- Email: [email protected]
Lesions in the basal ganglion and hippocampus on performance in a Wisconsin Card Sorting Test-like task in pigeons
Physiology & Behavior 85 (2005) 324 – 332
Lesions in the basal ganglion and hippocampus on performance in a Wisconsin Card Sorting Test-like task in pigeons Shigeru Watanabe* Department of Psychology, Keio University, Mita 2-15-45, Minato-Ku, Tokyo, Japan Received 24 October 2004; received in revised form 14 April 2005; accepted 25 April 2005
Abstract Previous studies of LPO (lobus parolfactorium) and hippocampal lesions in pigeons suggest function of cognitive flexibility in LPO and memory consolidation in hippocampus [Watanabe S. Effects of hippocampal lesions on repeated acquisition of spatial discrimination in pigeons. Behav Brain Res 2001;120:59 – 66. [40]; Watanabe S. Effects of LPO lesions on repeated acquisition of spatial discrimination in pigeons. Brain Behav Evol 2002;58:333 – 342. [41]]. Here, a test similar to the Wisconsin Card Sorting Test was applied to pigeons. The test consisted of four discriminations, namely red – green color discrimination and its reversal, left – right spatial discrimination and its reversal. In each trial stimuli were presented until the correct choice occurred. Ten successive correct trials without wrong response were defined as the criterion of discrimination. When the subjects reached the criterion in one discrimination, they were trained on one of three other discrimination tasks in the next session. These four discriminations were trained repeatedly in random sequence. After the birds have been well learned the WCST-like task, their hippocampus or lobus parolfactorium (LPO), the avian basal ganglion, was damaged. A sham lesion group received anesthesia only. Both lesions impaired the WCST-like test. Lesions of the LPO increased the number of errors, while the hippocampal lesions increased the number of trials to reach the criterion only. The number of errors reflects difficulty in finding the correct stimulus or cognitive flexibility, while the number of trials reflects difficulty in stable responding or memory consolidation. The present results suggest that LPO has the function of cognitive flexibility. D 2005 Elsevier Inc. All rights reserved. Keywords: Card sorting; Visual discrimination; Cognitive flexibility; Brain lesion
1. Introduction The Wisconsin Card Sorting Test (WCST) is a standard test used to examine cognitive flexibility in humans [2]. This test requires patients to sort cards according to different rule based on different perceptual dimensions such as color or shape. Thus, the patients need to shift their strategy of sorting (extradimensional shift). Ability of changing strategy according to contingency reflects cognitive flexibility. Three different regions in the telencephalon could be involved in this task. First, the frontal lobe is involved, because lesions in this region caused deficits in this task [3,6,25,34]. Studies using event related potentials (ERP) and near-infrared spectroscopy * Tel.: +81 3 5443 3896; fax: +81 3 5443 3897. E-mail address: [email protected]. 0031-9384/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2005.04.020
(NIRS) suggest that the prefrontalis is involved in the card sorting task [1,7]. Functional dissociation between the dorsolateral and inferomedial prefrontalis has also been reported. Lesions in the dorsolateral part resulted in difficulty in changing strategy [39], while lesions in the inferomedial part had no effect [37]. The second region that is involved is the basal ganglion. Patients with damage to the basal ganglia show deficits in the WCST [23]. Parkinson disease also results in deficits in the card sorting task [5]. Patients with Huntington’s disease showed impairment in reversal learning [18] and patients with Parkinson disease continuously employed simple notoptimal strategy probabilistic category learning in comparison with normal control subjects who adopted a complex optimal strategy through learning [36]. The prefrontalstriatal circuit seems to function as a system of cognitive flexibility in humans.
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
The hippocampus is the third region that has a role in this task, although earlier an study with patient HM demonstrated no impairment [24]. Hermann et al. [13] suggested that a kind of neural noise caused by temporal epilepsy propagated to the frontalis and disturbed WCST performance. Cronin-Golomb [4] also observed poor card sorting performance in patients with hippocampal sclerosis but supported the idea of the hippocampus as a comparator. More recently, Giovagnoli [9] examined 112 patients with temporal epilepsy and found severe deficits in patients with left hippocampal lesions. She explained the deficits with respect to the learning or associative function of hippocampus. Thus, there are cases of deficits in the WCST that are due to hippocampal dysfunction but different researchers have proposed different explanations for the deficits. The WCST has been examined by animal experiments also. In these experiments intradimensional shift and extradimentional shift of discrimination were examined. Cognitive flexibility in animals can be measured by reversal learning procedure but the subject has just two alternatives of discrimination, namely original learning or its reversal. Thus, the subjects need to attend to one particular stimulus dimension in reversal learning task. On the other hand, subjects have to attend to two or more stimulus dimensions in WCST-like task. Therefore, the subjects require more extensive flexibility in the WCST-like task. O’Reilly, Noelle, Braver and Cohen [26] developed a WCST-like task for monkeys and found that damage to the dorsolateral prefrontalis caused deficits in dimensional shift, whereas damage to the orbitofrontal part resulted in errors in specific features of the stimuli. These results agree to the prefrontal deficits observed in human patients. Interestingly, 6hydroxydopamine lesion of the prefrontal cortex in marmosets improved extradimensional shifts [32]. It was suggested that depletion of dopamine in neocortex elevated dopamine level in the striatum, and that the elevated dopamine enhanced ‘‘attention’’. As I described above, neuropsychological results suggests that the basal ganglion has not only motor function but also cognitive function. Animal researches also supported cognitive function of the basal ganglion [15,21]. Particularly, lesions in the basal forebrain result in deficits in serial reversal learning without disrupting acquisition of new learning in marmosets [33]. The deficits, deficits in cognitive flexibility, are similar to those caused by orbitofrontal lesions. Involvement of the dorsal striatum was demonstrated in rodents also [27,30]. I have previously reported that birds have pallio-striatal system that has a role in cognitive flexibility [41,42]. The task that I employed was a modified repeated acquisition task. Originally this procedure was developed for behavioral pharmacology to obtain baseline performance of acquisition. The task involved a choice among three keys in an operant chamber. One of the three keys was designated as correct key, and peck on which was rewarded by food. Every time the subjects learned the position of the correct key, they were
325
trained on different discriminations in which one of two previously incorrect keys became the correct key. Thus, the subjects had to learn new position discrimination every time they reached the criterion of discrimination. Because stable baseline of acquisition is obtained by repetition of this procedure, it gives a baseline performance of acquisition; hence we can compare acquisition before and after lesions in within subjects design. Lesions in the lobus parolfactorium (LPO or medial striatum), which is part of the avian basal ganglion, and the Wulst (or hyperpallium) caused deficits in this task [41,42]. On the other hand, damage to the ectostriatum (or entopallium), the major visual processing area in the avian telencephalon, did not impair learning of this task [42]. Thus, the Wulst-LPO system (pallio-striatal system) could be the avian counterpart to the mammalian prefrontal-striatal system of cognitive flexibility. Another area that resulted in deficits in this repeated acquisition task when damaged is the hippocampus [38]. The deficits caused by LPO lesion and those resulting from the hippocampal lesion are, however, qualitatively different. The birds with LPO lesions emitted many errors until they found the correct key, while those with the hippocampal damage required many trials to reach the criterion (10 consecutive correct choice). The former probably represents deficits in cognitive flexibility while the latter reflects deficits in memory consolidation. Repeated acquisition has characteristics in common with the card sorting task, because subjects are required to find a new correct response. However, repeated acquisition represents simple position discrimination even though the correct stimulus is changed. In other word, there is intradimensional shift but no extradimensional shift. I have therefore developed a card sorting analogue test for pigeons. The test is basically concurrent training of color and position discrimination using two colors and two keys, similar to earlier experiments by Schade and Bitterman [3fs5]. While Schade and Bitterman [35] changed tasks each day, subjects in the present experiment were trained on the one task until they reached the criterion then on the next task. Another important procedural difference between Schade and Bitterman [22] and the present experiment was contingency after incorrect response. Incorrect peck switched off keys in the formal procedure but the keys were lit until correct response occurred in the latter procedure. Subjects continue to peck the incorrect key if they cannot switch their choice. Thus, two behavioral indices, number of errors and number of trials, were available in the present procedure. Purpose of the repeated training is making baseline performance of this test before brain lesions. The test consists of four discrimination tasks, two position (left– right) discrimination tasks and two color (green – red) discrimination tasks. In the color discrimination tasks, positions of the key are irrelevant, so the pigeons had to ignore the key’s position. In the position discrimination tasks, color of the key is irrelevant, so the pigeons had to
326
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
The experimental chamber was a conventional operant chamber for pigeons (30 25 30 cm, MED) with two pecking keys (diameter 2 cm) on a front panel and a ceiling lamp. Green or red lamps (28 V, DC) could illuminate the keys. The distance between the keys was 4.5 cm. The chamber was placed in an enclosure box (40 63.5 60 cm). White noise (75 dB outside the enclosure box) was continuously presented during the experiment as masking noise. A computer with a MED-SKED system controlled the experiment.
color discrimination tasks and two position discrimination tasks. Every task was simultaneous discrimination of color or position. One of the keys was illuminated by a red light while the other by a green one. The position of the colors was changed for each trial in accordance with the Gellermann series. If the correct key was pecked, both keys were switched off and the hopper comes out. If the incorrect key was pecked, nothing happened until the correct key was pecked. For the red-green color discrimination task, a peck on the red key (correct key) was reinforced by 4 s access to the food hopper while peck on the green key was extinguished. Presentation of the hopper was followed by a 5 s blackout, which was the inter-trial interval. A peck on the green key (incorrect key) resulted in no reinforcement with both keys illuminated until the subject pecked the red key. This discriminative training procedure was reversed in the other color discrimination task (green – red discrimination). That is, a response to the green key was reinforced while that to the red key was extinguished. Because the position of the colored keys was irrelevant in these tasks, the subject had to attend to the color dimension. In the left – right position discrimination, pigeons were trained to peck the left key but not the right key. This discrimination paradigm was reversed in the other position discrimination task (right –left discrimination). In these position discrimination task either key was illuminated by red or green but the color was an irrelevant cue. Training of each task continued until the subject produced 10 correct choices in succession (criterion of discrimination) or had completed 80 trials (the maximum number of trials in one session). If the subject reached the criterion, one of the other three discrimination tasks was used in the next session. If the subject failed to reach the criterion within 80 trials, the same discrimination task was repeated in the next training session. Therefore, the subjects had to learn a different discrimination every time they accomplished the prior one. Shifts from one discrimination task to another were determined quasi-randomly. Because there were two color discrimination and two position discrimination tasks, both intradimensional reversal and extradimensional shift occurred. In the case of extradimensional shift, the subject has to attend to a new dimension of the discriminative stimulus. These training continued until the subjects had reached the criterion within 80 trials for all discrimination tasks. They then completed 12 additional tasks to obtain a baseline of acquisition. Then, the subjects were divided into two groups of four. One group received LPO lesions and the other hippocampal lesions. After 5 days recovery period, all subjects were retrained on this WCSTlike test until they had again successfully completed 12 discrimination tasks.
2.4. Procedure
2.5. Lesions
All subjects were trained to peck each key, and then trained on the WCST-like test. The test consisted of two
The subjects were injected with 50 mg/kg ketamine (Ketalar50, Sankyo, i.m.) then fixed in a stereotaxic
ignore the color of the key. Each time the subjects reached the criterion of discrimination, a new task was implemented. We can compare intradimensional shift (from red S+ to green S+, and from left S+ to right S+) with extradimensional shift (from position discriminations to color discriminations and vice versa). Subjects are repeatedly trained on these four discriminations to obtain baseline performance. The task has common features to WCST that tests cognitive flexibility but there are some differences between the task here and WCST. While conventional WCST for humans contains many discriminations, the task here contains only four discriminations and two discriminations in one stimulus dimension is reversal of the other. The intradimensional shift is always reversal of the preceding discrimination, while the extradimensional shift is not always reversal. For example, a sequence in which color discrimination with red S+ was trained first, then position discrimination with left S+, and then color discrimination with green S+ contains reversal of color discrimination. Whereas a sequence with red S+, position with left S+, then red S+ contains no color reversal. Here, I examined the effects of lesions to the LPO and hippocampus on the WCST like test in pigeons.
2. Materials and methods 2.1. Ethics The animals were treated in accordance with the Keio University guidelines for experimental animals. 2.2. Subjects Twelve experimentally naive homing pigeons (Columba livia) were used. They were maintained at 80% of their free feeding weights. Water was freely available in the living cages. 2.3. Apparatus
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
apparatus for the pigeon. Local anesthesia (Xylocaine) was injected in the incision sites. The LPO lesion was made bilaterally with a radio frequency generated by a lesion generator (Radionix, RF5). The coordinates were anterior 9.5 mm, lateral 1.5 mm and depth 6.5 mm from the dura, and anterior 11.0 mm, lateral 2.0 mm and depth 7.0 mm from the dura. The hippocampal lesions were made by aspiration under a microscope for microsurgery. The sham lesion group received just anesthesia. 2.6. Histology The subjects were injected with an overdose of pentobarbital (Nembutal), and then perfused with physiological saline followed by 4% formalin. Then, the brain was removed from the skull and kept in formalin for more than 5 days. The brain was then transferred into a 30% sucrose solution for 2– 3 days and sectioned at 50 Am with a cryostat. Every second section was mounted and stained with cresyl violet. The sections were examined under a microscope and the damages were reconstructed on a standard brain atlas using a computer.
327
consolidation. The other was based on the number of errors. If the subjects had difficulty in finding the correct key, they emitted more errors before correct choice. Thus, this is an index of behavioral rigidity. A mixed design ANOVA in which individuals were used as repeated measure was used for analysis. Because there were individual differences in baseline performance before lesion, performance after the lesion was divided by performance before lesions to evaluate effects of lesions. Because expectation of this ratio is 1.0, effects of lesion can be tested with a single group t-test with expectation of 1.0 in each group. Paired t-test for the number of errors and that of trials was also carried out to clarify difference between before and after the lesions. Because the training consisted of two different tasks, I could separate the data into two different ways, namely stimulus dimension (color task or position task) and dimensional shift (intradimension or extradimensional shift). After/before lesion ratio was calculated, then repeated measure ANOVA in which individuals were used as repeated measure was performed.
2.7. Data analysis
3. Results
Data obtained from 12 discrimination tasks before and after the lesions were used for the analysis. Two behavioral indices were used for analysis. One was the number of trials to reach the criterion. Because the criterion was 10 successive correct trials without incorrect response, this is an index of steady correct responding or memory
3.1. Histological results Figs. 1 and 2 show reconstruction of damage. In the LPO group, the medial LPO was damaged in every subject. Bird 413 had slightly asymmetrical damage and the ventral part of the LPO remained intact. Bird 621 had
Fig. 1. Reconstruction of damages to the LPO. ECT, ectostriatum; LPO, lobus parolfactorius; PA, paleostriatum augmentatum; PP, the paleostriatum primitivum.
328
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
Fig. 2. Reconstruction of damages to the hippocampus. APH, area parahippocampalis; CDL, corticoid dorsolateralis; HA, hyperstriatum accessorium; HIPP, hippocampus; HV, hyperstriatum ventrale.
the smallest damage. In the hippocampus group, the dorsomedial part was damaged in every subject. The area parahippocampalis was also damaged in each subject although some individual differences in the extent of the damage were observed. The caudal part of the hippocampus remained intact in Bird 412. Bird 612 had some damage to the neostriatum.
ence between them (paired t-test, t(3) = 0.85). Mean number of errors before and after the lesion in the LPO group was 41.7 and 127.2, respectively, and there was a significant difference between them (paired t-test, t(3) = 3.18, p < 0.05).
3.2. Behavioral results 3.2.1. Over-all performance Fig. 3 shows mean of the after/before ratio in each group. One-way ANOVA of the error ratio gives significant main effect of lesions ( F(2/11) = 5.79, p < 0.05). There was a significant difference between the sham group and the LPO (t(6) = 3.03, p < 0.05), but not between the sham and HIP lesion groups (t(6) = 0.73). There was also a significant difference between the LPO and HIP groups (t(6) = 3.35, p < 0.05). Thus, the LPO lesions, but not the hippocampal lesions, caused increment of the number of errors. Single group t-test of the after/before ratio (expectation = 1.0) shows a significant difference in the LPO group (t(3) = 2.95, p < 0.05) but not in the hippocampus or sham groups (t(3) = 0.59 and 0.98, respectively). Mean number of errors before and after the lesion in the sham group was 50.1 and 47.3, respectively, and there was no significant differ-
Fig. 3. Behavioral results. The upper panels show mean ratio of after/before lesion of the number of errors. The lower panel shows mean ratio of number of trials. Vertical bars indicate the standard errors. **p < 0.05.
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
Mean number of errors before and after the lesion in the HIP group was 51.9 and 66.0, respectively, and there was no significant difference between them (paired t-test, t(3) = 0.32). One-way ANOVA of the trial ratio did not give significant main effect of lesions ( F(2/11) = 1.39). Single group t-test of the after/before ratio shows a significant difference in the HIP group (t(3) = 4.03, p < 0.01) but not in other groups (t(3) = 1.24 and 0.01 for the sham and LPO group, respectively). Thus, the hippocampal lesions increased the number of trials to reach the criterion but the LPO lesions did not. Mean number of trials before and after the lesion in the sham group was 50.0 and 48.6, respectively, and there was no significant difference between them (paired t-test, t(3) = 0.34). Mean number of trials before and after the lesion in the LPO group was 48.9 and 64.3, respectively, and there was no significant difference between them (paired t-test, t(3) = 0.69). Mean number of trials before and after the lesion in the HIP group was 41.5 and 61.9, respectively, and there was a tendency of difference between them (paired t-test, t(3) = 2.72, p < 0.10). There was a significant difference in after/before ratio between the number of errors and that of trials in the LPO group (paired t-test, t(3) = 3.43, p < 0.05), but not in the HIP group (t(3) = 0.50). Thus, error specific deficit after the LPO lesions was clearly observed but trial specific deficit after the hippocampal lesions was not clear. 3.2.2. Comparison of intradimensional and extradimensional transfer Because two color discriminations and two position discriminations were randomly trained in the present training, we can separate intradimensional and extradimensional shifts from (n)th and (n + 1)th tasks. Performance of the subjects was divided into these two categories and after/ before lesions ratio was calculated. To analyze difference between the two kinds of shift in intact birds, results of all birds before surgery were combined. There was no statistically significant difference in the number of trials to reach the criterion between the intra- and extradimensional shifts (t(11) = 1.46), although mean of the intradimensional shift was slightly higher than that of the extradimensional shift (m = 51.0 and 43.7, respectively). Analysis of the number of errors neither gave a significant difference (t(11) = 0.14). Mean was 23.0 and 23.8 for intra- and extradimensional shifts, respectively. These examinations show that intact pigeons could learn both shift equally. Effects of lesions on each shift were examined by repeated measure two-way ANOVA using after/before ratio. ANOVA for the number of errors gives no significant main effect of the shift or the lesion groups ( F(1/23) = 2.84 and F(2/23) = 0.97, respectively). Their interaction was neither significant ( F(2/23) = 1.00). ANOVA for the number of trials shows that main effect of the shift and the groups was
329
not significant ( F(1/23) = 1.08 and F(2/23 = 1.89). There was no significant effect of their interaction ( F(2/ 23) = 0.82). Single group t-test (expectation = 1.0) for after/ before ratio in each group showed a significant difference in the number of trials in extradimensional shift in the HIP group (t(3) = 3.70, p < 0.03) and in the number of errors in intradimensional shift in the LPO group (t(3) = 7.31, p < 0.005). These examinations suggest that selective effects on intra- or extradimensional shift was not clear (Fig. 4). 3.2.3. Comparison of position and color discriminations. The present task consists of color and position discrimination. To analyze difference between the two kinds of transfer in intact birds, results of all birds before surgery were combined. In terms of number of errors, there is no statistically significant difference between the two discriminations (mean of position discrimination was 38.3 and that of the color discrimination 46.0, t(11) = 0.93). The color discrimination was, however, harder to learn than the position discrimination in terms of the number of trials to reach the criterion (mean of the position discrimination was 35.1 and that of the color task 50.2, t(11) = 2.28, p < 0.05). Repeated measure two-way ANOVA of the after/before ratio of the number of errors gives a significant effect of the stimulus dimension ( F(2/23) = 3.63, p < 0.05). There was no significant effect of the groups ( F(1/23) = 1.21) or their interaction ( F(2/23) = 2.05). Repeated measures twoway ANOVA of the after/before ratio of the number of trials gives no significant effect of the stimulus dimension ( F(1/23) = 0.90) or group neither interaction ( F(2/ 23) = 1.78 and 1.56, respectively). Single group t-test of the after/before ratio gives no significant difference in terms of the number of errors or in terms of the trials.
Fig. 4. Comparison of intradimensional and exradimensional shifts. The upper panels show mean ratio of after/before lesion of the number of errors. The lower panel shows mean ratio of number of trials. Vertical bars indicate the standard errors. **p < 0.05.
330
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
Fig. 5. Comparison of color and position discriminations. The upper panels show mean ratio of after/before lesion of the number of errors. The lower panel shows mean ratio of number of trials. Vertical bars indicate the standard errors. **p < 0.05.
Thus, selective deficits depending on stimulus dimensions were not clear (Fig. 5).
4. Discussion The present results demonstrated that (1) both LPO and hippocampal lesions impaired a WCST-like test in pigeons, (2) the LPO lesions increased the number of errors, while the hippocampal lesions deficits increased the number of trials to reach criterion and (3) the deficits were not selective to stimulus dimension nor type of transition of the tasks. The present results from the LPO lesions concur with my previous results from the repeated acquisition task [41]. Both experiments demonstrated deficits in searching for the correct stimulus after LPO lesions. Husband and Shimizu [14] also reported deficits in reversal learning after LPO lesion in pigeons. Lesions in area X, which is equivalent to the mammalian nucleus accumbens [19], resulted in repetition of fixed notes in the song of Bengalese finches [17] suggesting rigidity or perseveration of behavior patterns. These observations after area X damage are compatible with the present results. As described in Introduction, increment of the number of errors in this task suggest cognitive rigidity. The avian basal ganglion should have function comparable to mammalian basal ganglion, namely cognitive flexibility. The present deficits observed after hippocampal lesions are also consistent with previous data on repeated acquisition [40]. In the repeated acquisition task, the birds with hippocampal damage required more trials to learn spatial discrimination. The deficits were more severe in the number of trials than in the number of errors. The criterion of discrimination was 10 consecutive correct responding. If
memory is not well consolidated, the animal may have difficulty in stable correct responding. These observations suggest that what the hippocampal damage caused was not deficits in changing behavior or cognitive flexibility but formation of stability of acquired behavior or consolidation of memory. One unpredicted result in the present experiment was non-selective deficits in both the color and position tasks after hippocampal lesions. The hippocampal lesions did not caused deficits in repeated acquisition with added color cues [40]. Because the hippocampus seems to be specialized for the processing of spatial memory, position-specific deficits were expected. The present WCST-like test trained the subjects concurrently on the color and position tasks, thus, the interaction of learning both color and position discrimination might occur. After hippocampal lesions, memory of position discrimination might disturb acquisition of subsequent color discrimination. In the case of repeated acquisition with color cues [40], there was no such competition between the dimensions. The avian paleostriatal complex (or medial striatum) can be divided into three parts, namely the paleostriatum augmentatum (PA or lateral striatum), the paleostriatum primitivum (PP or globus pallidus), and the lobus parolfactorius (LPO). The LPO receives dopaminergic projections from the substantia nigra and sends GABAergic projections to the substantia nigra [11,22]. Met-enkephalin immunoreactivity studies also support a correspondence between the avian LPO-PA complex and the mammalian caudate-putamen [8]. Thus, Reiner et al. [31] suggested evolutionary homology between the avian LPO-PA and mammalian striatum. The medial LPO receives projections from the limbic caudolateral neostriatum, piriform cortex and archistriatum, whereas the lateral LPO receives projections from the Wulst, the pallium externum and the dorsal archistriatum [31]. These connection studies suggest that birds have an LPO-pallial system similar to the basal ganglio-cortical system in mammals. It should be noted, however, that there are still conflicting descriptions of the LPO subdivisions. In mammals, abnormal function of the interconnection between the frontal cortex and the basal ganglia is thought to be critical in such cognitive rigidity [10]. Perkel and Farries [28] suggested similarity between the mammalian basal ganglio-cortical system and songbirds’ neostriato-LPO (areaX) system in cognitive functions. Lesions in the Wulst caused deficits in repeated acquisition [41] and lesions in the neostriatum caudolaterale (NCL) also caused deficits in reversal learning in pigeons [12]. Wulst lesions caused deficits on shift from color discrimination to pattern discrimination [29] serial reversal of discrimination [20]. These results suggest that the birds have a pallio-striatal system similar to that in mammals. As described in Introduction, three regions of the human brain are involved in the WCST, namely, the prefrontalis, basal ganglion and hippocampus. Several studies based on mammalian experi-
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
ments suggested multiple brain memory system in which hippocampus and basal ganglion have different memory function [16,21]. The basal ganglion has function of response formation while the hippocampus has function of spatial expression or learning of relationship among stimuli. Although we need further research to clarify relationship between each mammalian brain structure and its avian counterpart, functional difference between the hippocampus and the basal ganglion was also observed in the present experiment. The present results from the WCST-like test, together with my previous experiments, suggest that the LPO (medial striatum), Wulst (pallium) and hippocampus in the pigeon brains also comprise a system for cognitive flexibility similar to those in human.
Acknowledgments The research was supported by Grant in Aids-for Scientific Research (09410028) and 21st Century Center of Excellence Program (D-1).
References [1] Barcelo F. Electrophysiological evidence of two different types of errors in the Wisconsin Card Sorting Test. Neuroreport 1999;10:1299 – 303. [2] Berg M. A simple objective test for measuring flexibility of thinking. J Gen Psychol 1948;39:15 – 22. [3] Corcoran R, Upton D. A role for the hippocampus in card sorting? Cortex 1993;29:293 – 304. [4] Cronin-Golomb A. Abstract thought in aging ad age-related disease. In: In Boller F, Grafman J, editors. Handbook of Neuropsychology, vol. 4. Amsterdam’ Elsevier; 1990. p. 279 – 310. [5] Dimitrov M, Grafman J, Soares AH, Clark K. Concept formation and concept shifting in frontal lesions and Parkinson’s disease patients assessed with the Californian Card Sorting Test. Neuropsychologia 1999;13:135 – 43. [6] Drewe EA. The effect of type and area of brain lesion on Wisconsin Card Sorting Test performance. Cortex 1974;10:159 – 70. [7] Fallgatter AJ, Strik WK. Frontal brain activation during the Wisconsin Card Sorting Test assessed with two-channel near-infrared spectroscopy. Eur Arch Psychiatry Clin Neurosci 1998;248:245 – 9. [8] Galatioto S, Abbate F, Laura R, Naccari F, Germana G. Morphological and immunohistochemical considerations on the basal ganglia in pigeon (Columba livia). Anat Histol Embryol 1998;27:173 – 8. [9] Giovagnoli AR. Relation of sorting impairment to hippocampal damage in temporal lobe epilepsy. Neuropsychologia 2001;39: 140 – 50. [10] Graybiel AM. The basal ganglia and cognitive pattern generator. Schizophr Bull 1997;23:459 – 69. [11] Hall K, Brauth SE, Kitt CA. Retrograde transport of [3H]GABA in the striatotegmental system of the pigeon. Brain Res 1984;310: 157 – 63. [12] Hartmann B, Guenkuerkun O. Selective deficits in reversal learning after neostriatum caudolaterale lesions in pigeons, possible behavioral equivalencies to the mammalian prefrontal system. Behav Brain Res 1998;96:125 – 33. [13] Hermann BP, Wyler AR, Richey ET. Wisconsin Card Sorting Test performance in patients with complex partial seizures of temporal lobe origin. J Clin Exp Neuropsychol 1988;10:467 – 76.
331
[14] Husband S, Shimizu T. Reversal learning after lesions in the presumptive nucleus accumbens in pigeons. Paper read at 10 th Ann Internl Conf Comp Cog; 2003. [15] Kesner RP, Bolland BL, Dakis M. Memory for spatial locations, motor responses, and objects: triple dissociation among the hippocampus, caudate nucleus, and extrastriate visual cortex. Exp Brain Res 1993;93:462 – 70. [16] Kim JJ, Baxter MG. Multiple brain-memory systems: the whole does not equal the sum of its parts. Trends Neurosci 2001;24: 324 – 30. [17] Kobayashi K, Uno H, Okanoya K. Partial lesions in the anterior forebrain pathway affect song production in adult Bengalese finches. Neuroreport 2001;84:483 – 8. [18] Lawrence AD, Sahakian BJ, Rogers RD, Hodges JR, Robbins TW. Discrimination, reversal, and shift learning in Huntington’s disease; mechanisms of impaired response selection. Neuropsychologia 1999;37:1359 – 74. [19] Lewis JW, Ryan SM, Arnold AP, Butcher LL. Evidence for a catecholaminergic projection to area X in the zebra finch. J Comp Neurol 1981;196:347 – 54. [20] Macphail EM, Reilly S. Double dissociation of effects on learning of medial versus lateral hyperstriatal lesions in pigeons. Behav Neurosci 1987;101:513 – 23. [21] McDonald RJ, White NM. A triple dissociation of memory systems: Hippocampus, amygdala, and dorsal striatum. Behav Neurosci 1993;107:3 – 22. [22] Metzger M, Jiang S, Wang J, Braun K. Organization of the dopaminergic innervation of forebrain areas relevant to learning combined immunohistochemical/retrograde tracing study in the domestic chick. J Comp Neurol 1996;376:1 – 27. [23] Milner B. Effect of different brain lesions on card sorting. Arch Neurol 1963;90:90 – 100. [24] Milner B, Corkin S, Teuber HL. Further analysis of the hippocampal amnestic syndrome: 14-year follow-up study of HM. Neuropsychologia 1968;6:215 – 34. [25] Nelson HE. A modified card sorting test sensitive to frontal lobe deficits. Cortex 1976;12:313 – 24. [26] O_Reilly RC, Noelle DC, Braver TS, Cohen JD. Prefrontal cortex and dynamic categorization tasks: representational organization and neuromodulatory control. Cereb Cortex 2002;12:246 – 57. [27] Palencia CA, Ragozzino ME. The influence of NMDA receptors in the dorsomedial striatum on response reversal learning. Neurobiol Learn 6 Mem 2004;82:81 – 9. [28] Perkel D, Farries MA. Complementary Fbottom-up_ and Ftop-down_ approaches to basal ganglia function. Curr Opin Neurobiol 2000; 10:725 – 31. [29] Powers AS, Halasz F, Williams S. The effects of lesions in telencephalic visual areas of pigeons on dimensional shifting. Physiol Behav 1982;29:1099 – 104. [30] Ragozzino ME, Jih J, Tzavos A. Involvement of the dorsomedial striatum in behavioral flexibility: role of muscarinic cholinergic receptors. Brain Res 2002;953:205 – 14. [31] Reiner A, Medina L, Veenman CL. Structural and functional evolution of the basal ganglia in vertebrates. Brain Res Rev 1998; 28:235 – 85. [32] Roberts AC, De Salvia MA, Wilkinson LS, Collins P, Muir JL, Evaritt BJ, et al. 6-Hydroxydopamine lesions of the prefrontal cortex in monkeys enhance performance on an analog of the Wisconsin Card Sorting Test: possible interaction with subcortical dopamine. J Neurosci 1994;14:2531 – 44. [33] Roberts AC, Robbins TW, Evaritt BJ, Jones GH, Sirkia TE, Wilkinson J, et al. The effects of excitotoxic lesions of the basal forebrain on the acquisition, retention and serial reversal of visual discriminations in marmosets. Neuroscience 1990;34:311 – 29. [34] Robinson AL, Heaton RK, Lehman RAW, Stillson DW. The utility of the Wisconsin Card Sorting Test in detecting and localizing frontal lesions. J Consult Clin Psychol 1980;48:605 – 14.
332
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
[35] Schade AF, Bitterman ME. Improvement in habit reversal as related to dimensional set. J Comp Physiol Psychol 1966;62:43 – 8. [36] Shohamy D, Myers CE, Onlaor S, Gluck MA. Role of the basal ganglia in category learning: How do patients with Parkinson’s disease learn? Behav Neurosci 2004;118:676 – 86. [37] Stuss DT, Levine B, Alexander MP, Hong J, Palumbo C, Hamer L, et al. Wisconsin Card Sorting Test performance in patients with focal frontal and posterior brain damage: effects of lesion location and test structure on separable cognitive processes. Neuropsychologia 2000; 38:388 – 402. [38] Veenman CL, Wild JM, Reiner A. Organization of the avian ‘‘corticostriatal’’ projection system: a retrograde and anterograde pathway tracing study in pigeons. J Comp Neurol 1995;354:87 – 126.
[39] Verin M, Partiot A, Pillon B, Malapani C, Aqid Y, Duboias B. Delayed response tasks and prefrontal lesions in man—evidence for self generated patterns of behavior with poor environmental modulation. Neuropsychologia 1993;31:1379 – 96. [40] Watanabe S. Effects of hippocampal lesions on repeated acquisition of spatial discrimination in pigeons. Behav Brain Res 2001;120:59 – 66. [41] Watanabe S. Effects of LPO lesions on repeated acquisition of spatial discrimination in pigeons. Brain Behav Evol 2002;58:333 – 42. [42] Watanabe S. Effects of Wulst and ectostriatal lesions on repeated acquisition of spatial discrimination in pigeons. Cogn Brain Res 2003;17:286 – 92.
Lesions in the basal ganglion and hippocampus on performance in a Wisconsin Card Sorting Test-like task in pigeons Shigeru Watanabe* Department of Psychology, Keio University, Mita 2-15-45, Minato-Ku, Tokyo, Japan Received 24 October 2004; received in revised form 14 April 2005; accepted 25 April 2005
Abstract Previous studies of LPO (lobus parolfactorium) and hippocampal lesions in pigeons suggest function of cognitive flexibility in LPO and memory consolidation in hippocampus [Watanabe S. Effects of hippocampal lesions on repeated acquisition of spatial discrimination in pigeons. Behav Brain Res 2001;120:59 – 66. [40]; Watanabe S. Effects of LPO lesions on repeated acquisition of spatial discrimination in pigeons. Brain Behav Evol 2002;58:333 – 342. [41]]. Here, a test similar to the Wisconsin Card Sorting Test was applied to pigeons. The test consisted of four discriminations, namely red – green color discrimination and its reversal, left – right spatial discrimination and its reversal. In each trial stimuli were presented until the correct choice occurred. Ten successive correct trials without wrong response were defined as the criterion of discrimination. When the subjects reached the criterion in one discrimination, they were trained on one of three other discrimination tasks in the next session. These four discriminations were trained repeatedly in random sequence. After the birds have been well learned the WCST-like task, their hippocampus or lobus parolfactorium (LPO), the avian basal ganglion, was damaged. A sham lesion group received anesthesia only. Both lesions impaired the WCST-like test. Lesions of the LPO increased the number of errors, while the hippocampal lesions increased the number of trials to reach the criterion only. The number of errors reflects difficulty in finding the correct stimulus or cognitive flexibility, while the number of trials reflects difficulty in stable responding or memory consolidation. The present results suggest that LPO has the function of cognitive flexibility. D 2005 Elsevier Inc. All rights reserved. Keywords: Card sorting; Visual discrimination; Cognitive flexibility; Brain lesion
1. Introduction The Wisconsin Card Sorting Test (WCST) is a standard test used to examine cognitive flexibility in humans [2]. This test requires patients to sort cards according to different rule based on different perceptual dimensions such as color or shape. Thus, the patients need to shift their strategy of sorting (extradimensional shift). Ability of changing strategy according to contingency reflects cognitive flexibility. Three different regions in the telencephalon could be involved in this task. First, the frontal lobe is involved, because lesions in this region caused deficits in this task [3,6,25,34]. Studies using event related potentials (ERP) and near-infrared spectroscopy * Tel.: +81 3 5443 3896; fax: +81 3 5443 3897. E-mail address: [email protected]. 0031-9384/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2005.04.020
(NIRS) suggest that the prefrontalis is involved in the card sorting task [1,7]. Functional dissociation between the dorsolateral and inferomedial prefrontalis has also been reported. Lesions in the dorsolateral part resulted in difficulty in changing strategy [39], while lesions in the inferomedial part had no effect [37]. The second region that is involved is the basal ganglion. Patients with damage to the basal ganglia show deficits in the WCST [23]. Parkinson disease also results in deficits in the card sorting task [5]. Patients with Huntington’s disease showed impairment in reversal learning [18] and patients with Parkinson disease continuously employed simple notoptimal strategy probabilistic category learning in comparison with normal control subjects who adopted a complex optimal strategy through learning [36]. The prefrontalstriatal circuit seems to function as a system of cognitive flexibility in humans.
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
The hippocampus is the third region that has a role in this task, although earlier an study with patient HM demonstrated no impairment [24]. Hermann et al. [13] suggested that a kind of neural noise caused by temporal epilepsy propagated to the frontalis and disturbed WCST performance. Cronin-Golomb [4] also observed poor card sorting performance in patients with hippocampal sclerosis but supported the idea of the hippocampus as a comparator. More recently, Giovagnoli [9] examined 112 patients with temporal epilepsy and found severe deficits in patients with left hippocampal lesions. She explained the deficits with respect to the learning or associative function of hippocampus. Thus, there are cases of deficits in the WCST that are due to hippocampal dysfunction but different researchers have proposed different explanations for the deficits. The WCST has been examined by animal experiments also. In these experiments intradimensional shift and extradimentional shift of discrimination were examined. Cognitive flexibility in animals can be measured by reversal learning procedure but the subject has just two alternatives of discrimination, namely original learning or its reversal. Thus, the subjects need to attend to one particular stimulus dimension in reversal learning task. On the other hand, subjects have to attend to two or more stimulus dimensions in WCST-like task. Therefore, the subjects require more extensive flexibility in the WCST-like task. O’Reilly, Noelle, Braver and Cohen [26] developed a WCST-like task for monkeys and found that damage to the dorsolateral prefrontalis caused deficits in dimensional shift, whereas damage to the orbitofrontal part resulted in errors in specific features of the stimuli. These results agree to the prefrontal deficits observed in human patients. Interestingly, 6hydroxydopamine lesion of the prefrontal cortex in marmosets improved extradimensional shifts [32]. It was suggested that depletion of dopamine in neocortex elevated dopamine level in the striatum, and that the elevated dopamine enhanced ‘‘attention’’. As I described above, neuropsychological results suggests that the basal ganglion has not only motor function but also cognitive function. Animal researches also supported cognitive function of the basal ganglion [15,21]. Particularly, lesions in the basal forebrain result in deficits in serial reversal learning without disrupting acquisition of new learning in marmosets [33]. The deficits, deficits in cognitive flexibility, are similar to those caused by orbitofrontal lesions. Involvement of the dorsal striatum was demonstrated in rodents also [27,30]. I have previously reported that birds have pallio-striatal system that has a role in cognitive flexibility [41,42]. The task that I employed was a modified repeated acquisition task. Originally this procedure was developed for behavioral pharmacology to obtain baseline performance of acquisition. The task involved a choice among three keys in an operant chamber. One of the three keys was designated as correct key, and peck on which was rewarded by food. Every time the subjects learned the position of the correct key, they were
325
trained on different discriminations in which one of two previously incorrect keys became the correct key. Thus, the subjects had to learn new position discrimination every time they reached the criterion of discrimination. Because stable baseline of acquisition is obtained by repetition of this procedure, it gives a baseline performance of acquisition; hence we can compare acquisition before and after lesions in within subjects design. Lesions in the lobus parolfactorium (LPO or medial striatum), which is part of the avian basal ganglion, and the Wulst (or hyperpallium) caused deficits in this task [41,42]. On the other hand, damage to the ectostriatum (or entopallium), the major visual processing area in the avian telencephalon, did not impair learning of this task [42]. Thus, the Wulst-LPO system (pallio-striatal system) could be the avian counterpart to the mammalian prefrontal-striatal system of cognitive flexibility. Another area that resulted in deficits in this repeated acquisition task when damaged is the hippocampus [38]. The deficits caused by LPO lesion and those resulting from the hippocampal lesion are, however, qualitatively different. The birds with LPO lesions emitted many errors until they found the correct key, while those with the hippocampal damage required many trials to reach the criterion (10 consecutive correct choice). The former probably represents deficits in cognitive flexibility while the latter reflects deficits in memory consolidation. Repeated acquisition has characteristics in common with the card sorting task, because subjects are required to find a new correct response. However, repeated acquisition represents simple position discrimination even though the correct stimulus is changed. In other word, there is intradimensional shift but no extradimensional shift. I have therefore developed a card sorting analogue test for pigeons. The test is basically concurrent training of color and position discrimination using two colors and two keys, similar to earlier experiments by Schade and Bitterman [3fs5]. While Schade and Bitterman [35] changed tasks each day, subjects in the present experiment were trained on the one task until they reached the criterion then on the next task. Another important procedural difference between Schade and Bitterman [22] and the present experiment was contingency after incorrect response. Incorrect peck switched off keys in the formal procedure but the keys were lit until correct response occurred in the latter procedure. Subjects continue to peck the incorrect key if they cannot switch their choice. Thus, two behavioral indices, number of errors and number of trials, were available in the present procedure. Purpose of the repeated training is making baseline performance of this test before brain lesions. The test consists of four discrimination tasks, two position (left– right) discrimination tasks and two color (green – red) discrimination tasks. In the color discrimination tasks, positions of the key are irrelevant, so the pigeons had to ignore the key’s position. In the position discrimination tasks, color of the key is irrelevant, so the pigeons had to
326
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
The experimental chamber was a conventional operant chamber for pigeons (30 25 30 cm, MED) with two pecking keys (diameter 2 cm) on a front panel and a ceiling lamp. Green or red lamps (28 V, DC) could illuminate the keys. The distance between the keys was 4.5 cm. The chamber was placed in an enclosure box (40 63.5 60 cm). White noise (75 dB outside the enclosure box) was continuously presented during the experiment as masking noise. A computer with a MED-SKED system controlled the experiment.
color discrimination tasks and two position discrimination tasks. Every task was simultaneous discrimination of color or position. One of the keys was illuminated by a red light while the other by a green one. The position of the colors was changed for each trial in accordance with the Gellermann series. If the correct key was pecked, both keys were switched off and the hopper comes out. If the incorrect key was pecked, nothing happened until the correct key was pecked. For the red-green color discrimination task, a peck on the red key (correct key) was reinforced by 4 s access to the food hopper while peck on the green key was extinguished. Presentation of the hopper was followed by a 5 s blackout, which was the inter-trial interval. A peck on the green key (incorrect key) resulted in no reinforcement with both keys illuminated until the subject pecked the red key. This discriminative training procedure was reversed in the other color discrimination task (green – red discrimination). That is, a response to the green key was reinforced while that to the red key was extinguished. Because the position of the colored keys was irrelevant in these tasks, the subject had to attend to the color dimension. In the left – right position discrimination, pigeons were trained to peck the left key but not the right key. This discrimination paradigm was reversed in the other position discrimination task (right –left discrimination). In these position discrimination task either key was illuminated by red or green but the color was an irrelevant cue. Training of each task continued until the subject produced 10 correct choices in succession (criterion of discrimination) or had completed 80 trials (the maximum number of trials in one session). If the subject reached the criterion, one of the other three discrimination tasks was used in the next session. If the subject failed to reach the criterion within 80 trials, the same discrimination task was repeated in the next training session. Therefore, the subjects had to learn a different discrimination every time they accomplished the prior one. Shifts from one discrimination task to another were determined quasi-randomly. Because there were two color discrimination and two position discrimination tasks, both intradimensional reversal and extradimensional shift occurred. In the case of extradimensional shift, the subject has to attend to a new dimension of the discriminative stimulus. These training continued until the subjects had reached the criterion within 80 trials for all discrimination tasks. They then completed 12 additional tasks to obtain a baseline of acquisition. Then, the subjects were divided into two groups of four. One group received LPO lesions and the other hippocampal lesions. After 5 days recovery period, all subjects were retrained on this WCSTlike test until they had again successfully completed 12 discrimination tasks.
2.4. Procedure
2.5. Lesions
All subjects were trained to peck each key, and then trained on the WCST-like test. The test consisted of two
The subjects were injected with 50 mg/kg ketamine (Ketalar50, Sankyo, i.m.) then fixed in a stereotaxic
ignore the color of the key. Each time the subjects reached the criterion of discrimination, a new task was implemented. We can compare intradimensional shift (from red S+ to green S+, and from left S+ to right S+) with extradimensional shift (from position discriminations to color discriminations and vice versa). Subjects are repeatedly trained on these four discriminations to obtain baseline performance. The task has common features to WCST that tests cognitive flexibility but there are some differences between the task here and WCST. While conventional WCST for humans contains many discriminations, the task here contains only four discriminations and two discriminations in one stimulus dimension is reversal of the other. The intradimensional shift is always reversal of the preceding discrimination, while the extradimensional shift is not always reversal. For example, a sequence in which color discrimination with red S+ was trained first, then position discrimination with left S+, and then color discrimination with green S+ contains reversal of color discrimination. Whereas a sequence with red S+, position with left S+, then red S+ contains no color reversal. Here, I examined the effects of lesions to the LPO and hippocampus on the WCST like test in pigeons.
2. Materials and methods 2.1. Ethics The animals were treated in accordance with the Keio University guidelines for experimental animals. 2.2. Subjects Twelve experimentally naive homing pigeons (Columba livia) were used. They were maintained at 80% of their free feeding weights. Water was freely available in the living cages. 2.3. Apparatus
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
apparatus for the pigeon. Local anesthesia (Xylocaine) was injected in the incision sites. The LPO lesion was made bilaterally with a radio frequency generated by a lesion generator (Radionix, RF5). The coordinates were anterior 9.5 mm, lateral 1.5 mm and depth 6.5 mm from the dura, and anterior 11.0 mm, lateral 2.0 mm and depth 7.0 mm from the dura. The hippocampal lesions were made by aspiration under a microscope for microsurgery. The sham lesion group received just anesthesia. 2.6. Histology The subjects were injected with an overdose of pentobarbital (Nembutal), and then perfused with physiological saline followed by 4% formalin. Then, the brain was removed from the skull and kept in formalin for more than 5 days. The brain was then transferred into a 30% sucrose solution for 2– 3 days and sectioned at 50 Am with a cryostat. Every second section was mounted and stained with cresyl violet. The sections were examined under a microscope and the damages were reconstructed on a standard brain atlas using a computer.
327
consolidation. The other was based on the number of errors. If the subjects had difficulty in finding the correct key, they emitted more errors before correct choice. Thus, this is an index of behavioral rigidity. A mixed design ANOVA in which individuals were used as repeated measure was used for analysis. Because there were individual differences in baseline performance before lesion, performance after the lesion was divided by performance before lesions to evaluate effects of lesions. Because expectation of this ratio is 1.0, effects of lesion can be tested with a single group t-test with expectation of 1.0 in each group. Paired t-test for the number of errors and that of trials was also carried out to clarify difference between before and after the lesions. Because the training consisted of two different tasks, I could separate the data into two different ways, namely stimulus dimension (color task or position task) and dimensional shift (intradimension or extradimensional shift). After/before lesion ratio was calculated, then repeated measure ANOVA in which individuals were used as repeated measure was performed.
2.7. Data analysis
3. Results
Data obtained from 12 discrimination tasks before and after the lesions were used for the analysis. Two behavioral indices were used for analysis. One was the number of trials to reach the criterion. Because the criterion was 10 successive correct trials without incorrect response, this is an index of steady correct responding or memory
3.1. Histological results Figs. 1 and 2 show reconstruction of damage. In the LPO group, the medial LPO was damaged in every subject. Bird 413 had slightly asymmetrical damage and the ventral part of the LPO remained intact. Bird 621 had
Fig. 1. Reconstruction of damages to the LPO. ECT, ectostriatum; LPO, lobus parolfactorius; PA, paleostriatum augmentatum; PP, the paleostriatum primitivum.
328
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
Fig. 2. Reconstruction of damages to the hippocampus. APH, area parahippocampalis; CDL, corticoid dorsolateralis; HA, hyperstriatum accessorium; HIPP, hippocampus; HV, hyperstriatum ventrale.
the smallest damage. In the hippocampus group, the dorsomedial part was damaged in every subject. The area parahippocampalis was also damaged in each subject although some individual differences in the extent of the damage were observed. The caudal part of the hippocampus remained intact in Bird 412. Bird 612 had some damage to the neostriatum.
ence between them (paired t-test, t(3) = 0.85). Mean number of errors before and after the lesion in the LPO group was 41.7 and 127.2, respectively, and there was a significant difference between them (paired t-test, t(3) = 3.18, p < 0.05).
3.2. Behavioral results 3.2.1. Over-all performance Fig. 3 shows mean of the after/before ratio in each group. One-way ANOVA of the error ratio gives significant main effect of lesions ( F(2/11) = 5.79, p < 0.05). There was a significant difference between the sham group and the LPO (t(6) = 3.03, p < 0.05), but not between the sham and HIP lesion groups (t(6) = 0.73). There was also a significant difference between the LPO and HIP groups (t(6) = 3.35, p < 0.05). Thus, the LPO lesions, but not the hippocampal lesions, caused increment of the number of errors. Single group t-test of the after/before ratio (expectation = 1.0) shows a significant difference in the LPO group (t(3) = 2.95, p < 0.05) but not in the hippocampus or sham groups (t(3) = 0.59 and 0.98, respectively). Mean number of errors before and after the lesion in the sham group was 50.1 and 47.3, respectively, and there was no significant differ-
Fig. 3. Behavioral results. The upper panels show mean ratio of after/before lesion of the number of errors. The lower panel shows mean ratio of number of trials. Vertical bars indicate the standard errors. **p < 0.05.
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
Mean number of errors before and after the lesion in the HIP group was 51.9 and 66.0, respectively, and there was no significant difference between them (paired t-test, t(3) = 0.32). One-way ANOVA of the trial ratio did not give significant main effect of lesions ( F(2/11) = 1.39). Single group t-test of the after/before ratio shows a significant difference in the HIP group (t(3) = 4.03, p < 0.01) but not in other groups (t(3) = 1.24 and 0.01 for the sham and LPO group, respectively). Thus, the hippocampal lesions increased the number of trials to reach the criterion but the LPO lesions did not. Mean number of trials before and after the lesion in the sham group was 50.0 and 48.6, respectively, and there was no significant difference between them (paired t-test, t(3) = 0.34). Mean number of trials before and after the lesion in the LPO group was 48.9 and 64.3, respectively, and there was no significant difference between them (paired t-test, t(3) = 0.69). Mean number of trials before and after the lesion in the HIP group was 41.5 and 61.9, respectively, and there was a tendency of difference between them (paired t-test, t(3) = 2.72, p < 0.10). There was a significant difference in after/before ratio between the number of errors and that of trials in the LPO group (paired t-test, t(3) = 3.43, p < 0.05), but not in the HIP group (t(3) = 0.50). Thus, error specific deficit after the LPO lesions was clearly observed but trial specific deficit after the hippocampal lesions was not clear. 3.2.2. Comparison of intradimensional and extradimensional transfer Because two color discriminations and two position discriminations were randomly trained in the present training, we can separate intradimensional and extradimensional shifts from (n)th and (n + 1)th tasks. Performance of the subjects was divided into these two categories and after/ before lesions ratio was calculated. To analyze difference between the two kinds of shift in intact birds, results of all birds before surgery were combined. There was no statistically significant difference in the number of trials to reach the criterion between the intra- and extradimensional shifts (t(11) = 1.46), although mean of the intradimensional shift was slightly higher than that of the extradimensional shift (m = 51.0 and 43.7, respectively). Analysis of the number of errors neither gave a significant difference (t(11) = 0.14). Mean was 23.0 and 23.8 for intra- and extradimensional shifts, respectively. These examinations show that intact pigeons could learn both shift equally. Effects of lesions on each shift were examined by repeated measure two-way ANOVA using after/before ratio. ANOVA for the number of errors gives no significant main effect of the shift or the lesion groups ( F(1/23) = 2.84 and F(2/23) = 0.97, respectively). Their interaction was neither significant ( F(2/23) = 1.00). ANOVA for the number of trials shows that main effect of the shift and the groups was
329
not significant ( F(1/23) = 1.08 and F(2/23 = 1.89). There was no significant effect of their interaction ( F(2/ 23) = 0.82). Single group t-test (expectation = 1.0) for after/ before ratio in each group showed a significant difference in the number of trials in extradimensional shift in the HIP group (t(3) = 3.70, p < 0.03) and in the number of errors in intradimensional shift in the LPO group (t(3) = 7.31, p < 0.005). These examinations suggest that selective effects on intra- or extradimensional shift was not clear (Fig. 4). 3.2.3. Comparison of position and color discriminations. The present task consists of color and position discrimination. To analyze difference between the two kinds of transfer in intact birds, results of all birds before surgery were combined. In terms of number of errors, there is no statistically significant difference between the two discriminations (mean of position discrimination was 38.3 and that of the color discrimination 46.0, t(11) = 0.93). The color discrimination was, however, harder to learn than the position discrimination in terms of the number of trials to reach the criterion (mean of the position discrimination was 35.1 and that of the color task 50.2, t(11) = 2.28, p < 0.05). Repeated measure two-way ANOVA of the after/before ratio of the number of errors gives a significant effect of the stimulus dimension ( F(2/23) = 3.63, p < 0.05). There was no significant effect of the groups ( F(1/23) = 1.21) or their interaction ( F(2/23) = 2.05). Repeated measures twoway ANOVA of the after/before ratio of the number of trials gives no significant effect of the stimulus dimension ( F(1/23) = 0.90) or group neither interaction ( F(2/ 23) = 1.78 and 1.56, respectively). Single group t-test of the after/before ratio gives no significant difference in terms of the number of errors or in terms of the trials.
Fig. 4. Comparison of intradimensional and exradimensional shifts. The upper panels show mean ratio of after/before lesion of the number of errors. The lower panel shows mean ratio of number of trials. Vertical bars indicate the standard errors. **p < 0.05.
330
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
Fig. 5. Comparison of color and position discriminations. The upper panels show mean ratio of after/before lesion of the number of errors. The lower panel shows mean ratio of number of trials. Vertical bars indicate the standard errors. **p < 0.05.
Thus, selective deficits depending on stimulus dimensions were not clear (Fig. 5).
4. Discussion The present results demonstrated that (1) both LPO and hippocampal lesions impaired a WCST-like test in pigeons, (2) the LPO lesions increased the number of errors, while the hippocampal lesions deficits increased the number of trials to reach criterion and (3) the deficits were not selective to stimulus dimension nor type of transition of the tasks. The present results from the LPO lesions concur with my previous results from the repeated acquisition task [41]. Both experiments demonstrated deficits in searching for the correct stimulus after LPO lesions. Husband and Shimizu [14] also reported deficits in reversal learning after LPO lesion in pigeons. Lesions in area X, which is equivalent to the mammalian nucleus accumbens [19], resulted in repetition of fixed notes in the song of Bengalese finches [17] suggesting rigidity or perseveration of behavior patterns. These observations after area X damage are compatible with the present results. As described in Introduction, increment of the number of errors in this task suggest cognitive rigidity. The avian basal ganglion should have function comparable to mammalian basal ganglion, namely cognitive flexibility. The present deficits observed after hippocampal lesions are also consistent with previous data on repeated acquisition [40]. In the repeated acquisition task, the birds with hippocampal damage required more trials to learn spatial discrimination. The deficits were more severe in the number of trials than in the number of errors. The criterion of discrimination was 10 consecutive correct responding. If
memory is not well consolidated, the animal may have difficulty in stable correct responding. These observations suggest that what the hippocampal damage caused was not deficits in changing behavior or cognitive flexibility but formation of stability of acquired behavior or consolidation of memory. One unpredicted result in the present experiment was non-selective deficits in both the color and position tasks after hippocampal lesions. The hippocampal lesions did not caused deficits in repeated acquisition with added color cues [40]. Because the hippocampus seems to be specialized for the processing of spatial memory, position-specific deficits were expected. The present WCST-like test trained the subjects concurrently on the color and position tasks, thus, the interaction of learning both color and position discrimination might occur. After hippocampal lesions, memory of position discrimination might disturb acquisition of subsequent color discrimination. In the case of repeated acquisition with color cues [40], there was no such competition between the dimensions. The avian paleostriatal complex (or medial striatum) can be divided into three parts, namely the paleostriatum augmentatum (PA or lateral striatum), the paleostriatum primitivum (PP or globus pallidus), and the lobus parolfactorius (LPO). The LPO receives dopaminergic projections from the substantia nigra and sends GABAergic projections to the substantia nigra [11,22]. Met-enkephalin immunoreactivity studies also support a correspondence between the avian LPO-PA complex and the mammalian caudate-putamen [8]. Thus, Reiner et al. [31] suggested evolutionary homology between the avian LPO-PA and mammalian striatum. The medial LPO receives projections from the limbic caudolateral neostriatum, piriform cortex and archistriatum, whereas the lateral LPO receives projections from the Wulst, the pallium externum and the dorsal archistriatum [31]. These connection studies suggest that birds have an LPO-pallial system similar to the basal ganglio-cortical system in mammals. It should be noted, however, that there are still conflicting descriptions of the LPO subdivisions. In mammals, abnormal function of the interconnection between the frontal cortex and the basal ganglia is thought to be critical in such cognitive rigidity [10]. Perkel and Farries [28] suggested similarity between the mammalian basal ganglio-cortical system and songbirds’ neostriato-LPO (areaX) system in cognitive functions. Lesions in the Wulst caused deficits in repeated acquisition [41] and lesions in the neostriatum caudolaterale (NCL) also caused deficits in reversal learning in pigeons [12]. Wulst lesions caused deficits on shift from color discrimination to pattern discrimination [29] serial reversal of discrimination [20]. These results suggest that the birds have a pallio-striatal system similar to that in mammals. As described in Introduction, three regions of the human brain are involved in the WCST, namely, the prefrontalis, basal ganglion and hippocampus. Several studies based on mammalian experi-
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
ments suggested multiple brain memory system in which hippocampus and basal ganglion have different memory function [16,21]. The basal ganglion has function of response formation while the hippocampus has function of spatial expression or learning of relationship among stimuli. Although we need further research to clarify relationship between each mammalian brain structure and its avian counterpart, functional difference between the hippocampus and the basal ganglion was also observed in the present experiment. The present results from the WCST-like test, together with my previous experiments, suggest that the LPO (medial striatum), Wulst (pallium) and hippocampus in the pigeon brains also comprise a system for cognitive flexibility similar to those in human.
Acknowledgments The research was supported by Grant in Aids-for Scientific Research (09410028) and 21st Century Center of Excellence Program (D-1).
References [1] Barcelo F. Electrophysiological evidence of two different types of errors in the Wisconsin Card Sorting Test. Neuroreport 1999;10:1299 – 303. [2] Berg M. A simple objective test for measuring flexibility of thinking. J Gen Psychol 1948;39:15 – 22. [3] Corcoran R, Upton D. A role for the hippocampus in card sorting? Cortex 1993;29:293 – 304. [4] Cronin-Golomb A. Abstract thought in aging ad age-related disease. In: In Boller F, Grafman J, editors. Handbook of Neuropsychology, vol. 4. Amsterdam’ Elsevier; 1990. p. 279 – 310. [5] Dimitrov M, Grafman J, Soares AH, Clark K. Concept formation and concept shifting in frontal lesions and Parkinson’s disease patients assessed with the Californian Card Sorting Test. Neuropsychologia 1999;13:135 – 43. [6] Drewe EA. The effect of type and area of brain lesion on Wisconsin Card Sorting Test performance. Cortex 1974;10:159 – 70. [7] Fallgatter AJ, Strik WK. Frontal brain activation during the Wisconsin Card Sorting Test assessed with two-channel near-infrared spectroscopy. Eur Arch Psychiatry Clin Neurosci 1998;248:245 – 9. [8] Galatioto S, Abbate F, Laura R, Naccari F, Germana G. Morphological and immunohistochemical considerations on the basal ganglia in pigeon (Columba livia). Anat Histol Embryol 1998;27:173 – 8. [9] Giovagnoli AR. Relation of sorting impairment to hippocampal damage in temporal lobe epilepsy. Neuropsychologia 2001;39: 140 – 50. [10] Graybiel AM. The basal ganglia and cognitive pattern generator. Schizophr Bull 1997;23:459 – 69. [11] Hall K, Brauth SE, Kitt CA. Retrograde transport of [3H]GABA in the striatotegmental system of the pigeon. Brain Res 1984;310: 157 – 63. [12] Hartmann B, Guenkuerkun O. Selective deficits in reversal learning after neostriatum caudolaterale lesions in pigeons, possible behavioral equivalencies to the mammalian prefrontal system. Behav Brain Res 1998;96:125 – 33. [13] Hermann BP, Wyler AR, Richey ET. Wisconsin Card Sorting Test performance in patients with complex partial seizures of temporal lobe origin. J Clin Exp Neuropsychol 1988;10:467 – 76.
331
[14] Husband S, Shimizu T. Reversal learning after lesions in the presumptive nucleus accumbens in pigeons. Paper read at 10 th Ann Internl Conf Comp Cog; 2003. [15] Kesner RP, Bolland BL, Dakis M. Memory for spatial locations, motor responses, and objects: triple dissociation among the hippocampus, caudate nucleus, and extrastriate visual cortex. Exp Brain Res 1993;93:462 – 70. [16] Kim JJ, Baxter MG. Multiple brain-memory systems: the whole does not equal the sum of its parts. Trends Neurosci 2001;24: 324 – 30. [17] Kobayashi K, Uno H, Okanoya K. Partial lesions in the anterior forebrain pathway affect song production in adult Bengalese finches. Neuroreport 2001;84:483 – 8. [18] Lawrence AD, Sahakian BJ, Rogers RD, Hodges JR, Robbins TW. Discrimination, reversal, and shift learning in Huntington’s disease; mechanisms of impaired response selection. Neuropsychologia 1999;37:1359 – 74. [19] Lewis JW, Ryan SM, Arnold AP, Butcher LL. Evidence for a catecholaminergic projection to area X in the zebra finch. J Comp Neurol 1981;196:347 – 54. [20] Macphail EM, Reilly S. Double dissociation of effects on learning of medial versus lateral hyperstriatal lesions in pigeons. Behav Neurosci 1987;101:513 – 23. [21] McDonald RJ, White NM. A triple dissociation of memory systems: Hippocampus, amygdala, and dorsal striatum. Behav Neurosci 1993;107:3 – 22. [22] Metzger M, Jiang S, Wang J, Braun K. Organization of the dopaminergic innervation of forebrain areas relevant to learning combined immunohistochemical/retrograde tracing study in the domestic chick. J Comp Neurol 1996;376:1 – 27. [23] Milner B. Effect of different brain lesions on card sorting. Arch Neurol 1963;90:90 – 100. [24] Milner B, Corkin S, Teuber HL. Further analysis of the hippocampal amnestic syndrome: 14-year follow-up study of HM. Neuropsychologia 1968;6:215 – 34. [25] Nelson HE. A modified card sorting test sensitive to frontal lobe deficits. Cortex 1976;12:313 – 24. [26] O_Reilly RC, Noelle DC, Braver TS, Cohen JD. Prefrontal cortex and dynamic categorization tasks: representational organization and neuromodulatory control. Cereb Cortex 2002;12:246 – 57. [27] Palencia CA, Ragozzino ME. The influence of NMDA receptors in the dorsomedial striatum on response reversal learning. Neurobiol Learn 6 Mem 2004;82:81 – 9. [28] Perkel D, Farries MA. Complementary Fbottom-up_ and Ftop-down_ approaches to basal ganglia function. Curr Opin Neurobiol 2000; 10:725 – 31. [29] Powers AS, Halasz F, Williams S. The effects of lesions in telencephalic visual areas of pigeons on dimensional shifting. Physiol Behav 1982;29:1099 – 104. [30] Ragozzino ME, Jih J, Tzavos A. Involvement of the dorsomedial striatum in behavioral flexibility: role of muscarinic cholinergic receptors. Brain Res 2002;953:205 – 14. [31] Reiner A, Medina L, Veenman CL. Structural and functional evolution of the basal ganglia in vertebrates. Brain Res Rev 1998; 28:235 – 85. [32] Roberts AC, De Salvia MA, Wilkinson LS, Collins P, Muir JL, Evaritt BJ, et al. 6-Hydroxydopamine lesions of the prefrontal cortex in monkeys enhance performance on an analog of the Wisconsin Card Sorting Test: possible interaction with subcortical dopamine. J Neurosci 1994;14:2531 – 44. [33] Roberts AC, Robbins TW, Evaritt BJ, Jones GH, Sirkia TE, Wilkinson J, et al. The effects of excitotoxic lesions of the basal forebrain on the acquisition, retention and serial reversal of visual discriminations in marmosets. Neuroscience 1990;34:311 – 29. [34] Robinson AL, Heaton RK, Lehman RAW, Stillson DW. The utility of the Wisconsin Card Sorting Test in detecting and localizing frontal lesions. J Consult Clin Psychol 1980;48:605 – 14.
332
S. Watanabe / Physiology & Behavior 85 (2005) 324 – 332
[35] Schade AF, Bitterman ME. Improvement in habit reversal as related to dimensional set. J Comp Physiol Psychol 1966;62:43 – 8. [36] Shohamy D, Myers CE, Onlaor S, Gluck MA. Role of the basal ganglia in category learning: How do patients with Parkinson’s disease learn? Behav Neurosci 2004;118:676 – 86. [37] Stuss DT, Levine B, Alexander MP, Hong J, Palumbo C, Hamer L, et al. Wisconsin Card Sorting Test performance in patients with focal frontal and posterior brain damage: effects of lesion location and test structure on separable cognitive processes. Neuropsychologia 2000; 38:388 – 402. [38] Veenman CL, Wild JM, Reiner A. Organization of the avian ‘‘corticostriatal’’ projection system: a retrograde and anterograde pathway tracing study in pigeons. J Comp Neurol 1995;354:87 – 126.
[39] Verin M, Partiot A, Pillon B, Malapani C, Aqid Y, Duboias B. Delayed response tasks and prefrontal lesions in man—evidence for self generated patterns of behavior with poor environmental modulation. Neuropsychologia 1993;31:1379 – 96. [40] Watanabe S. Effects of hippocampal lesions on repeated acquisition of spatial discrimination in pigeons. Behav Brain Res 2001;120:59 – 66. [41] Watanabe S. Effects of LPO lesions on repeated acquisition of spatial discrimination in pigeons. Brain Behav Evol 2002;58:333 – 42. [42] Watanabe S. Effects of Wulst and ectostriatal lesions on repeated acquisition of spatial discrimination in pigeons. Cogn Brain Res 2003;17:286 – 92.