Neurotoxic Lesions of the Rat Perirhinal Cortex ... - Semantic Scholar

Behavioral Neuroscience 2002, Vol. 116, No. 2, 232–240

Copyright 2002 by the American Psychological Association, Inc. 0735-7044/02/$5.00 DOI: 10.1037//0735-7044.116.2.232

Neurotoxic Lesions of the Rat Perirhinal Cortex Fail to Disrupt the Acquisition or Performance of Tests of Allocentric Spatial Memory P. Machin, S. D. Vann, J. L. Muir, and J. P. Aggleton Cardiff University Rats with neurotoxic lesions of the perirhinal cortex (n ⫽ 9) were compared with sham controls (n ⫽ 14) on a working memory task in the radial arm maze. Rats were trained under varying levels of proactive interference and with different retention intervals. Finally, performance was assessed when the maze was switched to a novel room. None of these manipulations differentially impaired rats with perirhinal lesions. Rats were next trained on delayed matching-to-place in the water maze. Even with retention delays of 30 min, there was no evidence of a deficit. Although interactions between the perirhinal cortex and hippocampus may be important for integrating object–place information, the perirhinal cortex is often not necessary for tasks that selectively tax allocentric spatial memory.

same tasks (Liu & Bilkey, 1998b, 1998c, 1999, 2001; Wiig & Bilkey, 1994a). A comparison of the respective surgeries shows that this difference is not related to lesion size, although at first it did appear that this difference might reflect the use of neurotoxic lesions versus conventional lesions (Ennaceur & Aggleton, 1997; Ennaceur, Neave, & Aggleton, 1996). This is because studies using neurotoxic lesions had consistently reported no lesion deficit on tests of spatial memory (Aggleton, Keen, Warburton, & Bussey, 1997; Bussey et al., 1999; Ennaceur & Aggleton, 1997; Ennaceur et al., 1996). Other evidence came from a diminishing of the T-maze alternation deficit in rats with excitotoxic lesions (Liu & Bilkey, 1998a). Nevertheless, significant deficits have been observed in recent studies using excitotoxins to examine radial arm maze (Liu & Bilkey, 1999) and Morris water maze (Liu & Bilkey, 2001) performance after perirhinal cortex lesions. Conversely, a number of studies using conventional perirhinal lesions either have failed to find spatial deficits (E. A. Gaffan, Eacott, & Simpson, 2000; Glenn & Mumby, 1998) or have found only transient impairments (Mumby & Glenn, 2000). Another suggestion is that these different lesion outcomes reflect differences in training protocols (Bilkey, 1999; Liu & Bilkey, 1999, 2001). Specifically, it has been argued that extended periods of habituation and pretraining might mask perirhinal lesion effects on tasks such as the radial arm maze and so render these tasks less sensitive. There is also evidence that working memory rather than reference memory versions of spatial tasks are especially sensitive to perirhinal lesions (Liu & Bilkey, 2001). Among these working memory tasks is delayed matching-to-place (DMP) in the water maze (Liu & Bilkey, 2001). This task is of interest because perirhinal lesion deficits at longer retention delays can be equivalent to those observed after hippocampectomy (Liu & Bilkey, 2001; but see Glenn & Mumby, 1998). It has also been argued that this task taxes episodic-like memory (Morris, 2001), making it especially appropriate when examining components of the medial temporal lobe memory system (Squire & Zola-Morgan, 1991). For these reasons, the present study reexamined the effects of perirhinal cortex lesions on spatial memory, paying particular regard to two factors. The first was the use of working memory

The perirhinal cortex (Areas 35 and 36) is one of the major sources of cortical inputs to the entorhinal cortex, which in turn projects to the hippocampus. The perirhinal cortex also has direct inputs to the hippocampus (Burwell & Amaral, 1998; Lavenex & Amaral, 2000; Witter, Groenewegen, Lopes da Silva, & Lohman, 1989; Witter, Naber, et al., 2000). Via these routes, the perirhinal cortex is assumed to be important for hippocampal function. Although the anatomical relationship between the perirhinal cortex and hippocampus remains uncontroversial, there is, however, considerable debate over the nature of their functional relationship. In the rat, much of this debate has focused on the importance of the perirhinal cortex for spatial memory (Aggleton, Vann, Oswald, & Good, 2000). The rodent hippocampus is vital for some forms of spatial processing (Morris, Garrud, Rawlins, & O’Keefe, 1982; O’Keefe & Nadel, 1978; Pearce, Roberts, & Good, 1998), and this is seen most dramatically when animals with hippocampal lesions are given tests of allocentric spatial memory. In such tests the animal uses the spatial relationships between distal landmarks to identify a given location. As a consequence, hippocampal lesions consistently impair allocentric tasks in the Morris water maze and the radial arm maze. The effects of perirhinal cortex lesions on the same tasks have, however, proved inconsistent. In some studies, loss of the perirhinal cortex has had no discernible effect on either acquisition or subsequent performance (Bussey, Dias, Amin, Muir, & Aggleton, 2001; Bussey, Muir, & Aggleton, 1999; Ennaceur & Aggleton, 1997; Glenn & Mumby, 1998; Kolb, Buhrmann, McDonald, & Sutherland, 1994; Mumby & Glenn, 2000). In contrast, other studies have reported mild, but significant, deficits on these

P. Machin, S. D. Vann, J. L. Muir, and J. P. Aggleton, School of Psychology, Cardiff University, Cardiff, Wales. This research was supported by the Wellcome Trust. We thank Tim Bussey, Moira Davies, Vanessa Marshall, and Jo Oswald for their assistance. Correspondence concerning this article should be addressed to J. P. Aggleton, School of Psychology, Cardiff University, Park Road, P.O. Box 901, Cardiff CF10 3YG, Wales. E-mail: [email protected] 232

PERIRHINAL CORTEX AND SPATIAL MEMORY

233

tasks that tax allocentric spatial memory; the second was the restriction of pretraining to maximize task sensitivity. In the first experiment, rats with perirhinal cortex lesions received limited pretraining before comparing the initial stages of radial arm maze learning. Next, radial arm maze performance was examined when the task was moved to a new room. In the second experiment, rats were trained on the DMP task in the water maze. Testing included the use of longer retention intervals (30 min) than those used in a previous study (maximum 180 s) that had found delay-dependent perirhinal deficits (Liu & Bilkey, 2001).

General Method Subjects The subjects were 28 male, pigmented rats (dark agouti strain; Harlan, Bicester, UK) weighing between 210 g and 260 g before surgery. The rats were housed in pairs and given free access to water throughout the study. All experiments were performed in accordance with UK Animals (Scientific Procedures) Act, 1986 and associated guidelines.

Surgery All rats received a 5-mg/kg subcutaneous injection of Rimadyl (Pfizer Animal Health, Sandwich, UK) and a 2.5-mg/kg intraperitoneal injection of diazepam (Phoenix Pharmaceuticals Ltd, Gloucester, UK). The rats were deeply anesthetized with a mixture of Isoflurane (Baker Norton, London, UK) and 6 ml oxygen and placed in a stereotaxic headholder (David Kopf Instruments, Tujunga, CA). The scalp was cut and retracted to expose the skull. The lesions were made by injecting a solution of 0.09 M N-methylD-aspartic acid (NMDA; Sigma, Poole, UK) dissolved in phosphate buffer (pH 7.2). Injections were made in three sites per hemisphere with a 1-␮l Hamilton syringe (Hamilton, Reno, NV). The stereotaxic coordinates of the lesion placements relative to ear-bar zero were, from the most anterior to the most posterior: AP ⫹4.2, L ⫾5.6; AP ⫹2.7, L ⫾6.1; AP ⫹1.3, L ⫾6.2. The depth, from bregma, at each site was 9.2 mm (most rostral), 9.4 mm, and 9.0 mm (most caudal). Bilateral injections of 0.26 ␮l of 0.09 M NMDA were made at each AP level, and the needle was then left in situ for 4 min. At the completion of all surgeries, the skin was sutured, and an antibiotic powder (Acramide; Dales Pharmaceuticals, Skipton, UK) was applied topically. The rats also received subcutaneous injections of 5 ml glucose– saline. The rats acting as surgical controls (shams) received the same procedure and drugs as those receiving lesions. This involved the needle being lowered into the same coordinates but without the injection of NMDA.

Histology On completion of the experiments, the rats were deeply anesthetized with sodium pentobarbital (1 mg/kg) and perfused transcardially with 0.1 M phosphate-buffered saline followed by 10% (wt/vol) Formol-saline. The brains were removed and postfixed in 10% Formol-saline and then transferred to 25% (wt/vol) sucrose overnight. Coronal sections were cut at 60 ␮m on a freezing microtome, and a one-in-three series of sections was mounted onto gelatin-coated slides and stained with cresyl violet, a Nissl stain.

Histological Analysis The nomenclature and borders of the perirhinal cortex are taken from Burwell and Amaral (1998). After histological analysis, 5 rats with partial perirhinal cortex lesions were excluded from the analyses. The perirhinal cortex lesion group (Peri) then contained 9 rats and the surgical sham group contained 14 rats. In all 9 Peri rats, there was extensive bilateral damage to

Figure 1. Coronal sections illustrating the extent of tissue loss in the cases with the largest (gray) and smallest (black) perirhinal cortex lesions. The numbers indicate the distance (in millimeters) of the sections from bregma. From The Rat Brain in Stereotaxic Coordinates (3rd ed., Figures 29, 33, 36, 39, and 42), by G. Paxinos and C. Watson, 1997, San Diego, CA: Academic Press. Copyright 1997 by Academic Press. Adapted with permission.

the perirhinal cortex, involving both banks of the rhinal sulcus (see Figure 1). The rostral limit of the lesions was level with the rostral third of the amygdala; the caudal limit reached the border with the postrhinal cortex. In 2 cases, the lesions extended into the postrhinal cortex. Within the lesions there was little evidence of neuronal sparing (see Figure 2), although there was some sparing of the deep layers of rostral perirhinal cortex in the case with the smallest lesion (see Figure 1). At the more rostral levels, the lesions often encroached ventrally to involve parts of the piriform cortex. This encroachment was marked in two cases. In 2 cases, there was partial, unilateral damage to the lateral nucleus of the amygdala.

234

MACHIN, VANN, MUIR, AND AGGLETON CA1 hippocampal field adjacent to the posterior perirhinal cortex (4 bilateral, 4 unilateral).

Experiment 1: Radial Arm Maze Method

Figure 2. Photomicrographs of Nissl-stained coronal sections showing the appearance of the rhinal sulcus in a normal rat (middle) and after an injection of N-methyl-D-aspartate (lower). In this case, the perirhinal cortex lesion extends ventrally to include parts of the piriform cortex in both hemispheres. The upper drawing shows the level and position of the areas depicted (–3.80 from bregma). From The Rat Brain in Stereotaxic Coordinates (3rd ed., Figure 33), by G. Paxinos and C. Watson, 1997, San Diego, CA: Academic Press. Copyright 1997 by Academic Press. Adapted with permission. The scale bars correspond to 1,000 ␮m.

In all cases, the lesion extended dorsally to encroach on the ventral part of Area TE, but there was more extensive damage to TE in 5 cases (two bilateral). In 8 cases, there was very restricted cell loss in that portion of the

Subjects and apparatus. Before training, the rats were food deprived to 85% of their free-feeding body weight while water remained available ad libitum. Testing was performed in an eight-arm radial maze. The maze consisted of an octagonal central platform (34 cm in diameter) with eight equally spaced radial arms (87 cm long, 10 cm wide). The floor of the central platform and the floors of the eight arms were made of wood, and clear Perspex (24 cm high) formed the walls of the arms. Close to the furthest end of each arm was a food well (2 cm in diameter and 0.5 cm deep). At the start of each arm was a clear Perspex guillotine door (12 cm high) that controlled access to and from the central platform. Each door was attached to a pulley system, enabling the experimenter to control access to the arms. The maze could be positioned in either of two rooms (295 cm ⫻ 295 cm or 255 cm ⫻ 330 cm), both of which were 260 cm high and contained salient visual cues such as geometric shapes and highcontrast stimuli. These cues were different in the two rooms. Behavioral training. Pretraining for the radial arm maze began 8 weeks after surgery and involved one habituation session in which the rats were allowed to explore the maze freely for 5 min with the guillotine doors raised and food pellets (45 mg; Purified Rodent Diet, P. J. Noyes, Lancaster, NH) scattered down the arms. The rats were then trained on the standard radial arm maze task (see below). A time limit of 10 min was placed on each trial (visiting eight different arms), so that trials lasting longer were terminated and not regarded as complete trials. Formal training lasted for 27 sessions and consisted of several stages. Before the radial arm maze task, all rats had been trained to magazine press for food reward pellets in response to a light in an automated, touchscreen apparatus (for details of apparatus, see Bussey, Muir, Everitt, & Robbins, 1997). Stage 1 (Sessions 1–14) was the standard working memory version of the radial arm maze task (Olton, Walker, & Gage, 1978) in which the subjects’ optimal strategy was to retrieve the reward pellets from all eight arms without reentering any previously entered arms. At the start of a trial, all eight arms were baited with two food pellets. The rat would make an arm choice and then return to the central platform; all the doors were then closed for about 5 s before they were opened again, permitting the rat to make another choice. This continued until all eight arms had been visited (i.e., a complete trial within 10 min). The number of sequential choice responses (successive choices involving immediately adjacent arms in a constant direction) was calculated. It was measured by giving the rat a score of ⫹1 (clockwise) or –1 (counterclockwise) if the arm was immediately adjacent to the previous choice and 0 for any other arm choice. A higher absolute score would therefore reflect the use of a sequential response strategy (Ennaceur & Aggleton, 1997; Olton & Samuelson, 1976). Stage 2 (Sessions 15–20) tested for the possible use of intramaze cues in performing the task. The start of the session was as before, but after the rat had made four different arm choices, it was removed from the maze. The rat was placed in a traveling box that had an aluminum top, base, and sides (10 cm long ⫻ 10 cm wide ⫻ 26 cm wide), which was also in the testing room. The maze was then rotated by 45°, and the remaining food pellets were moved so that they were still in the same allocentric locations but in different arms of the maze. The rat was then returned to the central platform after the 60 s that it took to rotate the maze, and the session continued until all reward pellets had been retrieved. Stage 3 (Sessions 21–25) increased task difficulty by increasing the degree of proactive interference. This was achieved by building up the number of trials the rats had each day so that for the first 3 days (Sessions 21–23) the rats were given two consecutive trials each day, with a 2-min

PERIRHINAL CORTEX AND SPATIAL MEMORY intertrial interval. On the following 2 days (Sessions 24 –25) they received three consecutive trials each day, again with a 2-min intertrial interval. Stage 4 (Sessions 26 and 27) was performed to determine the effects of novel environmental cues on the rats’ performance. For the final 2 days (Sessions 26 and 27) the rats were run in the same maze that had been used throughout, but this time it was in a novel room with different spatial landmarks (see the Subjects and Apparatus section). The rats were given three trials each day, again with a 2-min intertrial interval.

Results When performance from the first 14 days (Stage 1) of the radial arm maze task was analyzed for the number of correct arm entries out of the first eight (see Figure 3a), there was no effect of group (F ⬍ 1). There was also no group difference when the analysis used the total number of errors in each trial (see Figure 3b; F ⬍ 1). An analysis of the sequential choice responses also showed no group difference, F(1, 21) ⫽ 2.4, p ⬎ .10. To determine whether there was evidence of a group difference at the earliest stage of learning, the number of correct arm entries made in the first eight choices was compared over the first 10 trials. This analysis included those incomplete trials, in which the rat selected eight arms in 10 min but failed to visit every arm. By including these incomplete trials, we ensured that the earliest stages of training could be included. There was no evidence of a

Figure 3. Radial arm maze: Initial acquisition over 14 sessions (Stage 1), blocked in groups of two sessions. a: Mean (⫾ SEM) number of correct arm entries in the first eight choices. b: Mean (⫾ SEM) total number of errors per trial. Peri ⫽ perirhinal cortex lesions; Sham ⫽ surgical controls.

235

group difference (F ⬍ 1) or a Group ⫻ Trial interaction (F ⬍ 1), and consistent with this, analysis of the simple effects found no evidence for a group difference on any of the individual trials (1–10). Finally, the number of trials in which each rat failed to make eight or more arm choices within 10 min was compared (Peri median ⫽ 0.0, sham median ⫽ 1.5). This comparison showed that the Peri group had fewer of these incomplete sessions than the sham group, Mann–Whitney U(9, 14) ⫽ 25.0, p ⬍ .05. Similarly, the Peri group required fewer sessions to first select eight different arms within 10 min, 1.0 vs. 3.0 sessions, Mann–Whitney U(9, 14) ⫽ 14.0, p ⬍ .01. The next 6 days of testing (Stage 2) were analyzed separately, as these trials included a rotation of the maze after the first four choices. For these 6 days, there was no effect of group on the number of correct entries in the first eight entries for the entire trial, F(1, 21) ⫽ 2.3, p ⬎ .10, or the total number of errors, F(1, 21) ⫽ 3.8, p ⫽ .06. Neither was there a significant effect of group when the arm choices after the maze rotation were analyzed separately (see Figure 4), as reflected in the number of correct entries in the first four entries, F(1, 21) ⫽ 3.5, p ⬍ .08, and the total number of errors, F(1, 21) ⫽ 4.0, p ⬍ .06. For all of these comparisons, the Peri group made fewer errors than the sham controls. To determine whether the rats were using allocentric cues, the first arm choice after each maze rotation was also considered separately and a score of 1 was given for a correct entry and 0 for a reentry (maximum score of 6). At this stage, four arms had already been entered, so there was a 50% chance of getting this first arm entry correct. Inspection of the confidence intervals (95%) for the first entry after rotation showed that both groups performed above chance (M ⫾ SEM; sham 5.1 ⫾ 0.23, Peri 5.6 ⫾ 0.24). Furthermore, a comparison of the number of correct first arm entries after maze rotation showed no significant effect of group, Mann–Whitney U(9, 14)⫽ 46.0, p ⬎ .10. Stage 3 involved the rats performing two consecutive trials each day for 3 days (see Figure 5). There was a significant group difference when the number of correct entries in the first eight were considered, F(1, 21) ⫽ 4.5, p ⬍ .05, as the lesion group made more correct entries than the sham group. There was also a significant effect of trial, F(1, 21) ⫽ 33.0, p ⬍ .01, as both groups made fewer correct entries on the second trial. The same pattern of results was found for the total number of errors. Thus, there were main effects of group, F(1, 21) ⫽ 5.3, p ⬍ .05, and trial, F(1, 21) ⫽ 45.5, p ⬍ .01, with the lesion group making fewer errors than the sham group and both groups making more errors on the second trial. The last 2 days of Stage 3 involved the rats performing three consecutive trials each day (see Figure 5). There was no main group effect on either the number of correct entries in the first eight or the total number of errors (Fs ⬍ 1). There were significant main effects of trial on both the number of correct entries, F(1, 21) ⫽ 13.3, p ⬍ .01, and the total number of errors, F(1, 21) ⫽ 9.7, p ⬍ .01. Subsequent analysis showed that with both measures, there was a significant decline in performance between Trial 1 and Trials 2 and 3 ( p ⬍ .01). The arm choices for Sessions 24 –25 were also used to see whether either group had adopted the strategy of preferentially selecting the arm opposite to the previously selected one. It was found that 9% of the Peri group’s arm choices were of this sort, compared with 12% of the sham group’s choices (out of

236

MACHIN, VANN, MUIR, AND AGGLETON

Figure 4. Radial arm maze: Replacement after arm rotation (Stage 2). Mean (⫾ SEM) errors made in each trial when retrieving the last four pellets after maze rotation. Peri ⫽ perirhinal cortex lesions; Sham ⫽ surgical controls.

a minimum of 48 arm choices). The two groups did not differ significantly on this measure, t(21) ⫽ 1.16, nor did they show any evidence of reliance on this strategy. Stage 4 also involved rats performing three trials a day for 2 days in the same maze, but with the maze placed in a novel room (see Figure 5). There was no main effect of group when either the number of correct entries in the first eight or the total number of errors were considered (Fs ⬍ 1). There was, however, a main effect of trial on the number of correct entries, F(1, 21) ⫽ 6.0, p ⬍ .01, but not on the total number of errors, F(1, 21) ⫽ 2.2, p ⬎ .10. There was no main effect of group on all three trials of the first day in the novel room, in terms of either number of errors or correct arm entries (Fs ⬍ 1). Similarly, when performance on the first trial of the first day in the novel room was analyzed separately, there was no main effect of group on either the total number of errors, F(1, 21) ⫽ 1.3, p ⬎ .10, or correct entries (F ⬍ 1).

eight possible start positions. The rats received 2 days of pretraining, with four swims per day. During this pretraining, a curtain was drawn around the pool, and both the start position and platform position were changed for every forced swim. Each swim was terminated either when the rat located the submerged platform or after 120 s had elapsed. If the rat had not located the platform at the end of the 120 s, it was guided there by the experimenter and then had to remain on the platform for 30 s. For the 12 days of actual training, the curtain was removed from around the pool. The location of the platform remained constant across the four trials of a given day but varied between days. A different start position was used for each of the four trials within a session. Each trial terminated when the rat had located the platform or after 120 s had elapsed and the rat was left on the platform for 30 s. The next trial began immediately afterward, giving an intertrial interval of about 15 s. As the subjects received four trials a day with the same platform location, only the second trial of each day can be regarded as a specific test of working memory. On Days 13–16, the delay between the first and second trials was increased to 30 min, during which time the subject was returned to the home cage. The rats still received four trials a day with the same platform location, but after the second trial the intertrial interval for the two remaining trials was reduced to 15 s. The rats were transported between the holding room and water maze in an opaque, aluminum traveling box. They were also placed in the opaque holding box in between each trial.

Results The first 12 days involved the standard working memory task in the water maze, with a 15-s intertrial interval between the sample (Trial 1) and test (Trial 2), as well as the remaining two trials (see Figure 6). Analysis of the path lengths to reach the hidden platform revealed no group difference, F(1, 20) ⫽ 0.1, p ⬎ .50, although there was a significant main effect of trial, F(1, 20) ⫽ 83.3, p ⬍ .01. The trial effect reflected the much shorter path length on the second versus the first trial, showing that both groups were learning the new platform position. This pattern of learning was the same for both groups, as shown by the lack of a Group ⫻ Trial interaction (F ⬍ 1). Analysis of the simple effects confirmed that

Experiment 2: Water Maze Method Subjects and apparatus. The same rats were used as for the previous experiment except 1 from the perirhinal lesion group, which received no training because it became intolerant to being handled. For this experiment, both food and water were available ad libitum. The water maze (2 m in diameter, 60 cm deep) was made of white fiberglass and was mounted 58 cm above the floor. The pool was filled with water (24 ⫾ 1 °C) made opaque by the addition of nontoxic emulsion (Opacifier, Chesham Chemicals, Harrow, UK). An escape platform (10 cm in diameter, 2 cm below the water surface) could be placed in the pool. The pool was in a room (305 cm long ⫻ 396 cm wide) with salient cues on the walls, and a curtain was used to conceal the experimenter. Lighting was provided by four floormounted spotlights (500-W floodlights). The paths of the rats were tracked with a video camera suspended directly above the pool and were recorded on videotape. Data were collected and analyzed on-line with an HVS image analyzer connected to an Archimedes RISC computer (Acorn, Histon, UK) that used Watermaze software (Spooner, 1994). Behavioral training. The rats were tested on a working memory task in the water maze (delayed matching-to-place). Twelve platform positions that varied in their distance from the pool perimeter were used, along with

Figure 5. Radial arm maze: Repeated trials within a session (Stage 3) and switch to a new test room (Stage 4). The graph shows the mean (⫾ SEM) number of correct arm entries in the first eight choices. Stage 3 involved 3 days with two repeated trials, followed by 2 days with three consecutive trials. Stage 4 involved 2 days with three consecutive trials in a novel room. Peri ⫽ perirhinal cortex lesions; Sham ⫽ surgical controls.


Figure 6. Water maze: Acquisition and performance of the delayed matching-to-place task. The graph shows the mean (⫾ SEM) path length to find the hidden platform for each trial during successive blocks of 3 sessions (12 sessions in total). Trial 1 is the sample trial. Peri ⫽ perirhinal cortex lesions; Sham ⫽ surgical controls.

there were no group differences for any of the four trials (see Figure 6): maximum, F(1, 80) ⫽ 1.4, p ⫽ .24. Latency measures provided exactly the same pattern of results and, again, there was no evidence of any group differences (Fs ⬍ 1). The next stage (Days 13–16) involved a 30-min delay between Trial 1 (sample) and Trial 2 (working memory test), with the remaining two trials run as before (see Figure 7). Analysis of the path lengths again showed no overall group difference (F ⬍ 1) and no Group ⫻ Trial interaction (F ⬍ 1), although there was a significant main effect of trial, F(1, 20) ⫽ 44.1, p ⬍ .01. Subsequent analysis revealed a significant difference between Trials 1 and 2 ( p ⬍ .01), showing that rats had retained knowledge of the new platform position over a 30-min interval. There were, however, no group differences for any of the four trials when the simple effects were examined, maximum F(1, 65) ⫽ 1.9, p ⫽ .18. The same pattern of results was found when latency measures were analyzed, as again there was no overall group effect (F ⬍ 1) and no Group ⫻ Trial interaction (F ⬍ 1). Consistent with this, analysis of the simple effects showed that there were no group differences for any of the trials (Fs ⬍ 1).

237

Mumby & Glenn, 2000). In view of the complete nature of many of these lesions (e.g., Bussey et al., 1999) it is not realistic to suppose that this lack of effect simply reflects the consequences of incomplete surgery. In contrast, other studies have reported that perirhinal lesions in rats do produce spatial impairments (Liu & Bilkey 1998a, 1998b, 1998c, 1999, 2001; Wiig & Bilkey, 1994a, 1994b; Wiig & Burwell, 1998). The behavioral tests used in these studies include delayed nonmatching-to-position in a Skinner box (Wiig & Burwell, 1998), place learning in the Morris water maze (Liu & Bilkey 1998a, 2001; Wiig & Bilkey, 1994a), T-maze alternation (Wiig & Bilkey, 1994b), and working memory tested in a radial arm maze (Liu & Bilkey, 1998b, 2001). There is clearly an urgent need to resolve these different patterns of results for seemingly similar tests. For this reason the present study examined two factors that might account for some of these differences, namely testing protocol and task selection. The first of these arose from the suggestion that the effects of perirhinal cortex lesions are most evident when the test environment is novel (Bilkey, 1999; Liu & Bilkey, 1999, 2001). The second arose from the report that the disruptive effects of perirhinal cortex lesions on DMP in the water maze can be comparable to those of hippocampal lesions at longer (180-s) retention intervals (Liu & Bilkey, 2001; but see Glenn & Mumby, 1998). The results of the present study were very clear, as there was no evidence that perirhinal cortex lesions disrupted performance on either the radial arm maze or the water maze tasks. This lack of effect was found even when habituation and pretraining for the radial arm maze task were limited. Thus, the Peri rats required a median of 1 day (1 day habituation, plus 0 days pretraining) to make eight arm choices in 10 min and a median of 2 days (including habituation) to retrieve all eight pellets in less than 10 min. This can be compared with the 2 days of habituation used by Liu and Bilkey (1998b, 1999). This brief training period was possible because the subjects had already been trained on a very different task (in an operant chamber) and so had received exten-

General Discussion Reports on the effects of perirhinal cortex lesions on spatial memory are contradictory. In some studies, neurotoxic lesions of the perirhinal cortex or of the perirhinal cortex plus postrhinal cortex had no apparent effect on standard spatial memory tasks (T-maze alternation, radial arm maze, Morris water maze) that typically tax allocentric processing (Aggleton et al., 1997; Bussey, Dias, Amin, et al., 2001; Bussey, Dias, Redhead, Pearce, Muir, & Aggleton, 2001; Bussey, Duck, Muir, & Aggleton, 2000; Bussey et al., 1999; Ennaceur & Aggleton, 1997; Ennaceur et al., 1996). A similar lack of any apparent deficit after conventional lesions of the perirhinal cortex has also been reported on tasks such as spatial alternation and place learning in the Morris water maze (E. A. Gaffan et al., 2000; Glenn & Mumby, 1998; Kolb et al., 1994;

Figure 7. Water maze: Performance on delayed matching-to-place task with a 30-min delay between Trial 1 (sample) and Trial 2. The platform remained in the same position for all four trials, with a 15-s intertrial interval between Trials 2 and 3 and Trials 3 and 4. The graph shows the mean path length to find the hidden platform for each trial over four sessions. Peri ⫽ perirhinal cortex lesions; Sham ⫽ surgical controls.

238


sive handling and were familiar with the food rewards. In fact, the Peri group required significantly fewer pretraining sessions to complete eight arm choices than did the sham group. This difference means that, if anything, the room cues were more novel for the Peri group. Nevertheless, no evidence of a deficit was found, even in the very earliest trials. It might also be noted that during the habituation period, food pellets were scattered throughout the maze, as it has been suggested that a failure to do so might aid subjects with perirhinal lesions (Bilkey, 1999). The final stage of the radial arm maze task provided a more direct test of the impact of novel surroundings. For this stage, the maze was switched to a new room, and data were obtained from three trials within the very first session. Once again, there was no evidence that perirhinal cortex lesions impaired performance. It was also possible to show that the lesions did not increase sensitivity to proactive interference. This is because Stages 3 and 4 used multiple daily trials, which led to a highly significant increase in errors across trials within a session, consistent with a buildup of proactive interference. In fact, the rats with perirhinal cortex lesions made fewer errors under these conditions of high proactive interference than did the sham-lesioned rats. Experiment 2 examined the performance of rats with perirhinal cortex lesions on a DMP task in the Morris water maze. With short retention intervals (15 s), no deficit was found, a result consistent with previous findings (Glenn & Mumby, 1998; Liu & Bilkey, 2001). Unlike Glenn and Mumby (1998), who also used longer delays of up to 5 min and conditions of high proactive interference but still found no perirhinal cortex lesion deficit on this task, Liu and Bilkey (2001) did report marked deficits in rats with excitotoxic lesions. The latter deficit is noteworthy as it was delay dependent, becoming apparent at delays of 180 s. Furthermore, the impairment was not significantly different from that observed in rats with excitotoxic hippocampal lesions (Liu & Bilkey, 2001). The present study used a similar DMP task, but after initial acquisition (12 sessions), the retention interval between the sample and test was increased from 15 s to 30 min (4 sessions). Both groups of rats were still able to remember the location of the hidden platform after this extended period. Thus, even though the present study used a much longer retention interval (180 s vs. 30 min) than Liu and Bilkey (2001), there was still no evidence of a lesion-induced deficit. There are, however, a number of potentially important differences in the training procedures for these DMP tasks in the water maze. In the study by Glenn and Mumby (1998), in which perirhinal lesions had no apparent effect, the rats were trained before surgery. In the study by Liu and Bilkey (2001), rats were first trained for six sessions on a reference memory task in the water maze. Although the perirhinal cortex–lesioned rats showed a transient impairment on Days 1 and 2, their subsequent probe and reversal performance was normal. The rats were then switched to the DMP task for a final four sessions, within which two different conditions were used. Thus, unlike the present study, in which the rats were only trained on the DMP task in the water maze, the rats in the Liu and Bilkey (2001) study were required to change tasks in the same apparatus. Furthermore, they were only tested on each new DMP condition for a very limited period (two sessions each). As a consequence, it is not possible to determine whether the perirhinal deficit was one of location memory or a problem with adapting to new task variants.

When considering why the effects of perirhinal lesions appear to differ so much across studies, it is important to confirm whether the current tasks did indeed tax allocentric spatial memory. It was for this reason that Stage 2 of the radial arm maze task required the apparatus to be rotated after the first four arm choices. Both groups performed significantly above chance for the first arm choice when replaced in the apparatus, consistent with the use of allocentric cues. Furthermore, an analysis of arm choices failed to find evidence of reliance on an egocentric strategy. For the Morris water maze task, the rats were started from different release points for each trial. As a consequence, they could only use distal landmarks to learn the position of the platform as it varied from day to day, thus precluding the use of egocentric or idiothetic information. Consistent with this, the water maze DMP has been found to be highly sensitive to hippocampal dysfunction (Riedel et al., 1999; Steele & Morris, 1999). In contrast, tasks such as delayed nonmatching-to-position in an operant chamber (Wiig & Burwell, 1998) are unlikely to require allocentric cues (Aggleton, Neave, Nagle, & Sahgal, 1995; Chudasama & Muir, 1997). Likewise, in the T-maze alternation study by Liu and Bilkey (1998a), the rats were insensitive to the removal of distal room cues, revealing that they were not reliant on allocentric processing. Having made these exceptions, there remain other studies in which perirhinal cortex lesions impaired performance under conditions that do tax allocentric processing (e.g., Liu & Bilkey, 1998c, 2001). Perhaps the most parsimonious view is that basic mechanisms of spatial learning and memory that depend on allocentric cues are spared after perirhinal cortex removal. Indeed, on close inspection, those studies reporting spatial deficits on allocentric tasks often describe initial deficits that then disappear (Liu & Bilkey, 1998b, 1998c, 2001; Mumby & Glenn, 2000; Wiig & Bilkey, 1994a). Although these might reflect ceiling effects, it is also possible that these initial deficits are not spatial per se. For example, a perirhinal cortex lesion deficit was reported for the Morris water maze reference task, but it was confined to Day 1 (Liu & Bilkey, 1998c). During this session, the rats with perirhinal cortex lesions showed a greater tendency to swim close to the walls, which need not reflect a spatial deficit. Indeed, it is known that perirhinal cortex lesions result in abnormal patterns of object exploration and habituation (Bussey et al., 2000; Ennaceur et al., 1996; Liu & Bilkey, 2001), suggesting ways in which initial behavior could be disrupted through nonspatial deficits. In fact, a large body of evidence shows that the perirhinal cortex is crucial for object recognition and the use of object-based information (Meunier, Bachevalier, Mishkin, & Murray, 1993; Mumby & Pinel, 1994; Murray & Bussey, 1999; Suzuki, 1996). This involvement is also observed for the integration of object information with spatial information (Aggleton & Brown, 1999; Bussey & Aggleton, in press; Bussey, Dias, Amin, et al. 2001; Bussey, Dias, Redhead, et al., 2001; D. Gaffan & Parker, 1996). It is thus plausible that perirhinal cortex lesions interfere with initial task acquisition if potential test stimuli (e.g., floors and walls of runways) can be categorized as both discrete objects and components of spatial cues (Cassaday & Rawlins, 1997; Mumby & Glenn, 2000). Although this interference might lead to mild deficits, it could also lead to a paradoxical facilitation of performance on spatial tasks if the memory and resulting attention for specific objects disrupts task performance in normal subjects. In fact, a significant improvement in performance levels by rats with


perirhinal cortex lesions has been found for T-maze alternation with extended delays (Bussey et al., 2000; Bussey, Dias, Redhead, et al., 2001) and for the acquisition of the working memory radial arm maze task (Bussey et al., 1999). In the present study, the Peri group outperformed the controls on the initial part of Stage 3 of the radial arm maze task, when the rats received two consecutive trials within a single session. It can be seen that these facilitatory effects are confined to runway type tasks (T-maze, radial arm maze), where local “object” cues might capture attention in normal subjects and thus lead to incorrect choices. An alternative explanation for these findings is that perirhinal cortex lesions reduce sensitivity to proactive interference, reflecting a faster decay of prior, interfering trials. The problem with this view is that it fails to explain how the same lesions can leave intact the memory for places visited 30 min before, either in the water maze (Experiment 2) or in the radial arm maze (Bussey et al., 1999), yet increase forgetting rates. It is important to appreciate that this interpretation of the effects of perirhinal cortex lesions does not preclude the existence of vital, functional interactions with the hippocampus. Indeed, tasks such as those looking at object–place information provide some of the strongest evidence for when these two regions function together (Bussey & Aggleton, in press; Bussey et al., 2000; Bussey, Dias, Amin, et al., 2001; D. Gaffan & Parker, 1996). For this reason, a study by Liu and Bilkey (1999) is of especial interest, as they examined delayed nonmatching-to-sample in a radial arm maze task in which the salience of distal and local cues was varied on alternate trials. In fact, they found no evidence that this manipulation selectively influenced rats with perirhinal cortex lesions, but neither was there evidence that the lesioned rats performed worse than the controls on either condition (Liu & Bilkey, 1999). This leaves the possibility that this manipulation lacked sufficient sensitivity to detect a lesion effect and so leaves unresolved whether the key factor is the balance between spatial stimuli and objectbased stimuli (Bussey & Aggleton, in press). At the same time, great care needs to be taken to avoid falling into the circular argument that if there is a perirhinal deficit the task is influenced by object-based stimuli and if there is no deficit then it is purely “spatial.”

References Aggleton, J. P., & Brown, M. W. (1999). Episodic memory, amnesia and the hippocampal–anterior thalamic axis. Behavioral Brain Sciences, 22, 425– 489. Aggleton, J. P., Keen, S., Warburton, E. C., & Bussey, T. J. (1997). Extensive cytotoxic lesions of the rhinal cortices impair recognition but spare spatial alternation in the rat. Brain Research Bulletin, 43, 279 – 287. Aggleton, J. P., Neave, N., Nagle, S., & Sahgal, A. (1995). A comparison of the effects of medial prefrontal, cingulate cortex, and cingulum bundle lesions on tests of spatial memory: Evidence of a double dissociation between frontal and cingulum bundle contributions. Journal of Neuroscience, 15, 7270 –7281. Aggleton, J. P., Vann, S. D., Oswald, C. J. P., & Good, M. (2000). Identifying cortical inputs to the hippocampus that subserve allocentric spatial processes: A simple question with a complex answer. Hippocampus, 10, 466 – 474. Bilkey, D. K. (1999). Perirhinal cortex: Lost in space? Behavioral Brain Sciences, 22, 444 – 445. Burwell, R. D., & Amaral, D. G. (1998). Cortical afferents of the perirhi-

239

nal, postrhinal, and entorhinal cortices of the rat. Journal of Comparative Neurology, 398, 179 –205. Bussey, T. J., & Aggleton, J. P. (in press). The “what” and “where” of event memory: Independence and interaction within the temporal lobe. In A. Parker, T. J. Bussey & E. Wilding (Eds.), The cognitive neuroscience of memory formation: Encoding and retrieval. London: Psychology Press. Bussey, T. J., Dias, R., Amin, E., Muir, J. L., & Aggleton, J. P. (2001). Perirhinal cortex and place-object conditional place learning. Behavioral Neuroscience. 115, 776 –785. Bussey, T. J., Dias, R., Redhead, E. S., Pearce, J. M., Muir, J. L., & Aggleton, J. P. (2001). Intact negative patterning in rats with fornix or combined perirhinal and postrhinal cortex lesions. Experimental Brain Research, 134, 506 –519. Bussey, T. J., Duck, J. D., Muir, J. L., & Aggleton, J. P. (2000). Distinct patterns of behavioural impairments resulting from fornix transection or neurotoxic lesions of the perirhinal and postrhinal cortices in the rat. Behavioural Brain Research, 111, 187–202. Bussey, T. J., Muir, J. L., & Aggleton, J. P. (1999). Functionally dissociating aspects of event memory: The effects of combined perirhinal and postrhinal lesions on object and place memory in the rat. Journal of Neuroscience, 19, 495–503. Bussey, T. J., Muir, J. L., Everitt, B. J., & Robbins, T. W. (1997). Triple dissociation of anterior cingulate, posterior cingulate, and medial frontal cortices on visual discrimination tasks using a touchscreen testing procedure for the rat. Behavioral Neuroscience, 111, 1–17. Cassaday, H. J., & Rawlins, J. N. P. (1997). The hippocampus, objects, and their contexts. Behavioral Neuroscience, 111, 1228 –1244. Chudasama, Y., & Muir, J. L. (1997). A behavioural analysis of the delayed non-matching to position task: The effects of scopolamine, lesions of the fornix and of the prelimbic region on mediating behaviours by rats. Psychopharmacology (Berlin), 134, 73– 82. Ennaceur, A., & Aggleton, J. P. (1997). The effects of neurotoxic lesions of the perirhinal cortex combined to fornix transection on object recognition memory in the rat. Behavioural Brain Research, 88, 181–193. Ennaceur, A., Neave, N. J., & Aggleton, J. P. (1996). Neurotoxic lesions of the perirhinal cortex do not mimic the behavioural effects of fornix transection in the rat. Behavioural Brain Research, 80, 9 –25. Gaffan, D., & Parker, A. (1996). Interaction of perirhinal cortex with the fornix–fimbria; Memory for objects and “object-in-place” memory. Journal of Neuroscience, 16, 5864 –5869. Gaffan, E. A., Eacott, M. J., & Simpson, E. L. (2000). Perirhinal cortex ablation in rats selectively impairs object identification in a simultaneous visual comparison task. Behavioral Neuroscience, 114, 18 –31. Glenn, M. J., & Mumby, D. G. (1998). Place memory is intact in rats with perirhinal lesions. Behavioral Neuroscience, 112, 1353–1365. Kolb, B., Buhrmann, K., McDonald, R., & Sutherland, R. J. (1994). Dissociation of the medial prefrontal, posterior parietal, and posterior temporal cortex for spatial navigation and recognition memory in the rat. Cerebral Cortex, 6, 664 – 680. Lavenex, P., & Amaral, D. G. (2000). Hippocampal-neocortical interaction: A hierarchy of associativity. Hippocampus, 10, 420 – 430. Liu, P., & Bilkey, D. K. (1998a). Excitotoxic lesions centered on perirhinal cortex produce delay-dependent deficits in a test of spatial memory. Behavioral Neuroscience, 112, 512–524. Liu, P., & Bilkey, D. K. (1998b). Lesions of perirhinal cortex produce spatial memory deficits in the radial maze. Hippocampus, 8, 114 –121. Liu, P., & Bilkey, D. K. (1998c). Perirhinal cortex contributions to performance in the Morris water maze. Behavioral Neuroscience, 112, 304 –315. 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. Liu, P., & Bilkey, D. K. (2001). The effect of excitotoxic lesions centered

240


on the hippocampus or perirhinal cortex in object recognition and spatial memory tasks. Behavioral Neuroscience, 115, 94 –111. Meunier, M., Bachevalier, J., Mishkin, M., & Murray, E. A. (1993). Effects on visual recognition of combined and separate ablations of the entorhinal and perirhinal cortex in monkeys. Journal of Neuroscience, 13, 5418 –5432. Morris, R. G. M. (2001). Episodic-like memory in animals: Psychological criteria, neural mechanisms and the value of episodic-like tasks to investigate animal models of neurodegenerative disease. Philosophical Transactions of the Royal Society London, Series B, 1453–1465. 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. Mumby, D. G., & Glenn, M. J. (2000). Anterograde and retrograde memory for object discriminations and places in rats with perirhinal cortex lesions. Behavioural Brain Research, 114, 119 –134. Mumby, D. G., & Pinel, J. P. J. (1994). Rhinal cortex lesions and object recognition in rats. Behavioral Neuroscience, 108, 11–18. Murray, E. A., & Bussey, T. J. (1999). Functions of the perirhinal cortex. Trends in Neurosciences, 3, 142–151. O’Keefe, J., & Nadel, L. (1978). The hippocampus as a cognitive map. Oxford, England: Oxford University Press. Olton, D. S., & Samuelson, R. J. (1976). Remembrance of places passed: Spatial memory in rats. Journal of Experimental Psychology: Animal Behavior Processes, 2, 97–114. Olton, D. S., Walker, J. A., & Gage, F. H. (1978). Hippocampal connections and spatial discrimination. Brain Research, 139, 295–308. Paxinos, G., & Watson, C. (1997). The rat brain in stereotaxic coordinates (3rd ed.). San Diego, CA: Academic Press. Pearce, J. P., Roberts, A. D. L., & Good, M. (1998, November 5). Hippocampal lesions disrupt navigation based on cognitive maps but not heading vectors. Nature, 396, 75–77. Riedel, G., Micheau, J., Lam, A. G. M., Roloff, E. V. L., Martin, S. J.,

Bridge, H., et al. (1999). Reversible neural inactivation reveals hippocampal participation in several memory processes. Nature Neuroscience, 2, 898 –905. Spooner, R. (1994). Watermaze (Version 2. 12) [Computer software]. Edinburgh, Scotland: Richard Morris. Squire, L. R., & Zola-Morgan, S. (1991, September 20). The medial temporal lobe memory system. Science, 253, 1380 –1386. Steele, R. J., & Morris, R. G. M. (1999). Delay-dependent impairment of a matching-to-place task with chronic and intrahippocampal infusion of the NMDA-antagonist and D-AP5. Hippocampus, 9, 118 –136. Suzuki, W. A. (1996). The anatomy, physiology and functions of the perirhinal cortex. Current Opinion in Neurobiology, 6, 179 –186. Wiig, K. A., & Bilkey, D. K. (1994a). The effects of perirhinal cortical lesions on spatial reference memory in the rat. Behavioural Brain Research, 63, 101–109. Wiig, K. A., & Bilkey, D. K. (1994b). Perirhinal cortex lesions in rats disrupt performance in a spatial DNMS task. NeuroReport, 5, 1405– 1408. Wiig, K. A., & Burwell, R. D. (1998). Memory impairments on a delayed nonmatching-to-position task after lesions of the perirhinal cortex in the rat. Behavioral Neuroscience, 112, 827– 838. Witter, M. P., Groenewegen, H. J., Lopes da Silva, F. H., & Lohman, A. H. M. (1989). Functional organization of the extrinsic and intrinsic circuitry of the parahippocampal region. Progress in Neurobiology, 33, 161–253. Witter, M. P., Naber, P. A., Haeften, T. V., Machielsen, W. C. M., Rombouts, S. A. R. B., Barkhof, F., et al. (2000). Cortico-hippocampal communication by way of parallel parahippocampal–subicular pathways. Hippocampus, 10, 398 – 410.

Received May 22, 2001 Revision received September 10, 2001 Accepted October 8, 2001 䡲