Development of spontaneous neuronal activity in the caudate nucleus, globus pallidus-entopeduncular nucleus, and substantia nigra of the cat

Developmental Brain Research, 3 (1982) 429--441

429

Elsevier Biomedical Press

D E V E L O P M E N T OF SPONTANEOUS N E U R O N A L ACTIVITY IN T H E C A U D A T E NUCLEUS, GLOBUS P A L L I D U S - E N T O P E D U N C U L A R NUCLEUS, A N D SUBSTANTIA NIGRA OF T H E CAT

M. S. LEVINE, R. S. FISHER, C. D. HULL and N. A. BUCHWALD Mental Retardation Research Center, Brain Research Institute, School of Medicine, University of California, Los Angeles, CA 90024 (U.S.A.)

(Accepted August 2nd, 1981) Key words: development of basal ganglia -- spontaneous firing -- caudate nucleus - - globus

pallidus--entopeduncular nucleus - - substantia nigra

SUMMARY Spontaneous single unit activity was obtained from caudate (Cd), globus pallidus-entopeduncular nucleus (GP-Ento), and substantia nigra (SN) neurons in kittens of 1-60 days of age and adult cats. Five developmental trends were found in the spontaneous firing patterns of these neurons: (1) overall mean interspike intervals (ISis) decreased with age; (2) the occurrence of neurons with shorter mean ISis ( < 400 ms) increased with age; (3) the occurrence of neurons with burst activity increased with age; (4) burst activity became more complex with age; and (5) the rate of burst occurrence in neurons with burst activity increased with age. Neurons within each region of the basal ganglia had characteristic patterns of spontaneous activity. Furthermore, the developmental patterns of spontaneous neuronal activity were different in each structure. The spontaneous activity of GP-Ento and SN neurons matured before the spontaneous activity of Cd neurons. Thus, spontaneous firing may mature in the output nuclei of the basal ganglia prior to its maturation in the Cd.

INTRODUCTION This study is part of a continuing investigation of the electrophysiological and anatomical development of neurons in the basal ganglia of the cat. Previous reports elucidate the ontogeny of responses in striatal, pallidal, and nigral neurons to activation of their major afferents14,15,17, is. In adult cats under equivalent recording conditions, striatal and pallidal neurons have characteristic levels of spontaneous activity. Neurons in the caudate nucleus (Cd) fire at low rates ( < 1 spike/s) 3,7,1°,11. 0165-3806/82/0000-0000/$02.75 © Elsevier Biomedical Press

430 Neurons in the globus pallidus and entopeduncular nucleus (GP-Ento) fire faster (10-100 spikes/s)14,16. The spontaneous activity of cat substantia nigra (SN) neurons has not been described previously. In the adult rat, it has been shown that dopaminergic SN neurons fire more slowly (1-4 spikes/s) than non-dopaminergic neurons (4-10 spikes/s)5,8,21, 23. The present study was designed to describe the development of patterns of spontaneous neuronal activity in the Cd, GP-Ento, and SN and to compare the characteristics of spontaneous firing in the neurons of these structures. MATERIALS AND METHODS

Animals Extracellular single unit records of spontaneous activity were obtained from Cd, GP-Ento, and SN neurons of 83 kittens and 21 adult cats ( > 1 year old) obtained from the UCLA Mental Retardation Research Center cat breeding colony (Table I). The day of birth was considered as the first postnatal day. Age groups were chosen on the bases of morphological and previous physiological data which show rapid maturation during the first 3 weeks followed by slower changes that occur over several postnatal months1,2,14,15,17. Because of differences in the time-course of the development of spontaneous neuronal activity in the Cd, GP-Ento, and SN, animals were classified into 4 age groups for Cd neurons (1-20 days, 4-6 months, 8-11 months, and adults) and 3 age groups for GP-Ento and SN neurons (1-20 days, 21-60 days, and adults).

Surgery Surgical procedures have been detailed elsewhere 14,15,17. Briefly, anesthesia was induced in kittens with 2 ~ halothane-N20-O2 respiratory mixtures. Adult cats were anesthetized initially with sodium brevital, a short-acting barbiturate. Anesthesia was maintained in all animals by the continuous use of the respiratory mixture during the course of each experiment. All animals were paralyzed with gallamine triethiodide. Data were obtained from only one nucleus in any animal (Cd, GP-Ento, or SN). In cats older than 30 days, recordings were made simultaneously from both caudate nuclei.

Recording procedures Recording electrodes were glass micropipettes filled with 1.6 M potassium citrate (5-30 Mf~ resistance). Extracellular single unit neuronal activity was isolated, amplified, monitored on an oscilloscope, and recorded on FM tape for subsequent computer analysis. Data were obtained from 5-30 neurons per animal. Spontaneous activity was recorded from each neuron for 3-5 min. In animals in the youngest age group there were often variations in spike amplitude. For computer analysis, spikes of markedly different heights were analysed separately by using a window discriminator in the computer program. The discriminator also displayed spike waveform to provide an additional index to facilitate discrimination when more than one unit appeared on the same record.

431

Histology At the end of each experiment, recording electrode tip sites were marked by small electrolytic lesions. Animals were killed with an overdose of sodium pentobarbital. Brains were removed and fixed by immersion in buffered 10 ~o formalin. Recording electrode placements were determined by reconstruction from serial frozen sections stained with cresyl violet. Data analysis Three measurements of spontaneous neuronal activity were madeT: (1) the mean of all interspike intervals (ISis) generated by each neuron was determined. These means were averaged to produce a mean for all neurons recorded in each animal. These latter means were then averaged for all animals in each age group to produce an overall mean ISI for each nucleus of the basal ganglia. This measure allowed assessment of central tendency; (2) frequency distributions of mean ISis for all neurons in all animals of a given age group were also constructed for each of the 3 recording sites. These distributions provided a measure of the variation of mean ISis in neurons; (3) the occurrence of very fast firing patterns of 'bursts' was measured for each neuron. A burst was defined as at least 2 sequential spikes with an ISI of < 10 ms. Bursts were classed according to the sequential occurrence of ISis of < 10 ms. A neuron had a 1interval burst if at least 1 ISI of < 10 ms occurred. A neuron was considered to have both 1- and 2-interval bursts if there were at least 2 consecutive ISis of < 10 ms. The most complex bursts analyzed were 4-interval bursts. This analysis was stepwise and inclusive (i.e. neurons with 4-interval bursts also were considered to have 3-, 2-, and 1interval bursts). Both the frequency of occurrence of burst types and the rate of bursting were determined for each neuron. Burst rate was calculated by dividing the frequency of occurrence of bursting by the duration of the neuronal record (bursts/ min). Differences between overall mean ISis were assessed with appropriate analyses of variance and a posteriori contrasts 24. Differences between distributions of mean neuronal ISis and burst parameters for each group were determined with Z2-tests. Values given in the results are means ± S.E. unless otherwise noted. RESULTS Extracellular single unit records of spontaneous neuronal activity were obtained from 738 Cd neurons, 138 GP-Ento neurons, and 482 SN neurons in 83 kittens and 21 adult cats (Table I). In all neurons, the signal:noise ratio was better than 3:1. The durations of differentiated action potential waveforms tended to decrease with age in all nuclei of the basal ganglia (3.0-8.0 ms in 1-20 day old kittens vs 1.5-5.0 ms in adults for SN neurons).

Overall mean ISis The overall mean ISis of Cd, GP-Ento, and SN neurons decreased with age (P < 0.05, Fig. 1). Comparisons in 1- to 20-day-old kittens and adults showed that the

432 TABLE I

Numbers o f cats and units in each group Recording site Age

Caudate

1-20days 21-60 days 4-6months 8-11 m o n t h s Adults Totals

n (cats)

n (units)

27 -5 5 9 46

106 -113 185 334 738

L

Globuspallidusentopeduncular nucleus

Substantia nigra

n (cats)

n (units)

n (cats)

Total n (units)

Compacta Reticulata n (~nits) n (units)

40 53

12 18

157 245

93 126

64 119

55 274

25 208

8 8

.

.

.

.

.

.

.

.

.

.

.

.

7 23

45 138

5 35

80 482

CAUDATE

4000

2000

u

GLOBUS PALLIDUSENTOPEDUNCULAR

E 600

z

n

200

250C

BSTANTIA NIGRA

I50C

50C 1-20 days

21-60 days

4-6 mos

8-11 Adult mos (>12mos)

Fig. 1. Overall m e a n ISis ( ± S.E.) of Cd (top), G P - E n t o (center), a n d S N n e u r o n s (bottom) plotted as a function o f age. N o t e the differences in the scale of the ordinate for each nucleus.

433 overall mean ISis of the Cd, SN, and GP-Ento neurons decreased by 60-70 % during postnatal development (Cd: 4738 -4- 629 ms vs 1955 ± 226 ms, GP-Ento: 889 -4- 164 ms vs 346 -}- 63 ms, and SN: 2393 -/- 388 ms vs 730 ± 160 ms, respectively). Cd and GP-Ento neurons failed to show clear ontogenetic shifts in the 1- to 20-day period, but the overall mean ISis of SN neurons decreased significantly between 1-10 and 11-20 days of age (2943 i 443 ms vs 1295 -4- 364 ms, respectively; Fig. 5). At each age level that was directly comparable, the mean ISis of GP-Ento neurons were significantly shorter than those of SN neurons which in turn were shorter than those of Cd neurons (P < 0.05). In terms of mean ISis, there was a tendency for SN neurons to mature more rapidly than GP-Ento neurons ( < 6 0 days vs > 6 0 days) and for the latter to mature more rapidly than Cd neurons ( > 11 months).

Distributions o f mean ISis The distributions o f mean ISis were associated reliably with age for Cd, GPEnto, and SN neurons. There was a higher relative frequency of neurons with long mean ISis in kittens and a higher relative frequency of neurons with short mean ISis in adults (Fig. 2). CAUDATE

GLOBUSPALLIDUSENTOPEDUNCULAR NUCL.

SUBSTANTIA NIGRA

50 30

10

30

10 _

IZ LU ~.) rr W (I.

30

10

30

8-11 MONTHS

10 ADULT

ADULT

30 IC 200

600 t000 J400 1800 >2000 200 600 iO00 1400 1800 >2000 200 600 J000 1400 1800 >2000

MEAN ISl (msec) Fig. 2. Distributions of mean ISis of Cd (left), GP-Ento (center), and S N neurons (right) plotted as a function of age. Bin size = 200 ms. Ordinate represents percentage of total number of units in each bin.

434 The distributions of mean ISis in these neurons accounted for the differences in the overall mean ISis detailed above. Striatal neurons had broad distributions of mean ISis, and over half of these cells had mean ISis of > t000 ms throughout development. The distribution of mean ISis in Cd neurons was not adultlike until after 11 months of age. In contrast, GP-Ento neurons had a broad distribution of mean ISis in 1- to 20day-old kittens that skewed toward mean ISis of < 4 0 0 ms by 21-60 days. The latter distribution was similar to that seen in adult cats. The distribution of mean ISis of SN neurons was broad in 1- to 20-day-old kittens. It skewed to an adult profile where > 8 0 ~ of the mean ISis were < 6 0 0 ms by 21-60 days. For distributions of mean ISis, SN and GP-Ento neurons matured faster than Cd neurons. These trends were also apparent in the records of individual neurons. Fig. 3 consists of segments of recorded activity and ISI distributions for characteristic neurons in the basal ganglia of kittens and adult cats. The SN neurons were identified antidromically as nigrostriatal output neurons. In both kittens and adults, the GPEnto neurons fired most rapidly, SN neurons were intermediate, and the firing rates of Cd neurons were slowest. Spontaneous firing was more rapid in adults than in the kittens.

Burst activity Burst activity showed 3 changes as a function of age: the proportion of bursting neurons in each nucleus increased, the number of action potentials in each burst increased, and the rate of burst occurrence increased. As shown in Table II, at least 25 ~/o of the neurons in each nucleus had burst activity in each age group. Between 1- to 20-day-old kittens and adults, the proportions of Cd neurons with l-interval bursts increased by 32 70 while the proportions of GP-Ento and SN neurons increased by 43 ~o and 27 ~ , respectively. In adults, relatively more GP-Ento than SN neurons had linterval bursts and relatively more SN than Cd neurons had 1-interval bursts. Adult values were obtained by the 21- to 60-day period for SN neurons, after 60 days of GPEnto neurons, and by 8-11 months for Cd neurons. Since no record of neurons with 5 or more consecutive ISis of ~ 10 ms were obtained, all sequential burst patterns in these neurons were described as 1-, 2-, 3-, and 4-interval bursts. Table II shows that the proportions of Cd, GP-Ento, and SN neurons with 2-interval bursts increased significantly with age. Between 1- to 20-dayold kittens and adults increases for l-interval bursts occurred (28 ~ for Cd neurons, 27 ~o for GP-Ento neurons, and 32 ~ for SN neurons). More than 50 ~/o of the neurons with bursts in adults had 2-interval bursts. In adults, similar proportions of GP-Ento and SN neurons had 2-interval bursts (47 ~ and 43 ~o, respectively) while 32 70 of the Cd neurons had 2-interval bursts. Adult proportions of neurons with 2-interval bursts were attained before 60 days of age in the SN, after 60 days in GP-Ento, but only after 11 months of age in the Cd. Low proportions of Cd, GP-Ento, and SN neurons had 3- or 4-interval bursts in kittens of 1-20 days of age (0 ~ , 13 ~o, and 9 ~o, respectively). In adults, similar proportions of GP-Ento and SN neurons had 3- or 4-interval bursts while a smaller proportion of Cd neurons had these burst patterns. Adult levels of neurons with 3- or

435

CD .L..

~L

E

i_

f

Xls i = 4 8 6 7 rnsec

SN

sd = 8 8 4 5 msec 2 days

~lSl = 2 5 0 9 msec

GP

sd = 4 8 2 8 msec Anlidromic - 2 doys

I

l i J

tL]

1

I I [

Ill Ill II

Lit

I L

KITTENS

XIS I = 1 0 2 2 msec s . d : 4 3 3 msec 2 doys

, i,,,,~,,,,,,,

...................

~,,,,,

IIII11'""' ,,,,,,

I~i!ii

I II

'

il

IIII

I,I

Ill

lit

ADULTS

-XlSl

= 3 2 7 msec sd = 4:52 msec

Xtsl = 713 msec sd = 7 4 7 msec

XiSl = 1 9 9 8 rnsec sd = 1 7 7 9 msec

Antidromic

10%

5OO mse¢

Fig. 3. Oscillographic records (upper trace) and computer-generated ISI distributions (lower trace) of characteristic GP-Ento (left), SN (center), and Cd (right) neurons in kittens and adult cats. In each distribution, shorter intervals are represented on the left and longer intervals on the right (50 ms bin width, last bin is an overflow bin containing all ISis >2000 ms). Ordinate represents percentage of total number of interval in each bin. SN neurons were identified antidromically as nigrostriatal projection neurons.

436 T A B L E II

Development Of burst activity 1-Interval burst = 1 ISI < 10 m s ; 2-interval burst - 2 successive ISis <_ 10 m s ; 3-interval burst ::: 3 successive ISis _< 10 m s ; 4-interval burst : 4 successive ISis <_ 10 ms. D a t a given as percentage of units bursting.

Recording site

Age

Burst type 1-Interval

2-Interval

3-Interval

4-Interval

CD

1-20 days 4-6 months 8-11 m o n t h s Adult

25 45 55 57

4 17 18 32

0 4 9 16

0 4 3 l1

GP-ENTO

1-20 days 21-60 days Adult

35 40 78

20 11 47

13 0 28

0 0 28

SN

1-20 days 21-60 days Adult

38 71 65

11 47 43

6 32 31

3 29 24

36

GLOBUS PALLIDUS-~ ENTOPEDUNCULAR NUCLEUS

CAUDATE

......... SUBSTANTIA NIGRA

32 E 4-1

28 24

L

z 20 03

03 £E" CI3

16

T T

12

IX 8

T

4

1-20 21-60 DAYS D A Y S

4-6 MOS.

8-11 MOS

ADULT

1-20 21-60 DAYS D A Y S

4-6 MOS

8-11 MOS

ADULT

1-20 21-60 DAYS DAYS

4 -6 MOS

t _1 8 -11 MOS

ADULT

AGE

Fig. 4. M e a n rates o f 1-interval bursts ( b u r s t s / m i n ± S.E.) plotted as a function of age for Cd (left), G P - E n t o (center), a n d S N n e u r o n s (right).

437

A J

T

i

T

,

0 PARS

i

COMPACTA

3ooc

u~ E v)

200C

Z

LU

1ooo

1210

11-120

211-40 AGE( d o y s )

41'-60

Aduli'

B PARS COMPACTA

IPAI~S I~TICULATA g-e a ~ s

,o

11t

a:

-

-3

AI~ULT

I~AN IS~ (moec)

Fig. 5. A : overall m e a n ISis (4- S.E.) o f pars c o m p a c t a a n d pars reticulata S N n e u r o n s plotted as a function o f age. B: distributions o f m o a n ISis o f pars c o m p a c t a a n d pars reticulata n e u r o n s plotted as a function o f age. Bin size = 200 ms.

438 4-interval bursts were reached by 60 days of age in the SN, after 60 days of age in the GP-Ento, and after 11 months in the Cd. Between 1- to 20-day-old kittens and adults, l-interval burst rates increased by approximately 6 bursts/min, 15 bursts/min, and 13 bursts/rain for Cd, GP-Ento, and SN neurons, respectively. In adults, l-interval burst rates were highest in the GP-Ento, intermediate in the SN, and lowest in the Cd. Adult 1-interval burst rates were reached before 60 days of age in the SN, after 60 days in the GP-Ento, and after 11 months in the Cd. The rates of 2-interval bursts also tended to increase with age in the nuclei of the basal ganglia. This trend was not maintained for neurons with 3- or 4-interval bursts because of the relatively low incidence of such bursts in all age groups.

Spontaneous neuronal activity in the pars compacta and pars reticulata of the SN Histological reconstruction of recording sites showed that 274 neurons were located in the pars compacta, and 208 neurons were located in the pars reticulata of the SN. As we have reported previously, ~90~o of the pars compacta neurons had triphasic long-duration action potential waveforms, and ~ 80 ~ of the pars reticulata neurons had biphasic short-duration action potential waveforms 15. Most Cd and GPEnto neurons had action potentials that also were triphasic and of long-duration. The mean ISis of action potentials recorded from pars compacta neurons were longer than those of pars reticulata neurons at each age examined (Fig. 5A). These differences were statistically significant (P < 0.05) for all age groups except adults. Distribution of mean ISis for compacta and reticulata neurons were also different at all ages except adults (P < 0.05) (Fig. 5B). There were no apparent differences in the burst activity of pars compacta and pars reticulata neurons. Thus, the data from neurons in these areas were not analyzed separately. DISCUSSION

Development of spontaneous neuronal activity within the basal ganglia The results of this study show that spontaneous neuronal activity is present in the Cd, GP-Ento, and SN from the first day of postnatal life in the cat. Spontaneous neuronal activity has 5 major developmental trends in these nuclei: (1) interspike intervals decrease as a direct function of age; (2) proportionally more neurons with shorter mean ISis occur in adults than in young kittens; (3) the proportion of bursting neurons increases; (4) burst patterns become more complex with increasing age (more action potentials occur per burst in older animals); (5) in bursting neurons, the rate of burst occurrence increases with age. Spontaneous ISis and burst activity tend to mature simultaneously. Characteristic spontaneous activity in Cd, GP-Ento, and SN neurons While there are common trends in the development of spontaneous neuronal firing, each nucleus has characteristic activity. This has been shown previously in adult cats8,14-16, e3. Striatal neurons have the longest ISis and the least burst activity. Pallidal neurons have the shortest ISis and the greatest burst activity. Collectively,

439 nigral neurons have intermediate ISis and almost as much burst activity as pallidal neurons. Within the SN, ISis of pars reticulata neurons are very similar to those of pallidal neurons while the ISis of pars compacta neurons are nearly as long as those of striatal neurons. As shown in previous studies of the cat and the rat, pars compacta neurons usually have long-duration triphasic action potential waveforms while pars reticulata neurons have short-duration biphasic waveformss,21,23.

Developmental patterns of spontaneous neuronal activity in the nuclei of the basal ganglia For both spontaneous ISis and burst activity parameters, nigral neurons mature first, pallidal neurons next, and striatal neurons last. Thus, spontaneous neuronal activity is adult-like before 60 days of age in SN, after 60 days in the GP-Ento, and after 11 months in Cd. This indicates that spontaneous neuronal activity matures in the output nuclei of the basal ganglia (GP-Ento and SN) before it matures in the neostriatum. This suggests that, to a considerable extent, the spontaneous neuronal activity of the output nuclei develops independently of spontaneous neuronal activity in the neostriatum. Since the patterns of spontaneous neuronal activity have a markedly different time-course for Cd, GP-Ento, and SN neurons, it is improbable that the development of spontaneous activity is an artefact of an interaction of age with anesthetic, surgical, or recording procedures. If these interactions existed, the ontogenetic time-course of spontaneous neuronal activity would have been identical in all nuclei of the basal ganglia. Associations of arousal states with spontaneous neuronal activity are irrelevant in the present study because all animals were anesthetized. Factors in the development of spontaneous neuronal activity The current data indicate that the ISis and burst activity of striatal, pallidal, and nigral neurons display similar developmental shifts between 1-20 days of age and adulthood. These common trends suggest that there might be similar causes for the development of ISis in the 3 nuclei. Our studies on the development of electrophysiologically identified nigral output neurons show that maturational shifts in refractory periods are contemporaneous with the maturation of spontaneous neuronal activityxS. Refractory periods reach adult values (0.4-3.5 ms) at the same time (21-60 days) that spontaneous neuronal activity attains adult values in the SN. As refractory periods decrease with age, ISis also decrease (possibly because the neurons can generate action potentials at a higher rate). Similar trends may occur in Cd and GPEnto neurons although detailed measurements have not been made. The factors involved in the development of spontaneous activity in basal ganglia neurons are complex. A major factor, of course, is the influence of inputs to these neurons. For example, in both adult cats and kittens, partial denervation of striatum (by cortical ablation) clearly diminishes firing rates of Cd neurons 7. In terms of the proportions of responding neurons, initial response types (excitation or inhibition) and temporal characteristics of responses, action potentials evoked in Cd neurons are essentially mature by 50 days of age while spontaneous activity does not reach adult levels until after 11 months 17. Moreover, the extent to

440 which activation of the afferents produces postsynaptic excitation and inhibition is not clearly associated with spontaneous activity levels. In adults, stimulation of afferents to the Cd generates excitatory postsynaptic potentials (EPSPs) which are usually followed by inhibitory postsynaptic potentials (IPSPs)3,10,11. In the developing cat, IPSPs appear later than EPSPs. Thus, evoked IPSPs become more frequent in Cd neurons even though spontaneous activity increases with age. In contrast, GP-Ento and SN neurons of young kittens show both initial excitations and initial inhibitions after Cd stimulation. In the GP-Ento, the proportion of neurons with initial inhibitory responses increases with age until 180 days while spontaneous neuronal activity is mature by 60 days13,14. In the SN, only inhibitory evoked responses occur after 60 days of age while spontaneous activity matures earlierlS,1L The developmental increases of inhibitory influences on the neurons of the basal ganglia should diminish the level of spontaneous activity. Instead, spontaneous activity increases with age. It is possible that subthalamopallidal and subthalamonigral efferents may be primarily excitatory 4,9. These additional excitatory inputs could contribute to the high levels of spontaneous neuronal activity found in the GP-Ento and pars reticulata of the SN. Our results of increasing spontaneous firing rates during postnatal maturation are in agreement with other reports documenting such changes in visual cortex of kittens 2z, rat diencephalon t2, and cerebellum 2°. In contrast, a recent report indicates no change in spontaneous firing rates of neurons in the lateral hypothalamus or ventromedial hypothalamic nucleus in rats of 1-25 days 6. ACKNOWLEDGEMENTS This work was supported by U S P H S Grants H D 05958, 1T32MHI5345 and H D 04612.

REFERENCES 1 Adinolfi, A. M., The postnatal development of the caudate nucleus. A Golgi and electron microscopic study of kittens, Brain Res., 133 (1977) 251-266. 2 Adinolfi, A. M., Levine, M. S. and Hull, C. D., Early postnatal development of entopeduncular neurons and their neostriatal connections, Exp. NeuroL, 70 (1980) 463-474. 3 Buchwald, N. A., Price, D. D., Vernon, L. and Hull, C. D., Caudate intracellular response to thalamic and cortical inputs, Exp. Neurol., 38 (1973) 311-321. 4 Deniau, J. M., Hammond, C., Chevalier, G. and Feger, J., Evidence for branched subthalamic nucleus projections to substantia nigra, entopeduncular nucleus and globus pallidus, Neursci. Lett., 9 (1978) 117-121. 5 Dray, A., The physiology and pharmacology of mammalian basal ganglia, Neurobiology, 14 (1980) 221-336. 6 Fisher, R. S. and Almli, C. R., Postnatal ontogeny of hypothalamic extracellular unit activity in the rat, Neurosci. Abstr., 5 (1979) 159. 7 Garcia-Rill, E., Hull, C. D., Levine, M. S. and Buchwald, N. A., The spontaneous firing patterns of forebrain neurons. IV. Effects of bilateral and unilateral frontal cortical ablations on firing of caudate, globus pallidus and thalamic neurons, Brain Res., 165 (1979) 23-36. 8 Guyenet, P. G. and Aghajanian, G. K., Antidromic identification of dopaminergic and other output neurons of the rat substantia nigra, Brain Res., 150 (1978) 69-84.

441 9 Hammond, C., Deniau, J. M., Rizk, A. and Feger, J., Electrophysiological demonstration of an excitatory subthalamonigral pathway in the rat, Brain Res., 151 (1978) 235-244. 10 Hull, C. D., Bernardi, G. and Buchwald, N. A., Intracellular responses of caudate neurons to brain stem stimulation, Brain Res., 22 (1970) 163-179. 11 Hull, C. D., Bernardi, G., Price, D. D. and Buchwald, N. A., Intracellular responses of caudate neurons to temporally and spatially combined stimuli, Exp. NeuraL, 38 (1973) 324--336. 12 Hyarrinen, J., Analysis of spontaneous spike potential activity in the developing rabbit diencephalon, Acta physiol, scand., 68 (1966) 1-67. 13 Levine, M. S., Hull, C. D. and Buchwald, N. A., Pallidal and entopeduncular intracellular responses to striatal, cortical, thalamic, and sensory inputs, Exp. Neural., 44 (1974) 448--460. 14 Levine, M. S., Cherubini, E., Novack, G. D., Hull, C. D. and Buchwald, N. A., Development of responses of globus pallidus and entopeduncular nucleus neurons to stimulation of the caudate nucleus and precruciate cortex, Exp. Neural., 66 (1979) 479--492. 15 Levine, M. S., Adinolfi, A. M., Cospito, J. A., Fisher, R. S., Hull, C. D. and Buchwald, N. A., Anatomical and physiological studies on the development of substantia nigra in the cat, Anat. Rec., 199 (1981) 154A. 16 Noda, H., Manohar, S. and Adey, W. R., Responses of cat pallidal neurons to cortical and subcortical stimuli, Exp. NeuraL, 20 (1968) 585410. 17 Morris, R., Fuller, D. R. G., Hull, C. D. and Buchwald, N. A., Development of caudate neuronal responses to stimulation of the midbrain, thalamus, and cortex in the kitten, Exp. Neural., 57 (1977) 121-131. 18 Morris, R., Levine, M. S., Cherubini, E., Buchwald, N. A. and Hull, C. D., Intracellular analysis of the development of responses of caudate neurons to stimulation of cortex, thalamus, and substantia nigra in the kitten, Brain Res., 173 (1979) 471-487. 19 Preston, R. J., McCrea, R. A., Chang, H. T. and Kitai, S. T., Anatomy and physiology of substantia nigra and retrorubral neurons studied by extra- and intracellular recording and by horseradish peroxidase labeling, Neuroscience, 6 (1981) 331-344. 20 Puro, D. G. and Woodward, D. J., Maturation of evoked climbing fiber input to rat cerebellum Purkinje cells (I), Exp. Brain Res., 28 (1977) 85-100. 21 Ruffieux, A. and Schultz, W., Dopaminergic activation of reticulata neurons in the substantia nigra, Nature (Land.), 285 (1980) 240-241. 22 Wiesel, T. N. and Hubel, D. H., Single cell responses in striate cortex of kittens deprived of vision in one eye, J. NeurophysioL, 26 (1963) 1003-1017. 23 Wilson, C. J., Young, S. J. and Groves, P. M., Statistical properties of neuronal spike trains in the substantia nigra: cell types and their interactions, Brain Res., 136 (1977) 161-174. 24 Winer, B. J., Statistical Principles in Experimental Design, McGraw Hill, San Fransisco, 1971.