- Email: [email protected]
Role of hippocampus in polymodal-cue guided tasks in rats
Author’s Accepted Manuscript Role of hippocampus in polymodal-cue guided tasks in rats Maria Concetta Miniaci, Pellegrino Lippiello, Marcellino Monda, Pietro Scotto www.elsevier.com/locate/brainres
PII: DOI: Reference:
S0006-8993(16)30454-1 http://dx.doi.org/10.1016/j.brainres.2016.06.030 BRES44981
To appear in: Brain Research Received date: 15 October 2015 Revised date: 8 May 2016 Accepted date: 20 June 2016 Cite this article as: Maria Concetta Miniaci, Pellegrino Lippiello, Marcellino Monda and Pietro Scotto, Role of hippocampus in polymodal-cue guided tasks in rats, Brain Research, http://dx.doi.org/10.1016/j.brainres.2016.06.030 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Role of hippocampus in polymodal-cue guided tasks in rats Maria Concetta Miniaci1*, Pellegrino Lippiello1, Marcellino Monda2, Pietro Scotto1 1
Department of Pharmacy, University of Naples “Federico II”, 80131 Naples, Italy.
2
Department of Experimental Medicine, Second University of Naples, 80138 Naples, Italy
*
Corresponding author: Department of Pharmacy, University of Naples “Federico II”,
via D. Montesano, 49; 80131 - Naples, Italy. Tel. +39 081-678465; Fax +39 081-678403. [email protected] Abstract To examine how signals from different sensory modalities are integrated to generate an appropriate goal-oriented behavior, we trained rats in an eight-arm radial maze to visit a cue arm provided with intramaze cues from different sensory modalities, i.e. visual, tactile and auditory, in order to obtain a reward. When the same rats were then examined on test trials in which the cue arm contained one of the stimuli that the animals were trained with (i.e. light, sound or rough sheet), they showed a significant impairment with respect to the performance on the polymodal-cue task. The contribution of the dorsal hippocampus to the acquisition and retention of polymodal-cue guided task was also examined. We found that rats with dorsal hippocampal lesions before training showed a significant deficit in the acquisition of polymodal-cue oriented task that improved with overtraining. The selective lesion of the dorsal hippocampus after training disrupted memory retention, but the animals’ performance improved following retraining of the polymodal task. All hippocampal lesioned rats displayed an impaired performance on the unimodal test. These findings suggest that the dorsal hippocampus contributes to the processing of multimodal sensory information for the associative memory formation and consolidation.
ABBREVIATIONS
DH: dorsal hippocampus; pre-DH lesion: pre-training dorsal hippocampus lesion; postDH lesion: post-training dorsal hippocampus lesion Keywords: Dorsal hippocampus, learning, memory, polymodal cues, electrolytic lesions.
1. Introduction The distinction between elemental and configural models is a central issue in associative learning theory about how the hippocampus contributes to learning and memory (Harris, 2006; Iordanova et al., 2011; Pearce and Bouton, 2001; Rudy and Sutherland, 1989). According to the elemental theories, stimuli are processed separately, independent of whether they have been presented alone or compounded with other stimuli, meaning that that the associative strength of a compound is equal to the algebraic sum of the associative strength of its components (Rescorla and Wagner, 1972). The configural theories assume that conditioning with a compound results in a unitary representation of the compound as a configuration entering into an association with the reinforcement (Pearce, 2002). Interestingly, patients and animals with damage to the hippocampus are unable to solve a novel associative learning task whose solution required the acquisition of specific combinations of elements i.e. configuration, rather than single elements (Rudy and Sutherland, 1989; Schmajuk and DiCarlo, 1992). Negative patterning is a well-known discrimination task that has been proposed for comparing elemental and configural theories in animal models (Rescorla, 1972; Whitlow and Wagner, 1972). To solve the problem, animals have to learn to respond to two single reinforced stimuli (A+, B+) but not to their combination, which is non-reinforced (AB-). By using this paradigm, Rudy and Sutherland (1989) demonstrated that rats with hippocampal lesions behaved similarly
to their controls in the presence of a single stimulus, but were unable to generate the appropriate response when the two cues were associated. However, these results are in contrast with those reported by other authors who found little or even no effect of the hippocampal lesion on the acquisition of a configural associative task (Davidson et al., 1993; Gallagher and Holland, 1992; Whishaw and Tomie, 1991). Nevertheless, the hippocampus is considered a critical brain area where spatial and nonspatial information can be integrated into a unified event representation (Eichenbaum et al., 1999; Moser et al., 2008). Indeed, during new memory formation the hippocampus forms connections between sensory stimuli that can be stored and recalled later (Cohen and Eichenbaum, 1993; Gluck et al., 2005; O’Reilly and Rudy, 2001). To better understand how multiple stimuli combine to produce an appropriate behavior during associative learning, we developed a multisensory apparatus that allows us to measure the navigational behavior along with the precisely controlled presentation of visual, auditory and tactile stimuli. For this purpose, rats were trained in an eight-arm radial maze to visit a cue arm provided with intramaze cues from different sensory modalities, i.e. light, sound and rough sheet, in order to obtain a reward. At the end of the training, the same rats were examined on test trials in which the cue arm contained one of the stimuli, that the animals were trained with, to determine whether the animal’s performance was primarily controlled by the sum of the elements presented, as a compound stimulus, or by one specific sensory element of the compound. Since most of the highly processed information from the sensory cortices enters the hippocampus mainly through its dorsal section (Fanselow and Dong, 2010; Strange et al., 2014), we next determined the role played by dorsal hippocampus during acquisition and retention of the polymodal associative task. For this purpose, we examined the behavioral
effects of discrete electrolytic lesions of the dorsal hippocampus, before and after training of the polymodal cue-guided task.
2. Results
2.1 Histology A schematic representation of the extent of the electrolytic lesion of the dorsal hippocampus is shown in Fig. 1. The electrolytic lesions appeared to be complete within the targeted region with no damage to the surrounding tissue. The sham groups did not show any damage to any brain region.
2.2 Effect of the dorsal hippocampal lesion on acquisition Polymodal associative task. Dorsal hippocampal (pre-DH; n= 8) and sham (n= 8) lesioned rats were trained on the polymodal-cue guided task in the eight-arm radial maze (Fig. 2). During the polymodal training, each rat was required to visit the cue arm containing the visual (light), auditory (sound) and tactile (rough sheet) stimuli to obtain a reward. The position of the cue arm changed for each trial. Initially, both pre-DH and sham animals reached the cue arm taking random routes. However, the sham group learned the task more rapidly than the pre-DH group. As shown in Fig. 3, sham and pre-DH animals achieved the learning criterion after 18 and 25 daily sessions of training, respectively. A two-way ANOVA (lesion group x session) of the mean percentage of correct choices for the first 18 sessions indicated a significant difference between the sham and pre-DH groups [F(1, 14) = 61.8, p<0.01] and revealed also an effect of sessions [F(17, 238) = 17.69, p<0.01] and group x
session interaction [F(17, 238) = 2.4, p<0.01]. The two-way ANOVA (lesion group x session) of the mean errors for the first 18 sessions showed a significant difference between groups [F(1, 14) = 33.44, p<0.01] and sessions, [F(17, 238) = 4.20, p<0.01] but no significant lesion group x session interaction [F(17, 238) = 1.15, p = n.s.]. Concerning the latency, the effects of lesion group [F(1, 14) = 50.2, p<0.01] and session [F(I7, 238) = 10.5, p<0.01] were significant for the first 18 sessions as was the session x lesion group interaction [F(17, 238) = 2, p<0.05]. Unimodal associative test. One day after the animals completed the polymodal-cue training, each animal was tested on the unimodal-cue task in which the cue arm was signaled by one of the three stimuli at a time, i.e. light, sound or rough sheet. As shown in Fig. 4, both sham and pre-DH rats were impaired irrespective of which unimodal cue was used. The mean percentage of correct choices scored in the unimodal-cue task with light, sound and rough sheet was 37.5 (± 5.1), 37.5 (± 6.3) and 61.1 (± 5.9), respectively, in the sham group vs 43.0 (± 6.4), 22.2 (± 5.9) and 52.7 (± 3.5) sec, respectively, in the pre-DH group. A two-way ANOVA (group x session) was used to compare the mean errors in the unimodal-cue sessions with those scored in the last session of training with the polymodal cues by sham and pre-DH animals (Fig. 4A). This revealed a significant effect of session [F(3, 42) = 20.58, p<0.01] but no significant effect of lesion group [F(1, 14) = 0.13, p = n.s.] and lesion group x session interaction [F(3, 42) = 1.29, p = n.s.]. Tukey HSD test indicated that the performance with the unimodal cues was significantly different from to that observed in the polymodal-cue task (p<0.01). A three way ANOVA of the errors scored across the trials with the unimodal cues (lesion group x cue x trials) showed a significant effect of unimodal cue [F(2, 32) = 16.9, p<0.01], but no significant effect of
lesion [F(1, 32) = 0.07, p = n.s.] and trial [F(8, 256) = 0.43, p = n.s.]. In addition, the lesion group x cue interaction was significant [F(2, 32) = 4.7, p<0.05], whereas there was no significant difference for group x trial [F(8, 256) = 2.8, p = n.s.] and cue x trial interaction [F(16, 256) = 1.8, p = n.s.]. The time latency to complete the trial was significantly higher (p< 0.01) in the unimodal task than in the last session of training with the polymodal cues for both control and pre-DH rats (Fig. 4B).
2.3 Effect of dorsal hippocampal lesion on retention Polymodal associative task. To assess the role of the dorsal hippocampus in the retention, rats were first trained on the polymodal-cue task and then received lesions at the dorsal hippocampus (n= 8) or sham surgery (n= 7; Fig. 2). As shown in Fig. 5, the two groups of rats did not show any difference in the task acquisition before surgery. Following surgery, rats with lesions of DH appeared more impaired in the retention of the polymodal-cue task relative to sham controls. On the first day of retesting after surgery, post-DH lesioned rats scored a mean of 31.4 % (± 14.9) of correct choice compared to 84% (± 7.6) scored by sham rats (p<0.01). The sham group reached a stable performance (over 90% of correct choices) in the following 4 days; whereas the hippocampal rats had to relearn the polymodal-cue task reaching the asymptotic level on day 25. Unimodal associative test. When tested on the unimodal cue-oriented task, post-DH animals appeared to be significantly impaired than controls (Fig. 4). The mean percentage of correct choices in the presence of light, sound and rough sheet was 26.9 (± 8.3), 28.5 (± 5.9) and 49.1 (± 8.7), for the post-DH animals versus 30.3 (± 4.8), 27.6 (± 6.1) and 50.5 (± 4.3) for the
sham group. A two-way ANOVA (lesion group x session) was used to compare the mean errors in the unimodal-cue guided task with those scored in polymodal-cue guided task by sham and post-DH rats (Fig. 4A). Such analysis revealed a significant effect of session [F(3, 39) = 23.1, p<0.01] and lesion [F(1, 13) = 10.9, p<0.01] but no significant effect of lesion x session interaction [F(3, 39), p = n.s.]. Tukey HSD test indicated that the response to the unimodal cue was significantly different from that with the polymodal cues (p<0.01). A three way ANOVA (lesion group x cue x trials) of the errors scored across the trials with the unimodal cue showed a significant effect of cue [F(2, 30) = 10.8, p<0.01] and lesion group [F(1, 30) = 7.5, p<0.01]. The trial factor was not significant [F(8, 240) = 1.04, p = n.s.] as well as the lesion x cue [F(2, 30) = 0.0, p = n.s.], lesion x trial [F(8, 240) = 1.0, p = n.s.] and the cue x trial [F(16, 240) = 1.25, p = n.s.] interaction. The time latency to complete the trial was significantly higher (p< 0.01) in the unimodal task than in the last session of training with the polymodal cues for both control and post-DH rats (Fig. 4B).
3. Discussion The results of our study show that animals rely on the combination of polymodal cues to solve the task since their performance improves during the training with the three stimuli (light, sound, and rough sheet) but then drops dramatically when the animals are presented with a single cue. It is quite unlikely that the animals relied on other stimuli or strategies to solve the polymodal task, such as egocentric, head direction, spatial cues, as they were made
unavailable by the training protocol. If any head-direction supported navigation was eventually used by the animals, this would lead to errors or random performance. The decrease in the response to the elemental stimuli after the polymodal training can be a consequence of generalization decrement as predicted by most theories of associative learning ( on le et al., 2003; Pearce, 1994; Thorwart and Lachnit, 2010). According to these theories, animals trained with an AB compound and tested with B alone experience a greater change from training to test than those trained and tested with B alone. It is therefore assumed that the components within a compound may not activate the same association as when presented on its own. A clear example of generalization decrement has been described by Brandon et al (2000), using the rabbit conditioned eyeblink response. They found that rabbits trained with a compound of stimuli (light, tone and vibrations) showed a reduced level of conditional responding on test trial in which one cue was removed. Such decrement was less evident when the animals were trained with one cue and then tested with a compound. Interestingly, there were no differences among the groups across the acquisition sessions with one cue or multiple cues. Overall, we exclude that the results of the test trial in the presence of one cue can be interpreted as a generalization decrement. If this was the case the drop of performance in the presence of the single cue should be the same independently of the used stimuli. On the contrary, the performance with one specific cue of the compound was better than that with other cues, but in an unpredictable manner. For example, in the case of the control group (pre- and post-sham) the performance with the rough sheet was significantly better than that with light or sound, as indicated by the number of errors the animals made to reach the correct arm.
Based on our results, it is also hard to establish whether animals used elemental or configural strategies to solve the task. Many authors hypothesize that elemental processing occurs in the presence of stimuli that are very dissimilar, such as stimuli coming from different modalities (Kehoe et al., 1994; Myers et al., 2001; Rescorla and Coldwell 1995). On the other hand, configural processing is more likely to occur with similar stimuli, such as those coming from the same modality. For example, a summation effect has been reported in studies carried out in rats and rabbits using conditioned stimuli of different modalities (i.e. visual, tactile, auditory). No response summation has been found in studies using pigeon autoshaping with compounds of two auditory stimuli or two diffuse lights (Aydin and Pearce, 1995). Following DH lesion, we observed an overall deficit in the acquisition of polymodal-cue oriented task with respect to control animals. Pre-DH animals showed a slower rate of acquisition than controls, accompanied by consistently increased latency to approach the goal arm particularly in the early phase of training. Such deficit improved with overtraining and all the pre-DH lesioned animals ultimately achieved the learning criterion. This result suggests that the lesion of dorsal hippocampal circuits before acquisition might allow a takeover of the acquisition function by other regions that eventually operate in a less efficient way with respect to control conditions. Lesion of the dorsal hippocampus, one day after training, impaired memory of the previously learned polymodal task; whereas rats with intact hippocampi demonstrated excellent retention. Throughout the following days of postoperative training, the performance of the post-DH lesioned animals showed significant improvement and appeared even better than that of the pre-DH lesioned group without previous experience. These results suggest that certain spared memory of the polymodal task exists in the post-
DH animals. However, these spared memories appear to be less flexible than those formed in intact brains since the post-DH animals made more errors than control and pre-DH rats when they were switched to the unimodal test. In particular, pre-DH rats made more errors on test trials with the sound than with the other two stimuli, light or rough sheet. On the contrary, the performance of post-DH was poor with light and sound but less with rough sheet. By examining the differences between the two lesion groups during the test trials, we found that pre-DH performed better in the presence of light or rough sheet than post-DH, whereas there was no difference with sound. Our results confirm previous studies showing a profound retrograde amnesia following hippocampal damage in spatial tasks that was compensated with retraining (Kubie et al. 1999; Ramos, 2009). According to Kubie’s data, the improvement observed in the hippocampal animals during retraining was correlated to partial memory for the original preoperative information retained by the rats. Indeed, during the retraining, the performance of the lesioned group with previous training appeared to be better than that of the lesioned group without previous training. Lesion and genetic knockout studies indicate that hippocampus binds together the diverse elements of an event, allowing the recall of all memory when only subsets of initial cues are present. The ability to reconstruct entire memory patterns from partial cues is called pattern completion (McNaughton and Morris, 1987; Norman and O'Reilly, 2003; Rolls and Treves, 1998; Rolls, 2013). In particular, region CA3, with its dense recurrent collaterals, is thought to support pattern completion. Indeed, studies carried in mice lacking the N-methyl-D-aspartate (NMDA) receptor gene in the CA3 subregion showed that the performance of mutant mice was equivalent to that of controls on the Morris water maze task when all extramaze cues were present (Nakazawa et al., 2002). However,
when three out of the four extramaze cues were removed, control rodents showed the same level of recall as in the full-cue condition, whereas the mutant rodents were severely impaired on the task. The results of our study did not show any evidence of pattern completion strategy since control rats were unable to completely reconstruct the entire memory when only subsets of initial cues were presented during the unimodal test. Moreover, we did not find any further impairment under partial-cue conditions in the animals lesioned at the dorsal hippocampus before training. In summary, the results of our study suggest that dorsal hippocampus functions as an associative network that integrates multiple cues entering into association with the reinforcement (Fanselow, 2000; Lee and Kesner, 2004; O'Reilly and Rudy, 2001; Rolls, 1996). Such processing may represent the physiological basis of associative learning such as configural conditioning, conditioning to a context and eventually spatial navigation (Miniaci et al., 1999) particularly when the environment to which subjects are exposed is rich in stimuli.
5. Methods
5.1 Subjects Long-Evans male rats (250-350 g) were randomly assigned to 2 experimental groups according to the timing of the lesions with respect to polymodal associative training: pretraining dorsal hippocampus lesion (pre-DH) and post-training dorsal hippocampus lesion (post-DH). An additional group of rats was sham-operated before or after training to be used as controls for each experimental group. Animals were maintained on a natural light-
dark cycle (12-hour light and 12-hour dark cycle) and had free access to water. The food was restricted to maintain rats at 85% of their ad libitum body weight.
5.2 Surgery Rats were anesthetized by using sodium pentobarbital (30 mg/Kg i.p.) and placed in the stereotaxic frame. The cranium was exposed and holes were drilled to allow the passage to an electrode. Stereotaxic coordinates were calculated from the flat- skull, according to Paxinos and Watson (1986). Electrolytic lesions were made by passing 2 mA anodal current for 30 sec through a stainless steel electrode. For the dorsal hippocampus lesions, the coordinates were: -3.3 mm posterior to bregma, ±1.5 mm from the midline, 3.5mm ventral from the scalp. Sham-lesioned rats received the same treatment of the hippocampal lesioned animals with the exception that no electrical current was delivered. Following surgery, the skin was sutured and the rats were allowed 10 days to recover before behavioral testing was started.
5.3 Apparatus All rats were trained in a radial maze consisting of a plexiglass octagonal platform (55 cm in diameter) with eight radiating arms (61 cm long and 10 cm wide). The walls of the arms were made of opaque PCV. The maze was placed in a dimly lit and sound attenuated room. All arms contained a 30-lumen incandescent bulb and a small loudspeaker placed at their distal end. The flickering light (1 Hz) and the intermittent tone (50 dB) were activated in the cue arm before the rat left the central arena, through a remote control. A rough plexiglass sheet was used as the tactile cue. It covered the floor of the cue arm and
was transferred from a cue arm to another one. Animal behavior was recorded using a camera coupled with video tracking software (ANY-maze, Stoelting).
5.4 Behavioral procedure Training on the polymodal associative task. Sham and pre-training hippocampal lesioned rats (pre-DH) were first given 5 daily sessions of pre-training to facilitate habituation to the apparatus. At the beginning of each daily session, animals were individually placed in the center of the platform and were allowed to explore the environment until they entered the cue arm signaled by the three stimuli (light, sound and rough sheet) and containing few grains of popped rice. Rats were then moved toward the central platform for the next trial. The position of the cue arm was maintained constant within each session. Following pre-training, the same rats were trained on the polymodal associative task (Fig. 2). Each rat received a single training session per day that consisted of 10 trials, with an inter-trial interval of 30 sec. The position of the cue arm, signaled by the three stimuli, changed pseudo-randomly from trial to trial. The trial ended when the animals entered up to two-thirds of the cue arm resulting in a delivery of food or 15 min elapsed. At the end of each trial, the animal was removed from the maze and placed in a cage until the next trial started. To reduce as much as possible the olfactory cues, care was taken to clean the maze after each trial with 70% ethanol solution. To prevent the animals from seeing the speaker and light bulb, we placed them at the end of all arms, behind a gray panel. Training continued until learning criterion was achieved, i.e. over 90% of correct choices in the last five sessions. Unimodal associative test.
The day after rats completed the polymodal-cue training, each animal was tested on the unimodal associative task. Importantly, all groups of rats were tested on single stimulus trial only after they reached the learning criterion on the polymodal task. In the unimodal test, rats were required to visit the cue arm signaled by either light, sound or rough sheet. Each animal received three trials per day for each of the three unimodal cues, in a pseudorandom order, for three consecutive days. Retention of the polymodal associative task. Rats receiving hippocampal lesions (post-DH) or sham-operation after training were retrained on the polymodal associative task. Once a stable performance was obtained they were tested on the unimodal associative task, as described above. All experiments were approved by the Ethical Committee of the University of Naples (#13/125 of 2013) and were in agreement with the governmental guidelines.
5.5 Data collection and Analysis Three variables were considered for analysis: correct choices, latency and errors. The mean (± SEM) percentage of correct choices, number of errors and latency were calculated per block of ten trials. A correct choice was scored when the rat entered the cue arm on the first choice; the visits to the other arms were counted as errors. Latency was defined as the time taken by the rat to move from the starting location to the correct arm during each trial. Data were subjected to the analysis of variance with repeated measures (ANOVA) and multiple post hoc comparisons, whenever appropriate.
5.6 Histology
At the end of the behavioral experiments, rats were killed with pentobarbital (50 mg/Kg) and perfused with saline followed by 10% formalin. The brains were removed and stored in formalin. Brain slices (40 m thick) were obtained using a cryostat (Leica RM 2125), mounted on a glass slide and stained with cresyl violet. The sections were then examined for histological verification of the lesion placement.
ACKNOWLEDGMENTS We thank Elvira De Leonibus for the helpful discussions. We thank Angelo Russo and Giovanni Esposito for their valuable technical assistance. The present research was supported by Bank of Italy grant #744460/13 to Maria Concetta Miniaci. The funding sources had no involvement in the study design, collection, analysis or interpretation of data, in the writing of the report, or in the decision to submit for publication.
References Aydin, A., Pearce J.M., 1995. Summation in autoshaping with short-duration and longduration stimuli. Quarterly Journal of Experimental Psychology Section B: Comparative and Physiological Psychology. 48, 215–234. Brandon, S.E., Vogel, E.G., Wagner, A.R., 2000. A componential view of configural cues in generalization and discrimination in Pavlovian conditioning. Behavioral Brain Research. 110, 67-72 Cohen, N.J., Eichenbaum, H., 1993. Memory, Amnesia, and the Hippocampal System. Cambridge, MA: MIT Press. Davidson, T.L., McKernan, M.G., Jarrard, L.E., 1993. Hippocampal lesions do not impair negative patterning: a challenge to configural association theory. Behav Neurosci. 107, 227-34. Eichenbaum, H., Dudchenko, P., Wood, E., Shapiro, M., Tanila, H., 1999. The Hippocampus, Memory, and Place Cells. Is It Spatial Memory or a Memory Space? Neuron. 23, 209–226 Fanselow, M.S., 2000. Contextual fear, gestalt memories, and the hippocampus. Behav Brain Res. 110, 73-81. Fanselow, M.S., Dong, H.W., 2010. Are the dorsal and ventral hippocampus functionally distinct structures? Neuron. 65, 7-19. Gallagher, M., Holland, P.C., 1992. Preserved configural learning and spatial learning impairment in rats with hippocampal damage. Hippocampus. 2, 81-8.
Gluck, M. A., Myers, C.E., Meeter, M., 2005. Cortico-hippocampal interaction and adaptive stimulus representation: A neurocomputational theory of associative learning and memory. Neural Networks, 18, 1265-1279. on le , F , uinn, , Fanselow, S., 2003. Differential effects of adding and removing components of a context on the generalization of conditional freezing. Journal of Experimental Psychology: Animal Behavior Processes, 29, 78–83. Harris, J.A., 2006. Elemental representations of stimuli in associative learning. Psychol Rev. 113, 584-605. Iordanova, M.D., Burnett, D.J., Good, M., Honey, R.C., 2011. Pattern memory involves both elemental and configural processes: evidence from the effects of hippocampal lesions. Behav Neurosci. 125, 567-77. Kehoe, E. J., Horne, A.J., Horne, P.S., Macrae, M., 1994. Summation and configuration between and within sensory modalities in classical conditioning of the rabbit. Animal Learning & Behavior. 22, 19-26. Kubie, J.L., Sutherland, R.J., Muller, R.U., 1999. Hippocampal lesions produce a temporally graded retrograde amnesia on a dry version of the Morris swimming task. Psychobiology. 27, 313–330. Lee, I., Kesner, R.P., 2004. Differential contributions of dorsal hippocampal subregions to memory acquisition and retrieval in contextual fear-conditioning. Hippocampus. 14, 301-310. McNaughton, B.L., Morris, R.G.M., 1987. Hippocampal synaptic enhancement and information storage within a distributed memory system. Trends Neurosci. 10, 408–415. Miniaci, M.C., Scotto, P., Bures, J., 1999. Place navigation in rats guided by a vestibular and kinesthetic orienting gradient. Behav Neurosci. 113, 1115-26. Moser, E.I., Kropff, K., Moser, M., 2008. Place cells, grid cells, and the brain’s spatial representation system. Ann Rev Neurosci. 31, 69–89. Myers, K. M., Vogel, E. H., Shin, J., Wagner, A. R., 2001. A comparison of the RescorlaWagner and Pearce models in a negative patterning and a summation problem. Animal Learning & Behavior. 29, 36–45. Nakazawa, K., Quirk, M.C., Chitwood, R.A., Watanabe, M., Yeckel, M.F., Sun, L.D., Kato A., Carr, C.A., Johnston, D., Wilson, M.A., Tonegawa, S., 2002. Requirement for hippocampal CA3 NMDA receptors in associative memory recall. Science 297, 211–218. Norman, K A , O’Reilly, R C , 2003 odeling hippocampal and neocortical contributions to recognition memory: A complementary learning systems approach. Psychological Review. 110, 611–646. O'Reilly, R.C., Rudy, J.W., 2001. Conjunctive representations in learning and memory: principles of cortical and hippocampal function. Psychol Rev. 108, 311-45. O'Reilly, R.C., Rudy, J.W., 2001. Conjunctive representations in learning and memory: principles of cortical and hippocampal function. Psychol Rev. 108, 311-45. Pearce, J. M., 2002. Evaluation and development of a connectionist theory of configural learning. Animal Learning and Behavior, 30, 73–95. Pearce, J.M., 1994. Similarity and discrimination: a selective review and a connectionist model. Psychol Rev. 101, 587–607. Pearce, J.M., Bouton, M.E., 2001. Theories of associative learning in animals. Annu Rev Psychol. 52, 111-39.
Ramos, J.M., 2009. Remote spatial memory and the hippocampus: effect of early and extensive training in the radial maze. Learn Mem. 16, 554-563. Rescorla, R.A., Coldwell, S.E., 1995. Summation in autoshaping. Animal Learning & Behavior. 23, 314-326. Rescorla, R.A., Wagner, A.R., 1972. A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In A. H. Black & W. F. Prokasy (Eds.), Classical conditioning II: Current theory and research. New York, NY:Appleton-Century-Crofts. Rescorla, R.A., 1972. "Configural" conditioning in discrete-trial bar pressing. J Comp Physiol Psychol. 79, 307-17. Rolls, E.T., and Treves, A., 1998. Neural Networks and Brain Function. Oxford: Oxford University Press. Rolls, E.T., 1996. A theory of hippocampal function in memory. Hippocampus. 6, 601620. Rolls, E.T., 2013. The mechanisms for pattern completion and pattern separation in the hippocampus. Front Syst Neurosci. 7, 74. Rudy, J.W., Sutherland, R.J., 1989. The hippocampal formation is necessary for rats to learn and remember configural discriminations. Behav Brain Res. 34, 97-109. Schmajuk, N.A., DiCarlo, J.J., 1992. Stimulus configuration, classical conditioning, and hippocampal function. Psychol Rev. 99, 268-305. Strange, B.A., Witter, M.P., Lein, E.S., Moser, E.I., 2014. Functional organization of the hippocampal longitudinal axis. Nat Rev Neurosci. 15, 655-69. Thorwart, A., Lachnit, H., 2010. Generalization decrements: Further support for flexibility in stimulus processing. Learning & Behavior. 38, 367– 373. Whishaw, I.Q., Tomie, J.A., 1991. Acquisition and retention by hippocampal rats of simple, conditional, and configural tasks using tactile and olfactory cues: implications for hippocampal function. Behav Neurosci. 105, 787-97. Whitlow, J.W., Wagner, A.R., 1972. Negative patterning in classical conditioning: Summation of response tendencies to isolable and configural components. Psychonomic Science 27, 299-301.
Figure legends Fig. 1: Histological evaluation of the electrolytic dorsal hippocampal lesion. Photomicrographs of representative coronal brain sections stained with cresyl violet from a bilateral dorsal hippocampus lesioned (A) and sham (B) animals. (C) Histological reconstruction of representative electrolytic lesions of the dorsal hippocampus. The gray area represents the extent of the largest lesion, and the black area represents the extent of the smallest lesion. The coronal sections were taken from Paxinos and Watson (1986).
Fig. 2: Experimental procedures. To assess the role of the hippocampus in the acquisition of the polymodal-cue guided task (upper panel), rats were trained in the radial maze following either the hippocampal or sham lesion (surgery). During the polymodal associative training, the animals were required to visit the cue arm signaled by the three stimuli to obtain a reward. The day after the acquisition of the polymodal task, the same rats were tested on the unimodal cue-oriented task in which the cue arm was signaled by one of the three stimuli at a time. The position of the cue arm changed pseudo-randomly for each trial in both polymodal and unimodal tasks. To examine the role of the hippocampus in the retention (bottom panel), rats were first trained on the polymodal-cue oriented task and then received lesion at the dorsal hippocampal or sham surgery. The day after the reacquisition of the polymodal task, the same rats were tested on the unimodal cue-oriented task, as above.
Fig. 3: Effects of the dorsal hippocampal lesion on the acquisition of the polymodalcue oriented task. Percentage of correct choices (A), number of errors (B) and latency to complete the trial (C) by control sham animals and rats lesioned at the dorsal hippocampus (pre-DH) before the acquisition of polymodal associative task. Data are expressed as the mean ± SEM.
Figure 4. Performance in the unimodal test trials. The bar graphs show the number of errors (A) and latency to complete the trial (B) in the last session of the polymodal associative task (black column) and the unimodal associative test in the presence of a single stimulus (light, sound or rough sheet) by pre-sham, pre-DH, post-sham and postDH rats. Data are expressed as the mean ± SEM, *p< 0.05; ** p< 0.01.
Fig. 5: Effects of the dorsal hippocampal lesion on the retention of the polymodal-cue oriented task. Percentage of correct choices (A), number of errors (B) and latency to complete the trial (C) scored by rats that received sham or hippocampal lesion (post-DH) after polymodal training. Data are expressed as the mean ± SEM. The break in the x-axis corresponds to the period of recovery after surgery.
Highlights: When stimuli of different modalities are used in association to guide animal’s behavior, rats learn to respond to them as a compound stimulus. Selective lesion of the dorsal hippocampus impairs the acquisition and disrupts the retention of the polymodal-cue oriented task. The dorsal hippocampus contributes to the processing of multimodal sensory information for the associative memory formation and consolidation.
PII: DOI: Reference:
S0006-8993(16)30454-1 http://dx.doi.org/10.1016/j.brainres.2016.06.030 BRES44981
To appear in: Brain Research Received date: 15 October 2015 Revised date: 8 May 2016 Accepted date: 20 June 2016 Cite this article as: Maria Concetta Miniaci, Pellegrino Lippiello, Marcellino Monda and Pietro Scotto, Role of hippocampus in polymodal-cue guided tasks in rats, Brain Research, http://dx.doi.org/10.1016/j.brainres.2016.06.030 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Role of hippocampus in polymodal-cue guided tasks in rats Maria Concetta Miniaci1*, Pellegrino Lippiello1, Marcellino Monda2, Pietro Scotto1 1
Department of Pharmacy, University of Naples “Federico II”, 80131 Naples, Italy.
2
Department of Experimental Medicine, Second University of Naples, 80138 Naples, Italy
*
Corresponding author: Department of Pharmacy, University of Naples “Federico II”,
via D. Montesano, 49; 80131 - Naples, Italy. Tel. +39 081-678465; Fax +39 081-678403. [email protected] Abstract To examine how signals from different sensory modalities are integrated to generate an appropriate goal-oriented behavior, we trained rats in an eight-arm radial maze to visit a cue arm provided with intramaze cues from different sensory modalities, i.e. visual, tactile and auditory, in order to obtain a reward. When the same rats were then examined on test trials in which the cue arm contained one of the stimuli that the animals were trained with (i.e. light, sound or rough sheet), they showed a significant impairment with respect to the performance on the polymodal-cue task. The contribution of the dorsal hippocampus to the acquisition and retention of polymodal-cue guided task was also examined. We found that rats with dorsal hippocampal lesions before training showed a significant deficit in the acquisition of polymodal-cue oriented task that improved with overtraining. The selective lesion of the dorsal hippocampus after training disrupted memory retention, but the animals’ performance improved following retraining of the polymodal task. All hippocampal lesioned rats displayed an impaired performance on the unimodal test. These findings suggest that the dorsal hippocampus contributes to the processing of multimodal sensory information for the associative memory formation and consolidation.
ABBREVIATIONS
DH: dorsal hippocampus; pre-DH lesion: pre-training dorsal hippocampus lesion; postDH lesion: post-training dorsal hippocampus lesion Keywords: Dorsal hippocampus, learning, memory, polymodal cues, electrolytic lesions.
1. Introduction The distinction between elemental and configural models is a central issue in associative learning theory about how the hippocampus contributes to learning and memory (Harris, 2006; Iordanova et al., 2011; Pearce and Bouton, 2001; Rudy and Sutherland, 1989). According to the elemental theories, stimuli are processed separately, independent of whether they have been presented alone or compounded with other stimuli, meaning that that the associative strength of a compound is equal to the algebraic sum of the associative strength of its components (Rescorla and Wagner, 1972). The configural theories assume that conditioning with a compound results in a unitary representation of the compound as a configuration entering into an association with the reinforcement (Pearce, 2002). Interestingly, patients and animals with damage to the hippocampus are unable to solve a novel associative learning task whose solution required the acquisition of specific combinations of elements i.e. configuration, rather than single elements (Rudy and Sutherland, 1989; Schmajuk and DiCarlo, 1992). Negative patterning is a well-known discrimination task that has been proposed for comparing elemental and configural theories in animal models (Rescorla, 1972; Whitlow and Wagner, 1972). To solve the problem, animals have to learn to respond to two single reinforced stimuli (A+, B+) but not to their combination, which is non-reinforced (AB-). By using this paradigm, Rudy and Sutherland (1989) demonstrated that rats with hippocampal lesions behaved similarly
to their controls in the presence of a single stimulus, but were unable to generate the appropriate response when the two cues were associated. However, these results are in contrast with those reported by other authors who found little or even no effect of the hippocampal lesion on the acquisition of a configural associative task (Davidson et al., 1993; Gallagher and Holland, 1992; Whishaw and Tomie, 1991). Nevertheless, the hippocampus is considered a critical brain area where spatial and nonspatial information can be integrated into a unified event representation (Eichenbaum et al., 1999; Moser et al., 2008). Indeed, during new memory formation the hippocampus forms connections between sensory stimuli that can be stored and recalled later (Cohen and Eichenbaum, 1993; Gluck et al., 2005; O’Reilly and Rudy, 2001). To better understand how multiple stimuli combine to produce an appropriate behavior during associative learning, we developed a multisensory apparatus that allows us to measure the navigational behavior along with the precisely controlled presentation of visual, auditory and tactile stimuli. For this purpose, rats were trained in an eight-arm radial maze to visit a cue arm provided with intramaze cues from different sensory modalities, i.e. light, sound and rough sheet, in order to obtain a reward. At the end of the training, the same rats were examined on test trials in which the cue arm contained one of the stimuli, that the animals were trained with, to determine whether the animal’s performance was primarily controlled by the sum of the elements presented, as a compound stimulus, or by one specific sensory element of the compound. Since most of the highly processed information from the sensory cortices enters the hippocampus mainly through its dorsal section (Fanselow and Dong, 2010; Strange et al., 2014), we next determined the role played by dorsal hippocampus during acquisition and retention of the polymodal associative task. For this purpose, we examined the behavioral
effects of discrete electrolytic lesions of the dorsal hippocampus, before and after training of the polymodal cue-guided task.
2. Results
2.1 Histology A schematic representation of the extent of the electrolytic lesion of the dorsal hippocampus is shown in Fig. 1. The electrolytic lesions appeared to be complete within the targeted region with no damage to the surrounding tissue. The sham groups did not show any damage to any brain region.
2.2 Effect of the dorsal hippocampal lesion on acquisition Polymodal associative task. Dorsal hippocampal (pre-DH; n= 8) and sham (n= 8) lesioned rats were trained on the polymodal-cue guided task in the eight-arm radial maze (Fig. 2). During the polymodal training, each rat was required to visit the cue arm containing the visual (light), auditory (sound) and tactile (rough sheet) stimuli to obtain a reward. The position of the cue arm changed for each trial. Initially, both pre-DH and sham animals reached the cue arm taking random routes. However, the sham group learned the task more rapidly than the pre-DH group. As shown in Fig. 3, sham and pre-DH animals achieved the learning criterion after 18 and 25 daily sessions of training, respectively. A two-way ANOVA (lesion group x session) of the mean percentage of correct choices for the first 18 sessions indicated a significant difference between the sham and pre-DH groups [F(1, 14) = 61.8, p<0.01] and revealed also an effect of sessions [F(17, 238) = 17.69, p<0.01] and group x
session interaction [F(17, 238) = 2.4, p<0.01]. The two-way ANOVA (lesion group x session) of the mean errors for the first 18 sessions showed a significant difference between groups [F(1, 14) = 33.44, p<0.01] and sessions, [F(17, 238) = 4.20, p<0.01] but no significant lesion group x session interaction [F(17, 238) = 1.15, p = n.s.]. Concerning the latency, the effects of lesion group [F(1, 14) = 50.2, p<0.01] and session [F(I7, 238) = 10.5, p<0.01] were significant for the first 18 sessions as was the session x lesion group interaction [F(17, 238) = 2, p<0.05]. Unimodal associative test. One day after the animals completed the polymodal-cue training, each animal was tested on the unimodal-cue task in which the cue arm was signaled by one of the three stimuli at a time, i.e. light, sound or rough sheet. As shown in Fig. 4, both sham and pre-DH rats were impaired irrespective of which unimodal cue was used. The mean percentage of correct choices scored in the unimodal-cue task with light, sound and rough sheet was 37.5 (± 5.1), 37.5 (± 6.3) and 61.1 (± 5.9), respectively, in the sham group vs 43.0 (± 6.4), 22.2 (± 5.9) and 52.7 (± 3.5) sec, respectively, in the pre-DH group. A two-way ANOVA (group x session) was used to compare the mean errors in the unimodal-cue sessions with those scored in the last session of training with the polymodal cues by sham and pre-DH animals (Fig. 4A). This revealed a significant effect of session [F(3, 42) = 20.58, p<0.01] but no significant effect of lesion group [F(1, 14) = 0.13, p = n.s.] and lesion group x session interaction [F(3, 42) = 1.29, p = n.s.]. Tukey HSD test indicated that the performance with the unimodal cues was significantly different from to that observed in the polymodal-cue task (p<0.01). A three way ANOVA of the errors scored across the trials with the unimodal cues (lesion group x cue x trials) showed a significant effect of unimodal cue [F(2, 32) = 16.9, p<0.01], but no significant effect of
lesion [F(1, 32) = 0.07, p = n.s.] and trial [F(8, 256) = 0.43, p = n.s.]. In addition, the lesion group x cue interaction was significant [F(2, 32) = 4.7, p<0.05], whereas there was no significant difference for group x trial [F(8, 256) = 2.8, p = n.s.] and cue x trial interaction [F(16, 256) = 1.8, p = n.s.]. The time latency to complete the trial was significantly higher (p< 0.01) in the unimodal task than in the last session of training with the polymodal cues for both control and pre-DH rats (Fig. 4B).
2.3 Effect of dorsal hippocampal lesion on retention Polymodal associative task. To assess the role of the dorsal hippocampus in the retention, rats were first trained on the polymodal-cue task and then received lesions at the dorsal hippocampus (n= 8) or sham surgery (n= 7; Fig. 2). As shown in Fig. 5, the two groups of rats did not show any difference in the task acquisition before surgery. Following surgery, rats with lesions of DH appeared more impaired in the retention of the polymodal-cue task relative to sham controls. On the first day of retesting after surgery, post-DH lesioned rats scored a mean of 31.4 % (± 14.9) of correct choice compared to 84% (± 7.6) scored by sham rats (p<0.01). The sham group reached a stable performance (over 90% of correct choices) in the following 4 days; whereas the hippocampal rats had to relearn the polymodal-cue task reaching the asymptotic level on day 25. Unimodal associative test. When tested on the unimodal cue-oriented task, post-DH animals appeared to be significantly impaired than controls (Fig. 4). The mean percentage of correct choices in the presence of light, sound and rough sheet was 26.9 (± 8.3), 28.5 (± 5.9) and 49.1 (± 8.7), for the post-DH animals versus 30.3 (± 4.8), 27.6 (± 6.1) and 50.5 (± 4.3) for the
sham group. A two-way ANOVA (lesion group x session) was used to compare the mean errors in the unimodal-cue guided task with those scored in polymodal-cue guided task by sham and post-DH rats (Fig. 4A). Such analysis revealed a significant effect of session [F(3, 39) = 23.1, p<0.01] and lesion [F(1, 13) = 10.9, p<0.01] but no significant effect of lesion x session interaction [F(3, 39), p = n.s.]. Tukey HSD test indicated that the response to the unimodal cue was significantly different from that with the polymodal cues (p<0.01). A three way ANOVA (lesion group x cue x trials) of the errors scored across the trials with the unimodal cue showed a significant effect of cue [F(2, 30) = 10.8, p<0.01] and lesion group [F(1, 30) = 7.5, p<0.01]. The trial factor was not significant [F(8, 240) = 1.04, p = n.s.] as well as the lesion x cue [F(2, 30) = 0.0, p = n.s.], lesion x trial [F(8, 240) = 1.0, p = n.s.] and the cue x trial [F(16, 240) = 1.25, p = n.s.] interaction. The time latency to complete the trial was significantly higher (p< 0.01) in the unimodal task than in the last session of training with the polymodal cues for both control and post-DH rats (Fig. 4B).
3. Discussion The results of our study show that animals rely on the combination of polymodal cues to solve the task since their performance improves during the training with the three stimuli (light, sound, and rough sheet) but then drops dramatically when the animals are presented with a single cue. It is quite unlikely that the animals relied on other stimuli or strategies to solve the polymodal task, such as egocentric, head direction, spatial cues, as they were made
unavailable by the training protocol. If any head-direction supported navigation was eventually used by the animals, this would lead to errors or random performance. The decrease in the response to the elemental stimuli after the polymodal training can be a consequence of generalization decrement as predicted by most theories of associative learning ( on le et al., 2003; Pearce, 1994; Thorwart and Lachnit, 2010). According to these theories, animals trained with an AB compound and tested with B alone experience a greater change from training to test than those trained and tested with B alone. It is therefore assumed that the components within a compound may not activate the same association as when presented on its own. A clear example of generalization decrement has been described by Brandon et al (2000), using the rabbit conditioned eyeblink response. They found that rabbits trained with a compound of stimuli (light, tone and vibrations) showed a reduced level of conditional responding on test trial in which one cue was removed. Such decrement was less evident when the animals were trained with one cue and then tested with a compound. Interestingly, there were no differences among the groups across the acquisition sessions with one cue or multiple cues. Overall, we exclude that the results of the test trial in the presence of one cue can be interpreted as a generalization decrement. If this was the case the drop of performance in the presence of the single cue should be the same independently of the used stimuli. On the contrary, the performance with one specific cue of the compound was better than that with other cues, but in an unpredictable manner. For example, in the case of the control group (pre- and post-sham) the performance with the rough sheet was significantly better than that with light or sound, as indicated by the number of errors the animals made to reach the correct arm.
Based on our results, it is also hard to establish whether animals used elemental or configural strategies to solve the task. Many authors hypothesize that elemental processing occurs in the presence of stimuli that are very dissimilar, such as stimuli coming from different modalities (Kehoe et al., 1994; Myers et al., 2001; Rescorla and Coldwell 1995). On the other hand, configural processing is more likely to occur with similar stimuli, such as those coming from the same modality. For example, a summation effect has been reported in studies carried out in rats and rabbits using conditioned stimuli of different modalities (i.e. visual, tactile, auditory). No response summation has been found in studies using pigeon autoshaping with compounds of two auditory stimuli or two diffuse lights (Aydin and Pearce, 1995). Following DH lesion, we observed an overall deficit in the acquisition of polymodal-cue oriented task with respect to control animals. Pre-DH animals showed a slower rate of acquisition than controls, accompanied by consistently increased latency to approach the goal arm particularly in the early phase of training. Such deficit improved with overtraining and all the pre-DH lesioned animals ultimately achieved the learning criterion. This result suggests that the lesion of dorsal hippocampal circuits before acquisition might allow a takeover of the acquisition function by other regions that eventually operate in a less efficient way with respect to control conditions. Lesion of the dorsal hippocampus, one day after training, impaired memory of the previously learned polymodal task; whereas rats with intact hippocampi demonstrated excellent retention. Throughout the following days of postoperative training, the performance of the post-DH lesioned animals showed significant improvement and appeared even better than that of the pre-DH lesioned group without previous experience. These results suggest that certain spared memory of the polymodal task exists in the post-
DH animals. However, these spared memories appear to be less flexible than those formed in intact brains since the post-DH animals made more errors than control and pre-DH rats when they were switched to the unimodal test. In particular, pre-DH rats made more errors on test trials with the sound than with the other two stimuli, light or rough sheet. On the contrary, the performance of post-DH was poor with light and sound but less with rough sheet. By examining the differences between the two lesion groups during the test trials, we found that pre-DH performed better in the presence of light or rough sheet than post-DH, whereas there was no difference with sound. Our results confirm previous studies showing a profound retrograde amnesia following hippocampal damage in spatial tasks that was compensated with retraining (Kubie et al. 1999; Ramos, 2009). According to Kubie’s data, the improvement observed in the hippocampal animals during retraining was correlated to partial memory for the original preoperative information retained by the rats. Indeed, during the retraining, the performance of the lesioned group with previous training appeared to be better than that of the lesioned group without previous training. Lesion and genetic knockout studies indicate that hippocampus binds together the diverse elements of an event, allowing the recall of all memory when only subsets of initial cues are present. The ability to reconstruct entire memory patterns from partial cues is called pattern completion (McNaughton and Morris, 1987; Norman and O'Reilly, 2003; Rolls and Treves, 1998; Rolls, 2013). In particular, region CA3, with its dense recurrent collaterals, is thought to support pattern completion. Indeed, studies carried in mice lacking the N-methyl-D-aspartate (NMDA) receptor gene in the CA3 subregion showed that the performance of mutant mice was equivalent to that of controls on the Morris water maze task when all extramaze cues were present (Nakazawa et al., 2002). However,
when three out of the four extramaze cues were removed, control rodents showed the same level of recall as in the full-cue condition, whereas the mutant rodents were severely impaired on the task. The results of our study did not show any evidence of pattern completion strategy since control rats were unable to completely reconstruct the entire memory when only subsets of initial cues were presented during the unimodal test. Moreover, we did not find any further impairment under partial-cue conditions in the animals lesioned at the dorsal hippocampus before training. In summary, the results of our study suggest that dorsal hippocampus functions as an associative network that integrates multiple cues entering into association with the reinforcement (Fanselow, 2000; Lee and Kesner, 2004; O'Reilly and Rudy, 2001; Rolls, 1996). Such processing may represent the physiological basis of associative learning such as configural conditioning, conditioning to a context and eventually spatial navigation (Miniaci et al., 1999) particularly when the environment to which subjects are exposed is rich in stimuli.
5. Methods
5.1 Subjects Long-Evans male rats (250-350 g) were randomly assigned to 2 experimental groups according to the timing of the lesions with respect to polymodal associative training: pretraining dorsal hippocampus lesion (pre-DH) and post-training dorsal hippocampus lesion (post-DH). An additional group of rats was sham-operated before or after training to be used as controls for each experimental group. Animals were maintained on a natural light-
dark cycle (12-hour light and 12-hour dark cycle) and had free access to water. The food was restricted to maintain rats at 85% of their ad libitum body weight.
5.2 Surgery Rats were anesthetized by using sodium pentobarbital (30 mg/Kg i.p.) and placed in the stereotaxic frame. The cranium was exposed and holes were drilled to allow the passage to an electrode. Stereotaxic coordinates were calculated from the flat- skull, according to Paxinos and Watson (1986). Electrolytic lesions were made by passing 2 mA anodal current for 30 sec through a stainless steel electrode. For the dorsal hippocampus lesions, the coordinates were: -3.3 mm posterior to bregma, ±1.5 mm from the midline, 3.5mm ventral from the scalp. Sham-lesioned rats received the same treatment of the hippocampal lesioned animals with the exception that no electrical current was delivered. Following surgery, the skin was sutured and the rats were allowed 10 days to recover before behavioral testing was started.
5.3 Apparatus All rats were trained in a radial maze consisting of a plexiglass octagonal platform (55 cm in diameter) with eight radiating arms (61 cm long and 10 cm wide). The walls of the arms were made of opaque PCV. The maze was placed in a dimly lit and sound attenuated room. All arms contained a 30-lumen incandescent bulb and a small loudspeaker placed at their distal end. The flickering light (1 Hz) and the intermittent tone (50 dB) were activated in the cue arm before the rat left the central arena, through a remote control. A rough plexiglass sheet was used as the tactile cue. It covered the floor of the cue arm and
was transferred from a cue arm to another one. Animal behavior was recorded using a camera coupled with video tracking software (ANY-maze, Stoelting).
5.4 Behavioral procedure Training on the polymodal associative task. Sham and pre-training hippocampal lesioned rats (pre-DH) were first given 5 daily sessions of pre-training to facilitate habituation to the apparatus. At the beginning of each daily session, animals were individually placed in the center of the platform and were allowed to explore the environment until they entered the cue arm signaled by the three stimuli (light, sound and rough sheet) and containing few grains of popped rice. Rats were then moved toward the central platform for the next trial. The position of the cue arm was maintained constant within each session. Following pre-training, the same rats were trained on the polymodal associative task (Fig. 2). Each rat received a single training session per day that consisted of 10 trials, with an inter-trial interval of 30 sec. The position of the cue arm, signaled by the three stimuli, changed pseudo-randomly from trial to trial. The trial ended when the animals entered up to two-thirds of the cue arm resulting in a delivery of food or 15 min elapsed. At the end of each trial, the animal was removed from the maze and placed in a cage until the next trial started. To reduce as much as possible the olfactory cues, care was taken to clean the maze after each trial with 70% ethanol solution. To prevent the animals from seeing the speaker and light bulb, we placed them at the end of all arms, behind a gray panel. Training continued until learning criterion was achieved, i.e. over 90% of correct choices in the last five sessions. Unimodal associative test.
The day after rats completed the polymodal-cue training, each animal was tested on the unimodal associative task. Importantly, all groups of rats were tested on single stimulus trial only after they reached the learning criterion on the polymodal task. In the unimodal test, rats were required to visit the cue arm signaled by either light, sound or rough sheet. Each animal received three trials per day for each of the three unimodal cues, in a pseudorandom order, for three consecutive days. Retention of the polymodal associative task. Rats receiving hippocampal lesions (post-DH) or sham-operation after training were retrained on the polymodal associative task. Once a stable performance was obtained they were tested on the unimodal associative task, as described above. All experiments were approved by the Ethical Committee of the University of Naples (#13/125 of 2013) and were in agreement with the governmental guidelines.
5.5 Data collection and Analysis Three variables were considered for analysis: correct choices, latency and errors. The mean (± SEM) percentage of correct choices, number of errors and latency were calculated per block of ten trials. A correct choice was scored when the rat entered the cue arm on the first choice; the visits to the other arms were counted as errors. Latency was defined as the time taken by the rat to move from the starting location to the correct arm during each trial. Data were subjected to the analysis of variance with repeated measures (ANOVA) and multiple post hoc comparisons, whenever appropriate.
5.6 Histology
At the end of the behavioral experiments, rats were killed with pentobarbital (50 mg/Kg) and perfused with saline followed by 10% formalin. The brains were removed and stored in formalin. Brain slices (40 m thick) were obtained using a cryostat (Leica RM 2125), mounted on a glass slide and stained with cresyl violet. The sections were then examined for histological verification of the lesion placement.
ACKNOWLEDGMENTS We thank Elvira De Leonibus for the helpful discussions. We thank Angelo Russo and Giovanni Esposito for their valuable technical assistance. The present research was supported by Bank of Italy grant #744460/13 to Maria Concetta Miniaci. The funding sources had no involvement in the study design, collection, analysis or interpretation of data, in the writing of the report, or in the decision to submit for publication.
References Aydin, A., Pearce J.M., 1995. Summation in autoshaping with short-duration and longduration stimuli. Quarterly Journal of Experimental Psychology Section B: Comparative and Physiological Psychology. 48, 215–234. Brandon, S.E., Vogel, E.G., Wagner, A.R., 2000. A componential view of configural cues in generalization and discrimination in Pavlovian conditioning. Behavioral Brain Research. 110, 67-72 Cohen, N.J., Eichenbaum, H., 1993. Memory, Amnesia, and the Hippocampal System. Cambridge, MA: MIT Press. Davidson, T.L., McKernan, M.G., Jarrard, L.E., 1993. Hippocampal lesions do not impair negative patterning: a challenge to configural association theory. Behav Neurosci. 107, 227-34. Eichenbaum, H., Dudchenko, P., Wood, E., Shapiro, M., Tanila, H., 1999. The Hippocampus, Memory, and Place Cells. Is It Spatial Memory or a Memory Space? Neuron. 23, 209–226 Fanselow, M.S., 2000. Contextual fear, gestalt memories, and the hippocampus. Behav Brain Res. 110, 73-81. Fanselow, M.S., Dong, H.W., 2010. Are the dorsal and ventral hippocampus functionally distinct structures? Neuron. 65, 7-19. Gallagher, M., Holland, P.C., 1992. Preserved configural learning and spatial learning impairment in rats with hippocampal damage. Hippocampus. 2, 81-8.
Gluck, M. A., Myers, C.E., Meeter, M., 2005. Cortico-hippocampal interaction and adaptive stimulus representation: A neurocomputational theory of associative learning and memory. Neural Networks, 18, 1265-1279. on le , F , uinn, , Fanselow, S., 2003. Differential effects of adding and removing components of a context on the generalization of conditional freezing. Journal of Experimental Psychology: Animal Behavior Processes, 29, 78–83. Harris, J.A., 2006. Elemental representations of stimuli in associative learning. Psychol Rev. 113, 584-605. Iordanova, M.D., Burnett, D.J., Good, M., Honey, R.C., 2011. Pattern memory involves both elemental and configural processes: evidence from the effects of hippocampal lesions. Behav Neurosci. 125, 567-77. Kehoe, E. J., Horne, A.J., Horne, P.S., Macrae, M., 1994. Summation and configuration between and within sensory modalities in classical conditioning of the rabbit. Animal Learning & Behavior. 22, 19-26. Kubie, J.L., Sutherland, R.J., Muller, R.U., 1999. Hippocampal lesions produce a temporally graded retrograde amnesia on a dry version of the Morris swimming task. Psychobiology. 27, 313–330. Lee, I., Kesner, R.P., 2004. Differential contributions of dorsal hippocampal subregions to memory acquisition and retrieval in contextual fear-conditioning. Hippocampus. 14, 301-310. McNaughton, B.L., Morris, R.G.M., 1987. Hippocampal synaptic enhancement and information storage within a distributed memory system. Trends Neurosci. 10, 408–415. Miniaci, M.C., Scotto, P., Bures, J., 1999. Place navigation in rats guided by a vestibular and kinesthetic orienting gradient. Behav Neurosci. 113, 1115-26. Moser, E.I., Kropff, K., Moser, M., 2008. Place cells, grid cells, and the brain’s spatial representation system. Ann Rev Neurosci. 31, 69–89. Myers, K. M., Vogel, E. H., Shin, J., Wagner, A. R., 2001. A comparison of the RescorlaWagner and Pearce models in a negative patterning and a summation problem. Animal Learning & Behavior. 29, 36–45. Nakazawa, K., Quirk, M.C., Chitwood, R.A., Watanabe, M., Yeckel, M.F., Sun, L.D., Kato A., Carr, C.A., Johnston, D., Wilson, M.A., Tonegawa, S., 2002. Requirement for hippocampal CA3 NMDA receptors in associative memory recall. Science 297, 211–218. Norman, K A , O’Reilly, R C , 2003 odeling hippocampal and neocortical contributions to recognition memory: A complementary learning systems approach. Psychological Review. 110, 611–646. O'Reilly, R.C., Rudy, J.W., 2001. Conjunctive representations in learning and memory: principles of cortical and hippocampal function. Psychol Rev. 108, 311-45. O'Reilly, R.C., Rudy, J.W., 2001. Conjunctive representations in learning and memory: principles of cortical and hippocampal function. Psychol Rev. 108, 311-45. Pearce, J. M., 2002. Evaluation and development of a connectionist theory of configural learning. Animal Learning and Behavior, 30, 73–95. Pearce, J.M., 1994. Similarity and discrimination: a selective review and a connectionist model. Psychol Rev. 101, 587–607. Pearce, J.M., Bouton, M.E., 2001. Theories of associative learning in animals. Annu Rev Psychol. 52, 111-39.
Ramos, J.M., 2009. Remote spatial memory and the hippocampus: effect of early and extensive training in the radial maze. Learn Mem. 16, 554-563. Rescorla, R.A., Coldwell, S.E., 1995. Summation in autoshaping. Animal Learning & Behavior. 23, 314-326. Rescorla, R.A., Wagner, A.R., 1972. A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In A. H. Black & W. F. Prokasy (Eds.), Classical conditioning II: Current theory and research. New York, NY:Appleton-Century-Crofts. Rescorla, R.A., 1972. "Configural" conditioning in discrete-trial bar pressing. J Comp Physiol Psychol. 79, 307-17. Rolls, E.T., and Treves, A., 1998. Neural Networks and Brain Function. Oxford: Oxford University Press. Rolls, E.T., 1996. A theory of hippocampal function in memory. Hippocampus. 6, 601620. Rolls, E.T., 2013. The mechanisms for pattern completion and pattern separation in the hippocampus. Front Syst Neurosci. 7, 74. Rudy, J.W., Sutherland, R.J., 1989. The hippocampal formation is necessary for rats to learn and remember configural discriminations. Behav Brain Res. 34, 97-109. Schmajuk, N.A., DiCarlo, J.J., 1992. Stimulus configuration, classical conditioning, and hippocampal function. Psychol Rev. 99, 268-305. Strange, B.A., Witter, M.P., Lein, E.S., Moser, E.I., 2014. Functional organization of the hippocampal longitudinal axis. Nat Rev Neurosci. 15, 655-69. Thorwart, A., Lachnit, H., 2010. Generalization decrements: Further support for flexibility in stimulus processing. Learning & Behavior. 38, 367– 373. Whishaw, I.Q., Tomie, J.A., 1991. Acquisition and retention by hippocampal rats of simple, conditional, and configural tasks using tactile and olfactory cues: implications for hippocampal function. Behav Neurosci. 105, 787-97. Whitlow, J.W., Wagner, A.R., 1972. Negative patterning in classical conditioning: Summation of response tendencies to isolable and configural components. Psychonomic Science 27, 299-301.
Figure legends Fig. 1: Histological evaluation of the electrolytic dorsal hippocampal lesion. Photomicrographs of representative coronal brain sections stained with cresyl violet from a bilateral dorsal hippocampus lesioned (A) and sham (B) animals. (C) Histological reconstruction of representative electrolytic lesions of the dorsal hippocampus. The gray area represents the extent of the largest lesion, and the black area represents the extent of the smallest lesion. The coronal sections were taken from Paxinos and Watson (1986).
Fig. 2: Experimental procedures. To assess the role of the hippocampus in the acquisition of the polymodal-cue guided task (upper panel), rats were trained in the radial maze following either the hippocampal or sham lesion (surgery). During the polymodal associative training, the animals were required to visit the cue arm signaled by the three stimuli to obtain a reward. The day after the acquisition of the polymodal task, the same rats were tested on the unimodal cue-oriented task in which the cue arm was signaled by one of the three stimuli at a time. The position of the cue arm changed pseudo-randomly for each trial in both polymodal and unimodal tasks. To examine the role of the hippocampus in the retention (bottom panel), rats were first trained on the polymodal-cue oriented task and then received lesion at the dorsal hippocampal or sham surgery. The day after the reacquisition of the polymodal task, the same rats were tested on the unimodal cue-oriented task, as above.
Fig. 3: Effects of the dorsal hippocampal lesion on the acquisition of the polymodalcue oriented task. Percentage of correct choices (A), number of errors (B) and latency to complete the trial (C) by control sham animals and rats lesioned at the dorsal hippocampus (pre-DH) before the acquisition of polymodal associative task. Data are expressed as the mean ± SEM.
Figure 4. Performance in the unimodal test trials. The bar graphs show the number of errors (A) and latency to complete the trial (B) in the last session of the polymodal associative task (black column) and the unimodal associative test in the presence of a single stimulus (light, sound or rough sheet) by pre-sham, pre-DH, post-sham and postDH rats. Data are expressed as the mean ± SEM, *p< 0.05; ** p< 0.01.
Fig. 5: Effects of the dorsal hippocampal lesion on the retention of the polymodal-cue oriented task. Percentage of correct choices (A), number of errors (B) and latency to complete the trial (C) scored by rats that received sham or hippocampal lesion (post-DH) after polymodal training. Data are expressed as the mean ± SEM. The break in the x-axis corresponds to the period of recovery after surgery.
Highlights: When stimuli of different modalities are used in association to guide animal’s behavior, rats learn to respond to them as a compound stimulus. Selective lesion of the dorsal hippocampus impairs the acquisition and disrupts the retention of the polymodal-cue oriented task. The dorsal hippocampus contributes to the processing of multimodal sensory information for the associative memory formation and consolidation.