BRAIN RESEARCH

ELSEVIER Brain Research 697 (1995) 216-224

Research report

Effects of lesion of the inferior olivary complex in learning of the equilibrium behavior in the young rat during ontogenesis. I. Total lesion of

the inferior olive by 3-acetylpyridine

N. Jones a, T. Stelz a, C. Batini b j. Caston a, *

a UniversitO de Rouen Laboratoire de Neurophysiologie sensorielle Facultd des Sciences 76821 Mont-Samt-Aignan C6dex, France b Laboratoire de Physiologie de la motricit~, CHU Pitid-Salp~tri~re 91, Boulevard de l'H6pital 75013 Paris, France

Accepted 21 June 1995

Abstract

Young D A / H A N strain rats were submitted to an equilibrium test consisting in maintaining equilibrium upon a rotorod rotating at 10 or 20 rpm. They were either intact or lesioned, the lesion consisting in destruction of the inferior olivary complex (IOC) by 50-95 m g / k g i.p. administration of 3-acetylpyridine (3-AP) at day 15, followed, 2 to 4 h later, by i.p. injection of niacinamide (300 mg/kg) . All the 3-AP-treated animals included in this study were completely lesioned, the extent of the lesion being estimated by both the response of the rats to harmaline and histological controls at the end of the experiments. The IOC lesioned rats were either naive (tested at one given day) or trained every day (10 trials per day); among the latters, some were trained before and after the lesion, the others being trained either before or only after. Control rats were submitted to the same training schedule. Both quantitative (time during which the animals maintained the equilibrium upon the rotating rod) and behavioral data (strategy used by the animals to maintain equilibrium) were obtained. The results demonstrate that, compared to those of control rats, the quantitative and behavioral scores of the IOC lesioned animals were altered. Comparison of naive and trained animals shows that the impairment of the equilibrium behavior is not only due to the ataxia provoked by the IOC lesion but is also due to cognitive deficits. However, prelesion training facilitates the acquisition of a more efficient postlesion equilibrium behavior. From these results, it can be concluded that the olivo-cerebellar pathway is involved in the adaptation of motor behavior to the environmental conditions.

Keywords: Inferior olivary complex; Olivocerebellar pathway; 3-Acetylpyridine; Equilibrium behavior; Learning; Rat

I. Introduct ion

It is known for many years that the cerebellum is involved in classical conditioning, particularly in the con- ditioning of the nictitating membrane response [6,15,26,36,38,43] and in the adaptation of the vestibulo- ocular reflex to modifications of the visual field [10,19,21,29,40]. It is also involved in more complex motor learning [1,16,18,30,32,33,44], in instrumental learn- ing [9,11,13,17,25,34,42] as well as in the habituation of the acoustic startle response [27] and of the exploration behavior [12]. The cerebellum is also implicated in emo- tions (see [24], for review) and in man it would participate

* Corresponding author. Fax: (33) 35.14.63.49.

0006-8993/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0006-8993(95)00846-2

to language function [28] and all kinds of neural control, even mental control [20].

From Marr's theory [35] and Ito's model [19], cerebellar learning is based on modifiable synapses between parallel fibers and Purkinje cells of the cerebellar cortex when these cells receive conjoint inputs from climbing and parallel fibers. Moreover, electrophysiological experiments demonstrated that degeneration of the climbing fiber sys- tem due to lesion of the inferior olivary complex (IOC) induces functional modifications of Purkinje cell activity [2,4,5,7]. At last, it has been demonstrated that IOC lesion causes extinction of the classically conditioned nictitating membrane response [37] and impairs both learning and memory of spatial tasks [14]. It can therefore be assumed that the climbing fibers are essential in learning function of the cerebellum.

It has also been shown that the cerebellum is involved

N. Jones et al. /Bra in Research 697 (1995) 216-224 217

in the acquisition of motor skills during development [1,44]. Rats cerebellectomized by the 24th day, when the cerebellum is mature, and trained every day after the lesion, were able to maintain balance on a rotorod for a long time almost as well as controls [1]. By contrast, when cerebellectomy was done in young rats (by the 10th or the 15th postnatal day), when the cerebellum is still immature, the animals were unable to maintain balance for long on the rotorod even when they were trained every day after the lesion [1,44]. The cerebellum is obviously needed for the acquisition of motor skills, which once they are ac- quired can be conveyed in the absence of the cerebellum. In effect, in 15-day-old cerebellectomized rats, the postle- sion scores were greatly enhanced and the behavior of the rats on the rotating rod were similar to those of controls when the animals experienced a prelesion training on the rotorod which quantitatively and qualitatively increased the sensory cues used to maintain balance.

It can reasonably be assumed that the afferents imping- ing onto the Purkinje cells are involved in the learning function of the cerebellum. Given that the Purkinje cells of the cerebellar cortex receive two inputs, one from the climbing fibers, the other from the mossy fibers-granule cells-parallel fibers system, we wanted to know the re- spective role of both inputs in learning of the equilibrium behavior in the rat and, for now, the role of the climbing fibers since they have been shown to be essential in cerebellar learning. Preliminary unpublished results demonstrated that after degeneration of these fibers in adult rats, the animals were unable to maintain their equi- librium for long on a rotorod even when they were trained every day after the lesion, suggesting that, in adults, lesion of the olivocerebellar pathway has greater effects on bal- ance acquisition than destruction of the whole cerebellum. The aim of the present study was to see whether such a degeneration has the same effect when it was done in young animals whose cerebellum is still immature. Since the 15th postnatal day seems to be crucial in the develop- ment of the cerebellar mechanisms which sustain the equi- librium functions, lesions of the olivo-cerebellar system by 3-acetylpyridine (3-AP) was performed in 15 day-old ani- mals.

2. Materials and methods

2.1. Apparatus

The device used was a rotorod. It consisted in a wooden horizontal rod (3 cm in diameter, 40 cm long) covered with sticking plaster in order to increase roughness. It was rotated around its longitudinal axis at 10 or 20 revolutions per minute (rpm) by means of a DC electric motor. This rotorod was 18 cm above a landing platform covered with a thick sheet of soft plastic to cushion the fall of the animals.

Table 1 Different groups of rats and number of animals (in parentheses) in each group

Control groups IOC lesioned groups

NCN23 (10) NLN23 (6) NCN30 (6) NLN30 (6) NCT (10) NLT (6) TCN23 (15) TLN23 (I0) TEN30 (12) TLN30 (8) TCT (15) TLT (10)

NCN23 and NCN30: control rats which were naive and only tested at day 23 or at day 30, respectively; NCT: control rats which were naive before day 23 and trained every day from day 23; TCN23 and TCN30: control rats which were trained every day from day 10 to day 15 and which were not trained thereafter but only tested at day 23 or at day 30, respectively; TCT: control rats which were trained every day from day 10 to day 15 and every day also from day 23; NLN23 and NLN30: IOC lesioned rats which were naive and only tested at day 23 or at day 30, respectively; NLT: 10C lesioned rats which were naive before day 23 and trained every day from day 23; TLN23 and TLN30: IOC lesioned rats which were trained every day from day 10 to day 15 (before the lesion) and naive thereafter but only tested at day 23 or at day 30 (after the lesion), respectively; TLT: IOC lesioned rats which were trained every day from day 10 to day 15 (before the lesion) and every day also from day 23 (after the lesion).

2.2. Animals

The experiments were carried out in D A / H A N strain rats (pigmented rats), 10- or 15-day-old (depending on the experimental protocol) at the beginning of the study, day 1 being the day of birth. The young rats were housed with their mother in standard conditions: 12 h light-12 h dark, 22°C, food and water ad lib and were regularly weighed to control growth. IOC lesioned rats and controls were se- lected from different litters in order to avoid litter effects.

The IOC lesioned rats were given 3-AP (50 to 95 m g / k g i.p.) at day 15 followed 2 to 4 h later by niaci- namide (300 m g / k g i.p.) in order to block the 3-AP effects in other structures. The control rats were adminis- tered saline.

The extent of the IOC destruction was estimated by the responses of the animals to harmaline (administered i.p. at the dose of 20 mg /kg , 7 days after the injection of 3-AP) and, after completion of the experiments, by histological controls [23]. Were taken into account only the totally lesioned rats.

Totally IOC lesioned rats were divided into 6 groups according to the training schedule (Table 1).

- NLT rats (naive-lesioned-trained) were naive (not trained) before the lesion and trained every day on the rotorod from day 23 (that is 8 days after 3-AP injection).

- NLN rats (naive-lesioned-naive) were naive before and after the lesion; some of them were only tested on the rotorod by day 23 (NLN23), the others by day 30 (NLN30).

- TLT rats (trained-lesioned-trained) were trained every day before the lesion, from day 10 to day 15, and every day after the lesion, from day 23.

218 N. Jones et aL / Brain Research 697 (1995) 216-224

- TLN rats (trained-lesioned-naive) were not trained after the lesion but only tested by day 23 (TLN23) or by day 30 (TLN30).

The control rats were submitted to a similar protocol. - NCT rats (naive-control-trained) were not trained

before day 23 and trained every day from the 23rd day. - NCN rats (naive-control-naive) were only tested by

day 23 (NCN23) or by day 30 (NCN30). - TCT (trained-control-trained) were trained every day

from day 10 to day 15 and every day from day 23. - TCN (trained-control-naive) were trained before sham

injection and tested by day 23 (TCN23) or by day 30 (TEN30).

2.3. Training

The animals were placed on the rotorod and allowed to rotate until falling. From day 10 to day 15, the animals being too young to maintain their equilibrium for a long time, training was performed at a 10 rpm rotation rate. From day 23, the rats being older and maintaining their equilibrium better, the rotation rate was 20 rpm.

The rats were subjected every day (trained rats) or at a given day (naive rats) to a session of 10 trials given as follows: all the animals of the group were successively allowed to rotate for the first time; then, when the last rat had completed its first trial, the first one was allowed to rotate for the second time, etc. The time interval between two successive trials depended on the time during which all the animals of the group remained on the rotorod. The rats were placed on the rotating rod, their body axis being perpendicular to the rotation axis and the head directed against the direction of rotation; so, the animals had to progress forward to maintain equilibrium. Between trials, the rats were returned to their cage with their mother.

2.4. Data accumulation and statistical analysis

Quantitative and behavioral data were both obtained from the control and lesioned rats.

Quantitative data consisted in the time during which the animals maintained equilibrium upon the rotating rod (the chronometer was released when the rat was placed upon the rod and stopped when it fell down). Maximal score was arbitrarily fixed to 3 min (180 s) since it has been demonstrated, in previous experiments, that a rat which was able to maintain its equilibrium upon the rotating rod for 3 min could maintain it for a much longer time, 15-20 min, sometimes more [1]. However, the training was stopped when the score reached a plateau for several days even if the maximal score was not reached.

For each animal, the scores of the 10 daily trials were averaged (m). Then, the mean daily score (M), ___ S.E.M. (~r/~/n), of the n animals of each group was calculated ( M = E m / n ) . However, some rats reached the 180 s

maximal score more quickly than others. In order to prevent a bias, the mean daily score (M) had to be calculated on the same number (n) of animals. Therefore, for the quickly learning rats a fictitious daily score of 180 s was maintained for the following days until the last animal of the group was trained. Intra- and intergroup comparisons were made according to Wilcoxon and Mann-Whitney tests, respectively.

Behavioral data were obtained from the equilibrium behavior of the animals when rotating. For this purpose, the behavior of intact rats was analysed in experiments from video tapes in order to observe the strategy used by the rats to maintain their equilibrium. The relevant behav- ioral parameters were noted every 10 s during trials 2, 4, 6, 8 and 10. Two strategies, grasping (G) and walking (W), and three orientations of the animals relative to the rod were observed. However, as the animals' orientations did not seem a posteriori to be a useful parameter, only the activity profiles were taken into account in the results.

G behavior consisted for the animals to grasp on the rotorod and to be passively rotated or to slide, the animals being motionless and maintaining the same position while the rod was rotating. During W behavior, the rats walked asynchronously or synchronously upon the rotorod.

The daily frequencies of appearance of the two activity profiles were calculated for each rat and the values of the n rats of the group were averaged (means being given ±S.E.M.: ~r/v/n) in order to get the behavioral profile of the group for a given day. However, for individual rats, as training was interrupted the day (D) the rat reached the 180 s maximal score, the daily behavioral profile of D was fictitiously maintained during the following days. The behavioral profiles between two groups were statistically compared according to the X2 test.

2.5. Harmaline test

Seven days after the 3-AP injection, the animals were i.p. administered harmaline (20 mg/kg) and just after the injection were placed for 10 min on the membrane of a loud-speaker connected to a strip-chart recorder in order to quantify tremor. The total duration of tremor was mea- sured from the records during the last five minutes as described before [23].

2. 6. Histology

After completion of the behavioral study, the rats were administered an overdose of pentobarbital sodium and perfused intracardially, first with saline solution, then with 4% paraformaldehyde and 0.3% glutaraldehyde in phos- phate buffer. The brains were removed and placed in the same fixative solution for at least 10 days. Frozen sections were cut at 15-25 ~m, then stained with Cresyl violet and examined under a light microscope.

N. Jones et al. /Bra in Research 697 (1995) 216-224 219

3. Results

3.1. Maturation of the motor system and postlesion motor deficit

The animals, lesioned or not, were trained from postna- tal day 23. The motor performance on the rotorod of intact naive animals was therefore tested at day 23 and day 30. Fig. 1A shows that the score of the NCN30 group was significantly higher than that of NCN23. We interpreted this difference as due to an increasing maturation from day 23 to day 30, given that, in naive control rats, it has already been demonstrated that the scores regularly in- creased from day 18 to days 20, 22, 24, 26, 28 and 30 [1].

Fig. 1A also shows that the motor performance on the rotorod of the lesioned rats was lower than that of the controls, the difference being higher at day 30 (compare NCN23-30 and NLN23-30). Therefore, the poor motor performance of the lesioned animals was partially due to incomplete maturation, but mostly to the motor deficit produced by the destruction of the olivocerebellar system.

The behavioral profiles of NLN and NCN rats were significantly different at day 23 as well as at day 30. At day 23, the control rats maintained their equilibrium on the rotating rod by using mainly a walking behavior (they walked 71% of the time) and less frequently grasping (about 29% of the time) while the lesioned rats used grasping and were therefore passively rotated 90% of the time. At day 30, control and lesioned rats walked 94% and 24% of the time, respectively (Fig. 2). However, if walk- ing increased significantly with age in the control rats, its increase was not significant in the lesioned ones (Fig. 2). It is therefore clear that: (a) increased walking behavior in normal rats at day 30 indicates an increased maturation; (b) lesion of the IOC change the behavioral profile of the animals, the walking behavior being replaced by grasping.

3.2. Are the IOC lesioned rats able to learn to maintain their equilibrium when they are trained after the lesion ?

In control rats, late training rapidly compensates for the immature motor behavior as indicated by both scores (Fig. 1B) and strategies (Fig. 3A) of NCT.

To study the influence of the IOC lesion on learning of the equilibrium behavior, we observe the evolution of the scores and of the behavioral profiles of rats trained from day 23 (NLT) and compare them, at day 30, with those of rats which were naive both before and after the lesion (NLN30). In NLT, the scores increased from day to day and the animals reached the maximal score (180 s) by day 36, that is 2 weeks after the beginning of training (Fig. 1B). At day 29, the score was significantly greater than at day 23 ( P < 0.05) and also increased significantly from day 29 to day 31 ( P < 0.05). Moreover, at day 30, the mean score of NLT rats (97 + 12 s) was significantly greater than that of NLN30 rats (45 + 16 s) ( P < 0.05) (Fig. 1A, B). The score increase of the IOC lesioned rats is therefore due to both increasing maturation and postlesion training. One can therefore conclude that IOC lesioned rats are able to learn to maintain their equilibrium upon the rotorod.

However, such a learning was not as efficient as in control animals. In fact, comparison of NLT and NCT groups (Fig. 1B) shows that in the latter the animals reach the 180 s score by day 25, that is after 3 days training and that by the second and the third day of training their scores (171 + 7 s and 180 s, respectively) were significantly higher than those of NLT rats (28 + 8 s and 3 2 _ 8 s, respectively). Such score deficits in NLT rats are likely due, at least in part, to motor disabilities.

In the NLT, the behavioral profile changed from day to day (Fig. 3B). After two days training (day 25), the frequency of the walking behavior increased, the animals

~J o .=

L .

O t,I

180

160

140

120

100

8 0

6 0

4 0

A

NLN NCN • TLN A

TCN

20" []

0 , l~ ,~?~l~ , . , , , , , , , , , ' , ' ' , 10 13 16 19 22 25 28 31

a g e , in d a y s

B 180

• 160

140 qJ ~ 120

._¢ ~oo | ~ ao

O 6O

4 0

2O

TC T x* * "x

NLT H

NGT O - - O

TLT A.. . . .&

1 0 1 3 1 6 1 9 2 2 2 5 2 8 31 3 4 3 7 4 0

a g e , in d a y s

Fig. 1. Evolution of the scores of different groups of rats. Abscissae: age of the animals, in days; ordinates: score, in seconds. A: rats which were not trained from day 23, but tested at days 23 and 30. NLN: Naive-Lesioned-Naive; NCN: Naive-Control-Naive; TLN: Trained-Lesioned-Naive; TCN: Trained-Control-Naive; B: rats which were trained from day 23. TCT: Trained-Control-Trained; NLT: Naive-Lesioned-Trained; NCT: Naive-Control- Trained; TLT: Trained-Lesioned-Trained.

220 N. Jones et al. / Brain Research 697 (1995) 216-224

%

100 -

8 0 -

6 0 -

4O

2 0

0

G W 100

8 0

• N L N

[ ] N C N

6 O

% 4 0

20

0 ¸

G W Fig. 2. Mean behavioral profiles (±S.E.M.) of Naive-Control-Naive (NCN) and Naive-Lesioned-Naive (NLN) rats at days 23 and 30. Abscis- sae: activities of the animals upon the rotating rod (G: grasping; W: walking). Ordinates: frequency of appearance of the different activities.

6O

4O

2O

0

G W

~oo Day 30 r~

8o i i

6O

4 0

2O

I TLN

E]TCN

0 -7

G W Fig. 4. Mean behavioral profiles ( + S.E.M.) of Trained-Lesioned-Naive (TLN) and Trained-Control-Naive (TCN) rats at days 23 and 30. Abscis- sae, ordinates: same legends as Fig. 2.

walking about 50% of the time (while they walked 10% of the time only at day 23); it was still higher at day 30 and, when reaching the 180 s maximal score (day D on Fig. 3B), the frequency of walking was even greater, the rats walking 71% of the time.

Therefore, given that at day 25 NLT animals walked significantly more on the rotating rod than NLN ones at day 30, it can be concluded that the evolution of the behavioral profile in trained lesioned rats is more the result of training than the result of the maturation of the nervous structures which sustain this equilibrium behavior. In addi- tion, the behavioral profiles of NLT and of NCT are not significantly different when reaching the 180 s maximal score, but it has to be remembered that NCT were much younger (Fig. 3A,B).

3.3. Is the equilibrium behavior of the IOC lesioned rats" improved by a prelesion training?

The early training (or prelesion training) was performed from day 10 to day 15. At this age, control animals could not maintain equilibrium on the rotorod and training was ineffective (Fig. 1A). In addition, this early training did not significantly change the equilibrium behavior (compare TCN and NCN in Fig. 1A) nor the equilibrium strategies of intact animals when tested later at day 23 and day 30 (compare TCN23-30 and NCN23-30 in Figs. 4 and 2, respectively). Therefore, early training did not compensate for late maturation in normal rats.

The influence of prelesion training on the evolution of the equilibrium behavior in IOC lesioned rats is studied by

A B ~oo ~oo N L T • Day ?3

80 80

% 0/0 - ~ Day30 40 40

20 20 [ ] Day D

o o /I G W G W

Fig. 3. Mean behavioral profiles ( + S.E.M.) of Naive-Control-Trained (NCT) rats at days 23 and 25, and of Naive-Lesioned-Trained (NLT) rats at days 23, 25, 30 and at day D (when they reached the 180 s maximal score). Abscissae, ordinates: same legends as Fig. 2.

N. Jones et al. / Brain Research 697 (1995) 216-224 221

comparing the scores and the behavioral patterns of TLN and NLN groups of rats.

At day 23 and at day 30, the mean score of TLN rats (46___6 s at day 23 and, at day 30, 124___13 s) were significantly higher ( P < 0.01) than those of NLN rats (19 + 6 s at day 23 and 4 5 _ 16 s at day 30) (Fig. 1A). Therefore, early training reduced the postlesion motor impairment, although not completely. In fact, the mean scores of TCN rats at day 23 and at day 30 were respec- tively 8 7 _ 4 s and 175 + 2 s, while they were 46 + 6 s and 124 _ 13 s, respectively, in TLN animals, the differ- ence between the two groups being significant ( P < 0.01) (Fig. 1A). However, the behavioral profile of the prelesion trained animals (TLN) was not significantly different, at day 23 as well as at day 30, from that of naive lesioned rats (NLN). From Figs. 2 and 4 it can be seen that the animals of both groups used essentially grasping to main- tain their equilibrium (about 90-95% of the time at day 23 and 70-75% of the time at day 30). Therefore, the prele- sion training, which helps the IOC lesioned rats to main- tain their equilibrium on the rotating rod for a long time, does not accelerate the acquisition of a normal behavioral pattern (walking).

Obviously, the behavioral patterns of TLN rats are different from those of TCN ones (Fig. 4), the latter walking on the rotating rod significantly more often than the former (at day 23, TCN and TLN rats walked 58% and 5% of the time, respectively, these percentages being respectively equal to 91% and 26% at day 30).

3.4. What is the behavior o f the IOC lesioned rats when they were trained both before and after the lesion ?

Prelesion and postlesion trainings are both efficient in the development of the equilibrium behavior in the IOC lesioned rats. To know whether the effects of these train- ings can be summed up when they are combined, TLT are compared to TLN and to NLT.

From Fig. 1, it can be seen that the mean score of TLT rats (46 + 5 s at day 23) increased from day to day to

reach 180 s by day 36. From day 23 to day 29, the mean scores of TLT rats are significantly higher ( P < 0.01) than those of NLT animals; from day 30 to the end of training, the mean daily scores of both groups are not significantly different and they both reached 180 s 2 weeks after the beginning of training. Therefore, the scores of TLT were increased only in the first week of training. At day 30, the mean score of TLT (151 _ 10 s), of NLT (97 -t- 12 s) and of TLN rats (124 ___ 13 s) (Fig. 1) were not significantly different. In other words, the acquisition of sensorimotor skills was greater than in the animals submitted to a postlesion training only. From this viewpoint, it can be concluded that pre- and postlesion trainings summed up to increase the efficiency of postlesion equilibrium. Neverthe- less, the postlesion training is not absolutely necessary for the equilibrium behavior to be achieved providing the animals had experienced a prelesion training. Obviously, this is not the case in control animals since late training compensates for maturation but not the early training (Fig. 1).

The behavioral pattern of TLT rats differed from day 23 to day 25, the walking behavior increasing significantly (Fig. 5A), but did not change thereafter. At day 23, TLT and NLT rats (compare Fig. 5A and 3B) were similarly passively rotated (95% and 90% of the time, respectively: the difference was not significant). At day 25, the rats of both groups walked significantly more but, surprisingly, the walking behavior was much more increased in NLT than in TLT animals (the difference between the two groups was significant) (compare Fig. 3B and 5A). It can be concluded that whether the scores of TLT rats increased much more quickly after the IOC lesion than the scores of NLT rats, their behavior upon the rotating rod was more far from the normal than that of NLT animals. In addition, at day 30 TLT and TLN used essentially grasping to maintain their equilibrium (compare Fig. 5A and 4), while TCT and TCN used essentially walking (compare Fig. 5B and 4). Therefore, lesioned rats (a) improve their scores on the rotorod by using grasping and (b) improve grasping with early training.

A 100 r / 1 o o -

8 0 80 -

60 6 0 -

% °/o 4 0 4 0 -

20 | 2 0 "

) 0 0 -

G W

B

• Day 23

[ ] Day 25

Day30

G W Fig. 5. Mean behavioral profiles ( _+ S.E.M.) of Trained-Lesioned-Trained (TLT) and Trained-Control-Trained (TCT) rats at days 23, 25 and 30. Abscissae, ordinates; same legends as Fig. 2.

222 IV. Jones et al. / Brain Research 697 (1995) 216-224

3.5. Harmaline test and histological control

In response to harmaline, most of the animals studied exhibited no tremor or only a small tremor which lasted only 10% of the observed period (5 min). Moreover, in all the animals studied, the histological control at the end of the experiment demonstrated that the IOC was completely lesioned: no neuron or only a few scattered neurons could be observed in the IOC (see Fig. 2 in [23]). Given that were taken into account only the totally IOC lesioned rats, no correlation could be made between severity of IOC damage and degree of behavioral impairment across ani- mals.

4. Discussion

Given that the score and the frequency of walking of the IOC lesioned rats increased with postlesion training and were significantly higher than those of naive lesioned rats of the same age, it can be concluded that the postle- sion training is efficient in the restoration of the equilib- rium function from both quantitative and qualitative points of view. However, compared to that of controls (non-le- sioned rats), the equilibrium behavior of the IOC lesioned animals is impaired: their maximal score was delayed by 11 days with respect of controls and their behavioral profile was never similar to that of controls of the same age. Such an impairment is obviously due to the ataxia developed by the IOC lesioned animals; at day 23, as well as at day 30, the scores of naive lesioned rats were significantly lower than those of naive control ones and the former used essentially grasping to maintain their equilib- rium upon the rotating rod while the latter used mainly walking, and as both animals were not trained, the discrep- ancy can only be explained by the ataxia due to the lesion. Therefore, in spite of their motor deficits, the IOC lesioned animals were obviously able to learn if they were trained for a long time. It can be concluded that the IOC lesion would prevent the achievement of an efficient equilibrium behavior when the experimental situation (the rotorod) is a novel one. In other words, the lesion prevents use of the sensory informations the animals have experienced in their familiar environment (their housing cage) to directly elabo- rate an efficient equilibrium behavior when they are sub- mitted to the novel experimental situation for the first time. If the IOC lesion does not prevent learning of the equilib- rium behavior providing training is of long duration, such a learning is much less rapid than in intact animals and the role of the olivo-cerebellar pathway in the acquisition of the equilibrium behavior would be more to adapt the behavior to the environmental conditions than to initiate it. Such a conclusion well agrees with the role of the olivo- cerebellar system in the temporal organization of the movement [8,22,30,31,39,41] and, in that way, one can conclude that the olivo-cerebellar pathway is involved in

such an organization. Lesion of the IOC at day 15 has obviously less drastic effects than a cerebellectomy done at the same age since in 15-day-old cerebellectomized rats the postoperative training is inefficient in the acquisition of the equilibrium behavior, the scores of the animals reach- ing a plateau at about 20 s [44]. It is therefore likely that the mossy fibers-granule cells-parallel fibers pathway is also implicated in the acquisition of the equilibrium func- tion. Experiments have been undertaken to test this hy- pothesis. However, as lesion of the IOC in adult prevents the acquisition of an efficient equilibrium behavior (in preparation), it can be concluded that the mossy fibers- granule cell-parallel fibers pathway can be efficient alone in such an acquisition only when it is immature, and therefore still plastic, at the time of the IOC lesion.

The prelesion training (TLN animals) facilitates the acquisition of an efficient postlesion equilibrium behavior since the scores of the animals were significantly higher than those of naive ones (NLN rats) at day 23 and at day 30. However, the behavioral profiles of both groups were not significantly different and while in TLN as well as in NLN rats the frequency of walking was higher at day 30 than at day 23, both animals used essentially grasping to maintain their equilibrium. This grasping behavior was not due to the inability of the ataxic lesioned rats to walk. In effect, while walking is not so well coordinated as in controls, its frequency was rather high when the animals were trained after the lesion. During prelesion training, from day 10 to day 15, the scores of the young animals were very low (only a few seconds) and they maintained their equilibrium on the rotating rod by grasping only. Consequently, one can think that the animals have memo- rized this motor schema which, for that reason, is mainly used after the IOC lesion. If such an interpretation is correct, it would mean that between day 10 and day 15 (that is during the prelesion training), the olivo-cerebellar pathway is not needed for the animals to store the sensori- motor cues generated by the early experience on the rotating rod. Given that it has been demonstrated that, in adult rats, the olivo-cerebellar pathway [14] and the cere- bellum [12,13] are implicated in the memory of experi- ences done before the lesion, the fact that the cerebellum seems to play no role in memory of experiences in 10-15- day-old rats can be explained by its lack of maturity.

While it is hazardous, in this context, to correlate behavioral and neurophysiological data, the behavioral ef- fects of the IOC lesion can probably be explained by the degeneration of the climbing fibers that normally impinge onto the Purkinje cells of the cerebellar cortex. It has been demonstrated that the IOC lesion modifies the activity of the Purkinje cells. Of course, the complex spikes disappear after the lesion [2-5,7], but the frequency of the simple spikes, resulting from an activation of the mossy fibers- granule cells-parallel fibers pathway, is changed. Conse- quently, the activity of the deep cerebellar nuclei and of the vestibular nuclei, which receive inputs directly from

N. Jones et aL /Brain Research 697 (1995) 216-224 223

the Purkinje cells, is altered. The IOC therefore modulates the influence of the Purkinje cells on their target cells. If such a modulation affects the ventrolateral thalamic neu- rons which relay the cerebellar information to the cerebral cortex, one can estimate that the IOC lesion would affect learning, as well as the use of sensory informations to resolve a novel problem in a new context. In that way, it could well be that the equilibrium deficiencies of the animals after lesion of the IOC are to be explained in terms of cerebe l lo-cor t ica l relationships.

Acknowledgements

This work was partly 92.C.0756

supported by a MRE Grant

References

[1] Auvray, N., Caston, J., Reber, A. and Stelz, T., Role of the cerebellum in the ontogenesis of the equilibrium behavior in the young rat: a behavioral study, Brain Res., 505 (1989) 291-301.

[2] Bardin, J.M., Batini, C., Billard, J.M., Buisseret-Delmas, C., Con- rath-Verrier, M. and Corvaja, N., Cerebellar output regulation by the climbing and mossy fibers with and without the inferior olive, J. Comp. Neurol., 213 (1983) 464-477.

[3] Batini, C., Benedetti, F., Buisseret-Delmas, C.I Montarolo, P.G. and Strata, P., Metabolic activity of intracerebellar nuclei in the rat: effects of inferior olive inactivation, Exp. Brain Res., 54 (1984) 259-265.

[4] Batini, C. and Billard, J.M., Release cerebellar inhibition by climb- ing fiber deafferentation, Exp. Brain Res., 57 (1985) 370-380.

[5] Batini, C., Billard, J.M. and Daniel, H., Long term modification of cerebellar inhibition after inferior olive degeneration, Exp. Brain Res., 59 (1985) 404-409.

[6] Berthier, N.E. and Moore, J.W., Cerebellar Purkinje cell activity related to the classically conditioned nictitating membrane response, Exp. Brain Res., 63 (1986) 341-351.

[7] Billard, J.M. and Daniel, H., Persistent reduction of Purkinje cell inhibition on neurones of the cerebellar nuclei after climbing fiber deafferentation, Neurosci. Lett., 88 (1988) 21-26.

[8] Braitenberg, V., Is the cerebellar cortex a biological clock in the millisecond range? Brain Res., 25 (1967) 334-346.

[9] Buchtel, H.A., Visual learning deficits following cerebellar damage in rats, J. Comp. PhysioL Psychol., 72 (1970) 296-305.

[10] Collcwijn, H., The modifiability of the adult vestibulo-ocular reflex, Trends Neurosci., 2 (1979) 98-102.

[11] Dahhaoui, M., Caston, J., Auvray, N. and Reber, A., Role of the cerebellum in an avoidance conditioning task in the rat, Physiol. Behav., 47 (1990) 1175-1180.

[12] Dahhaoui, M., Caston, J., Lannou, J. and Avenel, S., Role of the cerebellum in habituation exploration behavior in the rat, PhysioL Behav., 52 (1992) 339-344.

[13] Dahhaoui, M., Lannou, J., Steiz, T., Caston, J. and Guastavino, J.M., Role of the cerebellum in spatial orientation in the rat, Behav. Neural Biol., 58 (1992) 180-189.

[14] Dahhaoui, M., Stelz, T. and Caston, J., Effects of lesion of the inferior olivary complex by 3-acetylpyridine on learning and mem- ory in the rat, J. Comp. PhysioL A, 171 (1992) 657-664.

[15] Desmond, J.E. and Moore, J.W., A brain stem region essential for classically conditioned but not unconditioned nictitating membrane response, PhysioL Behav., 28 (1982) 1029-1033.

[16] Glickstein, M. and Yeo, C., The cerebellum and motor learning, J. Cogn. Neurosci., 2 (1990) 69-78.

[17] Guillaumin, S., Dahhaoui, M. and Caston, J., Cerebellum and mem- ory: an experimental study in the rat using a passive avoidance conditioning test, PhysioL Behav., 49 (1991) 507-511.

[18] Houk, J.C. and Barto, A.G., Distributed sensorimotor learning. In G.E. Stelmach and J. Requin (Eds.), Tutorials in Motor Behavior. H, Elsevier Science Publishers, New York, 1992, pp. 71-100.

[19] Ito, M., Synaptic plasticity in the cerebellar cortex that may underlie the vestibulo-ocular adaptation. In A. Berthoz and G. Melvill Jones (Eds.), Adaptive Mechanisms Of Gaze Control Facts And Theories, Elsevier Science Publishers, New York, 1985, pp. 213-221.

[20] Ito, M., Movement and thought: identical control mechanisms by the cerebellum, Trends Neurosci., 16 (1993) 448-450.

[21] lto, M., Shlida, T., Yagi, N. and Yamamoto, M., The cerebellar modification of rabbit's horizontal vestibulo-ocular reflex induced by sustained head rotation combined with visual stimulation, Proc. Jpn. Acad., 50 (1974) 85-89.

[22] Ivry, R.B. and Keele, S.W., Timing functions of the cerebellum, J. Cogn. Neurosci., 1 (1989) 136-152.

[23] Jones, N., Le Marec, N., Stelz, T. and Caston, ,i., Effect of adminis- tration of 3-acetylpyridine followed by niacinamide injection on survival, extent of the inferior olivary complex lesion, and response to harmaline in the young rat, Brain Res., 656 (1994) 257=262.

[24] Lalonde, R. and Botez, M.I., The cerebellum and learning processes in animals, Brain Res. Rev., 15 (1990) 325-332.

[25] Lashley, K.S. and McCarthy, D.A., The survival of the maze habit after cerebellar injuries, J. Comp. Psychol., 6 (1926) 423-433.

[26] Lavond, D.G., Hembree, T.L. and Thompson, R.F., Effect of kainic acid lesions of the cerebellar interpositus nucleus on eyelid condi- tioning in the rabbit, Brain Res., 326 (1985) 179-182.

[27] Leaton, R.N. and Supple Jr., W.I., Cerebellar vermis: essential for long-term habituation in the acoustic startle response, Science, 232 (1986) 513-515.

[28] Leiner, H.C., Leiner, A.L. and Dow, R.S., Cognitive and language functions of the human cerebellum, Trends Neurosci., 16 (1993) 444-447.

[29] Lisberger, S.G., Role of the cerebellum during motor learning in the vestibulo-ocular reflex. Different mechanisms in different species? Trends Neurosci., 5 (1982) 437-441.

[30] Llinas, R. and PeUionisz, A., Cerebellar function and the adaptive feature of the central nervous system. In A. Berthoz and G. Melvill Jones (Eds.), Adaptive Mechanisms In Gaze Control Facts And Theories, Elsevier Science Publishers, New York, 1985, pp. 223- 231.

[31] Llinas, R. and Sasaki, K., The functional organization of the olivo- cerebellar system as examined by multiple Purkinje cell recordings, Eur. J. Neurosci., 1 (1989) 587-602.

[32] Llinas, R. and Walton, K., Place of the cerebellum in motor learn- ing. In M.A.B.Brazier (Ed.), Brain Mechanisms In Memory And Learning: From The Single Neuron To Man, 1BRO Monograph Series, VoL 4, Raven Press, New York, 1979, pp. 17-36.

[33] Llinas, R., Walton, K., Hillman, D.E. and Sotelo, C., Inferior olive: its role in motor learning, Science, 190 (1975) 1230-1231.

[34] Malyukova, I.V., Influence of the neocerebellum extirpation on situational conditioned reflexes in dog, Zh. Vyssh. Nervn. Deyat. im LP, Paviova, 13 (1963) 1052-1058.

[35] Marr, D.A., A theory of cerebellar cortex, J. Physiol., 202 (1969) 437-470.

[36] McCormick, D.A., Lavond, D.G., Clark, G.A., Kettner, R.E., Ris- ing, C.A. and Thompson, R.F., The engram found? Role of the cerebellum in classical conditioning of nictitating membrane and eyelid response, Bull Psychon. Soc., 18 (1981) 103-105.

[37] McCormick, D.A., Steinmetz, .I.E. and Thompson, R.F., Lesions of inferior olivary complex cause extinction of the classically condi- tioned eyeblink response, Brain Res., 359 (1985) 120-130.

[38] McCormick, D.A. and Thompson, R.F., Cerebellum: essential in-

224 N. Jones et al. / Brain Research 697 (1995) 216-224

volvement in the classically conditioned eyelid response, Science, 223 (1984) 296-299.

[39] Rawson, J.A., Wertheimer, S. and Rees, S., Modification of parallel fiber-Purkinje cell transmission by long-term activation of climbing fibres, Neurosci. Lett., 91 (1988) 14-18.

[40] Robinson, D.A., Adaptive gain control of vestibulo-ocular reflex by the cerebellum, J. Neurophysiol., 39 (1976) 954-969.

[41] Sasaki, K., Bower, J.M. and Llinas, R., Multiple Purkinje cell recording in rodent cerebellar cortex, Eur. J. Neurosci., 1 (1989) 572-586.

[42] Schneiderman Fish, B., Baisden, R.H. and Woodruff, M.L., Cerebel- lar nuclear lesions in rats: subsequent avoidance behavior and ascending anatomical connections, Brain Res., 166 (1979) 27-38.

[43] Yeo, C.H., Hardiman, M.S. and Glickstein, M., Discrete lesions of the cerebellar cortex abolish the classically conditioned nictating membrane response in the rabbit, Behav. Brain Res., 13 (1984) 261-266.

[44] Zion, C., Auvray, N., Caston, J., Reber, A. and Stelz, T., Effects of cerebellectomy at day 15 on the ontogenesis of the equilibrium behavior in the rat, Brain Res., 515 (1990) 104-110.