- Email: [email protected]
Rule-Based Learning Impairment in Rats with Lesions to the Dorsal Striatum
Neurobiology of Learning and Memory 72, 47– 61 (1999) Article ID nlme.1998.3905, available online at http://www.idealibrary.com on
Rule-Based Learning Impairment in Rats with Lesions to the Dorsal Striatum Barbara Van Golf Racht-Delatour and Nicole El Massioui Laboratoire de Neurobiologie de l’Apprentissage et de la Me´moire, CNRS - URA 1491, Bt. 446, Universite´ Paris-Sud, 91405 Orsay Cedex, France
The present study examined the effects of lesions to the dorsal striatum (DS) in Sprague-Dawley rats, when tested on the acquisition and successive shifts in the position of a goal arm in an eight-arm radial maze. In the procedure we used, rats had to retrieve the location of one baited arm among the eight arms of the maze after it had just been presented as a sample during a forced trial. After attainment of a fixed learning criterion, rats were submitted to five successive shifts in the goal location. Results showed that DS rats were able to learn the position of the goal arm during the acquisition phase as efficiently as sham-operated rats. In contrast, when the position of the goal arm was shifted, although DS rats were able to learn its new position, they made significantly more errors and required more sessions to reach criterion than sham-operated rats. These results suggested that both groups did not solve the task using the same behavioral strategy. The analysis of responses made suggested that sham-operated rats solved the task using the pairing rule between the forced and the free run (matching-to-sample rule), while DS rats solved the task using only visuospatial processing. These data therefore suggest that the dorsal striatum plays an important role in rule-learning ability. © 1999 Academic Press
INTRODUCTION The striatum is a complex structure known for its role in motor functions (Gerfen, 1992); however, it has also been shown to play an important role in learning and memory processes. In humans, neurodegenerative diseases involving the basal ganglia lead to pathologies, such as Parkinson’s or Huntington’s diseases, in which motor and cognitive deficits are apparent. The major cognitive impairments in both pathologies relate to retrieval processes (Mohr, Fabbrini, Ruggieri, Fedio, & Chase, 1988; Pillon, Deweer, Michon, Malapani, Agid, & Dubois, 1994; Faglioni, Scarpa, Botti, & Ferrari, 1995; Dubois & Pillon, 1997), as well as rule-finding abilities (Freedman & Oscar-Berman, 1986; Oscar-Berman, McNamara, & Freedman, 1991; Taylor & Saint Cyr, 1995). However, these diseases also result in procedural learning deficit (Heindel, Salmon, Shults, Walicke, & Butters, 1989; Knopman & Nissen, 1991; Saint-Cyr, Taylor, & Lang, 1988) and in an inability to elaborate or maintain new strategies in tasks requiring adaptation to novelty (Downes, Roberts, Sahakian, Evenden, Morris, & Robbins, 1989; Taylor, Saint-Cyr, & Lang, 1990; Owen, James, Leigh, Summers, Quinn, & Marsden, 1993; Owen, RobAddress correspondance and reprints request to Nicole El Massioui, Laboratoire de Neurobiologie de l’Apprentissage et de la Me´moire, CNRS-URA 1491, Bt. 446, Universite´ Paris Sud, 91405 Orsay Cedex, France. Fax (331) 69.15.77.26. E-mail: Nicole. [email protected]. 47
1074-7427/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
48
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
erts, Hodges, Summers, Polkey, & Robbins, 1993; Withaar, Brouwer, & Deelman, 1996). From the many studies carried out in the rat, where lesions have been made to different subregions of the striatum, it is generally believed that the basal ganglia and, in particular, the striatum play an important role in problem solving. Rats with lesions to the dorsal striatum have been shown to be mainly impaired in learning tasks where they have to make a specific motor response when presented with a reinforced discrete stimulus, such as signaled active avoidance in a two-way shuttle box (Green, Beatty, & Schwartzbaum, 1967; Winocur, 1974; Kirkby & Polgar, 1974; El Massioui & Van Golf Racht-Delatour, 1997), passive avoidance tasks (Winocur, 1974), conditional discriminations (Robbins, Giardini, Jones, Reading, & Sahakian, 1990; Reading, Dunnett, & Robbins, 1991), and win-stay procedures in the eight-arm radial maze (Packard, Hirsh, & White, 1989; McDonald & White, 1993). The difficulty to learn a stimulus-response (S-R) association seems to be the main impairment elicited by dorsal striatum damage. Furthermore, dorsal striatum appears to be involved in procedural learning and memory as all these experiments require that the rats initiate or ¨ berg & Divac, 1979), confirming a hypothesis withhold a specific response (O postulated by Philipps and Carr (1987) that the basal ganglia is necessary for processing procedural knowledge. Moreover, it has been shown that other subregions of the striatum are also important for learning. For example, rats with lesions to the lateral striatum show an inappropriate response to the discriminative stimuli in conditional discrimination tasks (Robbins et al., 1990; Reading et al., 1991). Similarly, lesions to the medial striatum result in a deficit in rats to select, develop, or access appropriate strategies in cue- and place-learning tasks in a Morris water maze (Whishaw, Mittleman, Bunch, & Dunnett, 1987). Furthermore, lesions to the caudate result in an inability to acquire a reference memory component of the task whereas the variable or working memory component was left intact (Colombo, Davis, & Volpe, 1989; Packard & White, 1990). The first group of authors discussed the observed impairments in terms of a deficit in tasks that required retention of rule information useful for improving performance across trials whereas the latter attributed the reference memory impairment to an inability to process correctly the stimulus-response association necessary to learn the baited arm’s position. Therefore, it seems likely that the basal ganglia, and more specifically the dorsal striatum, is necessary for problem solving and would consequently play an essential role in the acquisition and application of response rules. The aim of the present experiment was to determine whether lesions to the dorsal striatum would impair rats’ ability to use a behavioral shift in strategy in order to learn the successive transfer of a goal location in an eight-arm radial maze. The task consisted in associating the baited arm that had just been presented as a single sample during a forced trial, among the eight arms of the maze until the attainment of the learning criterion. This task, as well as the successive shifts in the goal arm location, could be solved either by visuospatial-based learning or by rule-based learning. The optimal strategy, the use of the pairing rule, would only allow a high level of performance.
LEARNING IMPAIRMENT AND DORSAL STRIATUM
49
MATERIAL AND METHODS Subjects Nineteen male Sprague-Dawley rats (Iffa Credo) weighing between 260 and 315 g were housed in pairs in a temperature-controlled colony room and maintained on a 12-h light-dark cycle (lights on 8:00 AM). On arrival in the laboratory, rats were allowed ad libitum access to food and water and handled on a daily basis. Following surgical procedures to lesion the dorsal striatum, rats underwent a food deprivation schedule to reduce their weight to 85% of their original free-feeding body weight. Apparatus The apparatus was a wooden eight-arm maze painted gray and elevated 70 cm above the floor surface. Each arm (60 3 12 cm) radiated from a central octagonal platform (30 cm diameter) and was equipped with a transparent perspex “tunnel” (44 cm long, 13 cm wide, and 13 cm high) and a food cup at the end. White perspex panels could be placed at the start of the tunnel to prevent an arm to be visited. The experimental room contained various extra maze cues. Surgery Rats were randomly assigned either to the experimental group in which lesions were made in the dorsal striatum (n 5 11) or to the sham-operated group (n 5 8). Injections of atropine sulfate 0.025% (Meram, 0.15 ml, ip) were given to rats to prevent respiratory problems, 15 min prior to anesthetizing with pentobarbital (50 mg/kg). Rats were then placed in a stereotaxic apparatus and the cranium was exposed. Holes were drilled in the skull and incisions made in the dura to make bilateral injections into the dorsal striatum. A stainless-steel cannulae mounted onto a microinjector on the stereotaxic frame was lowered into the dorsal striatum and bilateral lesions were made using N-methyl-D-aspartate (NMDA, Sigma; 37 mg/ml). Injections of 0.2 ml per site were made at the following coordinates relative to bregma (Paxinos & Watson, 1986) AP 1 1.6, ML 6 1.6, and DV 23.8 and AP 1 1.6, ML 6 3.2, and DV 23.8 for anterior sites; AP 1 0.2, ML 6 1.8, and DV 24.4 and AP 1 0.2, ML 6 3.8, and DV 24.4 for posterior sites. Sham-operated animals underwent operating procedures identical to those performed on dorsal striatum-lesioned rats, with the exception that the cannulae was not lowered into the brain. Animals with NMDA microinjections were given a single injection of Valium (Roche) 10 mg/kg following surgery to prevent seizure activity. One week after the end of surgery rats were progressively restricted in food, and behavioral testing began 1 week later. Procedure Rats were habituated to the maze 1 day before commencing the first phase of training. All arms were baited with chocolate cereal and animals were allowed to explore the maze until all arms had been visited or 10 min had elapsed. The number of arms visited and the order in which they were visited were recorded. Following this day, training sessions, consisting of four trials
50
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
divided into two phases, commenced. In the first phase (forced run), seven of the eight arms were blocked and rats were placed in the central platform and allowed to visit the baited goal arm (for each animal the goal arm was the forth entry it had made on the preliminary training). When rats had returned to the central platform, they were removed from the maze and placed in a holding box in the experimental room, and the remaining panels were removed for the second phase (free run). Rats had free access to all arms but only the arm baited during the forced trial was baited in the second phase. Rats completed the trial when they had visited the correct arm and the number of errors was scored as entries in unbaited arms. Three further trials were given per day (every forced run was followed by a free run), and to eliminate kinesthetic and olfactory strategies, rats were placed on the central platform facing either West, South, North, or East and the maze was rotated between each trial. Rats were tested every day with always the same goal arm position reinforced until they reached a criterion of an average number of errors, equal to or less than one, per session. After reaching the criterion, animals were tested on the same procedure using a different goal arm to test their ability to shift strategy. Five different shifts of the goal arm were tested in total. The number of arms separating the new goal location from the previous one as well as the direction was the same for all animals (three first shifts, clockwise 13, 13, 13; two last shifts, counterclockwise 22, 22). So, for example, if during the acquisition phase the reinforced arm was Number 1, the new goal locations for the five subsequent shifts were arms Numbers 4, 7, 2, 8, and 6. Histology On completion of the experiment, histological verification of the lesions was made. Rats were anesthetized with pentobarbital and perfused transcardially with isotonic saline buffer followed by 10% paraformaldehyde. The brains were removed and postfixed in a 16% sucrose solution for 3 days and stored at 220°C. Coronal brains sections cut at 40 mm and stained with Cresyl violet were used to assess the extent of the lesion. RESULTS Histology Figure 1 shows an example of a schematic reconstruction of the extent of lesion to the dorsal striatum. The lesions showed characteristic necrosis of tissue including glia cells. Although lesions had been made in 11 rats, 4 of them were discarded either because the lesion extended to the corpus callosum or was too restricted. The data from the remaining 7 rats were included in statistical analysis. Behavior Mean Number of Sessions and Errors to Reach Criterion (Fig. 2) Initial acquisition. Rats with lesions to the dorsal striatum had no difficulty in learning the initial visuo-spatial task, as they showed no difference in the number of sessions as well as in the number of errors to reach criterion as that shown by the sham-operated rats (F , 1 in both cases).
LEARNING IMPAIRMENT AND DORSAL STRIATUM
51
FIG. 1. NMDA lesions of the dorsal striatum. Maximum and minimum extent of the lesions are indicated by hatched areas (maximum, left-directed hatches; minimum, right-directed hatches). Numbers on the right-hand side of the figures represent coordinates relative to bregma (Paxinos & Watson, 1986).
Shifts of the reinforced goal arm. A comparison of performance between the initial acquisition phase and the first shift showed that both the sham-operated and the lesioned rats improved their performance as shown by a decrease in the number of sessions required to reach criterion (F(1, 7) 5 71.21, p , .001; and F(1, 6) 5 8.72, p , .05, respectively) and in the number of errors (F(1, 7) 5 56.85, p , .001; and F(1, 6) 5 28.51, p , .01, respectively). Over the five successive shifts, rats with striatal lesions needed significantly more sessions to reach criterion (F(1, 13) 5 9.14, p , .01) and made more errors than sham-operated rats (F(1, 13) 5 9.43, p , .01) with a significant groups by sessions interaction (F(4, 52) 5 2.64, p , .05), suggesting that sham-operated rats learned to find each new locations more rapidly than the rats with lesions to the dorsal striatum. However, although sham-operated rats were faster to learn the five shifts when compared to striatum-lesioned rats, they rapidly reached a plateau of performance which showed no further improve over shifts (F(4, 28) 5 1.42, ns for the number of sessions and 5 2.63, ns for the number of errors) as the case for SD rats (F(4, 24) 5 3.02, p , .05, for the number of sessions and 5 7.81, p , .01, for the number of errors).
52
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
FIG. 2. Mean number of sessions (1SEM; upper panel) and mean number of errors (1SEM; lower panel) made to reach the criterion on the initial phase of acquisition (A) and on the successive shifts of goal location (S1 to S5) in rats with lesions to the dorsal striatum and sham-operated animals.
Percentage of Rats Reaching Criterion Statistical analyses, using a Fisher test, showed that a significantly bigger percentage of sham-operated rats reached criterion on the first session of shift 2 and 3, compared with DS-lesioned rats (p , .05 and p , .025, respectively). There was no group difference in performance on the initial acquisition or the other shifts (Fig. 3). In addition, analysis to assess the number of sessions required for all rats (100%) in each group to reach the criterion showed, with a one-tail Student t test, that the sham-operated rats required significantly fewer sessions to reach criterion than the DS rats (p , .01). Type of Errors We classified the errors made by both groups of animals according to the angular distance separating the goal arm and the arm entered by each rat. Thus, 45° corresponds to an entry in one of the arms just adjacent to the correct
FIG. 3. Mean percentage of rats reaching the criterion on each session of acquisition of the baited goal arm location as well as on each session needed to learn the five shifts of the reinforced position in sham-operated and dorsal striatum-lesioned rats.
LEARNING IMPAIRMENT AND DORSAL STRIATUM
53
54
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
LEARNING IMPAIRMENT AND DORSAL STRIATUM
55
one; 90° means that there was one arm in between the baited arm and the visited one; 135° means that there were two arms in between the arm presented during the forced run and the one chosen during the free run; 180° corresponds to an opposite choice from the correct goal position. The total number of errors made at 45 or 90 or 135° was divided by 2 (as they corresponded either to clockwise or counterclockwise errors). The distribution of the different types of angular errors for both experimental groups shows that only the number of errors made at 135° during the first shift is significantly higher for the sham-operated animals (F(1, 7) 5 6.06, p , .05) and for the dorsal striatum-lesioned rats (F(1, 6) 5 16.89, p , .01). Furthermore, in rats with striatal lesions, there was a lesser rate of errors made in the opposite direction to the reinforced arm (180°) on shifts 1, 2, and 3 (p , .05 and ps , .01, respectively). Analyses of Strategies Four different analyses were carried out to assess whether rats were using particular types of strategies. (a) Correct choice on the first trial of the first session of each shift. To test whether rats were learning the rule of pairing the arm presented during the forced run with the one to choose in the free run, we analyzed the mean number of rats that made a correct choice on the first trial of the first session of each shift. The results showed that sham-operated animals made significantly more correct choices on this first trial than striatum-lesioned rats (F(1, 13) 5 13.62, p , .01; Fig. 4A). (b) Entries in the initially baited arm. To test whether rats showed a persevering behavior, we analyzed the percentage of visits in the initially baited arm with the total number of visits for each shift. The results showed that both the control rats and the striatum-lesioned rats decreased their number of entries in the initially baited arm across shifts (F(4, 52) 5 4.57, p , .005) with no intergroup difference (F , 1; Fig. 4B). (c) Entries into the previously baited arm. To test whether rats were able to adapt their behavior to the demand of the experiment, and therefore abandon the visit of the previous baited arm when a new shift occurred, we analyzed the mean percentage of entries in the previously baited arm to the total number of visits made during each shift (C 5 A/B 3 100; A 5 number of entries in the former baited arm for all sessions of one shift and B 5 total number of visits made during all sessions of each shift). Results showed that, across shifts, sham-operated rats made fewer entries into the previously reinforced goal arm than dorsal striatum-lesioned animals (F(1, 13) 5 4.92, p , .05). There was a
FIG. 4. Analysis of strategies used by rats with lesions to the dorsal striatum and shamoperated rats. (A) The mean percentage (6SEM) of animals making a correct choice on the first trial of each new shift of goal location for each experimental group. (B) The mean percentage of entries (6SEM) to the acquisition goal arm across the five successive shifts for both groups of animals. (C) The mean percentage of visits to the previously reinforced arm across the different shifts of the goal location in lesioned and sham-operated rats. (D) The mean percentage of entries (6SEM) in all previous goal arms across the five successive modifications of the baited arm location for both experimental groups.
56
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
significant decrease in the number of entries to the previously baited arm across shifts in sham-operated animals (F(4, 28) 5 7.05, p , .01), whereas dorsal striatum rats kept visiting the previously reinforced location on a steady percentage of trials from the first to the last shift (F , 1; Fig. 4C). Moreover, there was an interaction between the groups and the shifts (F(4, 52) 5 2.79, p , .05), confirming the differential evolution of performance across shifts as a consequence of the lesion. (d) Entries into all previously baited arms. We analyzed the mean percentage of entries in arms that had previously been baited across the shifts to the total number of visits made during each shift (C 5 A/B 3 100; A 5 number of entries in all former baited arms for all sessions of one shift and B 5 total number of visits made during all sessions of each shift). Across shifts, there was a progressive increase of the visits of previously baited arms since as shifts progressed there were more and more arms that had been previously reinforced; intergroup analyses showed no difference among groups (p . .05; Fig. 4D). DISCUSSION The results obtained in this experiment show that rats with lesions to the dorsal striatum were able to learn a fixed goal location in an eight-arm radial maze as rapidly as sham-operated rats. It therefore demonstrates that damage to the dorsal striatum neither disrupts the rats visuo-spatial abilities nor prevents them from successfully associating the goal arm with reward. These findings are consistent with previous studies showing preserved performance in spatial navigation in the Morris water maze (Packard & McGaugh, 1992; McDonald & White, 1994; Thullier, Lalonde, Mahler, Joyal, & Lestienne, 1996), and place learning in a radial-arm maze (McDonald & White, 1995) in rats with damage to the striatum. The impairment in the lesioned rats, in our experiment, occurred when the location of the goal arm was changed even though the learning rule remained the same. We found that the lesioned rats required more sessions and made more errors to learn the new reinforced locations than sham-operated rats. Despite the fact that DS rats required more training, they did in fact learn the task showing a significant decrease in the number of sessions and the number of errors needed to reach the criterion on the first shift when compared with the initial acquisition, as well as a slow decrease in the number of errors and the number of sessions required to learn the successive shifts in the goal arm. In contrast, the performance of the sham-operated rats showed a dramatic drop in the number of sessions required to learn the task and the number of errors they made by the first shift and had, in fact, reached an asymptotic level of performance at this point. Further indications as to whether the rats were actually using the pairing rule between the forced and the free trials to learn the task or were using some other strategy were given by additional analyses. As sham-operated rats faced new locations of the baited arm, they progressively reduced the number of entries they made into the previously baited arm, while increasing the number of correct choices they made on the first trial of each new goal location. This undoubtedly indicates that the sham-operated rats were able to use the infor-
LEARNING IMPAIRMENT AND DORSAL STRIATUM
57
mation given by the forced run before each free trial. As a matter of fact, independently of the strategy used to learn the initial task (only spatial or rule learning), they were able to very quickly base their responses only on the information gained by the forced run at each new change of the reinforced position. However, this appears not to be the case with the DS-lesioned rats. Between the first and the third shift, none of the DS rats was able to reach criterion on the first session and only 40% reached it on the first session of the last two shifts of the goal arm. This suggests that lesioned rats had to relearn the discrimination each time. Furthermore, these rats never increased their rate of correct response on the first trial of each new shift, which certainly indicates that they were not using the information given during the forced run to localize the reinforced arm. Therefore, it would appear that while sham-operated rats learned the rule of associating the forced run with the free run and could apply it to new “problems,” rats with damage to the dorsal striatum were unable to use the rule, improving their performance most likely with spatial learning, animals associating distal cues with food. Such an impairment in using some form of rule learning has also been shown by Robbins, Giardini, Jones, Reading, and Sahakian (1990) when they trained rats in a conditional task using discriminative stimuli for instrumental spatial responses. These results are consistent with others who found that lesioned rats were able to learn the initial visuo-spatial task in a water maze as well as control rats, but when the platform was relocated, the lesioned rats either searched for the platform in the previous location (McDonald & White, 1994) or used a combination of proximal and position response cues to locate the new position of the platform (Whishaw et al., 1987). Thus, striatum-lesioned rats relied on spatial cues and had difficulty using the strategy required to base their responses on different sets of cues in the same environment in much the same way as the rats in our experiment. The fact that the percentage of trials in which striatum-lesioned animals returned in the arm previously baited remained stable across the different shifts in the goal arm suggests that their inability to find rapidly the reinforced location was not due to let perseverating behavior. Even though the only statistical difference for both groups was a preference, on the first shift, in visiting the arms located at 135° from the goal position, which corresponds to the position of the initially reinforced goal arm, neither lesioned rats nor sham-operated rats showed a preference toward a particular angular entry. Furthermore, as lesioned rats required more sessions to reach criterion across shifts and therefore were presented more times a reinforced goal arm it could have been expected that they would show a perseverative behavior toward these former reinforced arms. The similar evolution of the percentage of visits to all previously reinforced positions for lesioned and control animals confirms the lack of perseverative behavior in rats without a functional striatum and is in contrast with some other results in the literature showing perseverative behavior in caudate-lesioned animals which have been interpreted as an inability to either “give up” or inhibit an ongoing action, or to a failure to initiate a new response (Kirkby, 1969; Devan, Goad, & Petri, 1996). The lack of perseverative behavior, in our experiment, indicates that dorsal striatum-lesioned animals had no specific difficulties in abandoning the previously correct spatial cues. Actually, in order to shift from one goal location to
58
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
another, successfully, in our procedure, rats had to abandon one fixed arrangement of spatial cues for another, and also to achieve rapidly the process of selecting visuo-spatial cues during the forced run and to recall the selected cues during the free trial when rats had to return to the arm visited during the forced trial. Therefore, the observed inability to use the information gained by the forced run, as well as the absence of perseveration in DS rats, could indicate a disruption of the working memory component: dorsal striatum rats would have difficulties withholding the information necessary to relocate the baited arm in between the free and the forced run. However, the possible implication of the dorsal striatum in working memory process has been ruled out by several authors showing that there was no deficit of caudate-lesioned animals in learning the standard eight-arm radial maze task (Becker, Walker, & Olton, 1980; Cook & Kesner, 1984; Packard et al., 1989). Moreover, caudatelesioned rats were not impaired in the working memory component, or in ability to discriminate among arms, of the four-arms-baited, four-arms-unbaited task (Packard & White, 1990) as animals rarely reentered an arm that had previously been visited within the daily trial. In contrast, the reference memory component of this task which required animals to discriminate between the sets of baited and unbaited arms was specifically impaired. Thus, in Packard and White’s experiment striatal lesioned rats were impaired at learning which of several arms were baited. Similarly, in our experiment, lesioned animals could learn the initial position of the reinforced arm based on a single set of environmental cues (matching-to-sample or win-stay rule), but were unable to transfer this rule to other sets of cues in the same environment. Even though a retrieval deficit could lead to such an impairment, a deficit in cue selection can also explain DS rats performances. Dorsal striatum could play an important role in the animal’s capacity to extract, during one forced trial, the necessary information. Lesion to the DS could, therefore, lead to a difficulty in selecting the relevant cues among all the ones available in the experimental situation. It seems likely that the rule-based learning impairment in rats with lesions to the dorsal striatum in our experiment could be interpreted as a deficit in the processing of identification or selection of relevant information which allows the association of one set of visuo-spatial cues with reinforcement, and thus the correct response. Such an hypothesis appears to be consistent with the results obtained (i) in rats, in Packard and McGaugh’s experiment (1992) showing a selective deficit in cue learning in a two-platform water maze and the performance deficits observed in dorsal striatum lesioned rats when tested in a 13-dish arena (Van Golf Racht-Delatour, 1997); and (ii) in primates for whom the DS would be involved in stimulus identification (for a review, see Wise, Murray, & Gerfen, 1996). In humans, attentional failures have commonly been used to explain cognitive deficits observed in patients suffering from Parkinson’s disease (Brown & Marsden, 1988; Wright, Burns, Geffen, & Geffen, 1990; Meco, Gasparani, & Doricchi, 1996; Whithaar, Brouwer, & Deelman, 1996). Moreover, deficits of visuo-spatial memory and rule learning in patients with striatal dysfunction are described and interpreted as an inability in reengaging attention on a new task and in maintaining a new rule (Saint Cyr, Taylor, & Lang, 1988; Partiot, Verin, Teixeira-Ferreira, Agid, & Dubois, 1996). Furthermore, most studies have stressed the similarities of cognitive deficits that follow Parkinson’s disease and those that follow frontal lobe dysfunctions giving further evidence
LEARNING IMPAIRMENT AND DORSAL STRIATUM
59
for a striato-frontal cooperation. Such data reflect the neuroanatomical findings that describe the frontal cortex-basal ganglia interconnection (Alexander, De Long, & Strick, 1986). The topographical organization of prefronto-striatal projections has also been demonstrated in rats (Groenewegen & Berendse, 1994) and according to the neuroanatomy, our site of lesion mainly receives projections from the dorsal part of the prefrontal cortex (anterior cingulate cortex and dorsal prelimbic cortex). Hence, such neuroanatomical data can account for the preserved spatial learning as well as the unimpaired fixed goal location task that has also been observed in prelimbic-lesioned rats while the acquisition of a variable-goal location task was disrupted (Delatour and Gisquet-Verrier; 1996). This parallel pattern of results gives further arguments supporting the already described basal ganglia and frontal cortex functional interactions, certainly involved in the behavioral flexibility necessary to solve new problems. REFERENCES Alexander, G. E., De Long, M. R., & Strick, P. L. (1986). Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annual Review of Neurosciences, 9, 357–381. Becker, J. T., Walker, J. A., & Olton, D. S. (1980). Neuroanatomical basis of spatial memory. Brain Research, 200, 307–320. Brown, R. G., & Marsden, C. D. (1988). Internal versus external cues and the control of attention in Parkinson’s disease. Brain, 111, 323–345. Colombo, P. J., Davis, H. P., & Volpe, B. T. (1989). Allocentric spatial and tactile memory impairments in rats with dorsal caudate lesions are affected by preoperative behavioral training. Behavioral Neuroscience, 103 (6), 1242–1250. Cook, D., & Kesner, R. P. (1988). Caudate nucleus and memory for egocentric localization. Behavioral and Neural Biology, 49, 332–343. Delatour, B., & Gisquet-Verrier, P. (1996). Prelimbic cortex specific lesions disrupt delayedvariable response tasks in the rat. Behavioral Neuroscience, 110 (6), 1282–1298. Devan, B. D., Goad, E. H., & Petri, H. L. (1996). Dissociation of hippocampal and striatal contributions to spatial navigation in the water maze. Neurobiology of Learning and Memory, 66, 305–323. Downes, J. J., Roberts, A. C., Sahakian, B. J., Evenden, J. L., Morris, R. G., & Robbins, T. W. (1989). Impaired extra-dimensional shift performance in medicated and unmedicated Parkinson’s disease: Evidence for a specific attentional dysfunction. Neuropsychologia, 27, 1329 – 1343. Dubois, B., & Pillon, B. (1997). Cognitive deficits in Parkinson’s disease. Journal of Neurology, 244, 2– 8. El Massioui, N., & Van Golf Racht-Delatour, B. (1997). Contrasting effects of central nucleus of the amygdala and dorsal striatum lesions on active avoidance learning and its contextual modification. Neuroscience Research Communications, 21 (2), 103–111. Faglioni, P., Scarpa, M., Botti, C., & Ferrari, V. (1995). Parkinson’s disease affects automatic and spares intentional verbal learning a stochastic approach to explicit learning processes. Cortex, 31, 597– 617. Freedman, M., & Oscar-Berman, M. (1986). Selective delayed-response deficits in Parkinson’s and Alzheimer’s disease. Archive of Neurology, 43, 886 – 890. Gerfen, C. R. (1992). The neostriatal mosaic: Multiple levels of compartmental organization in the basal ganglia. Annual Review of Neuroscience, 15, 285–320. Green, R. H., Beatty, W. W., & Schwartzbaum, J. S. (1967). Comparative effects of septohippocampal and caudate lesions on avoidance behavior in rats. Journal of Comparative and Physiological Psychology, 64 (3), 444 – 452.
60
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
Groenewegen, H. J., & Berendse, H. W. (1994). Anatomical relationships between the prefrontal cortex and basal ganglia in the rat. In A.M. Thierry (Ed.), Motor and cognitive function of the prefrontal cortex (pp. 57–77). Berlin: Springer-Verlang. Heindel, W. C., Salmon, D. P., Shults, C. W., Walicke, P. A., & Butters, N. (1989). Neuropsychological evidence for multiple implicit memory systems: A comparison of Alzheimer’s, Huntington’s and Parkinson’s disease patients. Journal of Neuroscience, 9, 582–587. Kirkby, R. J. (1969). Caudate nucleus lesions and perseverative behavior. Physiology and Behavior, 4, 451– 454. Kirkby, R. J., & Polgar, S. (1974). Active avoidance in the laboratory rat following lesions of the dorsal or ventral caudate nucleus. Physiological Psychology, 2 (3A), 301–306. Knopman, D. S., & Nissen, M. J. (1991). Procedural learning is impaired in Huntington’s disease: Evidence from the serial reaction time task. Neuropsychologia, 29, 245–254. McDonald, R. J., & White, N. M. (1993). A triple dissociation of memory systems: Hippocampus, amygdala, and dorsal striatum. Behavioral Neuroscience, 107 (1), 3–22. McDonald, R. J., & White, N. M. (1994). Parallel information processing in the water maze: Evidence for independent memory systems involving dorsal striatum and hippocampus. Behavioral and Neural Biology, 61, 260 –270. McDonald, R. J., & White, N. M. (1995). Hippocampal and nonhippocampal contributions to place learning in rats. Behavioral Neuroscience, 108 (4), 579 –593. Meco, G., Gasparani, M., & Dorricchi, F. (1996). Attentional functions in multiple system atrophy and Parkinson’s desease. Journal of Neurology, Neurosurgery and Psychiatry, 60, 393–398. Mohr, E., Fabbrini, G., Ruggieri, S., Fedio, P., & Chase, T. N. (1987). Cognitive concomitants of dopamine system stimulation in Parkinsonian patients. Journal of Neurology, Neurosurgery and Psychiatry, 50, 1192–1196. ¨ berg, R. G. E., & Divac, I. (1979). “Cognitive” functions of the neostriatum. In I. Divac & R. G. E. O ¨ berg (Eds.), The neostriatum (pp. 291–313). New York: Pergamon. O Oscar-Berman, M., McNamara, P., & Freedman, M. (1991). Delayed-response tasks: Parallels between experimental ablation studies and findings in patients with frontal lesions. In H. S. Levin, H. M. Eisenberg, & A. L. Benton (Eds.), Frontal lobe function and dysfunction (pp. 230 –255). Oxford: Oxford Univ. Press. Owen, A. M., James, M., Leigh, P. N., Summers, B. A., Quinn, N., & Marsden, C. D. (1993). Fronto-striatal cognitive deficits at different stages of Parkinson’s disease. Brain, 115, 1727– 1751. Owen, A. M., Roberts, A. C., Hodges, J. R., Summers, B. A., Polkey, C. E., & Robbins, T. R. (1993). Contrasting mechanisms of impaired attentional set-shifting in patients with frontal lobe damage or Parkinson’s disease. Brain, 116, 1159 –1175. Packard, M. G., Hirsh, R., & White, N. M. (1989). Differential effects of fornix and caudate nucleus lesions on two radial maze tasks: Evidence for multiple memory systems. Journal of Neuroscience, 9 (5), 1465–1472. Packard, M. G., & White, N. M. (1990). Lesions of the caudate nucleus selectively impair “reference memory” acquisition in the radial maze. Behavioral and Neural Biology, 53, 39 –50. Packard, M. G., & McGaugh, J. L. (1992). Double dissociation of fornix and caudate nucleus lesions on acquisistion of two water maze tasks: Further evidence for multiple memory systems. Behavioral Neuroscience, 106, 439 – 446. Partiot, A., Ve´rin, M., Pillon, B., Teixeira-Ferreira, C., Agid, Y., & Dubois, B. (1996). Delayed response tasks in basal ganglia lesions in man: Further evidence for a striato-frontal cooperation in behavioural adaptation. Neuropsychologia, 34 (7), 709 –721. Paxinos, G., & Watson, C. (1986). The rat brain in stereotaxic coordinates (2nd ed.). San Diego, CA: Academic Press. Philipps, A. G., & Carr, G. D. (1987). Cognition and the basal ganglia: A possible substrate for procedural knowledge. The Canadian Journal of Neurological Sciences, 14, 381–385. Pillon, B., Deweer, B., Michon, A., Malapani, C., Agid, Y., & Dubois, B. (1994). Are explicit memory disorders of progressive supranuclear palsy related to damage to striatofrontal circuits? Comparison with Alzheimer’s, Parkinson’s and Huntington’s diseases. Neurology, 44, 1264 – 1270.
LEARNING IMPAIRMENT AND DORSAL STRIATUM
61
Reading, P. J., Dunnett, S. B., & Robbins, T. W. (1991). Dissociable roles of the ventral, medial and lateral striatum on the acquisition and performance of a complex visual stimulus-response habit. Behavioural Brain Research, 45, 147–161. Robbins, T. W., Giardini, V., Jones G. H., Reading, P., & Sahakian, B. J. (1990). Effects of dopamine depletion from the caudate-putamen and nucleus accumbens septi on the acquisition and performance of a conditional discrimination task. Behavioural Brain Research, 38, 243–261. Saint-Cyr, J. A., Taylor, A. E., & Lang, A. E. (1988). Procedural learning and neostriatal dysfunction in man. Brain, 111, 941–959. Taylor, A. E., Saint-Cyr, J. A., & Lang, A. E. (1986). Frontal lobe dysfunction in Parkinson’s disease: The cortical focus of neostriatal outflow. Brain, 109, 279 –292. Taylor, A. E., Saint-Cyr, J. A., & Lang, A. E. (1990). Sub-cognitive processing in the frontocaudate ‘Complex Loop’: the role of the striatum. Alzheimer Disease and Associated Disorders, 4, 150 –160. Taylor, A. E., & Saint Cyr, J. A. (1995). The neuropsychology of Parkinson’s disease. Brain and Cognition, 28, 281–296. Thullier, F., Lalonde, R., Mahler, P., Joyal, C. C., & Lestienne, F. (1996). Dorsal striatum lesions in rats II: Effects on spatial and non-spatial learning. Archives of Physiology and Biochemistry, 104 (3), 307–312. Van Golf Racht-Delatour, B. (1997). Etude du roˆle du striatum dorsal dans les processus d’apprentissage et de me´moire. The`se de Doctorat en Sciences de l’Universite´ Paris XI-Orsay. Whishaw, I. Q., Mittleman, G., Bunch, S. T., & Dunnett, S. B. (1987). Impairments in the acquisition, retention and selection of spatial navigation strategies after medial caudate putamen lesions in rats. Behavioural Brain Research, 24, 125–138. Wise, S. P., Murray, E. A., & Gerfen, C. R. (1996). The frontal cortex-basal ganglia system in Primates. Critical Reviews in Neurobiology, 10, 317–356. Withaar, F. K., Brouwer, W. H., & Deelman, B. G. (1996). Attentional shifting in Parkinson’s disease: Planning or regulation? Brain and Cognition, 30 (3), 414 – 416. Winocur, G. (1974). Functional dissociation within the caudate nucleus of rats. Journal of Comparitive and Physiological Psychology, 86 (3), 432– 439. Wright, M. J., Burns, R. J., Geffen, G. M., & Geffen, L. B. (1990). Covert orientation of visual attention in Parkinson’s disease: An impairment in the maintenance of attention. Neuropsychologia, 28, 151–159.
Rule-Based Learning Impairment in Rats with Lesions to the Dorsal Striatum Barbara Van Golf Racht-Delatour and Nicole El Massioui Laboratoire de Neurobiologie de l’Apprentissage et de la Me´moire, CNRS - URA 1491, Bt. 446, Universite´ Paris-Sud, 91405 Orsay Cedex, France
The present study examined the effects of lesions to the dorsal striatum (DS) in Sprague-Dawley rats, when tested on the acquisition and successive shifts in the position of a goal arm in an eight-arm radial maze. In the procedure we used, rats had to retrieve the location of one baited arm among the eight arms of the maze after it had just been presented as a sample during a forced trial. After attainment of a fixed learning criterion, rats were submitted to five successive shifts in the goal location. Results showed that DS rats were able to learn the position of the goal arm during the acquisition phase as efficiently as sham-operated rats. In contrast, when the position of the goal arm was shifted, although DS rats were able to learn its new position, they made significantly more errors and required more sessions to reach criterion than sham-operated rats. These results suggested that both groups did not solve the task using the same behavioral strategy. The analysis of responses made suggested that sham-operated rats solved the task using the pairing rule between the forced and the free run (matching-to-sample rule), while DS rats solved the task using only visuospatial processing. These data therefore suggest that the dorsal striatum plays an important role in rule-learning ability. © 1999 Academic Press
INTRODUCTION The striatum is a complex structure known for its role in motor functions (Gerfen, 1992); however, it has also been shown to play an important role in learning and memory processes. In humans, neurodegenerative diseases involving the basal ganglia lead to pathologies, such as Parkinson’s or Huntington’s diseases, in which motor and cognitive deficits are apparent. The major cognitive impairments in both pathologies relate to retrieval processes (Mohr, Fabbrini, Ruggieri, Fedio, & Chase, 1988; Pillon, Deweer, Michon, Malapani, Agid, & Dubois, 1994; Faglioni, Scarpa, Botti, & Ferrari, 1995; Dubois & Pillon, 1997), as well as rule-finding abilities (Freedman & Oscar-Berman, 1986; Oscar-Berman, McNamara, & Freedman, 1991; Taylor & Saint Cyr, 1995). However, these diseases also result in procedural learning deficit (Heindel, Salmon, Shults, Walicke, & Butters, 1989; Knopman & Nissen, 1991; Saint-Cyr, Taylor, & Lang, 1988) and in an inability to elaborate or maintain new strategies in tasks requiring adaptation to novelty (Downes, Roberts, Sahakian, Evenden, Morris, & Robbins, 1989; Taylor, Saint-Cyr, & Lang, 1990; Owen, James, Leigh, Summers, Quinn, & Marsden, 1993; Owen, RobAddress correspondance and reprints request to Nicole El Massioui, Laboratoire de Neurobiologie de l’Apprentissage et de la Me´moire, CNRS-URA 1491, Bt. 446, Universite´ Paris Sud, 91405 Orsay Cedex, France. Fax (331) 69.15.77.26. E-mail: Nicole. [email protected]. 47
1074-7427/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
48
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
erts, Hodges, Summers, Polkey, & Robbins, 1993; Withaar, Brouwer, & Deelman, 1996). From the many studies carried out in the rat, where lesions have been made to different subregions of the striatum, it is generally believed that the basal ganglia and, in particular, the striatum play an important role in problem solving. Rats with lesions to the dorsal striatum have been shown to be mainly impaired in learning tasks where they have to make a specific motor response when presented with a reinforced discrete stimulus, such as signaled active avoidance in a two-way shuttle box (Green, Beatty, & Schwartzbaum, 1967; Winocur, 1974; Kirkby & Polgar, 1974; El Massioui & Van Golf Racht-Delatour, 1997), passive avoidance tasks (Winocur, 1974), conditional discriminations (Robbins, Giardini, Jones, Reading, & Sahakian, 1990; Reading, Dunnett, & Robbins, 1991), and win-stay procedures in the eight-arm radial maze (Packard, Hirsh, & White, 1989; McDonald & White, 1993). The difficulty to learn a stimulus-response (S-R) association seems to be the main impairment elicited by dorsal striatum damage. Furthermore, dorsal striatum appears to be involved in procedural learning and memory as all these experiments require that the rats initiate or ¨ berg & Divac, 1979), confirming a hypothesis withhold a specific response (O postulated by Philipps and Carr (1987) that the basal ganglia is necessary for processing procedural knowledge. Moreover, it has been shown that other subregions of the striatum are also important for learning. For example, rats with lesions to the lateral striatum show an inappropriate response to the discriminative stimuli in conditional discrimination tasks (Robbins et al., 1990; Reading et al., 1991). Similarly, lesions to the medial striatum result in a deficit in rats to select, develop, or access appropriate strategies in cue- and place-learning tasks in a Morris water maze (Whishaw, Mittleman, Bunch, & Dunnett, 1987). Furthermore, lesions to the caudate result in an inability to acquire a reference memory component of the task whereas the variable or working memory component was left intact (Colombo, Davis, & Volpe, 1989; Packard & White, 1990). The first group of authors discussed the observed impairments in terms of a deficit in tasks that required retention of rule information useful for improving performance across trials whereas the latter attributed the reference memory impairment to an inability to process correctly the stimulus-response association necessary to learn the baited arm’s position. Therefore, it seems likely that the basal ganglia, and more specifically the dorsal striatum, is necessary for problem solving and would consequently play an essential role in the acquisition and application of response rules. The aim of the present experiment was to determine whether lesions to the dorsal striatum would impair rats’ ability to use a behavioral shift in strategy in order to learn the successive transfer of a goal location in an eight-arm radial maze. The task consisted in associating the baited arm that had just been presented as a single sample during a forced trial, among the eight arms of the maze until the attainment of the learning criterion. This task, as well as the successive shifts in the goal arm location, could be solved either by visuospatial-based learning or by rule-based learning. The optimal strategy, the use of the pairing rule, would only allow a high level of performance.
LEARNING IMPAIRMENT AND DORSAL STRIATUM
49
MATERIAL AND METHODS Subjects Nineteen male Sprague-Dawley rats (Iffa Credo) weighing between 260 and 315 g were housed in pairs in a temperature-controlled colony room and maintained on a 12-h light-dark cycle (lights on 8:00 AM). On arrival in the laboratory, rats were allowed ad libitum access to food and water and handled on a daily basis. Following surgical procedures to lesion the dorsal striatum, rats underwent a food deprivation schedule to reduce their weight to 85% of their original free-feeding body weight. Apparatus The apparatus was a wooden eight-arm maze painted gray and elevated 70 cm above the floor surface. Each arm (60 3 12 cm) radiated from a central octagonal platform (30 cm diameter) and was equipped with a transparent perspex “tunnel” (44 cm long, 13 cm wide, and 13 cm high) and a food cup at the end. White perspex panels could be placed at the start of the tunnel to prevent an arm to be visited. The experimental room contained various extra maze cues. Surgery Rats were randomly assigned either to the experimental group in which lesions were made in the dorsal striatum (n 5 11) or to the sham-operated group (n 5 8). Injections of atropine sulfate 0.025% (Meram, 0.15 ml, ip) were given to rats to prevent respiratory problems, 15 min prior to anesthetizing with pentobarbital (50 mg/kg). Rats were then placed in a stereotaxic apparatus and the cranium was exposed. Holes were drilled in the skull and incisions made in the dura to make bilateral injections into the dorsal striatum. A stainless-steel cannulae mounted onto a microinjector on the stereotaxic frame was lowered into the dorsal striatum and bilateral lesions were made using N-methyl-D-aspartate (NMDA, Sigma; 37 mg/ml). Injections of 0.2 ml per site were made at the following coordinates relative to bregma (Paxinos & Watson, 1986) AP 1 1.6, ML 6 1.6, and DV 23.8 and AP 1 1.6, ML 6 3.2, and DV 23.8 for anterior sites; AP 1 0.2, ML 6 1.8, and DV 24.4 and AP 1 0.2, ML 6 3.8, and DV 24.4 for posterior sites. Sham-operated animals underwent operating procedures identical to those performed on dorsal striatum-lesioned rats, with the exception that the cannulae was not lowered into the brain. Animals with NMDA microinjections were given a single injection of Valium (Roche) 10 mg/kg following surgery to prevent seizure activity. One week after the end of surgery rats were progressively restricted in food, and behavioral testing began 1 week later. Procedure Rats were habituated to the maze 1 day before commencing the first phase of training. All arms were baited with chocolate cereal and animals were allowed to explore the maze until all arms had been visited or 10 min had elapsed. The number of arms visited and the order in which they were visited were recorded. Following this day, training sessions, consisting of four trials
50
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
divided into two phases, commenced. In the first phase (forced run), seven of the eight arms were blocked and rats were placed in the central platform and allowed to visit the baited goal arm (for each animal the goal arm was the forth entry it had made on the preliminary training). When rats had returned to the central platform, they were removed from the maze and placed in a holding box in the experimental room, and the remaining panels were removed for the second phase (free run). Rats had free access to all arms but only the arm baited during the forced trial was baited in the second phase. Rats completed the trial when they had visited the correct arm and the number of errors was scored as entries in unbaited arms. Three further trials were given per day (every forced run was followed by a free run), and to eliminate kinesthetic and olfactory strategies, rats were placed on the central platform facing either West, South, North, or East and the maze was rotated between each trial. Rats were tested every day with always the same goal arm position reinforced until they reached a criterion of an average number of errors, equal to or less than one, per session. After reaching the criterion, animals were tested on the same procedure using a different goal arm to test their ability to shift strategy. Five different shifts of the goal arm were tested in total. The number of arms separating the new goal location from the previous one as well as the direction was the same for all animals (three first shifts, clockwise 13, 13, 13; two last shifts, counterclockwise 22, 22). So, for example, if during the acquisition phase the reinforced arm was Number 1, the new goal locations for the five subsequent shifts were arms Numbers 4, 7, 2, 8, and 6. Histology On completion of the experiment, histological verification of the lesions was made. Rats were anesthetized with pentobarbital and perfused transcardially with isotonic saline buffer followed by 10% paraformaldehyde. The brains were removed and postfixed in a 16% sucrose solution for 3 days and stored at 220°C. Coronal brains sections cut at 40 mm and stained with Cresyl violet were used to assess the extent of the lesion. RESULTS Histology Figure 1 shows an example of a schematic reconstruction of the extent of lesion to the dorsal striatum. The lesions showed characteristic necrosis of tissue including glia cells. Although lesions had been made in 11 rats, 4 of them were discarded either because the lesion extended to the corpus callosum or was too restricted. The data from the remaining 7 rats were included in statistical analysis. Behavior Mean Number of Sessions and Errors to Reach Criterion (Fig. 2) Initial acquisition. Rats with lesions to the dorsal striatum had no difficulty in learning the initial visuo-spatial task, as they showed no difference in the number of sessions as well as in the number of errors to reach criterion as that shown by the sham-operated rats (F , 1 in both cases).
LEARNING IMPAIRMENT AND DORSAL STRIATUM
51
FIG. 1. NMDA lesions of the dorsal striatum. Maximum and minimum extent of the lesions are indicated by hatched areas (maximum, left-directed hatches; minimum, right-directed hatches). Numbers on the right-hand side of the figures represent coordinates relative to bregma (Paxinos & Watson, 1986).
Shifts of the reinforced goal arm. A comparison of performance between the initial acquisition phase and the first shift showed that both the sham-operated and the lesioned rats improved their performance as shown by a decrease in the number of sessions required to reach criterion (F(1, 7) 5 71.21, p , .001; and F(1, 6) 5 8.72, p , .05, respectively) and in the number of errors (F(1, 7) 5 56.85, p , .001; and F(1, 6) 5 28.51, p , .01, respectively). Over the five successive shifts, rats with striatal lesions needed significantly more sessions to reach criterion (F(1, 13) 5 9.14, p , .01) and made more errors than sham-operated rats (F(1, 13) 5 9.43, p , .01) with a significant groups by sessions interaction (F(4, 52) 5 2.64, p , .05), suggesting that sham-operated rats learned to find each new locations more rapidly than the rats with lesions to the dorsal striatum. However, although sham-operated rats were faster to learn the five shifts when compared to striatum-lesioned rats, they rapidly reached a plateau of performance which showed no further improve over shifts (F(4, 28) 5 1.42, ns for the number of sessions and 5 2.63, ns for the number of errors) as the case for SD rats (F(4, 24) 5 3.02, p , .05, for the number of sessions and 5 7.81, p , .01, for the number of errors).
52
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
FIG. 2. Mean number of sessions (1SEM; upper panel) and mean number of errors (1SEM; lower panel) made to reach the criterion on the initial phase of acquisition (A) and on the successive shifts of goal location (S1 to S5) in rats with lesions to the dorsal striatum and sham-operated animals.
Percentage of Rats Reaching Criterion Statistical analyses, using a Fisher test, showed that a significantly bigger percentage of sham-operated rats reached criterion on the first session of shift 2 and 3, compared with DS-lesioned rats (p , .05 and p , .025, respectively). There was no group difference in performance on the initial acquisition or the other shifts (Fig. 3). In addition, analysis to assess the number of sessions required for all rats (100%) in each group to reach the criterion showed, with a one-tail Student t test, that the sham-operated rats required significantly fewer sessions to reach criterion than the DS rats (p , .01). Type of Errors We classified the errors made by both groups of animals according to the angular distance separating the goal arm and the arm entered by each rat. Thus, 45° corresponds to an entry in one of the arms just adjacent to the correct
FIG. 3. Mean percentage of rats reaching the criterion on each session of acquisition of the baited goal arm location as well as on each session needed to learn the five shifts of the reinforced position in sham-operated and dorsal striatum-lesioned rats.
LEARNING IMPAIRMENT AND DORSAL STRIATUM
53
54
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
LEARNING IMPAIRMENT AND DORSAL STRIATUM
55
one; 90° means that there was one arm in between the baited arm and the visited one; 135° means that there were two arms in between the arm presented during the forced run and the one chosen during the free run; 180° corresponds to an opposite choice from the correct goal position. The total number of errors made at 45 or 90 or 135° was divided by 2 (as they corresponded either to clockwise or counterclockwise errors). The distribution of the different types of angular errors for both experimental groups shows that only the number of errors made at 135° during the first shift is significantly higher for the sham-operated animals (F(1, 7) 5 6.06, p , .05) and for the dorsal striatum-lesioned rats (F(1, 6) 5 16.89, p , .01). Furthermore, in rats with striatal lesions, there was a lesser rate of errors made in the opposite direction to the reinforced arm (180°) on shifts 1, 2, and 3 (p , .05 and ps , .01, respectively). Analyses of Strategies Four different analyses were carried out to assess whether rats were using particular types of strategies. (a) Correct choice on the first trial of the first session of each shift. To test whether rats were learning the rule of pairing the arm presented during the forced run with the one to choose in the free run, we analyzed the mean number of rats that made a correct choice on the first trial of the first session of each shift. The results showed that sham-operated animals made significantly more correct choices on this first trial than striatum-lesioned rats (F(1, 13) 5 13.62, p , .01; Fig. 4A). (b) Entries in the initially baited arm. To test whether rats showed a persevering behavior, we analyzed the percentage of visits in the initially baited arm with the total number of visits for each shift. The results showed that both the control rats and the striatum-lesioned rats decreased their number of entries in the initially baited arm across shifts (F(4, 52) 5 4.57, p , .005) with no intergroup difference (F , 1; Fig. 4B). (c) Entries into the previously baited arm. To test whether rats were able to adapt their behavior to the demand of the experiment, and therefore abandon the visit of the previous baited arm when a new shift occurred, we analyzed the mean percentage of entries in the previously baited arm to the total number of visits made during each shift (C 5 A/B 3 100; A 5 number of entries in the former baited arm for all sessions of one shift and B 5 total number of visits made during all sessions of each shift). Results showed that, across shifts, sham-operated rats made fewer entries into the previously reinforced goal arm than dorsal striatum-lesioned animals (F(1, 13) 5 4.92, p , .05). There was a
FIG. 4. Analysis of strategies used by rats with lesions to the dorsal striatum and shamoperated rats. (A) The mean percentage (6SEM) of animals making a correct choice on the first trial of each new shift of goal location for each experimental group. (B) The mean percentage of entries (6SEM) to the acquisition goal arm across the five successive shifts for both groups of animals. (C) The mean percentage of visits to the previously reinforced arm across the different shifts of the goal location in lesioned and sham-operated rats. (D) The mean percentage of entries (6SEM) in all previous goal arms across the five successive modifications of the baited arm location for both experimental groups.
56
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
significant decrease in the number of entries to the previously baited arm across shifts in sham-operated animals (F(4, 28) 5 7.05, p , .01), whereas dorsal striatum rats kept visiting the previously reinforced location on a steady percentage of trials from the first to the last shift (F , 1; Fig. 4C). Moreover, there was an interaction between the groups and the shifts (F(4, 52) 5 2.79, p , .05), confirming the differential evolution of performance across shifts as a consequence of the lesion. (d) Entries into all previously baited arms. We analyzed the mean percentage of entries in arms that had previously been baited across the shifts to the total number of visits made during each shift (C 5 A/B 3 100; A 5 number of entries in all former baited arms for all sessions of one shift and B 5 total number of visits made during all sessions of each shift). Across shifts, there was a progressive increase of the visits of previously baited arms since as shifts progressed there were more and more arms that had been previously reinforced; intergroup analyses showed no difference among groups (p . .05; Fig. 4D). DISCUSSION The results obtained in this experiment show that rats with lesions to the dorsal striatum were able to learn a fixed goal location in an eight-arm radial maze as rapidly as sham-operated rats. It therefore demonstrates that damage to the dorsal striatum neither disrupts the rats visuo-spatial abilities nor prevents them from successfully associating the goal arm with reward. These findings are consistent with previous studies showing preserved performance in spatial navigation in the Morris water maze (Packard & McGaugh, 1992; McDonald & White, 1994; Thullier, Lalonde, Mahler, Joyal, & Lestienne, 1996), and place learning in a radial-arm maze (McDonald & White, 1995) in rats with damage to the striatum. The impairment in the lesioned rats, in our experiment, occurred when the location of the goal arm was changed even though the learning rule remained the same. We found that the lesioned rats required more sessions and made more errors to learn the new reinforced locations than sham-operated rats. Despite the fact that DS rats required more training, they did in fact learn the task showing a significant decrease in the number of sessions and the number of errors needed to reach the criterion on the first shift when compared with the initial acquisition, as well as a slow decrease in the number of errors and the number of sessions required to learn the successive shifts in the goal arm. In contrast, the performance of the sham-operated rats showed a dramatic drop in the number of sessions required to learn the task and the number of errors they made by the first shift and had, in fact, reached an asymptotic level of performance at this point. Further indications as to whether the rats were actually using the pairing rule between the forced and the free trials to learn the task or were using some other strategy were given by additional analyses. As sham-operated rats faced new locations of the baited arm, they progressively reduced the number of entries they made into the previously baited arm, while increasing the number of correct choices they made on the first trial of each new goal location. This undoubtedly indicates that the sham-operated rats were able to use the infor-
LEARNING IMPAIRMENT AND DORSAL STRIATUM
57
mation given by the forced run before each free trial. As a matter of fact, independently of the strategy used to learn the initial task (only spatial or rule learning), they were able to very quickly base their responses only on the information gained by the forced run at each new change of the reinforced position. However, this appears not to be the case with the DS-lesioned rats. Between the first and the third shift, none of the DS rats was able to reach criterion on the first session and only 40% reached it on the first session of the last two shifts of the goal arm. This suggests that lesioned rats had to relearn the discrimination each time. Furthermore, these rats never increased their rate of correct response on the first trial of each new shift, which certainly indicates that they were not using the information given during the forced run to localize the reinforced arm. Therefore, it would appear that while sham-operated rats learned the rule of associating the forced run with the free run and could apply it to new “problems,” rats with damage to the dorsal striatum were unable to use the rule, improving their performance most likely with spatial learning, animals associating distal cues with food. Such an impairment in using some form of rule learning has also been shown by Robbins, Giardini, Jones, Reading, and Sahakian (1990) when they trained rats in a conditional task using discriminative stimuli for instrumental spatial responses. These results are consistent with others who found that lesioned rats were able to learn the initial visuo-spatial task in a water maze as well as control rats, but when the platform was relocated, the lesioned rats either searched for the platform in the previous location (McDonald & White, 1994) or used a combination of proximal and position response cues to locate the new position of the platform (Whishaw et al., 1987). Thus, striatum-lesioned rats relied on spatial cues and had difficulty using the strategy required to base their responses on different sets of cues in the same environment in much the same way as the rats in our experiment. The fact that the percentage of trials in which striatum-lesioned animals returned in the arm previously baited remained stable across the different shifts in the goal arm suggests that their inability to find rapidly the reinforced location was not due to let perseverating behavior. Even though the only statistical difference for both groups was a preference, on the first shift, in visiting the arms located at 135° from the goal position, which corresponds to the position of the initially reinforced goal arm, neither lesioned rats nor sham-operated rats showed a preference toward a particular angular entry. Furthermore, as lesioned rats required more sessions to reach criterion across shifts and therefore were presented more times a reinforced goal arm it could have been expected that they would show a perseverative behavior toward these former reinforced arms. The similar evolution of the percentage of visits to all previously reinforced positions for lesioned and control animals confirms the lack of perseverative behavior in rats without a functional striatum and is in contrast with some other results in the literature showing perseverative behavior in caudate-lesioned animals which have been interpreted as an inability to either “give up” or inhibit an ongoing action, or to a failure to initiate a new response (Kirkby, 1969; Devan, Goad, & Petri, 1996). The lack of perseverative behavior, in our experiment, indicates that dorsal striatum-lesioned animals had no specific difficulties in abandoning the previously correct spatial cues. Actually, in order to shift from one goal location to
58
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
another, successfully, in our procedure, rats had to abandon one fixed arrangement of spatial cues for another, and also to achieve rapidly the process of selecting visuo-spatial cues during the forced run and to recall the selected cues during the free trial when rats had to return to the arm visited during the forced trial. Therefore, the observed inability to use the information gained by the forced run, as well as the absence of perseveration in DS rats, could indicate a disruption of the working memory component: dorsal striatum rats would have difficulties withholding the information necessary to relocate the baited arm in between the free and the forced run. However, the possible implication of the dorsal striatum in working memory process has been ruled out by several authors showing that there was no deficit of caudate-lesioned animals in learning the standard eight-arm radial maze task (Becker, Walker, & Olton, 1980; Cook & Kesner, 1984; Packard et al., 1989). Moreover, caudatelesioned rats were not impaired in the working memory component, or in ability to discriminate among arms, of the four-arms-baited, four-arms-unbaited task (Packard & White, 1990) as animals rarely reentered an arm that had previously been visited within the daily trial. In contrast, the reference memory component of this task which required animals to discriminate between the sets of baited and unbaited arms was specifically impaired. Thus, in Packard and White’s experiment striatal lesioned rats were impaired at learning which of several arms were baited. Similarly, in our experiment, lesioned animals could learn the initial position of the reinforced arm based on a single set of environmental cues (matching-to-sample or win-stay rule), but were unable to transfer this rule to other sets of cues in the same environment. Even though a retrieval deficit could lead to such an impairment, a deficit in cue selection can also explain DS rats performances. Dorsal striatum could play an important role in the animal’s capacity to extract, during one forced trial, the necessary information. Lesion to the DS could, therefore, lead to a difficulty in selecting the relevant cues among all the ones available in the experimental situation. It seems likely that the rule-based learning impairment in rats with lesions to the dorsal striatum in our experiment could be interpreted as a deficit in the processing of identification or selection of relevant information which allows the association of one set of visuo-spatial cues with reinforcement, and thus the correct response. Such an hypothesis appears to be consistent with the results obtained (i) in rats, in Packard and McGaugh’s experiment (1992) showing a selective deficit in cue learning in a two-platform water maze and the performance deficits observed in dorsal striatum lesioned rats when tested in a 13-dish arena (Van Golf Racht-Delatour, 1997); and (ii) in primates for whom the DS would be involved in stimulus identification (for a review, see Wise, Murray, & Gerfen, 1996). In humans, attentional failures have commonly been used to explain cognitive deficits observed in patients suffering from Parkinson’s disease (Brown & Marsden, 1988; Wright, Burns, Geffen, & Geffen, 1990; Meco, Gasparani, & Doricchi, 1996; Whithaar, Brouwer, & Deelman, 1996). Moreover, deficits of visuo-spatial memory and rule learning in patients with striatal dysfunction are described and interpreted as an inability in reengaging attention on a new task and in maintaining a new rule (Saint Cyr, Taylor, & Lang, 1988; Partiot, Verin, Teixeira-Ferreira, Agid, & Dubois, 1996). Furthermore, most studies have stressed the similarities of cognitive deficits that follow Parkinson’s disease and those that follow frontal lobe dysfunctions giving further evidence
LEARNING IMPAIRMENT AND DORSAL STRIATUM
59
for a striato-frontal cooperation. Such data reflect the neuroanatomical findings that describe the frontal cortex-basal ganglia interconnection (Alexander, De Long, & Strick, 1986). The topographical organization of prefronto-striatal projections has also been demonstrated in rats (Groenewegen & Berendse, 1994) and according to the neuroanatomy, our site of lesion mainly receives projections from the dorsal part of the prefrontal cortex (anterior cingulate cortex and dorsal prelimbic cortex). Hence, such neuroanatomical data can account for the preserved spatial learning as well as the unimpaired fixed goal location task that has also been observed in prelimbic-lesioned rats while the acquisition of a variable-goal location task was disrupted (Delatour and Gisquet-Verrier; 1996). This parallel pattern of results gives further arguments supporting the already described basal ganglia and frontal cortex functional interactions, certainly involved in the behavioral flexibility necessary to solve new problems. REFERENCES Alexander, G. E., De Long, M. R., & Strick, P. L. (1986). Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annual Review of Neurosciences, 9, 357–381. Becker, J. T., Walker, J. A., & Olton, D. S. (1980). Neuroanatomical basis of spatial memory. Brain Research, 200, 307–320. Brown, R. G., & Marsden, C. D. (1988). Internal versus external cues and the control of attention in Parkinson’s disease. Brain, 111, 323–345. Colombo, P. J., Davis, H. P., & Volpe, B. T. (1989). Allocentric spatial and tactile memory impairments in rats with dorsal caudate lesions are affected by preoperative behavioral training. Behavioral Neuroscience, 103 (6), 1242–1250. Cook, D., & Kesner, R. P. (1988). Caudate nucleus and memory for egocentric localization. Behavioral and Neural Biology, 49, 332–343. Delatour, B., & Gisquet-Verrier, P. (1996). Prelimbic cortex specific lesions disrupt delayedvariable response tasks in the rat. Behavioral Neuroscience, 110 (6), 1282–1298. Devan, B. D., Goad, E. H., & Petri, H. L. (1996). Dissociation of hippocampal and striatal contributions to spatial navigation in the water maze. Neurobiology of Learning and Memory, 66, 305–323. Downes, J. J., Roberts, A. C., Sahakian, B. J., Evenden, J. L., Morris, R. G., & Robbins, T. W. (1989). Impaired extra-dimensional shift performance in medicated and unmedicated Parkinson’s disease: Evidence for a specific attentional dysfunction. Neuropsychologia, 27, 1329 – 1343. Dubois, B., & Pillon, B. (1997). Cognitive deficits in Parkinson’s disease. Journal of Neurology, 244, 2– 8. El Massioui, N., & Van Golf Racht-Delatour, B. (1997). Contrasting effects of central nucleus of the amygdala and dorsal striatum lesions on active avoidance learning and its contextual modification. Neuroscience Research Communications, 21 (2), 103–111. Faglioni, P., Scarpa, M., Botti, C., & Ferrari, V. (1995). Parkinson’s disease affects automatic and spares intentional verbal learning a stochastic approach to explicit learning processes. Cortex, 31, 597– 617. Freedman, M., & Oscar-Berman, M. (1986). Selective delayed-response deficits in Parkinson’s and Alzheimer’s disease. Archive of Neurology, 43, 886 – 890. Gerfen, C. R. (1992). The neostriatal mosaic: Multiple levels of compartmental organization in the basal ganglia. Annual Review of Neuroscience, 15, 285–320. Green, R. H., Beatty, W. W., & Schwartzbaum, J. S. (1967). Comparative effects of septohippocampal and caudate lesions on avoidance behavior in rats. Journal of Comparative and Physiological Psychology, 64 (3), 444 – 452.
60
VAN GOLF RACHT-DELATOUR AND EL MASSIOUI
Groenewegen, H. J., & Berendse, H. W. (1994). Anatomical relationships between the prefrontal cortex and basal ganglia in the rat. In A.M. Thierry (Ed.), Motor and cognitive function of the prefrontal cortex (pp. 57–77). Berlin: Springer-Verlang. Heindel, W. C., Salmon, D. P., Shults, C. W., Walicke, P. A., & Butters, N. (1989). Neuropsychological evidence for multiple implicit memory systems: A comparison of Alzheimer’s, Huntington’s and Parkinson’s disease patients. Journal of Neuroscience, 9, 582–587. Kirkby, R. J. (1969). Caudate nucleus lesions and perseverative behavior. Physiology and Behavior, 4, 451– 454. Kirkby, R. J., & Polgar, S. (1974). Active avoidance in the laboratory rat following lesions of the dorsal or ventral caudate nucleus. Physiological Psychology, 2 (3A), 301–306. Knopman, D. S., & Nissen, M. J. (1991). Procedural learning is impaired in Huntington’s disease: Evidence from the serial reaction time task. Neuropsychologia, 29, 245–254. McDonald, R. J., & White, N. M. (1993). A triple dissociation of memory systems: Hippocampus, amygdala, and dorsal striatum. Behavioral Neuroscience, 107 (1), 3–22. McDonald, R. J., & White, N. M. (1994). Parallel information processing in the water maze: Evidence for independent memory systems involving dorsal striatum and hippocampus. Behavioral and Neural Biology, 61, 260 –270. McDonald, R. J., & White, N. M. (1995). Hippocampal and nonhippocampal contributions to place learning in rats. Behavioral Neuroscience, 108 (4), 579 –593. Meco, G., Gasparani, M., & Dorricchi, F. (1996). Attentional functions in multiple system atrophy and Parkinson’s desease. Journal of Neurology, Neurosurgery and Psychiatry, 60, 393–398. Mohr, E., Fabbrini, G., Ruggieri, S., Fedio, P., & Chase, T. N. (1987). Cognitive concomitants of dopamine system stimulation in Parkinsonian patients. Journal of Neurology, Neurosurgery and Psychiatry, 50, 1192–1196. ¨ berg, R. G. E., & Divac, I. (1979). “Cognitive” functions of the neostriatum. In I. Divac & R. G. E. O ¨ berg (Eds.), The neostriatum (pp. 291–313). New York: Pergamon. O Oscar-Berman, M., McNamara, P., & Freedman, M. (1991). Delayed-response tasks: Parallels between experimental ablation studies and findings in patients with frontal lesions. In H. S. Levin, H. M. Eisenberg, & A. L. Benton (Eds.), Frontal lobe function and dysfunction (pp. 230 –255). Oxford: Oxford Univ. Press. Owen, A. M., James, M., Leigh, P. N., Summers, B. A., Quinn, N., & Marsden, C. D. (1993). Fronto-striatal cognitive deficits at different stages of Parkinson’s disease. Brain, 115, 1727– 1751. Owen, A. M., Roberts, A. C., Hodges, J. R., Summers, B. A., Polkey, C. E., & Robbins, T. R. (1993). Contrasting mechanisms of impaired attentional set-shifting in patients with frontal lobe damage or Parkinson’s disease. Brain, 116, 1159 –1175. Packard, M. G., Hirsh, R., & White, N. M. (1989). Differential effects of fornix and caudate nucleus lesions on two radial maze tasks: Evidence for multiple memory systems. Journal of Neuroscience, 9 (5), 1465–1472. Packard, M. G., & White, N. M. (1990). Lesions of the caudate nucleus selectively impair “reference memory” acquisition in the radial maze. Behavioral and Neural Biology, 53, 39 –50. Packard, M. G., & McGaugh, J. L. (1992). Double dissociation of fornix and caudate nucleus lesions on acquisistion of two water maze tasks: Further evidence for multiple memory systems. Behavioral Neuroscience, 106, 439 – 446. Partiot, A., Ve´rin, M., Pillon, B., Teixeira-Ferreira, C., Agid, Y., & Dubois, B. (1996). Delayed response tasks in basal ganglia lesions in man: Further evidence for a striato-frontal cooperation in behavioural adaptation. Neuropsychologia, 34 (7), 709 –721. Paxinos, G., & Watson, C. (1986). The rat brain in stereotaxic coordinates (2nd ed.). San Diego, CA: Academic Press. Philipps, A. G., & Carr, G. D. (1987). Cognition and the basal ganglia: A possible substrate for procedural knowledge. The Canadian Journal of Neurological Sciences, 14, 381–385. Pillon, B., Deweer, B., Michon, A., Malapani, C., Agid, Y., & Dubois, B. (1994). Are explicit memory disorders of progressive supranuclear palsy related to damage to striatofrontal circuits? Comparison with Alzheimer’s, Parkinson’s and Huntington’s diseases. Neurology, 44, 1264 – 1270.
LEARNING IMPAIRMENT AND DORSAL STRIATUM
61
Reading, P. J., Dunnett, S. B., & Robbins, T. W. (1991). Dissociable roles of the ventral, medial and lateral striatum on the acquisition and performance of a complex visual stimulus-response habit. Behavioural Brain Research, 45, 147–161. Robbins, T. W., Giardini, V., Jones G. H., Reading, P., & Sahakian, B. J. (1990). Effects of dopamine depletion from the caudate-putamen and nucleus accumbens septi on the acquisition and performance of a conditional discrimination task. Behavioural Brain Research, 38, 243–261. Saint-Cyr, J. A., Taylor, A. E., & Lang, A. E. (1988). Procedural learning and neostriatal dysfunction in man. Brain, 111, 941–959. Taylor, A. E., Saint-Cyr, J. A., & Lang, A. E. (1986). Frontal lobe dysfunction in Parkinson’s disease: The cortical focus of neostriatal outflow. Brain, 109, 279 –292. Taylor, A. E., Saint-Cyr, J. A., & Lang, A. E. (1990). Sub-cognitive processing in the frontocaudate ‘Complex Loop’: the role of the striatum. Alzheimer Disease and Associated Disorders, 4, 150 –160. Taylor, A. E., & Saint Cyr, J. A. (1995). The neuropsychology of Parkinson’s disease. Brain and Cognition, 28, 281–296. Thullier, F., Lalonde, R., Mahler, P., Joyal, C. C., & Lestienne, F. (1996). Dorsal striatum lesions in rats II: Effects on spatial and non-spatial learning. Archives of Physiology and Biochemistry, 104 (3), 307–312. Van Golf Racht-Delatour, B. (1997). Etude du roˆle du striatum dorsal dans les processus d’apprentissage et de me´moire. The`se de Doctorat en Sciences de l’Universite´ Paris XI-Orsay. Whishaw, I. Q., Mittleman, G., Bunch, S. T., & Dunnett, S. B. (1987). Impairments in the acquisition, retention and selection of spatial navigation strategies after medial caudate putamen lesions in rats. Behavioural Brain Research, 24, 125–138. Wise, S. P., Murray, E. A., & Gerfen, C. R. (1996). The frontal cortex-basal ganglia system in Primates. Critical Reviews in Neurobiology, 10, 317–356. Withaar, F. K., Brouwer, W. H., & Deelman, B. G. (1996). Attentional shifting in Parkinson’s disease: Planning or regulation? Brain and Cognition, 30 (3), 414 – 416. Winocur, G. (1974). Functional dissociation within the caudate nucleus of rats. Journal of Comparitive and Physiological Psychology, 86 (3), 432– 439. Wright, M. J., Burns, R. J., Geffen, G. M., & Geffen, L. B. (1990). Covert orientation of visual attention in Parkinson’s disease: An impairment in the maintenance of attention. Neuropsychologia, 28, 151–159.