Two experiments are reported in which rats with selective hippocampal lesions were tested on 2 prefrontal-dependent tasks. In Experiment 1, we compared the effects of lesions of the ventral hippocampus (vHC), dorsal hippocampus (dHC), and sham control surgery on the 5-choice reaction time task. Whereas rats with lesions of the dHC were indistinguishable from sham controls, those with vHC lesions showed increased premature responses and reduced accuracy throughout the experiment. The subsequent administration of systemic escitalopram (5 mg/kg), a selective serotonin reuptake inhibitor, reduced the number of premature responses in the vHC animals to control levels. In contrast, systemic injections of GBR 12909, a dopamine reuptake inhibitor, failed to ameliorate the impulsive deficit in the vHC group and, in addition, elevated perseverative responding in the vHC group only. In Experiment 2, we tested a separate group of rats with vHC lesions on a touchscreen visual discrimination and reversal learning task. Rats with vHC lesions acquired the visual discrimination as well as sham controls and showed normal inhibitory control of a previously reinforced response during reversal learning. These data support a role for the vHC in inhibitory control functions, especially in the inhibitory control of impulsive actions.
In addition to its well-established role in spatial learning and memory (for review, see Squire 1992), it has long been known that under certain conditions, hippocampal lesions induce a state of behavioral rigidity characterized by exaggerated and persistent responding. For example, rats with hippocampal lesions are resistant to extinction (Jarrard et al. 1964; Jarrard and Isaacson 1965; Swanson and Isaacson 1967; Rawlins et al. 1985), display reward-induced stereotypy (Devenport et al. 1981), exhibit abnormally high response rates (Clark and Isaacson 1965; Schmaltz and Isaacson 1966; Devenport 1979), and perseverate on a previously rewarded response (Kimble and Kimble 1965; Stevens and Cowey 1973; Nonneman et al. 1974). These observations emphasize an important role for the rodent hippocampus in the active inhibition of a behavioral response (McNaughton 1997; Gray and McNaughton 2000).
Hippocampal lesions produce behavioral effects similar to those produced by ventral prefrontal cortex (vPFC) lesions. For example, rats and mice with lesions to selective regions of the vPFC are also disinhibited, in that they act impulsively or perseverate compulsively in tasks that require the suppression of an inappropriate response (Nonneman et al. 1974; Ragozzino et al. 1999; Passetti et al. 2002; Chudasama and Robbins 2003; Chudasama et al. 2003; Bissonette et al. 2008). In rats, the ventral hippocampus (vHC) sends a unidirectional projection to the vPFC, but projections from the dorsal hippocampus (dHC) to the prefrontal cortex are weak (Jay and Witter 1991; Verwer et al. 1997; Ishikawa and Nakamura 2006). Furthermore, the vHC input to the vPFC converges with dopamine and serotonin neurons arising from the midbrain (Azmitia and Segal 1978; Sesack and Pickel 1990; Gasbarri et al. 1994), which are known to exert a powerful neuromodulatory influence over behavior (for review, see Robbins 2000). This organization of anatomical and neurochemical projections suggests that the vHC can influence activity in the vPFC and affect inhibitory control functions by modulating prefrontal dopamine and serotonin release. It is also possible that because both the vPFC and the vHC receive ascending dopamine and serotonin inputs, changes in prefrontal activity can facilitate the recovery of inhibitory control deficits induced by hippocampal lesions. In humans, selective serotonin reuptake inhibitors, which increase extracellular concentrations of monoamines including serotonin and dopamine in the prefrontal cortex (Bymaster et al. 2002; Li et al. 2002; Weikop et al. 2007), are used as cognitive-enhancing drugs in a variety of patient groups with reduced hippocampal volume (Vermetten et al. 2003; Gualtieri and Johnson 2007; Rozzini et al. 2010). If dopaminergic or serotonergic stimulation can promote recovery of inhibitory control deficits in vHC-lesioned rats, this approach would serve as a useful model for remediating cognitive deficits following long-term hippocampal damage.
In rodents, the temporary inactivation approach (e.g. intracranial infusion of muscimol) is useful for investigating the short-term behavioral effects of inactivating a region of interest. In contrast, the lesion approach is more appropriate for identifying behavioral deficits that may occur in the early or late phases of brain injury. In the present study, we created permanent hippocampal lesions in adult rats to model the effects of long-term brain damage in humans. Accordingly, we examined the effect of vHC lesions on 2 prefrontal-dependent inhibitory control tasks. First, we compared the effects of dorsal and ventral hippocampal lesions on rats’ ability to control premature (impulsive) and perseverative (compulsive) actions during performance of the 5-choice reaction time task (5-choice task). Subsequently, lesioned and control rats were systemically injected with varying doses of a selective dopamine or serotonin reuptake inhibitor to examine any potential beneficial effect of facilitating dopamine or serotonin neurotransmission on lesion-induced performance. Secondly, we evaluated whether ventral hippocampal lesions affected reversal learning, a test of inhibitory control of previously reinforced responses.
Materials and Methods
All subjects were male Long-Evans rats (Charles River, LaSalle, Quebec, Canada) housed in pairs in a temperature-controlled room (22°C), under diurnal conditions (12:12 h light:dark). Rats were food-deprived and maintained at 85% of their free-feeding weights throughout the experiment. All testing occurred at a regular time during the light period, and animals were 250–275 g at the start of behavioral training. All experimental procedures were approved by the McGill University Animal Care Committee, in accordance with the guidelines of the Canadian Council on Animal Care.
A photograph of the test apparatus for the 5-choice task is provided in Figure 1A. The test apparatus consisted of four 25 cm × 25 cm aluminum chambers (Lafayette Instruments, Lafayette, IN, USA). The rear wall of each chamber was concave and contained 9 apertures, each 2.5 cm2, 4 cm deep, and set 2 cm above floor level. Each aperture could be blocked with a metal cover when not required. For the present task, aperture numbers 1, 3, 5, 7, and 9 were open. Nosepoke responses in the apertures were detected by photocells located at the entrance of each aperture. Each aperture was illuminated by a standard 3 W bulb located at the rear of the aperture. The food magazine was located on the rear wall and was attached to a pellet dispenser. The food magazine could be illuminated with a white light emitting diode. In addition, food magazine entries were detected by photocells.
The 4 chambers were individually housed within sound-attenuating cabinets and were ventilated by low-level noise fans. Each chamber was illuminated by a 3 W house light mounted in the center of the roof. Sucrose pellets served as a food reward (Dustless Precision Pellets, Bioserve, NJ, USA). The apparatus and online data collection for each chamber were controlled by a Dell computer connected to an Animal Behavior Environmental Test system (Lafayette Instruments) using the Whisker control system for research (Cardinal and Aitken 2010).
All behavioral training and testing for the visual discrimination and reversal learning task was conducted in 4 automated touchscreen testing operant chambers (Lafayette Instruments). A photograph of the apparatus is provided in Figure 1B. The apparatus consisted of a standard operant chamber (23.5″W × 10.4″H × 13.5″D) and a touchscreen (see also Bussey et al. 2008). The 4 chambers were individually housed within sound-attenuating cabinets that were ventilated by low-level noise fans, which also served to mask extraneous background noise. Located on the ceiling of the sound-attenuating cabinet was a 3 W bulb house light, which illuminated the operant chamber.
The ceiling and 2 sidewalls of the operant chamber were made of clear Plexiglas (one of which served as the door). One side of the operant chamber was fitted with a 12″ touch-sensitive monitor measuring 11.9″W × 9.5″H × 2.4″D (ELO Touchsystems, CA, USA). The rear wall, located opposite to the touchscreen, was made of aluminum. Located centrally on the rear aluminum wall was a food magazine attached to a pellet dispenser. A light emitting diode illuminated the food magazine. Magazine entries were detected by photocells located at the entrance of the food magazine. Sucrose pellets served as a food reward (Dustless Precision Pellets, Bioserve, NJ, USA).
Computer graphic stimuli, comprising white geometric symbols on a black background, were presented on the touchscreen. A black Plexiglas mask was attached to the face of the screen approximately 1.5 cm from the surface of the display. This mask served to restrict the rat's access to the display except through response windows each measuring 2.05″L × 2.05″H. The apparatus and online data collection for each chamber were controlled by a Dell Optiplex desktop computer with custom software (Ryklin Software Inc., NY, USA).
5-Choice Reaction Time Task (5-Choice Task)
Habituation and magazine training
Rats were initially given 2 sessions (30 min), in which they were allowed to habituate to the testing chamber and collect sucrose pellets from the food magazine. Once rats were reliably retrieving and consuming pellets, they were ready for behavioral training.
Rats were trained to detect a brief visual stimulus presented randomly in 1 of the 5 spatial locations. At the beginning of each test session, the house light was illuminated and a free single pellet was delivered to the food magazine. The first trial was initiated when the rat collected the sucrose pellet from the food magazine. After a fixed 5 s intertrial interval (ITI), the light at the rear of one of the apertures was illuminated for a short period (0.5 s). Responses in this aperture during illumination and for 4.5 s afterwards (the limited hold period) were rewarded with the delivery of a single pellet, and a correct response was recorded. Responses in a nonilluminated hole during the signal period (incorrect response) and failures to respond within the limited hold period (omission) resulted in a period of darkness during which all lights were extinguished for 5 s (timeout). Responses in the apertures during the ITI were recorded as premature responses and resulted in a timeout period. Additional responses in the apertures during the limited hold period following a correct response were recorded as perseverative responses. A response in the food magazine after the delivery of a sucrose pellet, or after the timeout period, initiated the next trial.
During any one session, the light stimulus was presented an equal number of times in each of the 5 apertures in a random order. A daily session consisted of 100 trials or was terminated after 30 min of testing. For the first session of training, both the stimulus duration and the limited hold period were set at 60 s. These variables were altered on subsequent sessions according to the individual animal's performance until the target set of task parameters could be instituted. The target parameters were: stimulus duration, 0.5 s; ITI, 5 s; limited hold period, 5 s; and timeout period, 5 s. The animals were considered to have reached criterion when these target parameters were attained on 5 consecutive sessions: >70% correct responses and <20% omissions within the 30 min session time. Once rats had acquired this training criterion, they received vHC lesions, dHC lesions, or sham control surgery.
Two weeks following surgery, the rats were tested across 10 sessions on the standard baseline schedule of the task. At the completion of the baseline schedule, a variety of task manipulations were instituted to challenge attentional and inhibitory requirements. First, the stimulus duration was reduced from 0.5 to 0.25 s in one session. Secondly, rats were exposed to a session of variable long ITIs (4.5, 6.0, 7.5, and 9.0 s). The session length of the variable long ITI interval challenge was increased to 45 min. Thirdly, the ITI was reduced from 5 to 3 s for 3 consecutive sessions. Each schedule was preceded by at least one session, in which the standard parameters were restored to ensure that the rats returned to their baseline performance. Approximately 2 months postsurgery, performance was re-evaluated on the standard baseline schedule of the task for a further 10 days.
Drug preparation and administration
One week after the task manipulations described earlier, rats were tested on the baseline schedule of the task following an i.p. injection of a dopamine reuptake inhibitor, GBR 12909 (Sigma-Aldrich, Canada), at 3 doses (2.5, 5, and 10 mg/kg) and vehicle in a counterbalanced sequence. GBR 12909 was dissolved in sterile water and injected 20 min prior to behavioral testing. Ten days after the last injection of GBR 12909, rats were retested on the baseline schedule until stable performance was obtained (2 days). Rats then received i.p. injections of a serotonin reuptake inhibitor, escitalopram oxalate (NIMH Chemical Synthesis and Drug Supply Program, NIH, Bethesda, MD, USA), at 3 doses (2.5, 5, and 10 mg/kg) and vehicle administered in a counterbalanced sequence. Escitalopram was dissolved in 0.9% sterile saline and injected i.p. 20 min before testing. Drug test days were followed by a drug-free day of no testing. Rats were then tested on the baseline schedule until performance stabilized before the next drug treatment.
Several performance measures were calculated. (1) Accuracy was measured as the proportion of responses that were correct to the total number of responses (e.g. number of correct responses/total number of responses), expressed as a percentage (chance performance = 20%). Thus, this measured errors of commission without including errors of omission. (2) Errors of omission were defined as the proportion of trials in which no response was made to the total number of trials, expressed as a percentage. This measure reflects possible failures in detection and also motivational/motor deficits, depending on the overall pattern of results. (3) Premature responses were the number of nosepoke responses made in the apertures during the ITI. Such responses were inappropriate because the rat anticipated their occurrence. This measure reflects inhibitory mechanisms of impulse control. (4) Perseverative responses were additional responses in the apertures following a correct response. This measure reflects inhibitory mechanisms of compulsive behavior. (5) Response latency was defined as the time between the onset of the light stimulus and the point at which the animal made a correct nosepoke response. (6) Magazine latency was the time between the performance of the correct response and the time the rat entered the food magazine to collect its food reward.
Visual Discrimination and Reversal Learning Task
Habituation and pretraining
Rats were initially given one 30 min session to habituate to the testing chamber and consume free pellets available in the food magazine. Rats were then shaped to collect pellets that were delivered every 10 s concomitant with illumination of the food magazine. When rats were reliably retrieving 50 pellets within a 20 min session, they were trained to respond to the stimuli presented on the touchscreen. During this procedure, a white circle (3 cm in diameter) was randomly presented in 1 of the 3 locations (left, right, or center of the screen). The circle remained on the screen until the rat responded to it by touching it with a nosepoke response. Following the response, the rat was rewarded with a pellet concomitant with illumination of the food magazine. After a 5 s ITI, the rat initiated the next trial by making a food magazine entry, which led to the presentation of the stimuli on the screen. Once the rat was able to obtain 50 reward pellets within a 30 min session, the task contingencies were introduced.
Each session began with illumination of the house light and the food magazine light. After a 5 s ITI, the rat initiated the trial by making a food magazine entry. This resulted in the simultaneous presentation of 2 stimuli on the screen. One stimulus was positively associated with a sucrose pellet (the A+ stimulus). The other stimulus was never associated with a sucrose pellet (the B− stimulus). Both stimuli were counterbalanced across subjects. The same pair of stimuli was presented on every trial, and the left and right positions of the stimuli (i.e. which stimulus was on the left and which was on the right) were determined pseudorandomly. The rat was required to approach the touchscreen and to make a nosepoke touch response to one of the stimuli. The stimuli remained on the screen until the rat responded to it. A correct response to A+ was followed by the disappearance of the stimuli and the delivery of a sucrose pellet concomitant with illumination of the food magazine. The next trial was initiated by a food magazine entry after a 5 s ITI. An incorrect response to B− resulted in the disappearance of the stimuli from the screen and a 5 s timeout period during which all of the lights were extinguished. Rats were required to learn to respond to the correct, reinforced stimulus. In addition, rats were presented with correction trials such that after an incorrect response, the same stimulus configuration (i.e. the A+ and B− stimuli remained in the same left/right positions) was presented over successive trials until the rat responded correctly. Thus, each session consisted of 100 noncorrection trials, with up to an infinite number of correction trials (see “Performance measures” for details). Rats were required to learn to respond to the correct, reinforced stimulus to an average criterion of 85% accuracy for 2 consecutive days, for noncorrection trials only.
After acquisition of the initial discrimination task, the reward contingencies were reversed so that the previously nonrewarded stimulus (B−) became the rewarded stimulus (B+) and vice versa. All other parameters were the same as the acquisition phase of the task. Rats were required to attain an average criterion of 85% accuracy over 2 consecutive days.
The following performance measures were calculated: (1) number of sessions to attain criterion performance; (2) response latency, which was the time from stimulus onset to the time the rat made a choice response; and (3) magazine latency, which was the time from when the rat made a correct choice response to the time the rat entered the food magazine to collect its food reward. These measures were calculated on the basis of noncorrection trials only.
The number of errors made during correction and noncorrection trials was also calculated. During correction trials, the rewarded (i.e. A+) and nonrewarded (i.e. B−) stimuli remained in the same left/right positions over consecutive trials. Consequently, errors made during a correction trial can be directed to the specific stimulus (i.e. A+ or B−) or to a specific side (i.e. left or right side). Therefore, errors during correction trials provided a measure of generalized perseveration. These errors were referred to as “correction errors.”
After a “correct” correction trial, the rat was presented with a noncorrection trial that was spatially opposite to the previous correction trial. For example, if the rat responded correctly to an A+ B− correction trial, the following noncorrection trial would be B− A+. Furthermore, during consecutive noncorrection trials, the left/right configuration of the stimuli was presented pseudorandomly with the provision that the same stimulus configuration would not occur for more than 2 consecutive trials. Thus, unlike correction errors, which indicate general perseveration to the side and/or stimulus, this measure of perseveration indicates that the animal makes repeated responses to the visual stimulus irrespective of the side on which it was presented. Errors made during noncorrection trials were referred to as “noncorrection errors.”
During reversal learning, the number of noncorrection errors was also analysed according to 2 learning stages: errors committed before the attainment of chance level performance (39% correct for 100 trials) and errors committed after chance level performance (40–85%), to tease apart a specific impairment of response inhibition from new stimulus–reward learning, respectively. Errors committed during the early stages of learning, when the animal's performance is increasing from below chance levels, provide an indication of the subject's ability to extinguish the previously learned stimulus–reward association. Animals with impaired response inhibition would be expected to be impaired at this stage. In contrast, errors committed in the late stages of reversal, when the animal's performance is increasing to levels above chance, provide an indication of its ability to acquire the new stimulus–reward association (Jones and Mishkin 1972; Dias et al. 1996; Chudasama and Robbins 2003).
Rats received excitotoxic lesions of either the vHC or dHC or sham control surgery. All rats were anesthestized with isoflurane gas (4–5% induction and 1–3% maintenance) in conjunction with pure oxygen and placed in a head holder fitted with atraumatic ear bars in a stereotaxic frame (David Kopf Instruments, Tujanga, CA, USA). The incisor bar was set to −3.0 mm. The scalp was retracted, and holes were drilled into the skull to expose the target region of the brain. A 1 µL SGE microsyringe (Canadian Life Science, Ontario, Canada) was used to administer bilateral injections of 0.09 M N-methyl-d-aspartic acid (NMDA, Sigma-Aldrich, Canada) dissolved in 0.9% saline. The stereotaxic coordinates, according to the stereotaxic atlas of Paxinos and Watson (2005), are presented in Table 1. Dorsal–ventral readings were in reference to the dural surface. For both lesion groups, each injection was made over 2 min, and the needle remained in place for an additional 1 min for dispersion before it was retracted. Rats that served as shams for the vHC and dHC groups received the same surgical treatments, except that they were infused with saline. Following surgery, animals were administered an injection of Rimadyl (analgesic) at a dose of 5 mg/kg s.c. and Tribrissen (antibiotic) at a dose of 0.125 mL/kg s.c. To reduce transient seizure activity during postoperative recovery, rats were treated with 0.5 mg/kg i.p. midazolam 5 min before the end of surgery and monitored in a recovery cage that was void of extraneous sensory stimulation (e.g. excessive bright lights and loud noise).
|Region||Stereotaxic coordinates||Volume of neurotoxin (µL)|
|vHC||AP −4.6; ML ±5.0; DV −6.7||0.5|
|AP −4.7; ML ±4.4; DV −6.7||0.5|
|AP −4.8; ML ±4.6; DV −7.5||0.4|
|dHC||AP −2.4; ML ±1.4; DV −3.3||0.5|
|AP −2.7; ML ±1.6; DV −3.0||0.5|
|AP −3.2; ML ±2.6; DV −2.8||0.5|
|AP −3.6; ML ±3.0; DV −2.8||0.5|
|AP −4.0; ML ±1.4; DV −3.4||0.5|
|AP −4.4; ML ±4.4; DV −3.0||0.5|
|Region||Stereotaxic coordinates||Volume of neurotoxin (µL)|
|vHC||AP −4.6; ML ±5.0; DV −6.7||0.5|
|AP −4.7; ML ±4.4; DV −6.7||0.5|
|AP −4.8; ML ±4.6; DV −7.5||0.4|
|dHC||AP −2.4; ML ±1.4; DV −3.3||0.5|
|AP −2.7; ML ±1.6; DV −3.0||0.5|
|AP −3.2; ML ±2.6; DV −2.8||0.5|
|AP −3.6; ML ±3.0; DV −2.8||0.5|
|AP −4.0; ML ±1.4; DV −3.4||0.5|
|AP −4.4; ML ±4.4; DV −3.0||0.5|
Note: AP, anterior-posterior; dHC, dorsal hippocampus; DV, dorsal-ventral; ML, medial-lateral; vHC, ventral hippocampus.
At the conclusion of behavioral testing, the rats were perfused transcardially with 0.9% saline, followed by 4% formal saline. After dehydration by immersion in 20% sucrose, the brains were sectioned on a cryostat at 40 µm thickness. Every other section was mounted on glass slides and stained with cresyl violet. The sections were used to verify the location of the lesion and to assess the extent of lesion-induced neuronal loss.
Data for each variable were subjected to a repeated-measures analysis of variance (ANOVA) or a one-way ANOVA using PASW Statistical Software, version 18 (SPSS Inc., Chicago, IL, USA). Skewed data, which violated the distribution requirement of the ANOVA, were transformed appropriately (arcsine, square root, or logarithmic), as recommended by Winer (1971). Homogeneity of variance and covariance was assessed using Mauchly's test of sphericity. When data significantly violated this requirement for a repeated-measures design, the Huynh–Feldt epsilon was used to calculate a more conservative P-value for each F-ratio. Where F-ratios were significant (P < 0.05), the means were compared using least significant difference (LSD) post hoc comparisons. For the results obtained in the 5-choice task, the between-subjects factor consisted of 3 levels: sham controls (shams), vHC lesions, and dHC lesions. The within-subjects factors included session (10 days for baseline testing and 3 days for short 3.0 s ITI), stimulus duration at 2 levels (0.5 and 0.25 s), ITI at 4 levels (4.5, 6.0, 7.5, and 9.0 s), and drug at 4 doses (vehicle and 0.25, 5.0, and 10 mg/kg). For results obtained in the visual discrimination and reversal task, the between-subjects factor was at 2 levels (sham and vHC).
Cytoarchitectonic borders and nomenclature were taken from Paxinos and Watson (2005). Figures 2A and 2B provide a diagrammatic reconstruction of the lesion showing the extent of the ventral and dorsal hippocampal lesions, respectively, for rats tested in the 5-choice task. In this cohort, 4 rats from the dHC group had unilateral or incomplete damage to the dHC region. These animals were removed from subsequent behavioral analysis. The remaining 7 animals in the dHC group showed extensive cell loss in the dorsal portion of the hippocampus, bilaterally. The lesion extended rostrocaudally from anterior–posterior (AP) −1.92 mm to −4.68 mm from bregma and included the pyramidal cells of the cornu ammonis (CA) fields (CA1, CA2, and CA3), the dentate gyrus (molecular and granular layers), and the dorsal subiculum. In 1 animal, the lesion encroached the stria medullaris, but the habenula and adjacent midline nuclei (e.g. mediodorsal thalamus and paraventricular thalamus) remained intact. The fornix was also spared.
In the vHC group, 3 rats were discarded from the analysis: 1 animal showed extensive damage that included the subthalamic nucleus and the other 2 rats had minimal/incomplete damage to the vHC. The remaining 7 rats in the vHC group showed a bilateral vHC lesion as intended. The lesion started at AP −4.36 mm and extended caudally to −5.40 mm from bregma. The lesion included the pyramidal cells in the CA1–CA3 subfields, the dentate gyrus, and the ventral subiculum. In 5 rats, the vHC lesion encroached the ventral tip of the dHC, unilaterally or bilaterally. In all cases, the amygdala and the rhinal cortex lateral to the vHC were spared, as were the structures medial to vHC including the geniculate nuclei. Rats that served as sham controls for the vHC group (sham–vHC) and the dHC group (sham–dHC) did not show any neuronal damage. The final numbers in each group for behavioral analysis were as follows: vHC, n = 7; dHC, n = 7; sham–dHC, n = 4; and sham–vHC, n = 4.
Figure 2C provides a schematic representation of the extent of the vHC lesion for a separate cohort of rats that were tested in the visual discrimination and reversal learning task. Examination of the cresyl violet-stained sections revealed that 2 rats in this cohort had unilateral vHC damage, and 1 rat had minimal damage that was restricted to the inner pyramidal layer of the vHC. The data from these 3 rats were excluded from the statistical analysis. The remaining 8 rats showed extensive neuronal loss within the vHC that was similar to the vHC lesions obtained in Experiment 1. The lesion included the pyramidal cells of the CA fields (CA1, CA2, and CA3), the dentate gyrus (molecular and granular layers), and the ventral subiculum. The lesion started rostrally at AP −4.68 mm from bregma and continued until AP −5.40 mm. In 1 rat, the lesion encroached very slightly into the amygdala, unilaterally. In another rat, the lesion encroached into the most ventral tip of the dHC, unilaterally. In all cases, structures lateral to the vHC such as the parahippocampal region including entorhinal and perirhinal cortices remained intact. Sham rats did not show any neuronal damage. The final numbers in each group were vHC, n = 8 and shams, n = 8.
Effect of Ventral and Dorsal Hippocampal Lesions on Performance of the 5-Choice Task
The vHC and dHC lesions and sham control surgeries were performed after the rats had been trained to criterion performance in the 5-choice task. Two weeks following surgery, the 2 lesion groups and the sham control animals were retested in the same task. Subsequently, in separate sessions, we examined the effect of manipulating the duration of the stimulus and the timing of stimulus presentation on the behavior of the various groups. Two months after the lesion, we retested the animals again on the baseline version of the task to examine the long-term effects of the lesion. Each of these aspects of the study is described in turn in the following sections.
Postoperative behavior 2 weeks post-lesion
Preliminary analysis of the sham data revealed no significant difference between the 2 sham groups (sham–vHC and sham–dHC) across the 10 postoperative baseline sessions for accuracy (F1,6 = 0.06; P > 0.05), omissions (F1,6 = 0.17; P > 0.05), premature responses (F1,6 = 1.48; P > 0.05), or perseverative responses (F1,6 = 1.44; P > 0.05). Latency to respond or collect food was also within the normal range (all P > 0.05). Thus, all of these animals were treated as a single sham group for subsequent analysis.
Over the 10 postoperative sessions, all animals improved discriminative accuracy (F9,171 = 10.95; P < 0.01). However, as shown in Figure 3A, lesions of the vHC significantly reduced accuracy (F2,19 = 4.22; P < 0.05), relative to the dHC lesions (LSD, P < 0.01). Although rats in the vHC group were less accurate than sham controls overall, this difference just failed to reach significance (LSD, P = 0.06). Consistent with improved accuracy over time, the number of omissions declined with increasing number of sessions (F9,171 = 6.05; P < 0.01), but the groups did not differ significantly from each other (F2,19 = 6.06; P > 0.05). The number of premature responses also declined for all groups during the course of postoperative testing (F4,81 = 3.48; P < 0.01). However, a group × session interaction was obtained for premature responses (F9,81 = 3.64; P < 0.01). Post hoc comparisons confirmed that rats with vHC lesions made more premature responses relative to both dHC and sham groups (LSD, all P < 0.05, Fig. 3B). In contrast, Figure 3C shows that the number of perseverative responses did not differ between groups (F2,19 = 0.63; P > 0.05), session (F4,71 = 2.38; P > 0.05), or group × session interaction (F7,71 = 3.69; P > 0.05). There was no differential effect of lesion group on response latency (F2,19 = 0.98; P > 0.05) or magazine latency (F2,19 = 0.23; P > 0.05).
Effect of reducing the stimulus duration
Figure 4A shows that reducing the stimulus duration to 0.25 s resulted in a significant reduction in accuracy for all animals irrespective of the group (F1,19 = 24.72; P < 0.001). However, a significant lesion effect was obtained for this measure (F2,19 = 4.46; P < 0.05) because the vHC group was less accurate relative to the dHC group (LSD, P < 0.001) but not the sham group (LSD, P = 0.07). There was no group × stimulus duration interaction on this measure (F2,19 = 0.32; P > 0.05). Nor were there any statistical differences in terms of the number of omissions (all P > 0.05).
All rats showed almost equivalent rates of premature responses at both 0.5 and 0.25 s stimulus durations (F1,19 = 0.36; P > 0.05), but rats with vHC lesions maintained a significant increase in the number of premature responses relative to the sham and dHC groups (F2,19 = 4.04; P < 0.05; LSD, P < 0.05; Fig. 4B). As shown in Figure 4C, the number of perseverative responses was found to be consistently low among all groups irrespective of the stimulus duration (F1,19 = 0.06; P > 0.05) or lesion (F2,19 = 0.66; P > 0.05). No other significant effects were observed.
Effect of reducing ITI to 3.0 s for 3 consecutive sessions
Making the ITI short and highly predictable did not help improve the accuracy deficit in the vHC group of rats (F2,19 = 3.66; P < 0.05; Fig. 5A) across all 3 sessions (F2,38 = 1.13; P > 0.05). The number of omissions declined with increasing number of sessions (F2,38 = 13.53; P < 0.01), but the groups did not differ on this measure (F2,19 = 0.14; P > 0.05). Once again, however, as shown in Figure 5B, rats with vHC lesions continued to make more premature responses than animals of the other groups (F2,19 = 3.71; P < 0.05; LSD all P < 0.05), an effect which did not change over 3 consecutive sessions (F1,24 = 2.91; P > 0.05). The number of perseverative responses stayed within normal levels and did not change as a function of session (F2,38 = 0.52; P > 0.05) or lesion (F2,19 = 0.25; P > 0.05; Fig. 5C). Response and magazine latencies were also in the normal range (all P > 0.05).
Effect of variable long ITIs
In this manipulation, rats were exposed to long ITIs that were presented randomly and unpredictably across 1 test session. This resulted in an overall decrease in the accuracy for all groups irrespective of lesion (F2,45 = 5.43; P < 0.01). However, a significant group × ITI interaction was obtained (F5,45 = 4.35; P < 0.01), which was due to rats with vHC lesions that were less accurate at the 4.5, 6.0 and 9.0 s ITIs relative to both dHC and sham groups (LSD, all P < 0.05; Fig. 5D).
Figure 5E shows that premature responding significantly increased for all rats as ITI increased (F2,35 = 124.42; P < 0.001), but rats with vHC lesions made significantly more premature responses (F2,19 = 4.92; P < 0.05) when compared with rats in the sham and dHC groups (LSD, P < 0.01). In marked contrast, and consistent with previously reported effects, there was no differential effect of lesion on perseverative responses (F2,19 = 0.07; P > 0.05; Fig. 5F). Response and magazine latencies remained unaffected (all P > 0.05).
Postoperative behavior 2 months post-lesion
Approximately 2 months after surgery, the rats were retested for 10 days on the standard baseline schedule of the task to examine the stability of the observed deficits. Figure 3D shows that rats with vHC lesions maintained their impairment in discriminative accuracy (F2,19 = 4.42; P < 0.05) relative to rats with dHC lesions (LSD, P < 0.001), but they were not different from sham controls (LSD, P = 0.08). In addition, the vHC group continued to make more premature responses (F2,19 = 4.67; P < 0.05; LSD, all P < 0.05; Fig. 3E). For the first time in this experiment, however, a perseverative deficit emerged in the rats with vHC lesions (F2,19 = 3.57; P = 0.048; Fig. 3F). Nevertheless, the analysis did not reveal a main effect of session (F6,103 = 1.13; P > 0.05) or lesion × session interaction (F11,103 = 1.35; P > 0.05) on this measure. Latencies to respond and collect food reward did not differ between groups (P > 0.05).
Effects of Systemic Injections of Selective Dopamine and Serotonin Reuptake Inhibitors
Due to the persistent deficit in response accuracy and increased impulsivity observed in rats with vHC lesions, we administered reuptake inhibitors as stimulating agents, to examine the potential therapeutic effect of enhancing monaminergic transmission. Because commonly used selective serotonin reuptake inhibitors increase extracellular concentrations of dopamine as well as serotonin (Koch et al. 2002; Bymaster et al. 2002), we chose to administer reuptake inhibitors that were highly selective for the dopamine transporter (GBR 12909) and the serotonin transporter (escitalopram) (Andersen 1989; Owens et al. 2001).
Effects of systemic injections of GBR 12909
Figure 6A shows that rats with vHC lesions were less accurate in detecting the visual target relative to the other groups, across all doses, but this deficit was found not to be significant according to lesion group (F2,19 = 2.98; P = 0.07) or dose (F3,57 = 1.63; P > 0.05). Nonetheless, this finding suggests that the accuracy deficit in the vHC group was showing some signs of behavioral recovery. However, the number of premature responses remained elevated in the vHC group (F2,19 = 4.18; P < 0.05) at all doses (F2,44 = 4.68; P < 0.01). There was also a significant lesion × drug interaction on this measure (F5,44 = 2.47; P = 0.050). As shown in Figure 6B, the number of premature responses in the vHC group increased significantly at the medium 5 mg/kg dose (F2,19 = 3.35; P < 0.05) and at the high 10 mg/kg dose (F2,19 = 5.54; P < 0.01). In addition, consistent with their baseline performance, high premature responding was also maintained when the rats were injected with vehicle (F2,19 = 4.95; P < 0.01). At the low 2.5 mg/kg dose, the number of premature responses in the vHC group was not significantly elevated relative to the dHC and sham groups (F2,19 = 0.73; P > 0.05). We note, however, that the number of premature responses exhibited by the vHC group when on the low dose of GBR 12909, and when on vehicle, was almost equivalent (mean responses ± SEM vehicle, 53 ± 11 and 2.5 mg/kg, 45 ± 15), suggesting that the low dose of GBR 12909 failed to sufficiently recover the impulsive deficit.
Figure 6C shows that rats with vHC lesions became highly perseverative when injected with GBR 12909 (F2,19 = 3.43; P = 0.05) relative to both sham controls (LSD, P < 0.05) and dHC-lesioned rats (LSD, P = 0.05). Furthermore, consistent with their perseverative tendency during postoperative baseline performance (2 months after surgery), the vHC group remained perseverative when treated with vehicle. However, there was no effect of dose (F3,57 = 0.78; P > 0.05) or lesion × dose interaction (F6,57 = 0.73; P > 0.05) on this measure. Response and magazine latencies were not differentially impaired according to lesion group (all P > 0.05).
Effects of systemic injections of escitalopram
During this drug manipulation, response accuracy did not differ significantly between groups (F2,19 = 1.69; P > 0.05) at any dose (F3,57 = 0.74; P > 0.05; Fig. 6D). Nor did the groups differ in the number of omissions (F2,19 = 1.2; P > 0.05). For premature responses, both a dose effect (F2,43 = 3.67; P < 0.05) and a lesion group effect (F2,19 = 5.58; P < 0.01) were obtained. As shown in Figure 6E, escitalopram served to generally reduce the number of premature responses in the vHC group across all doses. In fact, the number of premature responses was lower for all 3 doses relative to vehicle (mean responses ± SEM vehicle, 46 ± 10; 2.5 mg/kg, 34 ± 8; 5 mg/kg, 19 ± 2; and 10 mg/kg, 34 ± 6). However, at the medium (5 mg/kg) dose of escitalopram, the impulsive deficit exhibited by the vHC group was abolished (F2,19 = 0.31; P > 0.05). Nonetheless, the vHC group continued to respond impulsively at the low 2.5 mg/kg dose (F2,19 = 4.51; P < 0.05), the high 10 mg/kg dose (F2,19 = 5.61; P < 0.05), and when injected with vehicle (F2,19 = 3.88; P < 0.05).
Despite the increased perseverative responses observed in the vHC group in the previous manipulation, there was not a main effect of lesion group (F2,19 = 1.62; P > 0.05) or dose (F3,57 = 0.60; P > 0.05; Fig. 6C) on this measure. The speed of response increased with dose (F3,57 = 3.10; P < 0.05) as did magazine latency (F2,48 = 21.46; P < 0.001), but again, the groups did not differ significantly on either measure (all P > 0.05).
Effect of Ventral Hippocampal Lesions on Visual Discrimination and Reversal Learning Task
Due to the persistent deficit in withholding impulsive responses and, to some extent, perseverative responses in the vHC group during the performance of the 5-choice task, separate groups of rats were prepared with vHC lesions or sham control surgery and then tested on a visual discrimination and reversal learning task. This additional study aimed to characterize the specificity of the inhibitory control deficit that was observed only in the vHC group. Once rats had successfully acquired the stimulus–reward visual discrimination, we reversed the stimulus–reward contingencies in order to examine the rat's ability to inhibit a perseverative response to a previously rewarded stimulus.
Acquisition of the visual discrimination
All rats successfully learned the visual discrimination; there was no significant difference between the vHC and the sham group with respect to the number of sessions required to attain criterion performance (F1,14 = 0.02; P > 0.05; Fig. 7A). As shown in Figure 7B, the number of noncorrection errors committed by the vHC-lesioned group remained within normal levels (F1,14 = 0.0001; P > 0.05). The vHC-lesioned rats committed more correction errors to reach criterion (Fig. 7C), but they did not differ significantly from the sham group (F1,14 = 2.928; P > 0.05). No significant main effect of group was observed for response latency for noncorrection trials (F1,14 = 0.888; P > 0.05; mean ± SEM in seconds, sham 2.83 ± 0.43 and vHC 4.21 ± 1.40) or for correction trials (F1,14 = 0.11; P > 0.05; mean ± SEM in seconds, sham 2.60 ± 0.21 and vHC 2.46 ± 0.39). Although rats with vHC lesions took longer to collect their food reward during noncorrection trials (F1,14 = 12.242; P < 0.05; mean ± SEM in seconds, sham 3.29 ± 0.27 and vHC 4.85 ± 0.39), their magazine latencies were not different from shams during correction trials (F1,14 = 4.154; P > 0.05; mean ± SEM in seconds, sham 3.85 ± 0.53 and vHC 5.07 ± 0.35).
Compared with the acquisition phase, rats in both groups needed more sessions to reach criterion when the reward contingencies were reversed (Fig. 7A). However, the groups did not differ significantly on this measure (F1,14 = 0.125; P > 0.05). Furthermore, as Figures 7B and 7C show, there was no significant effect of lesion group on the number of noncorrection errors (F1,14 = 0.037; P > 0.05) or correction errors committed to reach criterion performance (F1,14 = 0.051; P > 0.05). Finally, there were no group differences in response or magazine latency for noncorrection or correction trials (all P > 0.05).
The number of errors committed during the 2 stages of learning (before chance perseveration and learning errors) is shown in Figure 7D. A significant main effect of learning stage was obtained because more errors were committed by both groups of animals during the second stage of learning compared with the first stage (F1,14 = 44.947; P < 0.001). However, there was no effect of lesion (F1,14 = 0.037; P > 0.05) or lesion × stage interaction (F1,14 = 0.201; P > 0.05) on this measure.
In this study, we demonstrate that mechanisms of inhibitory response control are localized to the ventral subdivision of the rodent hippocampus. There were several key findings: (1) rats with vHC lesions were unable to inhibit the impulsive urge to make a response in the 5-choice task, a deficit that persisted over a period of 3 months of testing; (2) vHC lesions caused a reduction in response accuracy in the 5-choice task, an effect that recovered over time; (3) systemic treatment with escitalopram, a selective serotonin reuptake inhibitor, ameliorated the impulsive deficit at the 5 mg/kg dose; (4) systemic treatment with GBR 12909, a selective dopamine reuptake inhibitor, increased perseverative responding in the vHC group only; and (5) in the 5-choice task, there was a transient increase in perseverative responding in the vHC group 2 months after surgery, but the vHC group did not perseverate on the acquired visual discrimination during reversal learning. Furthermore, the deficits observed in the vHC group were not associated with speed or increased motivation; response and magazine latencies were generally unaffected. Together with the lack of deficit in rats with dHC lesions, these data suggest the selective involvement of the vHC in inhibitory control functions, especially the inhibitory control of impulsive actions.
When the target stimulus was presented following a 5 s ITI on every trial (i.e. baseline conditions), rats with dHC lesions and sham control surgery could reliably predict the onset of the stimulus and withhold responding in the apertures during the ITI. Rats with vHC lesions, however, responded prematurely in the holes during the ITI. Reducing the ITI to 3 s to increase the predictability of the stimulus (i.e. high event rate) failed to ameliorate the impulsive deficit, suggesting that rats with vHC lesions were unable to appropriately schedule their behavior during the ITI, in anticipation of the stimulus. In previous work, rats with a disrupted hippocampal system responded earlier than the scheduled time of reinforcement (Meck et al. 1984), suggesting that hippocampal lesions disrupt timing behavior. Hippocampal lesions also disrupt responding in differential reinforcement of low rates (DRL) of response schedules. In DRL, rats are trained to withhold responding for food until after a set time has elapsed (often >15 s). Rats with dorsal, ventral, or complete hippocampal lesions are highly inefficient in this task because they are less able to wait over a defined temporal interval (Bannerman et al. 1999). Unlike DRL, however, in which rats must estimate a relatively long time interval before making a response, in the 5-choice task, rats respond to a visual cue that signals the end of the time interval. As such, DRL is a more stringent measure of time estimation than the 5-choice task. This difference might explain why dHC lesions have been shown to produce impulsive-like responses in DRL (Bannerman et al. 1999), but not in the 5-choice task (present study). One possibility is that rats with dHC lesions are poor at estimating the passage of time and are therefore not able to space their responding appropriately, whereas rats with vHC lesions lack response inhibition and exhibit persistent responding. According to this view, both dorsal and ventral hippocampal lesions would produce the same behavioral effect in DRL (as in Bannerman et al. 1999), but for different reasons. In fact, introducing a signal that bisects the DRL interval improves performance in hippocampectomized rats (Braggio and Ellen 1976; Rawlins et al. 1983). Our data suggest that an intact vHC is necessary for the appropriate inhibition of impulsive actions when anticipating reward-related cues.
A disruption in the ability to effectively plan a response during the ITI might explain why the impulsive deficit exhibited by the vHC group was accompanied by a reduction in response accuracy. However, increasing the ITI to facilitate response preparation failed to improve response accuracy in vHC-lesioned rats. One possibility is that the accuracy deficit was indeed derived from the failure to withhold premature responses because the rats were engaged in exploring the holes rather than distributing their attention across the visual array. This would suggest that the tendency to impulsively over-respond during the ITI prevented these rats from correctly orienting to the target rather than impairing stimulus detection. First, in most cases, rats with vHC lesions performed as accurately as sham controls and were different from the dHC group who, if anything, out-performed the controls albeit nonsignificantly. Secondly, rats with vHC lesions were not “attentionally unaware” of the stimuli because there was no evidence of increased omissions in this group. In addition, there was some evidence of functional recovery of attention toward the end of behavioral testing. It would be informative to examine the effects of complete (dorsal and ventral) hippocampal lesions on the 5-choice task. Given that the dHC lesion seemed to facilitate accuracy in the 5-choice task whereas the vHC lesion impaired it, we might expect that the combination of the 2 lesion effects would be to cancel each other out. This speculation accords with the findings of Burk and Mair (2001) who showed that complete hippocampal lesions did not affect performance accuracy on a self-paced visual reaction time task. Unfortunately however, the self-paced visual reaction time task of Burk and Mair (2001) does not allow for the type of impulsive responding observed in the 5-choice task. Nevertheless, while there is no consistent evidence for hippocampal involvement in attentional function (Kirkby and Higgins 1998; Hahn et al. 2003; but see Le Pen et al. 2003), the current data suggest that the putative contribution of the vHC to attentional control warrants further investigation.
Similar to the effects of vHC lesions, rats with vPFC lesions become disinhibited in their behavior as well. Of the entire vPFC, only lesions of the infralimbic cortex (ILC) induce premature responding in the 5-choice task without affecting perseverative responding (Chudasama et al. 2003; see also Murphy et al. 2005). Because the vHC distributes its axons preferentially to the ILC and other vPFC regions including the prelimbic and orbitofrontal cortex (Jay and Witter 1991; Verwer et al. 1997; Ishikawa and Nakamura 2006), it is ostensibly capable of influencing the behavioral expression of its cortical terminal projections. Our data suggest that the vHC might directly influence the activity of the ILC with respect to the control of impulsive actions. Our findings do not speak to the specific mechanisms by which the vHC might modulate ILC function. One possibility is that the vHC regulates NMDA cortical transmission and that its removal is similar to the effect of local NMDA receptor antagonism of the ILC, which increases premature responding in the 5-choice task (Murphy et al. 2005; see also Higgins et al. 2003; Mirjana et al. 2004; Pozzi et al. 2011). Furthermore, the excitatory response of vPFC neurons that respond to electrical stimulation of the vHC is predominantly mediated by glutamatergic stimulation of AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) receptors (Jay et al. 1992). The effect of local vHC-glutamate antagonism in the 5-choice task is currently unknown, but the combined results from Murphy et al. (2005) and the present study strongly suggest a role for the vHC in the normal modulation of impulsive responding expressed by the ILC.
In contrast to their long-lasting impulsive deficit in the 5-choice task, there was no evidence of a perseverative deficit in the vHC group until they were tested again on the postoperative baseline schedule 2 months postsurgery. This effect could not be caused by factors related to the nature of the task (e.g. stimulus brevity, task difficulty, or poor temporal cues) and may have been a secondary manifestation of deficient inhibitory control associated with the long-term effect of the vHC lesion. This would explain why it was only the vHC group of rats that maintained an increase in perseverative responding under vehicle when counterbalanced with GBR 12909. Nevertheless, this perseverative effect was exaggerated at all doses when vHC-lesioned rats were treated systemically with GBR 12909. A similar increase in perseverative responses has been observed following lesions to the nucleus accumbens (NAcc) (Christakou et al. 2004). Indeed, in addition to its glutamatergic innervation of the vPFC, the vHC (CA1/subiculum) sends a glutamatergic projection to the NAcc that converges with dopaminergic projections arising from the ventral tegmental area (Totterdell and Smith 1989; Sesack and Pickel 1990). Stimulation of the ventral tegmental area, or the vHC, affects NAcc dopamine release, and this can alter the neural activity of NAcc neurons driven by the vHC input (Yang and Mogenson 1984, 1986; DeFrance et al. 1985; Pennartz et al. 1994; Floresco et al. 2001; Goto and Grace 2008). Thus, the increased perseveration exhibited by the vHC group might have been associated with an increase in the dopamine release in the NAcc caused by an absence of glutamatergic input from the vHC and was potentiated by GBR 12909. In addition, the vHC-lesioned rats maintained their impulsive tendency when injected with GBR 12909. Similar to these data are the findings that intra-accumbens stimulation of D1 and D2 receptors increases premature and perseverative responding, respectively (Pezze et al. 2007), whereas dopamine depletion of the NAcc reduces premature responding (Cole and Robbins 1989). Furthermore, hippocampal lesions enhance amphetamine-induced locomotor activity and extracellular dopamine release in the NAcc (Campbell et al. 1971; Wilkinson et al. 1993), suggesting that the hippocampus modulates the activity of the NAcc (Whishaw and Mittleman 1991). Specifically, lesions of the vHC, but not the dHC, enhance amphetamine-induced locomotion and increase dopaminergic activity in the NAcc (Lipska et al. 1991, 1992; but see Bannerman et al. 1999). One possibility is that GBR 12909 had an overall “activating” effect on the NAcc that was boosted by the vHC lesion, causing an overall increase in response vigor.
In contrast to the perseverative effect of GBR 12909, increasing serotonergic transmission with escitalopram attenuated the impulsive deficit induced by vHC lesions at the 5 mg/kg dose. Increased impulsivity was maintained in the vHC group at the low 2.5 mg/kg dose and the high 10 mg/kg dose, suggesting that there is an optimal “middle” dose at which serotonergic stimulation provides its therapeutic effect. Presumably, the vHC lesion altered monoamine levels in several intact subcortical structures, including the vPFC to which the vHC projects. Furthermore, serotonin reuptake inhibitors enhance extracellular 5-HT in the prefrontal cortex and hippocampus (Bymaster et al. 2002; Felton et al. 2003). In the present study, we cannot be certain of the locus of action or the precise receptor mechanism responsible for the reduction in impulsivity in the vHC group. Excitatory and inhibitory influences on impulsive actions are mediated via 5-HT2A receptors (Carli and Samanin 1992; Koskinen et al. 2000; Winstanley et al. 2004; Fletcher et al. 2007), particularly in the prefrontal cortex (Passetti et al. 2003; Winstanley et al. 2003; Liu et al. 2004). Therefore, it is possible that escitalopram exerted its effects on 5-HT receptors in the intact prefrontal cortex to compensate for the vHC lesion.
Perhaps, most surprising was the finding that the vHC lesions did not increase perseverative errors in the reversal learning task. Rats and monkeys with hippocampal lesions have been shown to perseverate on a previously rewarded response when reward contingencies are reversed (Thompson and Langer 1963; Kimble and Kimble 1965; Douglas and Pribram 1966; Mahut 1971; Jones and Mishkin 1972). The absence of an effect on reversal learning in the current study indicates that the vHC is not invariably involved in all forms of response control, but only under certain circumstances. Although we did not test dHC-lesioned rats on the reversal learning task, excitotoxic lesions of the dHC in rats do not yield spatial reversal learning deficits (Muñoz and Grossman 1981). In previous studies, lesions of the hippocampus involved aspiration or radiofrequency techniques. It is, therefore, possible that the previously reported reversal learning effects were associated with damage to fibers coursing through the hippocampus to its output structure, the vPFC. Indeed, processes engaged in reversal learning are susceptible to orbital prefrontal lesions (Dias et al. 1996; Chudasama and Robbins 2003; McAlonan and Brown 2003; Schoenbaum et al. 2003; Bissonette et al. 2008). Our data suggest, however, that reversal learning does not require the functional interaction of the vHC with the orbital prefrontal cortex.
Rats with ILC lesions are also impaired in reversal learning tasks, but in the absence of perseveration (Chudasama and Robbins 2003; see also Boulougouris et al. 2007). Instead, rats with ILC lesions show a specific deficit in learning the new stimulus–reward association when the stimulus–reward contingencies are reversed (Chudasama and Robbins 2003). This specific finding was not observed in rats with vHC lesions (Fig. 7D). Evidently, vHC lesions do not always produce the same effects as ILC lesions. Therefore, our data support a function for the vHC–ILC pathway in the control of premature responses in the 5-choice task, but not for the flexible adaptation of behavior to changes in stimulus–reward contingencies in reversal learning.
The present findings support the conclusion that inhibitory control functions are associated with the ventral, and not the dorsal, subregion of the hippocampus, especially in the inhibitory control of impulsive actions. This evidence is consistent with the effects of lesions to specific regions of the vPFC, a major target of vHC projections. Future studies will probe the nature of the interaction between the vHC and the vPFC in inhibitory control functions.
This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (grant nos 341600 and 345694) and the Canadian Foundation for Innovation (grant no. 14033).
We thank Sarah Coupland for her help with behavioral testing, Yuchen Liu and Laura Machan for their help with histology, and Sean Hassan with data analysis. We also thank David A. Leopold and Norman M. White for their helpful comments on the manuscript. A.R.A. was supported by an Alexander Graham Bell Canada Graduate Scholarship from the Natural Sciences and Engineering Research Council of Canada. Conflict of Interest: None declared.