic and nigroreticular neurons. METHODS

Unit recording and electrical stimulation Fifty male albino rats (230-270 g) were anaesthetized with chloral hydrate (400 mg/kg i.p.) and prepared for recording as previously described (Bunney et al.6). Extracellular action potentials were recorded differentially as previously described 39. Briefly, two-barrel electrodes (tip size 1 #m, tip separation 50/~m) were filled with NaC1 2 M saturated with fast green (impedance 4-8 Mr2). Electrical signals were fed into a high input impedance differential amplifier (Model DAM-5, W.P. Instruments). Neuronal action potentials were simultaneously displayed on a storage oscilloscope and fed into an electronic counter whose threshold was set so that it was triggered by the individual discharges. Integrated histograms, generated by the analog output of the electronic counter were plotted on a polygraph recorder. Electrical signals were also led into a Nicolet Signal Averager (Model 1072) to obtain poststimulus time histograms (PSTH). Concentric electrodes (NE-100, David Kopf Instruments) were used to stimulate various brain areas. Square pulses (100-500/~sec, 0.05-2.0 mA, 1 Hz or more) were generated by a digital pulse generator (W.P. Instruments series 800) and a photon coupled stimulus isolation unit (Model PC-3A, W.P. Instruments). Three criteria were used to ascertain the antidromic nature of a neuronal response: the constant latency of the response, the fact that the response could follow at a frequency of 300 Hz or more and the occurrence of collision when the stimulus was delivered within a certain time following a spontaneous spike (collision interval). Six barrel electrodes of the type previously described by Wang and Aghajaniana° were used to deliver iontophoretically GABA (0.2 M, pH 4.0) and DA (0.2 M, pH 3.8) while recording cell firing differentially. Histo logical pro cedures To mark the site of the stimulating electrode tip, small lesions were made by delivering a 0.5 mA positive direct current for 30 sec. The final recording site was marked by passing a 20 #A cathodal current through the recording barrel for 10 min. This resulted in the deposition of Fast Green in a discrete spot 3~. The animals were then perfused with 10~ buffered formalin. Serial sections (50 #m) were cut, then mounted and counterstained with neutral red. The dye spot was observed under a light microscope and served as a reference point for the location of each cell investigated. All stereotaxic coordinates are given according to the atlas of K6nig and Klippel (1970) 27. Tracing of nigrostriatal projections with horseradish peroxidase Horseradish peroxidase (HRP) was injected iontophoretically into various areas

72 o f the c a u d a t e - p u t a m e n as previously described :~. The animals were perfused t h r o u g h the left c a r d i a c ventricle 26 h later with a solution o f I "~i p a r a f o r m a l d e h y d e and I ",, g l u t a r a l d e h y d e in 0.05 M Tris buffer ( p H 7.6) c o n t a i n i n g 5 °ii sucrose. The hislochemical p r o c e d u r e used to d e m o n s t r a t e the intracellular presence o f H R P is a modification o f the original technique o f G r a h a m a n d K a r n o v s k y (1966) ~° and has been described previously z.

e/ ' ~ ,

A 9650

A

1

9410

A

2

¢" ~ -)

A t. . . .

~

..'"-,

10

4

8380

A

5

6

t"

5340

-'~

A t"

.'"" ,'

11

A

~" "

7890

1

A 7470

7

8

9

I.._ 15.

6360

.-i-

8920

3

) A

~ "

.'"

12

-'1

¢

'

13

4390

-'1

¢

"~t

--

'0"--

' 0"

14

15

--,

.

'

16

•

"-

.

."

17

Fig. I. Tracing of the nigrostriatal projection, with horseradish peroxidase. The injection sites are represented by dots approximately the size of the enzyme spread and are numbered from l to 17. The clusters of cells containing the enzyme are drawn and numbered in a representative frontal plane at the level of the substantia nigra in the corresponding order. Abbreviations: AC: nucleus accumbens; ac: nucleus amygdaloideus centralis; alp: nucleus amygdaloideus lateralis pars posterior; AVT: area ventralis tegmenti; CAI: capsula interna; CP: caudate-putamen; cl: claustrum: FMP: fasciculus medialis prosencephali; GP: globus pallidus; IP: nucleus interpeduncularis; LM: lemniscus medialis, NR: nucleus ruber; pma: pedoncutus corporis mammillaris; SNC; substantia nigra pars cornpacta; SNR: substantia nigra pars reticulata. All frontal planes refer to the Atlas of K6nig and Klippel.

73 RESULTS

(1) Topography oJ the nigrostriatal projection Following the ejection of HRP in discrete spots in the caudate-putamen, labeled cells were observed in the ipsilateral SN according to a precise topographical pattern. In the anterior two thirds of the structure, labeled cells were found almost exclusively in the medial pars compacta (Fig. 1). In the posterior third of the SN, labeled cells could be found throughout the nucleus. Whatever the location of the deposit in the striatum, these neurons were organized in longitudinally oriented clusters, running the whole length of the SN. The locations of the ejection sites and corresponding clusters of labeled cells are indicated schematically in Fig. 1. The clusters of labeled cells are thus represented in cross section in a typical frontal plane (A 1950 #m). Knowing this topography was of considerable importance to increase the chances of recording an antidromic response in the SN when the caudate-putamen was stimulated. In six animals in which horseradish peroxidase was ejected into the frontal motor cortex (coordinates: A, 4500#m-9000 #m; L, 500-3000 #m; depth 1.3 mm below the surface of the brain) no labeled cells were found in the SN. In two others, the enzyme was ejected into the medial forebrain bundle (MFB, coordinates: A, 3900 #m; L, 1500 #m; H, --2800 #m): less than 4 labeled cells were found in the whole SN in each case suggesting that in our experimental conditions the enzyme was not taken up significantly by fibers of passage.

(2) Cell types in the SN Extracellular recordings were made in the anterior and middle thirds of the SN where, in the rat, the pars compacta is best delineated. Two distinct populations of cells could be readily detected using criteria such as the duration, shape and size of the extracellular action potential in addition to the rate and pattern of cell firing. A first class of cells, subsequently referred to as type I cells, had a slow-bursting or regular firing rate (0.5-8 spikes/see). Their action potentials were unusually wide (4-5 msec), often displayed a distinct initial segment (Fig. 2) and a large late positive component. When audiomonitored, these cells exhibited a characteristic low-pitch sound. As has been discussed before 6, the unusual shape of their action potential could not be due to injury since they appeared this way when first encountered and they could be recorded for hours without any change in firing rate and spike shape. Upon histological examination, more than 95 700 of these cells were found within the pars compacta. The second type of cell, subsequently referred to as type II cells, was characterized by a much narrower action potential (2-3 msec) giving them a higher pitch sound when audiomonitored (Fig. 2). They did not display any significant late positive component, their firing rate was generally constant but varied from cell to cell over a wide range (0.1-60 spikes/see). They were found in a zone extending from the crus cerebri to the layer where type I cells are predominant. There was no significant difference in the amplitude of the extracellular action potentials of type I and type II cells. The initial positive component of both ranged between 0.05 and 0.6 mV.

74

Fig. 2. Characteristic shape of type 1 and type 11 cells action potentials. A: 3 different type 1 celts: Notice the unusual width of the action potential, the presence of a marked initial segment componcnl (left) and the large size of the late positive component (positivity upwards). B: 2 type II cells: These action potentials are much narrower, do not present a marked initial segment component. Thc late~positive component of the spike is very small when it exists.

All type I cells tested were extremely sensitive to the intravenous administration o f a p o m o r p h i n e as has been described 38. The ID 50 (dose producing 50 °/o inhibition o f the firing r a t e ) w a s 9.3 # g / k g ~ 1.6 (mean ~ S.E.M., n ~ 16). Following a dose in the range o f the I D 50, total recovery of the initial firing rate occurred within 20 or 30 min. Subsequent administration of the same low dose produced the same effect. The administration o f high doses of a p o m o r p h i n e i.v. (0.2 mg/kg) produced a transient total inhibition of cell firing followed 20 min later by recovery. However, after an initial high dose of this drug, tachyphylaxis developed to subsequent administrations as has been reported 38. Type II ceils were unresponsive to tow doses of a p o m o r p h i n e (20 #g/kg). Even higher doses usually did not alter their firing rate (9 cells out of 11).

(3) Effect of the electrical stimulation of the caudate-putamen on the activity of S N eel& In 25 animals, a stimulating electrode was implanted in various locations in the caudate-putamen and the firing of type I and II ceils subsequently recorded. In 17 cases, the electrode was located close to site 8 (see Fig. 1), in 3 cases a r o u n d site 2, in 3 cases between sites 5 and 6, in one case close to site 9; in one case, the stimulating electrode was situated in site 16.

75

Fig. 3. Antidromic activation of type I neurons. A: this type I cell was spontaneously firing at 3.5 spikes/sec. Antidromic activation was elicited by stimulating the caudate-putamen with pulses of 500/~sec duration delivered with an intensity of 0.65 mA. Upper picture: the stimulus (arrow) was delivered 18 msec after a spontaneously occurring spike(s). Lower picture: the stimulus was delivered 15 msec after (10 superimposed sweeps). Calibration: 0.2 mV, 5 msec; a : antidromic spike. B: this type I cell was spontaneously firing at 5.2 spikes/sec. The stimulus (500 ~sec, 0.65 mA) was delivered 12 msec (upper trace) or 10 msec (lower trace) after a spontaneous spike (10 superimposed sweeps). Calibration : 0.2 mV, 5 msec. The stimulating electrode was placed as in A. Note that a is an initial segment component and not a field potential since collision occurs. C: this type I cell was stimulated antidromically from the caudate-putamen and was firing spontaneously at 0.8 spikes/sec. The stimulation was delivered at 0.9 mA (500/~sec pulse width). The 10 consecutive sweeps illustrate the different shapes of the antidromic action potential. Calibration: 0.5 mV, 5 msec. D: this type I cell (spontaneous firing 0.3 spikes/sec) was stimulated antidromically from the caudate nucleus with pulses of 500/~sec duration delivered at 0.6 mA (10 sweeps). Calibration: 0.2 mV, 2 msec.

A t o t a l o f 144 t y p e I cells were a c t i v a t e d a n t i d r o m i c a l l y (total s a m p l e : 250). Usually, o n l y the initial segment c o m p o n e n t o f the spike was triggered (Fig. 3B) a n d o n l y occasionally a full spike d e v e l o p e d (Fig. 3C). A few cells r e s p o n d e d 100 o f the time with full spikes (Fig. 3A). U s u s a l l y these were the n e u r o n s t h a t c o u l d be a c t i v a t e d with the lowest c u r r e n t (0.15-0.4 m A , 500 #sec). All o f the a b o v e responses were c o n s i d e r e d to be a n t i d r o m i c because they met the criteria o f fixed latency a n d collision with s p o n t a n e o u s spikes (Fig. 3A a n d B). T h e collision interval was always greater t h a n the delay o c c u r r i n g between the stimulus a n d the a n t i d r o m i c a l l y e v o k e d spike (latency o f the a n t i d r o m i c response) b y a p p r o x i m a t e l y 0.5 msec (Fig. 3 A a n d B). These a n t i d r o m i c responses c o u l d follow a train o f stimuli delivered at u p to 350 H z (typical e x a m p l e in Fig. 4A). A t 200 Hz, only the initial segment c o m p o n e n t r e m a i n e d even with cells t h a t r e s p o n d e d to 1 H z a n t i d r o m i c s t i m u l a t i o n b y a full spike 100

76

Fig. 4. Minimum delay between two propagated impulses in type I and type II ceils. A: this type I cell was antidromically activated from the caudate-putamen with trains of two stimuli (500/1see, i.~ mA) delivered at 390 Hz (left) and 650 Hz (right). This cell was firing spontaneously at 0.8 spikes/see Its antidromic response was essentially limited to the initial segment component of the spike. Calibration: 0.1 mV, 2 msec. Superposition of 20 sweeps. Note failure of second antidromic spike to occur when stimulus is delivered at 650 Hz, B: this type I1 cell was antidromically activated by stimulating the caudate-putamen with trains of three stimuli (500/1see duration, 0.6 mA intensity) delivered at 920 Hz (left) and 1150 Hz (right). Calibration: 0.2 mY, I reset. o f the time. A n t i d r o m i c a l l y driven type 1 cells were f o u n d p r e d o m i n a n t l y in a longitudinal cluster t o p o g r a p h i c a l l y related to the location o f the stimulating electrode. This t o p o g r a p h y was s u p e r i m p o s a b l e with t h a t observed in the horseradish peroxidase experiments. G A B A a n d DA were a p p l i e d i o n t o p h o r e t i c a l l y to 9 a n t i d r o m i c a l l y identified t y p e I cells. G A B A (30-100 nA) a n d D A (30--100 nA) consistently inhibited these cells, a n d p r o d u c e d an increase in the a m p l i t u d e o f the extracellularly recorded action potential. Q u a n t i t a t i v e d a t a concerning the c o n d u c t i o n velocity o f t y p e I neurons are s u m m a r i z e d in Fig. 5. As can be seen, the latencies o f the a n t i d r o m i c responses increase linearly with the distance between the stimulating electrode and the rec o r d i n g site except when the s t i m u l a t i n g electrode was i m p l a n t e d in the tail o f the c a u d a t e - p u t a m e n , because the n i g r o s t r i a t a l fibers enter the s t r i a t u m at the level o f the globus pallidus a n d then travel posteriorly. The average c o n d u c t i o n velocity calculated f r o m Fig. 5 is 0.58 m/sec. A total o f 27 t y p e I[ cells were also a n t i d r o m i c a l l y activated f r o m various l o c a t i o n s in the a n t e r i o r h a l f o f the c a u d a t e - p u t a m e n (0.1 to 2 m A , 200 #sec pulses). These cells were l o c a t e d in or i m m e d i a t e l y b e l o w the t y p e I cell layer (pars c o m p a c t a ) .

77 - -

I

I

T~----V-

I

q-

1

....

| ---

'

I

I

• type I cells type II cells 5

14

](tail)

12

32~p CP

s cr

6

28/GP

2

4

6

mm

8

10

Fig. 5. Conduction velocity of type I and type II cells. The abscissa indicates the straight line dis~rlc e between the stimulating electrode and the recording site in our 250 g rats. The conduction time (mean ± S.E.M.) is plotted in ordinate. Data for type I cells fit a straight line except for the point corresponding to the tail of the caudate for obvious anatomical reasons. The numbers refer to the size of the sampling. Abbreviations: CP: caudate-putamen; GP: globus pallidus; MFB: median forebrain bundle; Vm: nucleus ventralis medialis thalami. Their antidromic spikes never displayed an IS-SD spike dissociation. The response occurred at a fixed latency (2.2 ± 0.16 msec, mean 4- S.D. for 17 cells activated from the same locus in the striatum) and collision could be induced (Fig. 6). These antidromically driven type II cells could follow faithfully a train of stimuli delivered at up to 800 Hz without exhibiting an IS-SD dissociation (Fig. 4). The average conduction velocity of these cells was 2.8 m/sec (Fig. 5). Four such cells were tested for their sensitivity to the intravenous administration of apomorphine up to 0.3 mg/kg and found unresponsive. The vast majority of type II cells could not be activated antidromically (157 cells tested) by stimulating the caudate-putamen. The two types of antidromically activated cells could also be readily distinguished by their P S T H obtained during repetitive stimulation at 1 Hz. The firing of type I cells was suppressed for a considerable period starting 4-6 msec after the stimulus (mean ± S.E.M. for 12 cells: 149 4- 26 msec). The firing of type I I cells was suppressed for a much shorter period, generally interrupted by a very intense orthodromic activation (mean duration of the inhibition and S.E.M. for 7 cells: 61 q- 14 msec; one other cell was not inhibited; mean latency of the orthodromic activation and S.E.M. for 6 cells: 23.7 + 2.2 msec, two other cells were not activated). A representative example of both types of PSTH is given in Fig. 7.

78

Fig. 6. Type II cells antidromically activated from the striatum or from the thalamus. A: this type 1I cell (spontaneous firing rate 13 spikes/sec) was antidromically activated from the striatum. The stimulus (2.5 mA, 200 msec) was delivered 3 msec (upper trace) or 2.5 msec (lower trace) following a spontaneous spike (10 sweeps). The initial part of the artefact was suppressed, a stands for antidromic spike, s for spontaneous spike, arrow points at beginning of stimulus. B: this type II cell was activated from the nucleus ventralis medialis thalami. The cell (spontaneous first rate: 50spikes/sec) was stimulated with 200/~sec pulses (0.35 mA) delivered 1.6 msec (upper trace, ten sweeps) or 1.3 msec (lower trace, ten sweeps) following a spontaneous spike. Calibration for all traces: 1 msec, 0.2 mV.

(4) Electrical stimulation of the ventromedial nucleus of the thalamus In 6 animals, a s t i m u l a t i n g electrode was i m p l a n t e d in o r within 0.3 m m o f the nucleus ventralis medialis t h a l a m i ; the extracellular action potentials o f b o t h type I a n d t y p e II cells were then r e c o r d e d in the SN from its medial to lateral margins. N o t y p e I cell could be A D - a c t i v a t e d o u t o f a total o f 59 even when c u r r e n t s as high as 5 m A were used. By c o n t r a s t m o s t type II cells could be driven a n t i d r o mically (92 o u t o f 115) with pulses o f 200 #sec d u r a t i o n (l Hz, 0.15-1.5 mA). The latency o f the response was 0.85-5 msec (mean ~ S.E.M. for 73 cells: 2.1 ~ 0.1 msec). This result is essentially in a g r e e m e n t with D e n i a u et al. 12. N o I S - S D dissociation was ever observed. In each case, the response occurred at a fixed latency, a n d collision with s p o n t a n e o u s spikes could be o b t a i n e d (Fig. 6). T h e collision interval was identical to the latency o f the a n t i d r o m i c response within 0.2 msec. These cells were inhibited by i o n t o p h o r e t i c a l l y a p p l i e d G A B A (30-100 n A ) b u t their firing rate was n o t altered by D A (up to 100 nA, 10 cells tested). In two animals, two stimulating electrodes were s i m u l t a n e o u s l y i m p l a n t e d , one in the c a u d a t e nucleus a n d the o t h e r in the Vm. In these experiments a t o t a l o f 7 type l I cells could be activated a n t i d r o m i c a U y from the c a u d a t e - p u t a m e n . Five o f t h e m

79

B

A

C

..... J ~ a u l u ~ J t u u _ J !

.5 s ~ Fig. 7. PSTHs of type I and type II cells activated from the caudate-putamen or the thalamus. A: this type I cell (spontaneous firing rate 3.5 spikes/sec) was activated antidromically from the caudateputamen with a latency of 15 msec. Top trace: control; lower trace: stimulation at 1 Hz (1.5 mA, 500/tsec pulses). B: this type II cell (spontaneous firing rate 13 spikes/sec) was activated antidromically from the caudate-putamen with a latency of 1.7 msec. Top trace: control. Lower trace: stimulation at 1 Hz (500 ktsec pulses 0.35 mA). C: this type II cell (spontaneous firing rate 62 spikes/sec) was activated antidromically from the thalamus with a latency of 2.2 msec. Top trace: control; bottom trace: stimulation at 1 Hz (1.1 mA, 200/~sec pulses). In A, B and C the PSTH was obtained with the superposition of 50 stimuli.

could also be antidromically driven from the Vm using rather high intensities (0.2 msec, 0.75 to 2 mA).

(5) Electrical stimulation of the motor cortex, the globus pallidus, the pontine reticular formation and the medial forebrain bundle In 4 animals a stimulating electrode was implanted at a depth of 1.3 mm below the pial surface in a zone corresponding to the motor cortex as defined by Hall and Lindholm 2z. A total of 67 type I and 108 type II ceils were recorded in the substantia nigra from its medial to its lateral edge. No cell of either kind could be antidromically driven even with currents as high as 3 mA and pulses of 0.5 msec duration. Many type II cells were orthodromically driven with latencies of either 4-7 msec or 22-30 msec. Type I cells were not orthodromically activated. In one rat, a stimulating electrode was implanted in the globus pallidus (coordinates A, 5800/~m; L, 3200 /~m; H, --1200/~m) through which the nigrostriatal fibers pass aT. Seven type I cells could be antidromically activated (total sampling: 24 cells). Type II cells were not tested in this experiment. In one animal, a stimulating electrode was implanted in the middle of the nucleus reticulopontis (coordinates = P, 300 #m; L, 500 #m; H, --3700/~m) and a second electrode was placed in the Vm. No type I cell could be antidromically activated from the nucleus reticulopontis (17 cells tested). Four type II cells out of 27 could be antidromically driven from this nucleus (latency: 1.2; 1.7; 3.5; 1.3 msec) with 200/~sec pulses delivered with an intensity of 0.7-1.5 mA. These cells were

80 TABLE 1 Selective interruption of axonal conduction in type I neurons following the injection of 6-OHDA in the vicinity of the MFB

A stimulating electrode was implanted in the dorsal part of the MFB (coordinates A, 4110/~m; L, 1500/~m; H--2500/~m) and extracellular recordings of single cell activity were made in the substantia nigra. A number of type I and type II neurons were sampled (figure between brackets), most of which could be antidromically activated. 6-OH-DA (6/tg, 2/~1) was then slowly injected close to the MFB between the stimulating and recording sites (coordinates A, 3300/~m; L, 1300/~m; H--2500 tan). Recording of antidromically activated cells was resumed in the substantia nigra 15 rain later. Exp.

Number of A D-activated cells and total sampling Before 6-OH-DA

1

2 3 4 Totals

After 6-OH-DA

Type 1

Type H

Type 1

Type II

10 8 5 11 34

6 4 -3 13

0 3 9 1 13

7 8 -3 18

(10) (9) (5) (13) (37)

(9) (7) (4) (20)

(5) (15) (19) (13) (52)

(10) (11) (10) (31)

located close to the crus cerebri. These four cells could also be driven antidromically from the thalamus with comparable latencies (1.5; 1.7; 1.9; 1.3 msec). In 5 rats, a stimulating electrode was implanted in the dorsal edge of the MFB (A, 4100 # m ; L, 1500 # m ; H, --2500 #m) and the extracellular action potentials of type I and type II cells were recorded. Most type I cells (43 out of 47) and a great many type II cells (35 out of 55) could be antidromically activated in these experiments with 300 #sec pulses delivered at 1 Hz with an intensity of 0.2-l.5 mA. In 4 of these animals, after a sufficient number of type I and type II cells had been sampled, 6-hydroxydopamine (6/zg, 2 #1) was slowly injected in the vicinity of the MFB between the stimulating electrode and the recording site using a microliter syringe fitted with a 28-gauge needle. (coordinates: A, 3300 # m ; L, 1300 # m ; H, --2500 #m). Fifteen to thirty minutes later, the recording of nigral cells was resumed. In Table I, four experiments are recorded in which upon histological examination, the injection of 6-hydroxydopamine was found adequately close to the MFB but did not cause any significant non-specific damage. Following the injection of the neurotoxin, a significantly lower proportion of type I cells could be activated antidromically (25 ?,~i instead of 9 2 ~ , P < 0.001, Z2 test). However, the same proportion of type II cells could still be activated (59 ~ instead of 65 ~ ; this difference was not significant). DISCUSSION Type I cells appear to be identical to the presumed D A neurons initially described by Bunney et al. in 1973 and in a number of subsequent publications (for a review see Bunney and Aghajanian 1976) 7. The present work provides further evidence that these cells constitute the DA neurons of the nigrostriatal pathway. Both previous

81 data and the new information in the present study which support this conclusion can be summarized as follows: (1) Type I cells are located mainly in the SN pars compacta (at least in the anterior two thirds of which this layer is clearly identifiable in the rat). Their location is thus identical to that of DA cell bodies as evidenced by histofluorescence histology 11. (2) Type I cells cannot be detected following the injection of 6-hydroxydopamine immediately above the SN 6. This procedure produced a disappearance of DA but not other neurons in the substantia nigra. (3) Various drugs modify their firing rate in a way which is entirely compatible with their effect on the release of DA in the striatum as demonstrated by neuropharmacological techniquesl,9,19 (neuroleptics, DA receptor agonists such as apomorphine and amphetamine). (4) Type I cells do project to the striatum in a topographical manner which parallels the organization of the dopaminergic projection as evidenced by histofluorescence and retrograde tracing techniques. Indeed these cells could be driven antidromically from the striatum but not from the thalamus or the cortex in the present study. (5) The electrophysiological characteristics of type I cells (slow conduction velocity, 0.58 m/sec, high threshold of excitability) are compatible with available data describing the axons of DA cells as fine and non-myelinated 23. (6) In the rat ane3thetized with chloral hydrate, the conduction velocity of the noradrenaline-containing neurons of the locus coeruleus was recently found to be strikingly similar to the figure reported here for DA axons (0.57 m/sec)L (7) The vast majority of the other spontaneously active cells found in the substantia nigra (type I cells) can be antidromically driven from the thalamus but type II cells cannot. (8) 6-Hydroxydopamine specifically blocks nerve conduction in type I cells but not type II axons. This neurotoxin was also found to block very rapidly and specifically axonal conduction in noradrenergic neurons 2. (9) Type II neurons could be antidromically activated from all areas of the brain where nigrostriatal DA fibers are known to pass (MFB, globus pallidus and caudateputamen) but none of these cells could be driven from the cortex, the thalamus and the pons. This, in our view, constitutes strong evidence that type I neurons are the DA cells of the nigrostriatal pathway as suggested previously by Bunney et al. 6. The above data in turn validates the classification of spontaneously active nigral neurons in two classes according to the criteria chosen in the present study (spike duration and shape, firing rate and inhibition by intravenous administration of low doses of apomorphine). The fact that a small minority of type II cells could be activated antidromically from the caudate nucleus raises once more the interesting possibility of the existence of a second, non-DA nigrostriatal pathway16,17,2L Indeed the characteristics of these cells are entirely different from those of the DA neurons. Their conduction velocity is 5 times that of type I cells (2.8 m/sec instead of 0.58), their excitability threshold

82 is lower and the antidromic spikes never exhibit the IS-SD dissociation characteristic of the dopaminergic cells. In addition their average firing rate is far in excess of that of type I cells and they can follow higher frequencies of stimulation (900 Hz vs. 350). Finally, most of these cells also could be antidromically activated from the thalamus while no type I cell could be. It is thus tentatively suggested that this fast conducting nigrostriatal pathway rather than the slow conducting DA neurons could be responsible for some of the short latency excitations usually observed in the striatum following the electrical stimulation of the SN10.~6,32,33. Our results also suggest that the nigrostriatal non-DA neurons may send collaterals to the thalamus: however, further studies will be necessary to eliminate the possibility of current spread from the thalamus to the internal capsule where nigrostriatal fibers (including the more excitable non-DA axons) are presumably located. The present study demonstrates, however, that type II cells located in the pars reticulata do project both to the ventromedial thalamus (or at least run through this structure) and to the pons and thus exhibit some degree of branching. On the other hand, the existence of a nigrocortical projection has been described with a number of techniques in the cat 4,31. At this stage, it is impossible to rule out completely the possibility that the type lI cells that could be antidromically driven from the caudate-putamen in the present study were in fact nigrocortical neurons whose axons were stimulated via the scattered bundles that compose the internal capsule of the rat. The possibility that the cells recorded from in the present study might be nigrocortical cells is certainly compatible with the finding of MolinaNegro 3a. Indeed, most of the antidromically driven type II cells were found in the vicinity of the compacta in the present study. However, no evidence was found in the present study for the existence of a nigrocortical pathway; when horseradish peroxidase was ejected in the frontal cortex of the rat, it was not retrogradely transported to the SN except when the diffusion of the enzyme had been such as to contaminate the caudate-putamen. Similarly negative results were obtained with injections of horseradish peroxidase in the prefrontal cortex of the rat (Bunney, personal communication). Still, it is conceivable that in the rat, this nigrocortical projection might innervate cortical areas other than the prefrontal and frontal cortex and thus the problem is in need of further investigation. In conclusion, most spontaneously active cells in the SN of the rat anaesthetized with chloral hydrate are output neurons. DA cells present several characteristics that permit them to be distinguished from other cell types in the area. Most other spontaneously active cells belong to the thalamic projection which is not dopaminergic. Some non-DA cells were also shown to project to the pons and to the telencephalon. Further experiments will be required to determine whether this non-DA telencephalic projection innervates the striatum or another telencephalic structure. ACKNOWLEDGEMENTS We thank N. Margiotta and A. Lorette for their excellent technical assistance and L. Fields for typing this manuscript.

83 The work was supported by USPHS Grant MH-17871, the State of Connecticut, and the CNRS, France (salary to P. Guyenet).

REFERENCES 1 Aghajanian, G. K., Bunney, B. S. and Kuhar, M. J., Use of single unit recording in correlating transmitter turn-over with impulse flow in monoamine neurons. In A. J. Mandell (Ed.), New Concepts in Neurotransmitter Regulation, Plenum Press, New York, 1973, pp. 115-135. 2 Aghajanian, G. K., Cedarbaum, J. M. and Wang, R. Y., Evidence for norepinephrine-mediated collateral inhibition of locus coeruleus neurons, Brain Research, 136 (1977) 570-577. 3 Aghajanian, G. K. and Wang, R. Y., Habenular and other midbrain raphe afferents demonstrated by a modified retrograde tracing technique, Brain Research, 122 (1977) 229-242. 4 Avendano, C., Reinoso-Suarez, F. and Llamas, A., Projections to gyrus sigmoidus from the sub stantia nigra as revealed by horseradish peroxidase retrograde transport technique, Neurosci. Lett., 2 (1976) 61-65. 5 B6dard, P., Larochelle, H., Parent, A. and Poirier, L. J., The nigrostriatal pathway: a correlative study based on neuroanatomical and neurochemical criteria in the cat and the monkey, Exp. NeuroL, 25 (1969) 365-377. 6 Bunney, B. S., Waiters, J. R., Roth, R. H. and Aghajanian, G. K., Dopaminergic neurons: Effect of antipsychotic drugs and amphetamine on single cell activity, J. PharmacoL exp. Ther., 185 (1973) 560-571. 7 Bunney, B. S. and Aghajanian, G. K., Dopaminergic influence in the basal ganglia: Evidence for striatonigral feed-back stimulation. In M. D. Yahr (Ed.), The Basal Ganglia, Raven Press, New York, 1976, pp. 249-267. 8 Carpenter, M. B., Anatomical organization of the corpus striatum and related nuclei. In M.D. Yahr (Ed.), The Basal Ganglia, Raven Press, New York, Vol. 55, 1976, pp. 1-36. 9 Carlsson, A. and Lindqvist, M., Effect of chtorpromazine or haloperidol on the formation of 3-methoxytyramine and normetanephrine in mouse brain, Acta pharmacol. (Kbh.), 20 (1963) 140-144. 10 Connor, J. D., Caudate nucleus neurons: correlation of the effect of substantia nigra stimulation with iontophoretic dopamine, J. Physiol. (Lond.), 208 (1970) 691-703. 1l DahlstrSm, A. and Fuxe, K., Evidence for the existence of monoamine-containing neurones in the central nervous system, Acta physioL scand., 62, Suppl. 232 (1964) 1-55. 12 Deniau, J. M., Feger, J. and Le Guyader, C., Striatal evoked inhibition of identified nigro-thalamic neurons, Brain Research, 104 (1976) 152-156. 13 Dray, A., G/Snye, T. J. and Oakley, N. R., Caudate stimulation and substantia nigra activity in the rat, J. Physiol. (Lond.), 259 (1976) 825-849. 14 Dray, A. and Staughan, S. W., Synaptic mechanisms in the substantia nigra, J. Pharm. Pharmacol., 28 (1976) 400-405. 15 Faull, R. L. M. and Mehler, W. R., Studies of the fiber connections of the substantia nigra using the method of the retrograde transport of horseradish peroxidase. In Society for Neuroscience Sixth Annual Meeting, Vol. 2, part 1, Society for Neuroscience, Bethesda, 1976, Abstract 84, pp. 62. 16 Feltz, P. and de Champlain, J., Persistence of caudate unitary responses to nigral stimulation after destruction and functional impairment of the striatal dopaminergic terminals, Brain Research, 43 (1972) 595-600. 17 Fibiger, H. C., Pudritz, R. E., McGeer, P. L. and McGeer, E. G., Axonal transport in nigrostriatal and nigrothalamic neurons: effects of medial forebrain bundle lesions and 6-hydroxydopamine, J. Neurochem., 19 (1972) 1697-1708. 18 Frigyesi, T. L. and Machek, J., Basal ganglia-diencephalon synaptic relations in the cat. II. IntraceUular recordings from dorsal thalamic neurons during low-frequency stimulation of the caudato-thalamic projection system and the nigrothalamic pathway, Brain Research, 27 (1971) 59-78. 19 Glowinski, J., Axelrod, J. and Iversen, L. L., Regional studies of catecholamines in the rat brain. IV: effect of drugs on the disposition and metabolism of 3 H-dopamine, J. Pharmacol., 153 (1966) 30-41. 20 Graham, Jr. R. C. and Karnovsky, M. J., The early stages of absorption of horseradish peroxi-

84

21 22 23

24 25 26 27 28 29 30 3l 32 33 34 35

36 37 38

39 40

dase in the proximal tubules of mouse kidney; ultrastructural cytochemistry by a new techniquc. J. Histochem., 14 (1966) 291-302. Groves, P. M., Wilson, C. J., Young, S. J. and Rebec, G. V., Self inhibition by dopaminergic neurons, Science, 190 (1975) 522-529. Hall, R. D. and Lindholm, E. P., Organization of motor and somatosensory neocortex in the ~lbino rat, Brain Research, 66 (1974) 23-28. H6kfelt, T. and Ungerstedt, U., Specificity of 6-hydroxydopamine induced degeneration of central monoamine neurons: an electron and fluorescence microscopic study with special reference to intracerebral injection on the nigro-striatal dopamine system, Brain Research, 60 (1973) 269 297. Hopkins, D. A. and Niessen, L. W., Substantia nigra projections to the reticular formation and central gray in the rat, cat and monkey, Neurosci. Lett., 2 (1976) 253-259. Kaelber, W. W., and Afifi, A. K., Nigro-amygdaloid fiber connections in the cat, Amer. J. Anot., 148 (1977) 129-135. Kitai, S. T., Sugimori, M. and Kocsis, J. D., Excitatory nature of dopamine in the nigro-caudate pathway. Exp. Brain Res., 21 (1976) 351-362. K6nig, J. F. R. and Klippel, R. A., The Rat Brain: ,4 Stereotaxic Atlas, Krieger Publ., Huntington, New York, 1970. LaVail, J. H. and La Vail, M. M., Retrograde axonal transport in the central nervous system, Science, 176 (1972) 1416-1417. Ljungdahl, A., Hokfelt, T., Goldstein, M. and Park, D., Retrograde peroxidase tracing of neurons combined with transmitter histochemistry, Brain Research, 84 (1975) 313-319. Mettler, F. A., Nigrofugal connections in the primate brain, J. comp. Neurol., 138 (1970) 291-320. Molina-Negro, P., Connexiones nigro-corticales, Ann. Anat., 18 (1969) 285-354. Purpura, D. P., Physiological organization of the basal ganglia, In M. D. Yahr (Ed.), The Basal Ganglia, Raven Press, New York, 1976, pp. 91-114. Richardson, T. L., Miller, J. J. and McLennan, H., Mechanisms of excitation and inhibition in the nigrostriatal system, Brain Research, 127 (1977) 219-234. Rinvik, E., Grofova, I., and Ottersen, O. P., Demonstration of nigrotectal and nigroreticular projections in the cat by axonal transport of proteins, Brain Research, 112 (1976) 388-394. Siggins, G. R., Hoffer, B. J., Bloom, F. E. and Ungerstedt, U., Cytochemical and electrophysiological studies of dopamine in the caudate nucleus. In M. D. Yahr (Ed.), The Basal Ganglia, Vol. 55, Raven Press, New York, 1976, pp. 227-248. Thomas, R. C. and Wilson, V. J., Precise localization of Renshaw cells with a new marking technique, Nature (Lond.), 206 (1965) 211-213. Ungerstedt, U., Stereotaxic mapping of the monoamine pathways in the rat brain, Acta physiol. scand., 267 (1971) 1~,8. Waiters, J. R., Bunney, B. S. and Roth, R. H., Pirebedil and apomorphine; Pre- and postsynaptic effects on dopamine synthesis and neuronal activity, In D. B. Calne, T. N. Chase and A. Barbeau (Eds.), Advances in Neurology, Vol. 9, Raven Press, New York, pp. 273-284. Wang, R. Y. and Aghajanian, G. K., Inhibition of neurons in the amygdala by dorsal raphe stimulation: mediation through a direct serotoninergic pathway, Brain Research, 120 (1977) 85-102. Wang, R. Y. and Aghajanian, G. K., Recording of Single Unit Activity During Electrical Stimulation and Microiontophoresis : A Method of Minimizing Stimulus Artifacts, Electroenceph. clin. NeurophysioL, 43 (1977) 434~,37.