Selective Memory Impairments Produced by Transient Lidocaine ...

Behavioral Neuroscience 1994. Vol. 108, No. 3. 456-468

Copyright 1994 by the American Psychological Association, Inc. 0735-7044;1>4/S3.IH>

Selective Memory Impairments Produced by Transient Lidocaine-Induced Lesions of the Nucleus Accumbens in Rats

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Jeremy K. Seamans and Anthony G. Phillips Reversible lidocaine-induced lesions of the nucleus accumbens (N.Acc.) impaired performance on the spatial win-shift, but not on the cued win-stay, radial arm maze task. Pretraining lesions on the former task did not affect foraging for 4 pellets during either the training or test phases. In contrast, lesions given prior to the test phase significantly disrupted retrieval of 4 pellets on the 8-arm maze. Comparable deficits also were observed in rats trained to forage for 4 pellets on an 8-arm maze without prior win-shift experience. State-dependent drug effects were ruled out by replicating the disruptive effects of lidocaine infusions into the N.Acc. on spatial win-shift performance in rats receiving this treatment prior to both training and test phases. These results suggest that the N.Acc. may interact with the hippocampus to guide foraging behavior requiring memory of previous spatial locations on a maze.

Over the past decade, there has been a growing interest in the function of the nucleus accumbens (N.Acc.) as a neural interface between the limbic and motor systems of the brain (Mogenson, Jones, & Yim, 1980; Mogenson & Yang, 1991). Recently, Cools, Van Den Bos, Ploeger, and Ellenbroek (1991) advanced the idea that noradrenaline in the N.Acc. may gate input received from limbic regions. In a similar vein, dopamine activity in the N.Acc. may have a gain-amplifying effect on afferent input from the amygdala (Cador et al., 1991). Likewise, the glutamatergic input to the N.Acc. from the hippocampus (Groenewegen et al., 1991; Pennartz & Katai, 1991) may also be modulated by dopamine activity. Stimulation of discrete areas of the hippocampus activates cells in the N.Acc. (Pennartz, Dolleman-Van Der Weel, Katai, & Lopes DA Silva, 1992; Pennartz & Katai, 1991), whereas microinjection of dopamine into the N.Acc. inhibits these excitatory responses (Yang & Mogenson, 1984, 1985). DeFrance, Sikes, & Chronister (1985) have postulated that this effect of dopamine may serve to increase the signal-to-noise ratio of hippocampal input to the N.Acc. Although the behavioral manifestations of this interaction between the hippocampus and N.Acc. remain unclear, damage to either region produces deficits on a variety of tasks used to assess spatial memory, such as acquisition of a spatial Morris water maze task or spatial reversal in the T-maze (Annett, McGregor, & Robbins, 1989; Brandeis, Brandys, & Yehuda, 1989; Goodlett, Nichols, Halloran, & West, 1989; Isaacson, 1982; Sutherland & Rodriguez, 1989). Scheel-Kruger and

Willner (1991) reported that microinjections of glutamate antagonists into the N.Acc. produce impairments in a spatial water maze task both during initial training and when the task is well learned. Microinjection of a glutamate antagonist into the N.Acc. also leads to more entries into a consistent set of unbaited arms in a radial arm maze, thereby indicating a reference memory impairment (Schacter, Yang, Innis, & Mogenson, 1989). However, neither fornix nor CA1 lesions had an effect on reference memory on either a 17- or 8-arm maze, respectively (Davis, Baranowski, Pulsinelli, & Volpe, 1986; Olton & Pappas, 1979). As the reference memory component of this task is sensitive to lesions of the striatum (Colombo, Davis, & Volpe, 1989; Packard & White, 1990), it may also be argued that the N.Acc. plays a synergistic role with the dorsal striatum in guiding specific behaviors. Indeed, disruption of dorsal or ventral striatal function results in impairments in spatial reversal in the T-maze (Annett et al., 1989; Divac, 1971) and spatially mediated behavior in the Morris water maze (ScheelKuger & Willner, 1991; Sutherland & Rodriguez, 1989; Whishaw, Mittleman, Bunch, & Dunnett, 1987). Collectively, these data suggest that the N.Acc. may be involved in memory processes mediated by either the caudate nucleus or the hippocampal formation. The key to determining the specific contribution of the N.Acc. to hippocampal or dorsal striatal function lies in the use of tasks that are influenced selectively by lesions of each of these brain regions. One candidate is the spatial win-shift eight-arm radial maze task developed by Olton and Samuelson (1976) and Packard, Regenold, Quirion, and White (1990). This task draws on many of the functions ascribed to the hippocampus, such as working memory, spatial memory, and behavioral flexibility (Hirsh, 1970; O'Keefe & Nadel, 1978; Olton & Pappas, 1979). An important finding is that this task is disrupted by lesions of the hippocampus, but not by damage to the striatum or amygdala (McDonald & White, 1993). In contrast, efficient performance of the cued win-stay task (Packard, Hirsh, & White, 1989; Packard & White, 1991) is influenced by several functions ascribed to the striatum, including association of a response with reinforcement, refer-

Jeremy K. Seamans and Anthony G. Phillips, Department of Psychology, University of British Columbia, Vancouver, British Columbia, Canada. This research was supported by a grant from the Natural Sciences and Engineering Council of Canada to Anthony G. Phillips. We thank Norman White and Robert McDonald for many helpful suggestions, and we thank Stan Floresco and Tim Bussey for assistance with behavioral testing. Correspondence concerning this article should be addressed to Anthony G. Phillips, Department of Psychology, University of British Columbia, Vancouver, British Columbia, Canada V6T 12.4. Electronic mail may be sent to seamans@'unixg.ubc.ca.

456


NUCLEUS ACCUMBENS AND MEMORY

ence memory, habit learning, and memory for a visual stimulus (Colombo et al., 1989; Cook & Kesner, 1988; Hikosaka & Sakamoto, 1986; Mitchell, Channell, & Hall, 1985; Packard et al., 1989; Packard & McGaugh, 1992; Packard & White, 1990). Furthermore, this task is disrupted by lesions of the dorsal striatum, but not by lesions of the hippocampus or amygdala (McDonald & White, 1993). The present series of experiments used spatial win-shift and cued win-stay procedures to determine the possible contribution of the N.Acc. to behaviors mediated selectively by the hippocampus or the caudate nucleus. Transient lesions of the N.Acc. were produced by intracranial injections of the local anesthetic lidocaine, a technique used to examine the role of the hippocampal-MLR (mesencephalic locomotor region) pathway in locomotion (Mogenson & Yang, 1991) and to block the activity of cerebellar nuclei transiently during the performance of a conditioned eyeblink task (Chapman, Steinmetz, Sears, & Thompson, 1990; Knowlton & Thompson, 1988; Welsh & Harvey, 1991). Others (Perez-Ruiz & Prado-Akala, 1989) have demonstrated that intrastriatal lidocaine infusions produce impairments in a passive avoidance task. Lidocaine is thought to act as a local anesthetic by reducing ionic permeability (Hille, 1992). On the basis of the effects of lidocaine infusions into the fornix on the magnitude of evoked potentials in the N.Acc., onset of the action of lidocaine is approximately 1 min, whereas the anesthetic effects of the drug dissipate in 10 to 15 min (Boeijinga, Mudler, Pennartz, Manshanden, & Lopes Da Silva, 1993). There are different estimates of the functional spread of lidocaine within the brain that appear to depend on the rate of infusion. Using an infusion rate of 1 jil/min, Welsh and Harvey (1991) estimated that the functional spread is 1.4 mm in the cerebellum from the site of infusion. The functional spread of lidocaine in the occulomotor nucleus was estimated to be 0.5 mm with an infusion rate of 4 jxl/15 min (Albert & Madryga, 1980). The present study used an infusion rate of 1 (il/2 min to ensure optimal spread within the N.Acc. Experiment 1 was conducted to examine the effects of intra-N.Acc. lidocaine infusions on the performance of a spatial win-shift task and a cued win-stay task. In Experiment 2, transient N.Acc. lesions were delivered prior to the training phase or prior to the test phase of the spatial win-shift task to determine effects on the acquisition or retrieval of newly acquired information. In Experiment 3 we examined the effects of transient N.Acc. lesions on a random foraging task. As this task was identical to the test phase of the delayed spatial win-shift task, it permitted an assessment of the effects of transient N.Acc. lesions on acquisition or use of information solely within the test phase. Experiment 4 provided a necessary control for state-dependency phenomena, given that the rats in Experiments 1 and 2 were either trained in a drug-free state and tested under the influence of lidocaine or vice versa.

General Method Subjects The subjects were 104 male Long-Evans rats (350-550 g; Wilmington, MA) housed individually in a temperature- and light-controlled (12-hr light-dark cycle) colony. All subjects were given free access to

457

water and were maintained at 85% of their free-feeding weight with servings of approximately 25-30 g of Purina Lab Chow pellets once daily. All rats were tested between 3 and 8 p.m. daily.

Surgery Before surgery rats were anesthetized with 100 mg/kg ketamine hydrochloride and 7 mg/kg xylazine. We implanted 22-gauge stainless steel guide cannulas bilaterally with standard stereotaxic procedures in the ventromedial region of the N.Acc. The stereotaxic coordinates were AP = +1.9 mm and ML = ±1.4 mm from bregma and DV = -6.2 mm from cortex (Paxinos & Watson, 1982). Twenty-six-gauge steel obturators flush with the guide cannulas remained in place until the injections were made. Each rat convalesced for at least 10 days prior to behavioral testing.

Injection Procedure All rats received injections in their home cages in the testing rooms. We delivered 1 u.1 of 20-mg/ml lidocaine hydrochloride (Astra Phamaceuticals) or 1 (xl isotonic saline via 26-gauge inner cannulas protruding 0.8 mm beyond the guide at a rate of 1 u.1/2 min by a Sage Instruments Model 341 syringe pump. Injection cannulas were left in place for an additional 1 min to allow for diffusion. Each rat remained in its home cage for another 2 min before being placed on the maze.

Apparatus Two similar eight-arm radial mazes were used for all behavioral testing. Both mazes had octagonal center platforms measuring 40 cm in diameter. Radiating out from these platforms were eight arms each (50 x 9 cm) with cylindrical food cups at the ends of the maze. Maze 1 had eight 6-W lightbulbs mounted on Plexiglas sheets (6 x 10 cm) located directly above the food cups. Metal doors (9 x 13 cm), which laid flat against each arm, could be raised to block the entrance to each arm of Maze 1, and pieces of opaque plastic (9 x 13 cm) were used to block arms in Maze 2. In Experiment 1, Maze 1 was used for both the spatial win-shift and cued win-stay tasks. For the spatial win-shift task, the maze was elevated 40 cm off the ground and surrounded by several extra maze cues. For the cued win-stay task, the maze was placed on a large opaque cylindrical platform surrounded on all sides by curtains. Rats were monitored by an overhead video camera. The lights at the ends of the arms were operated by a remote switch box. All other experiments were conducted in Maze 2. This maze was elevated 40 cm off the ground and surrounded by several extra maze cues.

Histology After behavioral testing, the rats were sacrificed in a carbon dioxide chamber. Brains were removed and fixed in a 10% formalin solution. The brains were then frozen and sliced in 40-u.m sections prior to mounting and staining with the Kluver-Barrera method (Luna, 1960). Figure 1 indicates schematically the region of the N.Acc. in which the injection sites were located bilaterally. Cannula placements were largely confined within the medial N.Acc. Two rats in Experiment 1, 4 rats from Experiment 2, and 4 rats from Experiment 4 were not included in the data analysis because tissue damage was observed ventral to the N.Acc. No disturbances in foraging behavior were observed in subjects where subsequent histological analysis revealed tissue damage exclusively outside the N.Acc.

458

JEREMY K. SEAMANS AND ANTHONY G. PHILLIPS


AP +2.2mm

AP+1.7mm

AP+1.2mm

AP+0.7mm

Figure 1. Schematic representation of the infusion sites for all rats. Cannulas were placed bilaterally in regions indicated by hatching on the right side of each plate. N.Acc. = nucleus accumbens; co = corpus callosum; ac = anterior commissure; CPu = caudate putamen. Original computer-generated drawings using reference points from Paxinos and Watson (1982).

maze for 10 min with no food available. Subsequently, they received five Bioserv pellets (Holton Industries, Frenchtown, NJ) in their home cage. Daily learning trials were divided into a training phase and a test phase, with a 5-min delay between phases. Before the training phase, a set of four arms, chosen randomly, were blocked, and a food pellet was placed in the food cups of the four remaining open arms. During the training phase, the rat was required to retrieve each pellet from the four baited arms within 10 min. On Day 1 of training, four pellets were also placed on the center platform, and one pellet was placed in the center of the four open arms and in each food cup. After the first training session, food was placed only in the food cups. During the test phase of each daily trial, all arms were open, but only the arms blocked during the training phase were baited. We gave the rats a maximum of 10 min to retrieve the four pellets during the test phase. Criterion performance was defined as retrieval of all four pellets during the training phase, and four pellets in five or fewer choices during the test phase for 3 consecutive days. The day after criterion performance was attained, the rats again foraged for pellets in four randomly selected open arms and were then injected with either saline or lidocaine into the N.Acc. as described in the Injection Procedure section. The test phase began after a 5-min delay. The number of visits to unbaited arms during the test phase was recorded for the day prior to and on the day of the injection. A correct choice was scored as an entry into a baited arm and consumption of the food pellet. For the cued win-stay task, the maze was placed on a large circular platform surrounded by curtains with lights at the ends of the arms to signal the presence of food. The initial 2-day adaption period was similar to the spatial win-shift task. Each subsequent day of training consisted of two similar phases separated by a 5-min delay. This procedure was used to mimic the temporal design of the spatial win-shift task and to allow time for an injection between phases. Therefore, this procedure allowed a direct comparison between performance during a transient N.Acc. lesion to drug-free performance 5 min earlier. On each daily trial, four randomly selected arms were lit and baited. Lights were turned off following consumption of each pellet. After retrieving all four pellets, the rats were removed from the maze for a 5-min delay. After the delay, the same four arms were lit and baited again. Criterion performance was attained when eight pellets could be obtained in 10 or fewer choices over two daily sessions for 3 consecutive days. The day after reaching criterion performance, rats received either saline or lidocaine injections into the N.Acc. during the period between the two training sessions. The number of incorrect choices (visits to unlit arms) was recorded for the preinjection and injection trial.

Results

Experiment 1: The Effects of Intra-N.Acc. Lidocaine Infusions on Spatial Win-Shift and Cued Win-Stay Behavior The purpose of this experiment was to examine the effects of transient N.Acc. lesions on the spatial win-shift and cued win-stay tasks to determine whether this structure may interact with the hippocampus or caudate nucleus to guide radial arm maze behavior. Method The delayed spatial win-shift and cued win-stay paradigms in the present study were adapted from Packard et al. (1989) and Packard and White (1991). The procedure for the spatial win-shift task was as follows. On the first 2 days of testing, rats were acclimatized to the

The mean number of trials to reach criterion was 10 in the spatial win-shift task and 18 in the cued win-stay task. The effects of intra-N.Acc. lidocaine injections on spatial win-shift behavior are shown in Figure 2. A two-way repeated measures analysis of variance (ANOVA) was performed on the scores obtained the during the test phase on the day prior to the injection and on the day of the injection. A significant group effect, F(\, 18) = 9.05, p = .008, and a Day x Group interaction effect, F(l, 18) = 24.91, p < .05, were observed. Simple main effect analyses yielded no significant difference between the groups on the day prior to the injection: saline, M = 0.6; lidocaine, M = 0.3 errors; F(\, 18) = 1.8, p = .19. There was a significant difference between the groups on the day of the injection: saline, M = 1.6, lidocaine, M = 5.4 errors; F(\, 18) = 25.37,p < .001.

NUCLEUS ACCUMBENS AND MEMORY

459

day prior to injection day of injection


101

Lidocaine Group

Saline Group

Figure 2. Visits (M ± SEm) to unbaited arms during the test phase of the spatial win-shift task in Experiment 1 the day prior to (hatched bar) and the day of (solid bar) lidocaine or saline injections into the nucleus accumbens prior to the test phase. **p < .01.

The effects of intra-N.Acc. lidocaine infusions on cued win-stay behavior are shown in Figure 3. A two-way repeated measures ANOVA revealed no significant difference between the saline and lidocaine groups overall, F(l, 14) = 1.168,p = .298, nor were there significant group effects between the trial prior to the injection versus the trial immediately following the injection, F(l, 14) = 0.331,p = .574. Discussion The results of this study demonstrate that intra-N.Acc. lidocaine infusions have no observable effects on motor func-

tions. Rats with transient N.Acc. lesions were indistinguishable from control rats in their ability to move from one location to another on the maze in either task. Although the N.Acc. does play a role in motivated behaviors (Phillips, Pfaus, & Blaha, 1991), the lidocaine-lesioned rats did not seem to be less motivated than controls. They always searched for and consumed the food pellets regardless of the number of choices needed to find them. Transient lesions of the medial N.Acc. did not impair cued win-stay behavior in an eight-arm radial maze. In contrast, this manipulation did produce a severe impairment in performance

day prior to injection day of injection 10E