- Email: [email protected]
Training Method Dramatically Affects the Acquisition of a Place Response in Rats with Neurotoxic Lesions of the Hippocampus
Neurobiology of Learning and Memory 77, 109–118 (2002) doi:10.1006/nlme.2000.3997, available online at http://www.idealibrary.com on
BRIEF REPORT Training Method Dramatically Affects the Acquisition of a Place Response in Rats with Neurotoxic Lesions of the Hippocampus Juan M. J. Ramos Departamento de Psicologı´a Experimental y Fisiologı´a del Comportamiento, Facultad de Psicologı´a, and Instituto de Neurociencias Federico Olo´riz, Universidad de Granada, 18071 Granada, Spain Published online September 10, 2001
A considerable number of studies have demonstrated that hippocampal damage impairs the acquisition of a place response in rats. In Experiment 1, using a fourarm plus-shaped maze, we replicated this finding. Experiment 2 showed, however, that hippocampally damaged rats can learn a place response just as well as control rats when, during the training, a salient intramaze landmark indicates the position of the goal (the west arm). After reaching criterion, the hippocampal and control groups performed the task with the same degree of mastery during a transfer test in which the intramaze signal used during the acquisition was removed. In Experiment 3, the intramaze cue was substituted by an egocentric cue. The results revealed that both control and lesioned subjects learned the spatial problem well. However, a transfer test showed that control rats learned the task using a place response strategy but hippocampally lesioned animals used a rigid, hyperspecific strategy. Taken together, these results suggest that special training procedures which encourage variability in response versus perseveration make it possible to overcome the acquisition deficit normally observed in hippocampal rats. 䉷 2001 Elsevier Science Key Words: place learning; amnesia; spatial learning and hippocampus; hippocampus; neurotoxic lesions; maze.
A variety of findings have led to the proposal that the hippocampus is a necessary component of the neural system supporting spatial learning and navigation. First, lesion This research was supported by Grant PB96-1425 from the Ministerio de Educacio´n y Cultura, Direccio´n General de Ensen˜anza Superior, Spain. I thank Juan Carlos Rodrı´guez for helping to build the apparatus. Address correspondence and reprint requests to Juan M. J. Ramos, Departamento de Psicologı´a Experimental y Fisiologı´a del Comportamiento, Facultad de Psicologı´a, Campus de Cartuja, Universidad de Granada, 18071 Granada, Spain. E-mail: [email protected] 109
1074-7427/01 $35.00 䉷 2001 Elsevier Science All rights reserved.
110
JUAN M. J. RAMOS
studies have shown that hippocampally damaged rats are severely impaired in the acquisition of navigational tasks based on the creation of a complex representation of the environment (Morris, Garrud, Rawlins, & O’Keefe, 1982; Packard, Hirsh, & White, 1989; Packard & McGaugh, 1992). Second, electrophysiological studies have found place cells within the rat hippocampus, which fire when the animal is in a specific location within its environment (O’Keefe & Speakman, 1987). However, recent studies in rats have shown that extrahippocampal regions have crucial functions in spatial information processing. Supporting this idea, some studies have shown that lesions in the entorhinal cortex (Nagahara, Otto, & Gallagher, 1995), perirhinal cortex (Liu & Bilkey, 1999), cingulate cortex (Sutherland, Whishaw, & Kolb, 1988), and anterior thalamus (Aggleton, Hunt, Nagle, & Neave, 1996) deteriorate the acquisition of spatial tasks based on a cartographic/allocentric strategy. Also, electrophysiological studies have shown the existence of location selective cells and head direction cells in various extrahippocampal structures (Muller, Ranck, & Taube, 1996; Sharp, 1999). It has been suggested that in hippocampally lesioned rats, head cells are capable of creating and maintaining a novel representation of the animal’s environmental context (Golob & Taube, 1997). Also, studies involving humans using functional magnetic resonance imaging have shown an association between the mnemonic encoding of new place information and increased activity in the parahippocampal cortex but not in the hippocampus itself (Aguirre, Detre, Alsop, & D’Esposito, 1996; Epstein, Harris, Stanley, & Kanwisher, 1999; for review see Maguire, Burgess, & O’Keefe, 1999). Taken together, these data raise the possibility that the hippocampus is not a primary region in cartographic learning. Hippocampal lesions certainly impair the acquisition of a cartographic task (Morris, Garrud, Rawlins, & O’Keefe, 1982). However, according to various authors, a large set of nonspatial deficits have been observed after hippocampal lesions, which could contribute to the acquisition deficit (Whishaw & Jarrard, 1996; Day, Weisend, Sutherland, & Schallert, 1999). For example, the behavior of hippocampally damaged rats lacks variability and fails to alternate when resolving a spatial problem (Hirsh, 1970; Whishaw, Cassel, & Jarrard, 1995; Oliveira, Bueno, Pomarico, & Gugliano, 1997). Therefore, the objective of this study was to investigate whether rats with extensive lesions to the dorsal hippocampus can learn a place response at the same rate as control rats when special training conditions designed to minimize the possible nonspatial deficits are used. For this purpose, using a four-arm plus-shaped maze, some aspects of the training procedure were modified in order to encourage the formation of place responses. Specifically, an intramaze cue was introduced that clearly indicated the location of the goal arm. As hippocampally lesioned rats are extremely resistant to modifying their learning strategy during training in a wooden maze (Hirsh, 1970), we hypothesized that the intramaze cue would help the lesioned animals to alter their learning strategy more easily, discouraging the use of perseverative, nonplace, inappropriate strategies. The results suggest that only under the preceding training conditions do hippocampally damaged rats show normal acquisition. Experiment 1 was performed to replicate, within the stimulatory context of our laboratory and particular spatial task, previous findings that hippocampal rats are impaired in the acquisition of a place response when trained with traditional procedures. Nineteen naive male Wistar rats, weighing 270–310 g, were individually housed in a room with constant temperature and a 12:12 h light–dark cycle. The animals were food-deprived to 85% of
PLACE LEARNING WITHOUT HIPPOCAMPUS
111
normal body weight during the training period. Rats were randomly assigned to a hippocampus (HIP, n ⫽ 11) or to a control (CON, n ⫽ 8) group. Under sodium pentobarbital anesthesia (50 mg/kg), the rats were placed in a David Kopf stereotaxic apparatus. The HIP subjects received bilateral injections of NMDA (Sigma, U.S.A., PBS, pH 7.4, 0.077 M) in eight sites of the dorsal hippocampus in relation to the interaural zero point (Paxinos & Watson, 1998): AP ⫽ 5.9, L ⫽ ⫾1, V ⫽ 6.7; AP ⫽ 5.9, L ⫽ ⫾2, V ⫽ 6.7; AP ⫽ 4.8, L ⫽ ⫾1.5, V ⫽ 6.7; AP ⫽ 4.8, L ⫽ ⫾3, V ⫽ 6.7. The neurotoxin was administered in a 0.4-l volume at each site through a 30-gauge stainless steel cannula attached to a 5-l Hamilton microsyringe. Delivery of the solution was carried out with a Harvard Apparatus pump set (Model 22, Molliston, U.S.A.) at an infusion rate of 30 l/h. The cannula was left in situ for an additional 3 min before being withdrawn. The rats in the CON group underwent the same surgical procedure except that vehicle injections were administered at the same eight coordinates. After a 10-day recovery period, all rats were handled on 7 successive days for 5 min each. On the following day the behavioral training began, using a four-arm plus-shaped maze as apparatus. Each arm of the maze was 60 ⫻ 10 cm. They were connected to an octagonal central platform 35 cm in diameter. A schematic diagram of the maze and cues in the testing room has been presented elsewhere (Ramos, 1998). During the training procedure animals received eight trials per session and one session per day. The training of each rat finished when the animal reached a learning criterion of at least 14 correct trials on 2 consecutive days (87%). At the beginning of a trial, the rat was placed at the end of one of the arms used for starting (the south, north, and east), with its back to the central platform. The order in which the different starting arms were used was randomized in each daily session. Two 45-mg food pellets (P.J. Noyes Company Inc., UK) was placed in the food cup located at the end of the west arm. After a choice was made and the subject passed the midway point of the chosen arm, a wooden cube measuring 10 ⫻ 10 ⫻ 10 cm was placed by the experimenter just behind the rat. In this way the animal remained at the end of the chosen arm for 5–7 s. Then the rat was picked up and confined in a box for an intertrial interval of 30 s. The maze was rotated 90⬚ in a clockwise direction from trial to trial in order to prevent the animals from using olfactory signals to reach the goal arm. After completing the behavioral procedure, the rats were deeply anesthetized with sodium pentobarbital and perfused intracardially with saline followed by 10% Formalin. Several days later, the brains were frozen and sliced at 50 m. Coronal sections were stained with cresyl violet to evaluate the extent of the lesion with the aid of an Olympus CH-30 microscope. The histological analysis in this experiment as well as those mentioned below revealed appropriately positioned bilateral lesions (Figs. 1A and 1B). At the most rostral level of the hippocampus, cell loss affected all of the Ammon’s horn (i.e., fields CA1–CA3) in most of the animals; however, more limited cell loss was observed in the dentate gyrus. At the level of the ventromedial nucleus, the hippocampal lesion presented losses of between 80 and 100% of the cells of the CA1–CA3 fields, with polymorph layer dentate gyrus cells intact in 40% of the animals. Finally, at the level of the mammillary nuclei, CA1 and the inner blade dentate gyrus were partial and bilaterally damaged in 70% of the animals. However, CA2 and CA3 were only partially damaged in a third of the subjects. No sign of cell loss was detected at any further posterior levels.
112
JUAN M. J. RAMOS
FIG. 1. (A) Photomicrograph of a coronal brain section from a representative animal with selective damage to the hippocampus (left) and from a sham-injected rat (right). (B) Coronal sections showing the largest (gray) and smallest (central white area) hippocampal lesions. Anteroposterior coordinates in relation to the interaural zero point (Paxinos & Watson, 1998).
As illustrated in Figs. 2A and 2B, results indicated that the HIP group presented a serious deterioration in the acquisition of the cartographic task. A one-way ANOVA revealed that the number of incorrect trials to criterion was significantly higher in the HIP than in the CON group (HIP ⫽ 38.2 ⫾ 4.1 vs. CON ⫽ 23.5 ⫾ 2.4, F(1, 17) ⫽ 7.87, p ⬍ .01, Fig. 2A). Consistently, the number of days required to reach criterion was also significantly higher in the HIP group (HIP ⫽ 11.2 ⫾ 1.1 vs. CON ⫽ 8.1 ⫾ 0.7, F(1, 17) ⫽ 4.90, p ⬍ .04, Fig. 2B). In Experiment 2, the traditional training method was substituted by another one devised in my lab for the purpose of encouraging the formation of place responses versus nonplace strategies. Thus, the goal of Experiment 2 was to demonstrate that hippocampal rats are capable of acquiring a cartographic task at a rate and level of performance similar to that of the CON group if an appropriate training method which discourages the use of nonplace strategies is used. A total of 16 naive male Wistar rats with neurotoxic lesions in the dorsal hippocampus and 9 naive sham-operated rats were used. The only difference
PLACE LEARNING WITHOUT HIPPOCAMPUS
113
FIG. 2. (A) The mean number of errors before criterion for the HIP and CON groups without an intramaze cue. (B) The mean number of days before criterion without an intramaze cue. (C) The mean number of errors before criterion when an intramaze cue orients the animals toward the goal arm. (D) The mean number of days before criterion with an intramaze cue. (E) The performance of the HIP and CON groups during the transfer test (without intramaze cue) 1 day after reaching criterion. Vertical bars represent SEM.
between the first and second experiments was related to the behavioral procedure, which was identical to that followed in Experiment 1 except in two aspects. First, throughout the training period a piece of sandpaper (10 ⫻ 60 cm; 00 thickness) was placed on the floor of the goal arm. Second, the day after reaching criterion, each animal underwent a transfer test. During this test, the sandpaper was removed so that the only stimulus available to the rats to solve the test was the configuration of extramaze landmarks. The transfer test consisted of eight trials. The order in which the different starting arms were used was randomized, and it was the same for all the animals. The performance of the HIP and CON groups was similar. Two one-way analyses of variance revealed no significant differences between groups as far as the number of incorrect responses before reaching criterion (HIP ⫽ 21.9 ⫾ 1.5 vs. CON ⫽ 25.4 ⫾ 1.3, F(1, 23) ⫽ 2.46, p ⬍ .13, Fig. 2C) nor in the mean number of days to criterion (HIP ⫽ 6.5 ⫾ 0.24 vs. CON ⫽ 7.0 ⫾ 0.33, F(1, 23) ⫽ 1.16, p ⬍ .29, Fig. 2D). Importantly,
114
JUAN M. J. RAMOS
1 day after reaching criterion, when the transfer test was carried out, both groups performed similarly, and no significant differences were found in the mean percentage of correct responses during the transfer (HIP ⫽ 86.4% ⫾ 3.14 vs. CON ⫽ 88.7% ⫾ 4.9, F(1, 2) ⫽ 0.18, p ⬍ .67, Fig. 2E). So, although it appears that the presence of the sandpaper facilitated the acquisition, once the rats reached the learning criterion, during the transfer test, the sandpaper was no longer important in the expression of this learning. In the third experiment, the training method facilitated the development of nonplace response strategies. Therefore, in this experiment, we hypothesized that, in contrast to Experiment 2, hippocampal rats should show reduced acquisition of a place response. Subjects were 18 naive male Wistar rats (8 lesioned and 10 controls). The only difference between this experiment and the two preceding ones was related to the behavioral procedure, which was identical except in two aspects. First, during the training the animals always started from the same arm (half of the subjects from the south and the other half from the north) and the reward was invariably placed in the food cup located at the end of the west arm. Second, the day after reaching the criterion, each animal underwent a transfer test. During this test, which contained 10 consecutive trials, each rat started randomly from the two arms from which it had not started during the training phase. The west arm was still the goal arm with two pellets in its food cup. For each trial of the transfer it was recorded whether the trial was correct or incorrect. If it was incorrect, the error was classified according to the two following categories: congruent versus noncongruent errors. The error was considered congruent when, having reached the central platform, the animal made a turn identical to (or congruent with) the turn that it made in order to access the west arm during the training. The noncongruent error was defined as any not classified as congruent. Results showed that the groups differed in rate of learning. A one-way ANOVA indicated that hippocampal rats made significantly more errors than the controls (HIP ⫽ 43.2 ⫾ 6.5 vs. CON ⫽ 28.6 ⫾ 1.8, F(1, 16) ⫽ 5.88, p ⬍ .02) and needed more days than the controls to reach the criterion (HIP ⫽ 11.6 ⫾ 1.1 vs. CON ⫽ 8.9 ⫾ 0.3, F(1, 16) ⫽ 6.45, p ⬍ .02). Importantly, in the transfer test the mean percentage of correct responses was substantially higher in the sham-operated group (HIP ⫽ 17.6 ⫾ 6.2 vs. CON ⫽ 73.0 ⫾ 4.7, F(1, 16) ⫽ 52.65, p ⬍ .000002, Fig. 3A). It is interesting to note that, as shown in Fig. 3A, the performance of the hippocampal group was significantly below that of chance level (33.3% vs. 17.6%, t7 ⫽ 2.55, p ⬍ .05). These results suggest that during the training period the lesioned subjects learned to orient themselves in a highly hyperspecific and rigid way and that the strategy elicited by the training procedure was not cartographic. In an attempt to confirm the preceding idea, the types of errors made during the transfer test were analyzed. As shown in Fig. 3B, the number of congruent errors made by the lesioned group was much higher than that of control subjects (6.5 ⫾ 1.1 vs. 1.6 ⫾ 0.5, F(1, 16) ⫽ 18.47, p ⬍ .0005). In contrast, a one-way ANOVA revealed no significant differences between the two groups in terms of the number of noncongruent errors (1.7 ⫾ 0.5 vs. 1.1 ⫾ 0.2, F(1, 16) ⫽ 1.34, p ⬍ .26, Fig. 3C). The main finding is that the hippocampus is not essential for learning a place response when, during the training, an intramaze cue helps the lesioned rats to alter their learning strategy more easily, encouraging a place strategy and deterring nonplace, inappropriate, strategies. These data are consistent with the results obtained by other authors using the
PLACE LEARNING WITHOUT HIPPOCAMPUS
115
FIG. 3. (A) Mean percentage of correct responses obtained throughout the 10 trials of the transfer test 1 day after criterion. (B) Average of congruent errors made by HIP and CON rats during the transfer test. (C) Average of noncongruent errors made by HIP and CON rats during the transfer. Vertical bars represent SEM.
Morris water maze, who have demonstrated minimal deficit in the acquisition of a place response when special training procedures are used (Whishaw, Cassel, & Jarrard, 1995; Whishaw & Jarrard, 1996; Day, Weisend, Sutherland, & Schallert, 1999). However, to our knowledge, no previous investigation has demonstrated rapid place learning acquisition in hippocampal rats using a procedure like the one used in this study. For example, in the study of Day, Weisend, Sutherland, and Schallert (1999), different submerged platform sizes were used in descending order. Also, a Plexiglas corridor (a three-sided, U-shaped box) was placed between the wall of the pool and the platform, thus preventing the reinforcement of thigmotaxic circling behavior. Thus, in that study the corridor between the wall and the platform was performing the same facilitating function as the sandpaper in Experiment 2 of our series, reducing inefficient strategies and allowing lesioned rats to use a more efficient spatial strategy. In the Whishaw and Jarrard (1996) study, the spatial deficit was overcome, mainly by pretraining the lesioned rats to swim to a visible platform from different starting points.
116
JUAN M. J. RAMOS
Several studies have shown that hippocampal rats manifest a profound deficit in behavioral flexibility (Bunsey & Eichenbaum, 1996). Other investigations have revealed that hippocampal rats perseverate on previously reinforced responses when such behavior is not followed by reward (Hirsh, 1970). Thus, special training procedures like those used in this and other studies (Whishaw & Jarrard, 1996; Day, Weisend, Sutherland, & Schallert, 1999) may facilitate the acquisition of a place task in hippocampal rats because, in part, they encourage the behavioral flexibility of the animals. This view is in agreement with the data obtained in Experiments 2 and 3 of our series. Thus, in Experiment 2, the sandpaper helps the lesioned animals to modify their learning strategy more easily and prevents the hippocampal rats from using less effective prepotent nonspatial behavior to solve the task. This encourages behavioral flexibility in the animals, which leads to the development of appropriate strategies to solve the task using a place strategy. However, in Experiment 3, the training procedure encourages the development of perseverative, rigid responses (always turning to the left or to the right). This reduces the behavioral flexibility of the lesioned animals, making it very difficult to develop appropriate strategies to solve the task using a place or cartographic strategy. In conclusion, the data of the present study support the view that the hippocampus is not absolutely necessary for spatial learning. The data also suggest that the hippocampus is involved in providing the behavioral flexibility necessary to acquire complex spatial tasks quickly and nonrigidly. However, in this last function, in addition to the hippocampus, other forebrain structures are involved (see, for example, Ragozzino, Wilcox, Raso, & Kesner, 1999). The results of the present study agree with recent studies which suggest that extrahippocampal structures are important components of a brain system involved in allocentric/ allothetic processing. First, electrophysiological studies have identified location selective cells (Burwell, Shapiro, O’Mally, & Eichenbaum, 1998; Sharp, 1999) and head direction cells (Muller, Ranck, & Taube, 1996) in several extrahippocampal regions. It has been suggested that extrahippocampal location-related cells (Sharp, 1997) and extrahippocampal direction-related cells (Golob & Taube, 1997; Kubie, Sutherland, & Muller, 1999) are capable of creating a complex representation of the animal’s environmental context. Second, some studies have shown that lesions in different extrahippocampal regions (see, for example, Sutherland, Whishaw, & Kolb, 1988; Aggleton, Hunt, Nagle, & Neave, 1996; Liu & Bilkey, 1999) produce a deficit in the acquisition of spatial tasks. Thus, it may be that the use of special training procedures in hippocampal rats favors the enlistment of extrahippocampal regions involved in spatial processing. Finally, it is unlikely that the rats in Experiment 2 learned the task using a guidance versus an allocentric type of strategy. Several studies have shown that hippocampal rats learn a spatial task more quickly than normal rats when a salient cue consistently indicates the location of the goal (Packard, Hirsh, & White, 1989; Pearce, Roberts, & Good, 1998). Therefore, if in our experimental room a salient extramaze cue had been responsible for the rats of Experiment 2 learning the task by guidance, then the lesioned animals of Experiment 1 would not have shown any deficit during the acquisition; however, a profound deficit was evident. REFERENCES Aggleton, J. P., Hunt, P. R., Nagle, S., & Neave, N. (1996). The effect of selective lesions within the anterior thalamic nuclei on spatial memory in the rat. Behavioural Brain Research, 81, 189–198.
PLACE LEARNING WITHOUT HIPPOCAMPUS
117
Aguirre, G. K., Detre, J. A., Alsop, D. C., & D’Esposito, M. (1996). The parahippocampus subserves topographical learning in man. Cerebral Cortex, 6, 823–829. Bunsey, M., & Eichenbaum, H. (1996). Conservation of hippocampal memory function in rats and humans. Nature, 379, 255–257. Burwell, R. D., Shapiro, M. C., O’Mally, M. T., & Eichenbaum, H. (1998). Positional firing properties of perirhinal cortex neurons. Neuroreport, 9, 3013–3018. Day, L. B., Weisend, M., Sutherland, R. J., & Schallert, T. (1999). The hippocampus is not necessary for a place response but may be necessary for pliancy. Behavioral Neuroscience, 113, 914–924. Epstein, R., Harris, A., Stanley, D., & Kanwisher, N. (1999). The parahippocampal place area: Recognition, navigation, or encoding? Neuron, 23, 115–125. Golob, E. J., & Taube, J. S. (1997). Head direction cells and episodic spatial information in rats without a hippocampus. Proceedings of National Academy of Sciences, USA, 94, 7645–7650. Hirsh, R. (1970). Lack of variability or perseveration: Describing the effect of hippocampal ablation. Physiology and Behavior, 5, 1249–1254. Kubie, J. L., Sutherland, R. J., & Muller, R. U. (1999). Hippocampal lesions produce a temporally graded retrograde amnesia on a dry version of the Morris swimming task. Psychobiology, 27, 313–330. Liu, P., & Bilkey, D. K. (1999). The effect of excitotoxic lesions centered on the perirhinal cortex in two versions of the radial arm maze task. Behavioral Neuroscience, 113, 672–682. Maguire, E. A., Burges, N., & O’Keefe, J. (1999). Human spatial navigation: Cognitive maps, sexual dimorphism, and neural substrates. Current Opinion in Neurobiology, 9, 171–177. Morris, R. G. M., Garrud, P., Rawlins, J. N. P., & O’Keefe, J. (1982). Place-navigation impaired in rats with hippocampal lesions. Nature, 297, 681–683. Muller, R. U., Ranck, J. B. Jr., & Taube, J. S. (1996). Head direction cells: properties and functional significance. Current Opinion in Neurobiology, 6, 196–206. Nagahara, A. H., Otto, T., & Gallagher, M. (1995). Entorhinal-perirhinal lesions impair performance of rats on two versions of place learning in the Morris water maze. Behavioral Neuroscience, 109, 3–9. O’Keefe, J., & Speakman, A. (1987). Single unit activity in the rat hippocampus during a spatial memory task. Experimental Brain Research, 68, 1–27. Oliveira, M. G. M., Bueno, O. F. A., Pomarico, A. C., & Gugliano, E. B. (1997). Strategies used by hippocampaland caudate-putamen-lesioned rats in a learning task. Neurobiology of Learning and Memory, 68, 32–41. 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, 1465–1472. Packard, M. G., & McGaugh, J. L. (1992). Double dissociation of fornix and caudate nucleus lesions on acquisition of two water maze tasks: Further evidence for multiple memory system. Behavioral Neuroscience, 106, 439–446. Paxinos, G., & Watson, C. (1998). The rat brain in stereotaxic coordinates, 4th ed. New York: Academic Press. Pearce, J. M., Roberts, A. D. L., & Good, M. (1998). Hippocampal lesions disrupt navigation based on cognitive maps but not heading vectors. Nature, 396, 75–77. Ragozzino, M. E., Wilcox, C., Raso, M., & Kesner, R. P. (1999). Involvement of rodent prefrontal cortex subregions in strategy switching. Behavioral Neuroscience, 113, 32–41. Ramos, J. M. J. (1998). Retrograde amnesia for spatial information: a dissociation between intra and extramaze cues following hippocampus lesions in rats. European Journal of Neuroscience, 10, 3295–3301. Sharp, P. E. (1997). Subicular cells generate similar spatial firing patterns in two geometrically and visually distinctive environments—Comparisons with hippocampal place cells. Behavioral Brain Research, 85, 71–92. Sharp, P. E. (1999). Complimentary roles for hippocampal versus subicular/entorhinal place cells in coding place, context, and events. Hippocampus, 9, 432–443. Sutherland, R. J., Whishaw, I. Q., & Kolb, B. (1988). Contribution of cingulate cortex to two forms of spatial learning and memory. Journal of Neuroscience, 8, 1863–1872.
118
JUAN M. J. RAMOS
Whishaw, I. Q., Cassel, J-C., & Jarrard, L. E. (1995). Rats with fimbria-fornix lesions display a place response in a swimming pool: a dissociation between getting there and knowing where. Journal of Neuroscience, 15, 5779–5788. Whishaw, I. Q., & Jarrard, L. E. (1996). Evidence for extrahippocampal involvement in place learning and hippocampal involvement in path integration. Hippocampus, 6, 513–524.
BRIEF REPORT Training Method Dramatically Affects the Acquisition of a Place Response in Rats with Neurotoxic Lesions of the Hippocampus Juan M. J. Ramos Departamento de Psicologı´a Experimental y Fisiologı´a del Comportamiento, Facultad de Psicologı´a, and Instituto de Neurociencias Federico Olo´riz, Universidad de Granada, 18071 Granada, Spain Published online September 10, 2001
A considerable number of studies have demonstrated that hippocampal damage impairs the acquisition of a place response in rats. In Experiment 1, using a fourarm plus-shaped maze, we replicated this finding. Experiment 2 showed, however, that hippocampally damaged rats can learn a place response just as well as control rats when, during the training, a salient intramaze landmark indicates the position of the goal (the west arm). After reaching criterion, the hippocampal and control groups performed the task with the same degree of mastery during a transfer test in which the intramaze signal used during the acquisition was removed. In Experiment 3, the intramaze cue was substituted by an egocentric cue. The results revealed that both control and lesioned subjects learned the spatial problem well. However, a transfer test showed that control rats learned the task using a place response strategy but hippocampally lesioned animals used a rigid, hyperspecific strategy. Taken together, these results suggest that special training procedures which encourage variability in response versus perseveration make it possible to overcome the acquisition deficit normally observed in hippocampal rats. 䉷 2001 Elsevier Science Key Words: place learning; amnesia; spatial learning and hippocampus; hippocampus; neurotoxic lesions; maze.
A variety of findings have led to the proposal that the hippocampus is a necessary component of the neural system supporting spatial learning and navigation. First, lesion This research was supported by Grant PB96-1425 from the Ministerio de Educacio´n y Cultura, Direccio´n General de Ensen˜anza Superior, Spain. I thank Juan Carlos Rodrı´guez for helping to build the apparatus. Address correspondence and reprint requests to Juan M. J. Ramos, Departamento de Psicologı´a Experimental y Fisiologı´a del Comportamiento, Facultad de Psicologı´a, Campus de Cartuja, Universidad de Granada, 18071 Granada, Spain. E-mail: [email protected] 109
1074-7427/01 $35.00 䉷 2001 Elsevier Science All rights reserved.
110
JUAN M. J. RAMOS
studies have shown that hippocampally damaged rats are severely impaired in the acquisition of navigational tasks based on the creation of a complex representation of the environment (Morris, Garrud, Rawlins, & O’Keefe, 1982; Packard, Hirsh, & White, 1989; Packard & McGaugh, 1992). Second, electrophysiological studies have found place cells within the rat hippocampus, which fire when the animal is in a specific location within its environment (O’Keefe & Speakman, 1987). However, recent studies in rats have shown that extrahippocampal regions have crucial functions in spatial information processing. Supporting this idea, some studies have shown that lesions in the entorhinal cortex (Nagahara, Otto, & Gallagher, 1995), perirhinal cortex (Liu & Bilkey, 1999), cingulate cortex (Sutherland, Whishaw, & Kolb, 1988), and anterior thalamus (Aggleton, Hunt, Nagle, & Neave, 1996) deteriorate the acquisition of spatial tasks based on a cartographic/allocentric strategy. Also, electrophysiological studies have shown the existence of location selective cells and head direction cells in various extrahippocampal structures (Muller, Ranck, & Taube, 1996; Sharp, 1999). It has been suggested that in hippocampally lesioned rats, head cells are capable of creating and maintaining a novel representation of the animal’s environmental context (Golob & Taube, 1997). Also, studies involving humans using functional magnetic resonance imaging have shown an association between the mnemonic encoding of new place information and increased activity in the parahippocampal cortex but not in the hippocampus itself (Aguirre, Detre, Alsop, & D’Esposito, 1996; Epstein, Harris, Stanley, & Kanwisher, 1999; for review see Maguire, Burgess, & O’Keefe, 1999). Taken together, these data raise the possibility that the hippocampus is not a primary region in cartographic learning. Hippocampal lesions certainly impair the acquisition of a cartographic task (Morris, Garrud, Rawlins, & O’Keefe, 1982). However, according to various authors, a large set of nonspatial deficits have been observed after hippocampal lesions, which could contribute to the acquisition deficit (Whishaw & Jarrard, 1996; Day, Weisend, Sutherland, & Schallert, 1999). For example, the behavior of hippocampally damaged rats lacks variability and fails to alternate when resolving a spatial problem (Hirsh, 1970; Whishaw, Cassel, & Jarrard, 1995; Oliveira, Bueno, Pomarico, & Gugliano, 1997). Therefore, the objective of this study was to investigate whether rats with extensive lesions to the dorsal hippocampus can learn a place response at the same rate as control rats when special training conditions designed to minimize the possible nonspatial deficits are used. For this purpose, using a four-arm plus-shaped maze, some aspects of the training procedure were modified in order to encourage the formation of place responses. Specifically, an intramaze cue was introduced that clearly indicated the location of the goal arm. As hippocampally lesioned rats are extremely resistant to modifying their learning strategy during training in a wooden maze (Hirsh, 1970), we hypothesized that the intramaze cue would help the lesioned animals to alter their learning strategy more easily, discouraging the use of perseverative, nonplace, inappropriate strategies. The results suggest that only under the preceding training conditions do hippocampally damaged rats show normal acquisition. Experiment 1 was performed to replicate, within the stimulatory context of our laboratory and particular spatial task, previous findings that hippocampal rats are impaired in the acquisition of a place response when trained with traditional procedures. Nineteen naive male Wistar rats, weighing 270–310 g, were individually housed in a room with constant temperature and a 12:12 h light–dark cycle. The animals were food-deprived to 85% of
PLACE LEARNING WITHOUT HIPPOCAMPUS
111
normal body weight during the training period. Rats were randomly assigned to a hippocampus (HIP, n ⫽ 11) or to a control (CON, n ⫽ 8) group. Under sodium pentobarbital anesthesia (50 mg/kg), the rats were placed in a David Kopf stereotaxic apparatus. The HIP subjects received bilateral injections of NMDA (Sigma, U.S.A., PBS, pH 7.4, 0.077 M) in eight sites of the dorsal hippocampus in relation to the interaural zero point (Paxinos & Watson, 1998): AP ⫽ 5.9, L ⫽ ⫾1, V ⫽ 6.7; AP ⫽ 5.9, L ⫽ ⫾2, V ⫽ 6.7; AP ⫽ 4.8, L ⫽ ⫾1.5, V ⫽ 6.7; AP ⫽ 4.8, L ⫽ ⫾3, V ⫽ 6.7. The neurotoxin was administered in a 0.4-l volume at each site through a 30-gauge stainless steel cannula attached to a 5-l Hamilton microsyringe. Delivery of the solution was carried out with a Harvard Apparatus pump set (Model 22, Molliston, U.S.A.) at an infusion rate of 30 l/h. The cannula was left in situ for an additional 3 min before being withdrawn. The rats in the CON group underwent the same surgical procedure except that vehicle injections were administered at the same eight coordinates. After a 10-day recovery period, all rats were handled on 7 successive days for 5 min each. On the following day the behavioral training began, using a four-arm plus-shaped maze as apparatus. Each arm of the maze was 60 ⫻ 10 cm. They were connected to an octagonal central platform 35 cm in diameter. A schematic diagram of the maze and cues in the testing room has been presented elsewhere (Ramos, 1998). During the training procedure animals received eight trials per session and one session per day. The training of each rat finished when the animal reached a learning criterion of at least 14 correct trials on 2 consecutive days (87%). At the beginning of a trial, the rat was placed at the end of one of the arms used for starting (the south, north, and east), with its back to the central platform. The order in which the different starting arms were used was randomized in each daily session. Two 45-mg food pellets (P.J. Noyes Company Inc., UK) was placed in the food cup located at the end of the west arm. After a choice was made and the subject passed the midway point of the chosen arm, a wooden cube measuring 10 ⫻ 10 ⫻ 10 cm was placed by the experimenter just behind the rat. In this way the animal remained at the end of the chosen arm for 5–7 s. Then the rat was picked up and confined in a box for an intertrial interval of 30 s. The maze was rotated 90⬚ in a clockwise direction from trial to trial in order to prevent the animals from using olfactory signals to reach the goal arm. After completing the behavioral procedure, the rats were deeply anesthetized with sodium pentobarbital and perfused intracardially with saline followed by 10% Formalin. Several days later, the brains were frozen and sliced at 50 m. Coronal sections were stained with cresyl violet to evaluate the extent of the lesion with the aid of an Olympus CH-30 microscope. The histological analysis in this experiment as well as those mentioned below revealed appropriately positioned bilateral lesions (Figs. 1A and 1B). At the most rostral level of the hippocampus, cell loss affected all of the Ammon’s horn (i.e., fields CA1–CA3) in most of the animals; however, more limited cell loss was observed in the dentate gyrus. At the level of the ventromedial nucleus, the hippocampal lesion presented losses of between 80 and 100% of the cells of the CA1–CA3 fields, with polymorph layer dentate gyrus cells intact in 40% of the animals. Finally, at the level of the mammillary nuclei, CA1 and the inner blade dentate gyrus were partial and bilaterally damaged in 70% of the animals. However, CA2 and CA3 were only partially damaged in a third of the subjects. No sign of cell loss was detected at any further posterior levels.
112
JUAN M. J. RAMOS
FIG. 1. (A) Photomicrograph of a coronal brain section from a representative animal with selective damage to the hippocampus (left) and from a sham-injected rat (right). (B) Coronal sections showing the largest (gray) and smallest (central white area) hippocampal lesions. Anteroposterior coordinates in relation to the interaural zero point (Paxinos & Watson, 1998).
As illustrated in Figs. 2A and 2B, results indicated that the HIP group presented a serious deterioration in the acquisition of the cartographic task. A one-way ANOVA revealed that the number of incorrect trials to criterion was significantly higher in the HIP than in the CON group (HIP ⫽ 38.2 ⫾ 4.1 vs. CON ⫽ 23.5 ⫾ 2.4, F(1, 17) ⫽ 7.87, p ⬍ .01, Fig. 2A). Consistently, the number of days required to reach criterion was also significantly higher in the HIP group (HIP ⫽ 11.2 ⫾ 1.1 vs. CON ⫽ 8.1 ⫾ 0.7, F(1, 17) ⫽ 4.90, p ⬍ .04, Fig. 2B). In Experiment 2, the traditional training method was substituted by another one devised in my lab for the purpose of encouraging the formation of place responses versus nonplace strategies. Thus, the goal of Experiment 2 was to demonstrate that hippocampal rats are capable of acquiring a cartographic task at a rate and level of performance similar to that of the CON group if an appropriate training method which discourages the use of nonplace strategies is used. A total of 16 naive male Wistar rats with neurotoxic lesions in the dorsal hippocampus and 9 naive sham-operated rats were used. The only difference
PLACE LEARNING WITHOUT HIPPOCAMPUS
113
FIG. 2. (A) The mean number of errors before criterion for the HIP and CON groups without an intramaze cue. (B) The mean number of days before criterion without an intramaze cue. (C) The mean number of errors before criterion when an intramaze cue orients the animals toward the goal arm. (D) The mean number of days before criterion with an intramaze cue. (E) The performance of the HIP and CON groups during the transfer test (without intramaze cue) 1 day after reaching criterion. Vertical bars represent SEM.
between the first and second experiments was related to the behavioral procedure, which was identical to that followed in Experiment 1 except in two aspects. First, throughout the training period a piece of sandpaper (10 ⫻ 60 cm; 00 thickness) was placed on the floor of the goal arm. Second, the day after reaching criterion, each animal underwent a transfer test. During this test, the sandpaper was removed so that the only stimulus available to the rats to solve the test was the configuration of extramaze landmarks. The transfer test consisted of eight trials. The order in which the different starting arms were used was randomized, and it was the same for all the animals. The performance of the HIP and CON groups was similar. Two one-way analyses of variance revealed no significant differences between groups as far as the number of incorrect responses before reaching criterion (HIP ⫽ 21.9 ⫾ 1.5 vs. CON ⫽ 25.4 ⫾ 1.3, F(1, 23) ⫽ 2.46, p ⬍ .13, Fig. 2C) nor in the mean number of days to criterion (HIP ⫽ 6.5 ⫾ 0.24 vs. CON ⫽ 7.0 ⫾ 0.33, F(1, 23) ⫽ 1.16, p ⬍ .29, Fig. 2D). Importantly,
114
JUAN M. J. RAMOS
1 day after reaching criterion, when the transfer test was carried out, both groups performed similarly, and no significant differences were found in the mean percentage of correct responses during the transfer (HIP ⫽ 86.4% ⫾ 3.14 vs. CON ⫽ 88.7% ⫾ 4.9, F(1, 2) ⫽ 0.18, p ⬍ .67, Fig. 2E). So, although it appears that the presence of the sandpaper facilitated the acquisition, once the rats reached the learning criterion, during the transfer test, the sandpaper was no longer important in the expression of this learning. In the third experiment, the training method facilitated the development of nonplace response strategies. Therefore, in this experiment, we hypothesized that, in contrast to Experiment 2, hippocampal rats should show reduced acquisition of a place response. Subjects were 18 naive male Wistar rats (8 lesioned and 10 controls). The only difference between this experiment and the two preceding ones was related to the behavioral procedure, which was identical except in two aspects. First, during the training the animals always started from the same arm (half of the subjects from the south and the other half from the north) and the reward was invariably placed in the food cup located at the end of the west arm. Second, the day after reaching the criterion, each animal underwent a transfer test. During this test, which contained 10 consecutive trials, each rat started randomly from the two arms from which it had not started during the training phase. The west arm was still the goal arm with two pellets in its food cup. For each trial of the transfer it was recorded whether the trial was correct or incorrect. If it was incorrect, the error was classified according to the two following categories: congruent versus noncongruent errors. The error was considered congruent when, having reached the central platform, the animal made a turn identical to (or congruent with) the turn that it made in order to access the west arm during the training. The noncongruent error was defined as any not classified as congruent. Results showed that the groups differed in rate of learning. A one-way ANOVA indicated that hippocampal rats made significantly more errors than the controls (HIP ⫽ 43.2 ⫾ 6.5 vs. CON ⫽ 28.6 ⫾ 1.8, F(1, 16) ⫽ 5.88, p ⬍ .02) and needed more days than the controls to reach the criterion (HIP ⫽ 11.6 ⫾ 1.1 vs. CON ⫽ 8.9 ⫾ 0.3, F(1, 16) ⫽ 6.45, p ⬍ .02). Importantly, in the transfer test the mean percentage of correct responses was substantially higher in the sham-operated group (HIP ⫽ 17.6 ⫾ 6.2 vs. CON ⫽ 73.0 ⫾ 4.7, F(1, 16) ⫽ 52.65, p ⬍ .000002, Fig. 3A). It is interesting to note that, as shown in Fig. 3A, the performance of the hippocampal group was significantly below that of chance level (33.3% vs. 17.6%, t7 ⫽ 2.55, p ⬍ .05). These results suggest that during the training period the lesioned subjects learned to orient themselves in a highly hyperspecific and rigid way and that the strategy elicited by the training procedure was not cartographic. In an attempt to confirm the preceding idea, the types of errors made during the transfer test were analyzed. As shown in Fig. 3B, the number of congruent errors made by the lesioned group was much higher than that of control subjects (6.5 ⫾ 1.1 vs. 1.6 ⫾ 0.5, F(1, 16) ⫽ 18.47, p ⬍ .0005). In contrast, a one-way ANOVA revealed no significant differences between the two groups in terms of the number of noncongruent errors (1.7 ⫾ 0.5 vs. 1.1 ⫾ 0.2, F(1, 16) ⫽ 1.34, p ⬍ .26, Fig. 3C). The main finding is that the hippocampus is not essential for learning a place response when, during the training, an intramaze cue helps the lesioned rats to alter their learning strategy more easily, encouraging a place strategy and deterring nonplace, inappropriate, strategies. These data are consistent with the results obtained by other authors using the
PLACE LEARNING WITHOUT HIPPOCAMPUS
115
FIG. 3. (A) Mean percentage of correct responses obtained throughout the 10 trials of the transfer test 1 day after criterion. (B) Average of congruent errors made by HIP and CON rats during the transfer test. (C) Average of noncongruent errors made by HIP and CON rats during the transfer. Vertical bars represent SEM.
Morris water maze, who have demonstrated minimal deficit in the acquisition of a place response when special training procedures are used (Whishaw, Cassel, & Jarrard, 1995; Whishaw & Jarrard, 1996; Day, Weisend, Sutherland, & Schallert, 1999). However, to our knowledge, no previous investigation has demonstrated rapid place learning acquisition in hippocampal rats using a procedure like the one used in this study. For example, in the study of Day, Weisend, Sutherland, and Schallert (1999), different submerged platform sizes were used in descending order. Also, a Plexiglas corridor (a three-sided, U-shaped box) was placed between the wall of the pool and the platform, thus preventing the reinforcement of thigmotaxic circling behavior. Thus, in that study the corridor between the wall and the platform was performing the same facilitating function as the sandpaper in Experiment 2 of our series, reducing inefficient strategies and allowing lesioned rats to use a more efficient spatial strategy. In the Whishaw and Jarrard (1996) study, the spatial deficit was overcome, mainly by pretraining the lesioned rats to swim to a visible platform from different starting points.
116
JUAN M. J. RAMOS
Several studies have shown that hippocampal rats manifest a profound deficit in behavioral flexibility (Bunsey & Eichenbaum, 1996). Other investigations have revealed that hippocampal rats perseverate on previously reinforced responses when such behavior is not followed by reward (Hirsh, 1970). Thus, special training procedures like those used in this and other studies (Whishaw & Jarrard, 1996; Day, Weisend, Sutherland, & Schallert, 1999) may facilitate the acquisition of a place task in hippocampal rats because, in part, they encourage the behavioral flexibility of the animals. This view is in agreement with the data obtained in Experiments 2 and 3 of our series. Thus, in Experiment 2, the sandpaper helps the lesioned animals to modify their learning strategy more easily and prevents the hippocampal rats from using less effective prepotent nonspatial behavior to solve the task. This encourages behavioral flexibility in the animals, which leads to the development of appropriate strategies to solve the task using a place strategy. However, in Experiment 3, the training procedure encourages the development of perseverative, rigid responses (always turning to the left or to the right). This reduces the behavioral flexibility of the lesioned animals, making it very difficult to develop appropriate strategies to solve the task using a place or cartographic strategy. In conclusion, the data of the present study support the view that the hippocampus is not absolutely necessary for spatial learning. The data also suggest that the hippocampus is involved in providing the behavioral flexibility necessary to acquire complex spatial tasks quickly and nonrigidly. However, in this last function, in addition to the hippocampus, other forebrain structures are involved (see, for example, Ragozzino, Wilcox, Raso, & Kesner, 1999). The results of the present study agree with recent studies which suggest that extrahippocampal structures are important components of a brain system involved in allocentric/ allothetic processing. First, electrophysiological studies have identified location selective cells (Burwell, Shapiro, O’Mally, & Eichenbaum, 1998; Sharp, 1999) and head direction cells (Muller, Ranck, & Taube, 1996) in several extrahippocampal regions. It has been suggested that extrahippocampal location-related cells (Sharp, 1997) and extrahippocampal direction-related cells (Golob & Taube, 1997; Kubie, Sutherland, & Muller, 1999) are capable of creating a complex representation of the animal’s environmental context. Second, some studies have shown that lesions in different extrahippocampal regions (see, for example, Sutherland, Whishaw, & Kolb, 1988; Aggleton, Hunt, Nagle, & Neave, 1996; Liu & Bilkey, 1999) produce a deficit in the acquisition of spatial tasks. Thus, it may be that the use of special training procedures in hippocampal rats favors the enlistment of extrahippocampal regions involved in spatial processing. Finally, it is unlikely that the rats in Experiment 2 learned the task using a guidance versus an allocentric type of strategy. Several studies have shown that hippocampal rats learn a spatial task more quickly than normal rats when a salient cue consistently indicates the location of the goal (Packard, Hirsh, & White, 1989; Pearce, Roberts, & Good, 1998). Therefore, if in our experimental room a salient extramaze cue had been responsible for the rats of Experiment 2 learning the task by guidance, then the lesioned animals of Experiment 1 would not have shown any deficit during the acquisition; however, a profound deficit was evident. REFERENCES Aggleton, J. P., Hunt, P. R., Nagle, S., & Neave, N. (1996). The effect of selective lesions within the anterior thalamic nuclei on spatial memory in the rat. Behavioural Brain Research, 81, 189–198.
PLACE LEARNING WITHOUT HIPPOCAMPUS
117
Aguirre, G. K., Detre, J. A., Alsop, D. C., & D’Esposito, M. (1996). The parahippocampus subserves topographical learning in man. Cerebral Cortex, 6, 823–829. Bunsey, M., & Eichenbaum, H. (1996). Conservation of hippocampal memory function in rats and humans. Nature, 379, 255–257. Burwell, R. D., Shapiro, M. C., O’Mally, M. T., & Eichenbaum, H. (1998). Positional firing properties of perirhinal cortex neurons. Neuroreport, 9, 3013–3018. Day, L. B., Weisend, M., Sutherland, R. J., & Schallert, T. (1999). The hippocampus is not necessary for a place response but may be necessary for pliancy. Behavioral Neuroscience, 113, 914–924. Epstein, R., Harris, A., Stanley, D., & Kanwisher, N. (1999). The parahippocampal place area: Recognition, navigation, or encoding? Neuron, 23, 115–125. Golob, E. J., & Taube, J. S. (1997). Head direction cells and episodic spatial information in rats without a hippocampus. Proceedings of National Academy of Sciences, USA, 94, 7645–7650. Hirsh, R. (1970). Lack of variability or perseveration: Describing the effect of hippocampal ablation. Physiology and Behavior, 5, 1249–1254. Kubie, J. L., Sutherland, R. J., & Muller, R. U. (1999). Hippocampal lesions produce a temporally graded retrograde amnesia on a dry version of the Morris swimming task. Psychobiology, 27, 313–330. Liu, P., & Bilkey, D. K. (1999). The effect of excitotoxic lesions centered on the perirhinal cortex in two versions of the radial arm maze task. Behavioral Neuroscience, 113, 672–682. Maguire, E. A., Burges, N., & O’Keefe, J. (1999). Human spatial navigation: Cognitive maps, sexual dimorphism, and neural substrates. Current Opinion in Neurobiology, 9, 171–177. Morris, R. G. M., Garrud, P., Rawlins, J. N. P., & O’Keefe, J. (1982). Place-navigation impaired in rats with hippocampal lesions. Nature, 297, 681–683. Muller, R. U., Ranck, J. B. Jr., & Taube, J. S. (1996). Head direction cells: properties and functional significance. Current Opinion in Neurobiology, 6, 196–206. Nagahara, A. H., Otto, T., & Gallagher, M. (1995). Entorhinal-perirhinal lesions impair performance of rats on two versions of place learning in the Morris water maze. Behavioral Neuroscience, 109, 3–9. O’Keefe, J., & Speakman, A. (1987). Single unit activity in the rat hippocampus during a spatial memory task. Experimental Brain Research, 68, 1–27. Oliveira, M. G. M., Bueno, O. F. A., Pomarico, A. C., & Gugliano, E. B. (1997). Strategies used by hippocampaland caudate-putamen-lesioned rats in a learning task. Neurobiology of Learning and Memory, 68, 32–41. 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, 1465–1472. Packard, M. G., & McGaugh, J. L. (1992). Double dissociation of fornix and caudate nucleus lesions on acquisition of two water maze tasks: Further evidence for multiple memory system. Behavioral Neuroscience, 106, 439–446. Paxinos, G., & Watson, C. (1998). The rat brain in stereotaxic coordinates, 4th ed. New York: Academic Press. Pearce, J. M., Roberts, A. D. L., & Good, M. (1998). Hippocampal lesions disrupt navigation based on cognitive maps but not heading vectors. Nature, 396, 75–77. Ragozzino, M. E., Wilcox, C., Raso, M., & Kesner, R. P. (1999). Involvement of rodent prefrontal cortex subregions in strategy switching. Behavioral Neuroscience, 113, 32–41. Ramos, J. M. J. (1998). Retrograde amnesia for spatial information: a dissociation between intra and extramaze cues following hippocampus lesions in rats. European Journal of Neuroscience, 10, 3295–3301. Sharp, P. E. (1997). Subicular cells generate similar spatial firing patterns in two geometrically and visually distinctive environments—Comparisons with hippocampal place cells. Behavioral Brain Research, 85, 71–92. Sharp, P. E. (1999). Complimentary roles for hippocampal versus subicular/entorhinal place cells in coding place, context, and events. Hippocampus, 9, 432–443. Sutherland, R. J., Whishaw, I. Q., & Kolb, B. (1988). Contribution of cingulate cortex to two forms of spatial learning and memory. Journal of Neuroscience, 8, 1863–1872.
118
JUAN M. J. RAMOS
Whishaw, I. Q., Cassel, J-C., & Jarrard, L. E. (1995). Rats with fimbria-fornix lesions display a place response in a swimming pool: a dissociation between getting there and knowing where. Journal of Neuroscience, 15, 5779–5788. Whishaw, I. Q., & Jarrard, L. E. (1996). Evidence for extrahippocampal involvement in place learning and hippocampal involvement in path integration. Hippocampus, 6, 513–524.