- Email: [email protected]
Effects of intra-accumbens NMDA and AMPA receptor antagonists on short-term spatial learning in the Morris water maze task
Behavioural Brain Research 179 (2007) 43–49
Research report
Effects of intra-accumbens NMDA and AMPA receptor antagonists on short-term spatial learning in the Morris water maze task Valentina Ferretti a,b,c , Francesca Sargolini a,b,1 , Alberto Oliverio b,c,d , Andrea Mele b,c,d , Pascal Roullet a,∗ a
Centre de Recherches sur la Cognition Animale, Universit´e Paul Sabatier, CNRS-UMR 5169, 118 route de Narbonne, 31062 Toulouse Cedex 9, France b Dipartimento di Genetica e Biologia Molecolare, Universit` a di Roma ‘La Sapienza’, P.le Aldo Moro 5, 00185 Roma, Italy c Centro di Neurobiologia-D. Bovet, Universit` a di Roma ‘La Sapienza’, P.le Aldo Moro 5, 00185 Roma, Italy d Istituto di Neuroscienze, CNR-CERC, Roma, Italy Received 5 September 2006; received in revised form 4 January 2007; accepted 9 January 2007 Available online 13 January 2007
Abstract Glutamatergic transmission within the nucleus accumbens (Nac) is considered to subserve the transfer of different types of information from the cortical and limbic regions. In particular, it has been suggested that glutamatergic afferences from the hippocampus and the prefrontal cortex provide the main source of contextual information to the Nac. Accordingly, several authors have demonstrated that the blockade of glutamate receptors within the Nac impairs various spatial tasks. However, the exact role of the different classes of glutamate receptors within the Nac in short-term spatial memory is still not clear. In this study we investigated the involvement of two major classes of glutamate receptors, NMDA and AMPA receptors, within the Nac in the acquisition of spatial information, using the Morris water maze task. Focal injections of the NMDA antagonist, AP-5 (0.1 and 0.15 g/side), and the AMPA antagonist, DNQX (0.005, 0.01 g/side), were performed before a massed training phase, and mice were tested for retention immediately after. NMDA and AMPA receptor blockade induced no effect during training. On the contrary, injection of the two glutamatergic antagonists impaired spatial localization during the probe test. These data demonstrate an involvement of the Nac in short-term spatial learning. Moreover, they prove that within this structure the short-term processing of spatial information needs the activation of both NMDA and AMPA receptors. © 2007 Elsevier B.V. All rights reserved. Keywords: Glutamate; AP-5; DNQX; Spatial memory; Short-term memory; Mice
1. Introduction In recent years there has been much interest on the possible role of the nucleus accumbens (Nac) in mediating cognitive functions. This structure receives projections from different brain regions such as the hippocampus, the prefrontal cortex and the amygdala [14,17,42], all involved in different kinds of learning [5,10]. In particular the activity of the cortico-hippocampal network (including hippocampal, parahippocampal and prefrontal cortical areas) is essential for spatial
∗
Corresponding author. Tel.: +33 5 61 55 65 69; fax: +33 5 61 55 61 54. E-mail address: [email protected] (P. Roullet). 1 Present address: Laboratoire de Neurobiologie de la Cognition, Universit´ e de Provence, CNRS-UMR 6155, Marseille, France. 0166-4328/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bbr.2007.01.009
navigation [15,26,29,30,38]. Thus, it seems conceivable that the Nac, which is directly connected with those cortical and hippocampal regions might contribute to spatial learning and memory [33,36,39]. This hypothesis is sustained by a certain amount of experimental evidence. For example, cell firing according to location in space, similar to that shown by ‘place cells’ in the hippocampus, has been reported in the Nac [20,24]. Moreover, several studies have provided behavioural evidence that manipulations of the Nac impair performance in spatial learning tasks [2,11,23,31,32,37,39]. However, it is not completely clear which role the Nac might play. Indeed several factors could contribute to spatial navigation: information associated with different sensory modalities, reference systems (i.e. egocentric and allocentric), or motivation. Thus, the impairments observed after lesions or pharmacological manipulation of the Nac could be due in deficits to a variety of
44
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
cognitive abilities (for a discussion, see [2,7,25,37]). A further issue that should be considered is whether the Nac activation may be required during specific stages of spatial information processing—acquisition, memory consolidation and recall. In this regard, microinfusions of selective agonists and antagonists of specific receptor classes are extremely useful tools. In fact, they allow manipulation of the receptors’ activity within a limited time-window, thus permitting precise analysis of the effects induced during the different stages of memory. Furthermore, they enable the investigation of the specific role of the different receptor subtypes. Receptor binding studies have shown high levels of the different subtypes of glutamate receptors on the Nac neurons [1,40]. In particular, both NMDA and AMPA receptors are localized on the medium spiny neurons within this structure [6,35]. In light of the fact that the putative neurotransmitter for most of the inputs impinging on the Nac is glutamate, it has been suggested that NMDA and AMPA receptors could mediate the transmission of information to the Nac [32,37]. Interestingly, recent data point to a functional dissociation between the NMDA and AMPA receptors within the Nac in the modulation of spatial memory. For example, it has been demonstrated that NMDA, but not AMPA receptors are necessary for consolidating spatial information in an open field task, designed to estimate the ability of rodents to detect and react to object displacement [31]. Similar conclusions have been drawn on the basis of the effects observed using the Morris water maze [34], as well as non-spatial tasks, such as object recognition [33] and one-trial inhibitory avoidance [7]. Unfortunately, it is still not clear whether a similar difference between the two receptors of the Nac also exists in the short-term processing of spatial information. Recent reports provide evidence that AMPA but not NMDA receptors are involved in the acquisition of the one-trial inhibitory avoidance and object recognition tasks [7,33], thus suggesting that AMPA but not NMDA receptor activation is needed for the acquisition/encoding of information. On the other hand, a study by Maldonado-Irizarry and Kelley [23] demonstrated that pre-training focal injections within the Nac of both NMDA and AMPA receptor antagonists impaired rats in a spatial food-search task, but NMDA receptor blockade seemed more disruptive. Similarly, intra-accumbens administration of NMDA, but not AMPA antagonists, affect the acquisition of a Pavlovian approach task [8]. The aim of this study was therefore to further investigate the involvement of the two classes of glutamate receptors located in the Nac on short-term processing of spatial information. For this purpose, focal administrations of AP-5 and DNQX (NMDA and AMPA receptor antagonists, respectively) were performed intra-accumbens immediately before training, in the Morris water maze. The behavioural paradigm assesses the ability of mice to navigate in an environment using distal cues and requires the constitution of complex associations among stimuli. In addition, the entire behavioural procedure lasted 45 min, thus excluding the possibility that mice have to rely on long-term consolidated information in order to navigate in the maze.
2. Materials and methods 2.1. Animals A total of 95 CD1 male mice (IFFA CREDO, Lyon, France) were used in this study. Upon arrival, the animals were housed in groups of five in standard breeding cages (21 cm × 21 cm × 12 cm) placed in a rearing room at a constant temperature (22 ± 1 ◦ C) under diurnal conditions (light–dark: 08:00–20:00), with food and water ad libitum. At the time of surgery, they were approximately 8–9-weeks-old and their weights ranged from 35 to 40 g. All experiments were run in the afternoon, between 14:00 and 18:00 h. Every possible effort was made to minimise animal suffering and all procedures were in strict accordance with European community laws and regulations on the use of animals in research and NIH guidelines on animal care.
2.2. Surgery Mice underwent surgery 1 week after their arrival. They were anaesthetised with chloral hydrate (400 mg/kg) and placed in a stereotaxic frame (David Kopf Instruments, Tujunga, CA, USA) with mouse adapter and lateral ear bars. After placing the animals on the stereotaxic apparatus the head skin was cut longitudinally and bilateral guide cannulae (0.5 mm in diameter) were fixed on the calvarium with dental acrylic. The following coordinates with lambda and bregma in the same horizontal plane were used: anterior to bregma, +1.7 mm; lateral to midline, ±1 mm; ventral from the dura, +2.3 mm, according to Franklin and Paxinos [13]. Mice were then left in their home cage for a recovery period of 1 week.
2.3. Apparatus The circular swimming pool (110 cm in diameter and 30 cm in height) was made of ivory-coloured PVC, filled with water (25 ± 1 ◦ C) made opaque with Lytron 631, to 15 cm below the edge of the wall. Four start positions were located equidistantly around the edge of the maze, dividing it into four equal quadrants. During training, a circular goal platform painted white (9 cm of diameter) was laid in the centre of one quadrant, 15 cm from the wall. The platform had a rough (metal grid) surface providing sufficient grip for the animals to climb on top of it. The apparatus was placed in a separate room and surrounded by white curtains containing several extra-maze cues. It was illuminated by a white light (60 W) and had a video camera placed overhead and connected to a video recorder and monitor.
2.4. Procedure The general procedure consisted of three different phases: a familiarization phase, a massed training phase and a probe test [12,34]. Several extra-maze visual cues, approximately 50–100 cm away from the pool, were attached to the walls surrounding the apparatus. The mice were required to navigate to the invisible platform using the spatial cues available in the room. On the first day, the mice were individually submitted to a single familiarization session of three trials, with the platform protruding 0.5 cm above the surface of the water. This familiarization procedure is used to allow the mice to acquire the procedural components of the task (the presence of a platform, the handling), which is especially important when a massed training protocol is used. The session started with the mouse standing on the platform for 60 s. At the beginning of each trial the mice were introduced into the maze facing the wall at one of the four designated starting points and allowed to swim freely or until they reached the platform. Mice failing to find the platform within a fixed period of 60 s were gently guided by hand to the platform and a maximum escape latency of 60 s was recorded. After the animals climbed onto the platform, they were allowed to remain on it for additional 60 s, and subsequently replaced in the maze from a different starting position. The starting positions were determined in a pseudorandom order, such that each was only used once in a single session. The same spatial location of the platform was used for all the animals. Training started the next day. The mice were submitted to four consecutive sessions of three trials. The procedure was the same as for the familiarization phase, except for the platform, which was submerged 0.5 cm beneath the surface of the water. The platform was left in the same position during familiarization and
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49 training. Animals were introduced into the maze from different starting points and the sequence of starting locations randomised. Testing was conducted in sets of two mice, such that following a session the first mouse was placed in a holding cage while the second was run; following that session the first mouse was re-tested. Immediately after the last training session, the mice were submitted to a single probe test. The platform was removed and the mice, starting from the centre of the pool, were allowed 60 s to search for the platform.
2.5. Drugs Selective and reversible NMDA and AMPA receptor blockade was obtained through direct infusions into the Nac of (−)-AP-5 (0.1 and 0.15 g/side) and DNQX (0.005 and 0.01 g/side), competitive antagonists of the NMDA and the AMPA receptors respectively. AP-5 was dissolved in saline solution (0.9% NaCl in distilled water), and DNQX was dissolved in a solution of 2% DMSO in distilled water. Mice were assigned to a drug treatment group and received injections 10 min before the first training trial. Drug injections were performed by lowering into the guide cannula a 9 mm long injector connected with polyethylene tubing to a 1.0 l Hamilton syringe. The volume injected was always 0.2 l. Each injection lasted 2 min and the injector was left in place for an additional minute to allow diffusion. Before the injections mice were manipulated daily in order to habituate them to handling. On the injection day they were hand-held to insert the injector while during the injection they where allowed to move freely in the cage. Drug-treated animals were always compared to control mice injected with the same volume of vehicle solution, under the same conditions.
2.6. Data collection and statistics Data collection was performed using a video camera suspended above the pool and interfaced with a video tracking system (Ethovision® 2.3, Noldus Information Technology, Wageningen, The Netherlands). Several parameters of behavioural performance were recorded. The average swimming speed was calculated during all sessions. During training sessions, path length to mount onto the platform was recorded in each trial. During the probe test the number of annulus crossings was scored, i.e. the number of times a mouse crossed an ideal circle (14 cm of diameter) located around each of the four possible platform positions in the four quadrants. We used this behavioural parameter instead of the percent time spent in the training quadrant during the probe test, which has been reported to overestimate animals’ performance in term of spatial ability [4]. All data were analysed by analysis of variance (ANOVA). For the training phase, the path length for each session (average of three trials) was analysed with a repeated measure ANOVA analysis (between factor: treatment, three levels; within factor: sessions, four levels). When a significant session effect was found, a further repeated measure ANOVA analysis (within factor: sessions, four levels) was conducted for each group. The number of annulus crossings in each of the four quadrants during the probe test was analysed using a two-way ANOVA design (first factor: treatment, three levels; second factor: quadrants, four levels). A one-way (treatment: three levels) ANOVA design was employed to analyse possible differences in swimming speed. Post hoc multiple comparisons analysis was carried out when allowed, using Tukey’s honestly significant difference (HSD) test.
2.7. Histology At the completion of the experiment, mice were sacrificed by an overdose of chloral hydrate, the brains were removed and frozen at −20 ◦ C. Cannula placements were determined by examination of serial coronal sections (25 m) stained with Cresyl Violet.
3. Results 3.1. Cannula placement verification Fig. 1 shows a schematic representation of injector placements for the two experiments. Only animals showing a correct
45
placement within the Nac were included in the statistical analysis. 3.2. Experiment 1: NMDA receptors blockade Fig. 2A shows the path length to reach the platform during the training phase in saline (n = 10), AP-5 0.1 (n = 9) and AP-5 0.15 (n = 9) groups. Both saline and AP-5 groups decreased their path length to find the platform along the sessions. The repeated measures ANOVA analysis showed a significant session effect (F3,75 = 16.98, p < 0.001), no treatment effect (F2,25 = 1.789, p = 0.188) and no interaction between the two factors (F6,75 = 0.307, p = 0.931). A further ANOVA analysis showed a significant effect of session for each of the three groups (saline: F3,27 = 4.495, p = 0.011; AP-5 0.1: F3,24 = 4.880, p = 0.009; AP-5 0.15: F3,24 = 8.534, p = 0.001). During probe test, similarly to that observed during training sessions, the path lengths to reach the annulus in the target quadrant did not differ between saline and AP-5-injected mice (F2,25 = 0.002, p = 0.998) (data not shown). On the other hand, control animals crossed the annulus located in the correct quadrant significantly more than those located in the remaining three quadrants (Fig. 2B). Pre-training AP-5 injections significantly decreased the number of annulus crossings in the correct quadrant. Two-way ANOVA showed a significant quadrant effect (F3,100 = 19.089, p < 0.001), no treatment effect (F2,100 = 0.047, p = 0.954), and a significant interaction between the two factors (F6,100 = 3.050, p = 0.009) indicating a different profile of quadrants’ exploration in the three groups of mice. Post hoc comparison revealed a significant difference between the correct quadrant and the other three quadrants for control animals (p < 0.001). For both groups of AP-5 treated mice, no significant difference among all four quadrants was revealed even though, in the AP-5 0.1 group, there was a marginal effect on the difference between the correct quadrant and the left and opposite quadrants (p = 0.055). Note that post hoc comparison also showed a significant difference for annulus crossing in the correct quadrant between the control animals and the animals treated with the highest dose of AP-5 (p = 0.025). A further comparison between the annulus crossing in the correct quadrant in the first and in the second halves of the probe test revealed a significant decrease over time (F3,150 = 2.35, p = 0.034) but no difference among the experimental groups (F6,150 = 1.595, p = 0.152) (data not shown). Table 1 shows the effects of pre-training AP-5 injections on swimming speed during the probe test. No significant difference was observed between saline and AP-5-injected mice (F2,25 = 1.492, p = 0.244). 3.3. Experiment 2: AMPA receptors blockade Fig. 3A shows the path length to reach the platform during the training phase in saline (n = 9), DNQX 0.005 (n = 9) and DNQX 0.01 (n = 8) groups. Both vehicle and DNQX groups decreased the distance to find the platform across the four sessions. The repeated measure ANOVA analysis showed a significant session effect (F3,69 = 19.93, p < 0.001), no treatment
46
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
Fig. 1. Schematic representation of cannula placements in the two experiments. Each symbol represents the site of injection for one animal. (A) Experiment 1: pre-training AP-5 injections: (*) saline; () AP-5 0.1 g/side; (䊉) AP-5 0.15 g/side. (B) Experiment 2: pre-training DNQX injections. (*) vehicle; () DNQX 0.005 g/side; () DNQX 0.01 g/side.
effect (F2,23 = 1.288, p = 0.295), and no interaction between the two factors (F6,69 = 0.435, p = 0.853). A further ANOVA analysis showed a significant effect of session for each of the three groups (vehicle: F3,24 = 11.526, p < 0.001; DNQX 0.005: F3,24 = 4.778, p = 0.009; DNQX 0.01: F3,21 = 5.331, p = 0.007). Fig. 3B represents the number of annulus crossing in the four quadrants during the probe test, by vehicle- and DNQX-injected mice. The control group as well as the 0.005 DNQX-treated mice crossed the annulus in the correct quadrant significantly Table 1 Mean swimming speed ± S.E.M. in probe test, for the two experiments Group
Speed
Experiment 1 Saline (n = 10) AP-5 0.1 g/side (n = 9) AP-5 0.15 g/side (n = 9)
20.05 ± 2.86 22.64 ± 3.10 21.64 ± 2.87
Experiment 2 Vehicle (n = 9) DNQX 0.005 g/side (n = 9) DNQX 0.01 g/side (n = 8)
20.87 ± 1.59 20.72 ± 1.65 19.93 ± 1.15
more often, compared to the remaining three quadrants. The same did not occur for the mice injected with the highest dose of DNQX. The two-way ANOVA showed a significant quadrant effect (F3,92 = 20.158, p < 0.001) and no treatment effect (F2,92 = 0.030, p = 0.970). The three groups of mice displayed a different profile of exploration as demonstrated by the interaction between the quadrant and treatment factors (F6,92 = 2.365, p = 0.036). Post hoc comparison revealed a significant difference between the correct quadrant and the remaining three quadrants for both control animals (correct versus right and opposite p < 0.001; correct versus left p = 0.028) and mice injected with the lower dose of DNQX (correct versus left and opposite p < 0.001; correct versus right p = 0.028). For the DNQX 0.01 group, we observed a significant difference only between the correct and the opposite quadrant (p = 0.039), but no difference between the correct and the two adjacent quadrants. We performed a further analysis on the annulus crossings during the first and the second halves of the probe test (data not shown). No significant difference was found among the experimental groups (F6,132 = 0.721, p = 0.633). The number of annulus crossings in the target quadrant decreased over time, however this did not reach statistical significance (F3,132 = 2.52, p = 0.061). Sim-
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
Fig. 2. Experiment 1: effect of pre-training intra-accumbens AP-5 administrations on mouse performance during training and probe test, in the Morris water maze. (A) Mean of path length (cm) during training sessions in mice injected with saline, AP-5 0.1 g/side and AP-5 0.15 g/side. (B) Mean number of annulus crossings in the four quadrants, during the probe test for saline, AP-5 0.1 g/side groups and AP-5 0.15 g/side groups. (C) representative paths taken by animals treated with saline, AP-5 0.1 g/side and AP-5 0.15 g/side. * p ≤ 0.05 correct vs. opposite, right, left quadrants, within groups. + p ≤ 0.05, correct quadrant saline vs. AP-5 groups.
ilarly to that observed in the first experiment, an ANOVA on the path length to first entry in annulus revealed no treatment effect (F2,22 = 1.683, p = 0.209). Table 1 reports the effects of pre-training DNQX injections on swimming speed during probe test. No significant differences were observed between saline and DNQX-injected mice (F2,23 = 0.106, p = 0.900). 4. Discussion The present study demonstrates that intra-accumbens focal administrations of the NMDA competitive antagonist, AP-5, and the AMPA antagonist, DNQX, impair the ability of mice to find the hidden platform in a short-term version of the Morris
47
Fig. 3. Experiment 2: effect of pre-training intra-accumbens DNQX administrations on mouse performance during training and the probe test, in the Morris water maze. (A) Mean of path length during training sessions in mice injected with vehicle, DNQX 0.005 g/side groups and DNQX 0.01 g/side groups. (B) Mean number of annulus crossing in the four quadrants for vehicle, DNQX 0.005 g/side groups, and DNQX 0.01 g/side groups. (C) Representative paths taken by animals treated with vehicle, DNQX 0.005 g/side and DNQX 0.01 g/side. (*) p ≤ 0.05 correct vs. opposite, right, left quadrants, within groups.
water maze [12,27,34]. The results demonstrate that the Nac is necessary for short-term place navigation, and this requires the activation of both glutamate receptor subtypes. In both experiments, control and drug-injected mice progressively decreased the path length to reach the platform on the training phase. During the probe test, control mice crossed the annulus more often in the quadrant where the platform was previously located. On the contrary AP-5 and DNQX administrations dose-dependently decreased the number of annulus crossings in the correct quadrant. Interestingly in the two experiments a decrease in path length along the sessions was always observed independently of the effect on annulus crossing. This dissociation between the two measures was also observed within the probe trial, in which the path length to reach the annulus did not differ among experimental groups. Several authors have reported
48
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
a similar dissociation on the two parameters following different CNS manipulations. For example, it has been shown that temporary or permanent inactivation of the hippocampus, does not affect escape latency on training days but induces a significant deficit on the probe test, in both rats [28] and mice [19]. It has been suggested by different authors that a decrease in the path length or in the latencies does not necessarily represent a good assessment of spatial navigation abilities [19]. In fact, path length or latencies could improve with the acquisition of the procedural components of the task. On the contrary, the time spent or the number of annulus crossings in the target quadrant on probe test are considered to be directly linked to the acquisition of the correct spatial localization of the platform and could give a more appropriate measurement of spatial learning [19,21,27,41]. In our experiments, AP-5 and DNQX injected mice decreased the number of correct annulus crossings during the probe test but the learning performances along the training session is the same in the different groups showing that experimental mice certainly used another strategy to find the platform. The use of a different strategy seems also confirmed by the differences in the representative paths of control and of experimental animals during the probe test (Figs. 2C and 3C). A further caveat to be considered is that the number of annulus crossings during the probe test may represent an index of behavioural flexibility. According to this view, it is possible that during the probe test, control mice actively search for the platform in the correct spatial location and persist in their fruitless search inside the target sector. In contrast, mice injected with the two glutamate antagonists first searched for the platform in the correct quadrant and quickly extend their search to the entire pool. However, this does not seem to be the case because we observed a similar decrease in the number of annulus crossings during the first and the second halves of the probe test in all groups. This demonstrates that NMDA and AMPA receptor blockade in the Nac do not directly affect the behavioural flexibility. Rather, the activation of both NMDA and AMPA receptors within this structure is needed to maintain information for short periods of time. The results of the present study confirm but also extend previous findings in the literature. Several reports have investigated the role of the Nac, and in particular the contribution of AMPA and NMDA receptors within this structure in the different steps of memory formation. They provided convincing evidence on the role of both receptor subtypes in memory consolidation and retrieval. In fact, independently of the behavioural test, posttraining administrations of NMDA but not AMPA antagonist impair the ability to perform the task 24 h later [7,31,34], demonstrating that NMDA receptor activation is needed in the early stages of consolidation. On the contrary, AMPA but not NMDA antagonist injections before testing affect performance in tasks based on a large variety of cognitive abilities [7,8,31,37]. This suggests that AMPA but not NMDA receptors are involved in information retrieval. Based on these observations, the current hypothesis states that NMDA receptor activation in the Nac is needed when plastic changes are required in order to maintain and stabilize the memory trace (i.e. consolidation), while AMPA receptors are necessary when this information needs to be utilized.
In the light of the short time lapse between the acquisition phase and the test, the procedure used in the present study did not require any consolidation process for the task to be performed. In fact, even though on the one hand it seems conceivable that the consolidation processes started as soon as the animals began training, on the other hand distinct biochemical and neuroanatomical substrates underlie short and long-term memory processes [16]. It is thus unlikely that long-term processes are engaged in animals that have to provide an efficient response in a short delay. For the same reason, based on the current hypothesis on memory, the involvement of an active process of retrieval can also be excluded. Rather, in order to perform the task, the animals have to construct a spatial map of the environment, keep this information on a short-term buffer and interface it with the motor system. Pre-training administrations do not allow to dissect these processes and in the present study we did not perform any pre-test administration to distinguish encoding and maintenance from the use of information. However, as mentioned before, NMDA receptor activation does not seem to be necessary for using information in order to guide behaviour [3,8,31,37]. On the contrary, pre-test focal administrations of AMPA antagonists both in the accumbens [7,31] and in the hippocampus [3] have been demonstrated to impair performance in different behavioural tasks. Therefore, it seems conceivable to hypothesise that NMDA receptor activation within the accumbens is needed only to acquire and maintain spatial information. On the contrary, AMPA receptors within this structure may be required either to interface the cognitive process with the motor system or for both the acquisition-maintenance and the use of such information. Experimental evidence seems to support this latter hypothesis, but manipulations allowing better time resolution will be needed to address this issue more fully. A wide body of evidence is now available in the literature as regards to the possible biological mechanisms underlying long-term storage of information [9,18,22]. Instead, less is known about the putative changes necessary for short-term maintenance. It is interesting in this regard to compare the effects induced by intra-accumbens pre-training administrations of NMDA and AMPA antagonists in the Morris water maze with those observed in one-trial inhibitory avoidance task [7]. In fact, blockade of AMPA antagonists affected the ability to solve both tasks, while the NMDA antagonist impaired performance only in the water maze. This suggests that within the nucleus accumbens the need for the activation of the NMDA but not the AMPA receptors depends upon the kind of information to be processed. Further neuropharmacological studies based on a larger number of tasks requiring different cognitive processes will help to elucidate this issue. However, on the basis of these two studies solely, it could be speculated that during short-term information processing, the NMDA receptor might come into play only when complex associations among stimuli are required, such as in the Morris water maze. Acknowledgements This study was supported by an I.S.E.R.M. fellowship to FS, a fellowship from the University of Rome ‘La Sapienza’ to VF,
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
F.I.R.B. and PRIN grants by MIUR to AO and AM, and a grant from the University of Rome ‘La Sapienza’ to AO. We thank Peter Winterton and Anne-Marie Mc Gauran for proofreading this manuscript. References [1] Albin RL, Makowiec RL, Hollingsworth Z, Dure IVLS, Penney JB, Young AB. Excitatory amino acid binding sites in the basal ganglia of the rat: a quantitative autoradiographic study. Neuroscience 1991;46:35–48. [2] Annett LE, McGregor A, Robbins TW. The effects of ibotenic acid lesions of the nucleus accumbens on spatial learning and extinction in the rat. Behav Brain Res 1989;31:231–42. [3] Bast T, da Silva BM, Morris RG. Distinct contributions of hippocampal NMDA and AMPA receptors to encoding and retrieval of one-trial placememory. J Neurosci 2005;25:5845–56. [4] Blokland A, Geraerts E, Been M. A detailed analysis of rats’ spatial memory in a probe trial of a Morris task. Behav Brain Res 2004;154(1):71–5. [5] Burns LH, Annett L, Kelley AE, Everitt BJ, Robbins TW. Effects of lesions to amygdala, ventral subiculum, medial prefrontal cortex, and nucleus accumbens on the reaction to novelty: implication for limbic-striatal interactions. Behav Neurosci 1996;110:60–73. [6] Daw NW, Stein PS, Fox K. The role of NMDA receptors in information processing. Annu Rev Neurosci 1993;16:207–22. [7] De Leonibus E, Costantini VJ, Castellano C, Ferretti V, Oliverio A, Mele A. Distinct roles of the different ionotropic glutamate receptors within the nucleus accumbens in passive-avoidance learning and memory in mice. Eur J Neurosci 2003;18(8):2365–73. [8] Di Ciano P, Cardinal RN, Cowell RA, Little SJ, Everitt BJ. Differential involvement of NMDA, AMPA/kainate, and dopamine receptors in the nucleus accumbens core in the acquisition and performance of pavlovian approach behavior. J Neurosci 2001;21(23):9471–7. [9] Dudai Y. Molecular bases of long-term memories: a question of persistence. Curr Opin Neurobiol 2002;12:211–6. [10] Floresco SB, Seamans JK, Phillips AG. Differential effects of lidocaine infusions into the ventral CA1/subiculum or the nucleus accumbens on the acquisition and retention of spatial information. Behav Brain Res 1996;81:163–71. [11] Floresco SB, Seamans JK, Phillips AG. Selective roles for hippocampal, prefrontal cortical and ventral striatal circuits in radial-arm maze tasks with or without a delay. J Neurosci 1997;17:1880–90. [12] Florian C, Roullet P. Hippocampal CA3-region is crucial for acquisition and memory consolidation in Morris water maze task in mice. Behav Brain Res 2004;154(2):365–74. [13] Franklin BJ, Paxinos G. The mouse brain in stereotaxic coordinates. San Diego: Academic Press; 1997. [14] Friedman DP, Aggleton JP, Saunders RC. Comparison of hippocampal, amygdala, and perirhinal projections to the nucleus accumbens: combined anterograde and retrograde tracing study in the Macaque brain. J Comp Neurol 2002;450:345–65. [15] Hok V, Save E, Lenck-Santini PP, Poucet B. Coding for spatial goals in the prelimbic/infralimbic area of the rat frontal cortex. Proc Natl Acad Sci USA 2005;102(12):4602–7. [16] Izquierdo I, Medina JH, Vianna MR, Izquierdo LA, Barros DM. Separate mechanism for short- and long-term memory. Behav Brain Res 1999;103(1):1–11. [17] Jay TM, Witter MP. Distribution of hippocampal CA1 and subicular efferents in the prefrontal cortex of the rat studied by means of anterograde transport of Phaseolus vulgaris leucoagglutinin. J Comp Neurol 1991;313:574–86. [18] Kandel ER. The molecular biology of memory storage: a dialogue between genes and synapses. Science 2001;294:1030–8. [19] Lassalle JM, Bataille T, Halley H. Reversible inactivation of the hippocampal mossy fiber synapses in mice impairs spatial learning, but neither consolidation nor memory retrieval, in the Morris navigation task. Neurobiol Learn Mem 2000;73:243–57.
49
[20] Lavoie AM, Mizumori SJ. Spatial, movement- and reward-sensitive discharge by medial ventral striatum neurons of rats. Brain Res 1994;638:157–68. [21] Lipp HP, Wolfer DP. Genetically modified mice and cognition. Curr Opin Neurobiol 1998;8:272–80. [22] MacGaugh JL. Memory—a century of consolidation. Science 2000;287: 248–51. [23] Maldonado-Irizarry CS, Kelley AE. Excitatory amino acid receptors within nucleus accumbens subregions differentially mediate spatial learning in the rat. Behav Pharmacol 1995;6:527–39. [24] Martin PD, Ono T. Effects of reward anticipation, reward presentation, and spatial parameters on the firing of single neurons recorded in the subiculum and nucleus accumbens of freely moving rats. Behav Brain Res 2000;116:23–38. [25] Mele A, Avena M, Roullet P, De Leonibus E, Mandillo S, Sargolini F, et al. Nucleus accumbens dopamine receptors in the consolidation of spatial memory. Behav Pharmacol 2004;15(5–6):423–31. [26] Morris RGM, Garrud P, Rawlins JNP, O’Keefe J. Place navigation in rats with hippocampal lesions. Nature 1982;297:681–3. [27] Morris RGM. Developments of a water-maze procedure for studying spatial learning in the rat. J Neurosci Meth 1984;11:47–60. [28] Moser E, Moser MB, Andersen P. Spatial learning impairment parallels the magnitude of dorsal hippocampal lesions, but is hardly present following ventral lesions. J Neurosci 1993;13:3916–25. [29] O’Keefe J, Dostrovski J. The hippocampus as a spatial map: preliminary evidence from unit activity in the freely moving rat. Brain Res 1971;34:171–5. [30] O’Keefe J, Nadel L. The hippocampus as a cognitive map. Oxford: Clarendon Press; 1978. [31] Roullet P, Sargolini F, Oliverio A, Mele A. NMDA and AMPA antagonist infusions into the ventral striatum impair different steps of spatial information processing in a nonassociative task in mice. J Neurosci 2001;21:2143–9. [32] Sargolini F, Roullet P, Oliverio A, Mele A. Effects of lesions to the glutamatergic afferents to the nucleus accumbens in the modulation of reactivity to spatial and non-spatial novelty in mice. Neuroscience 1999;93:855– 67. [33] Sargolini F, Roullet P, Oliverio A, Mele A. Effects of intra-accumbens focal administrations of glutamate antagonists on object recognition memory in mice. Behav Brain Res 2003;138:153–63. [34] Sargolini F, Florian C, Oliverio A, Mele A, Roullet P. Differential involvement of NMDA and AMPA receptors within the nucleus accumbens in consolidation of information necessary for place navigation and guidance strategy, in mice. Learn Mem 2003;10:285–92. [35] Sesack SR, Pickel VM. In the rat medial nucleus accumbens, hippocampal and catecholaminergic terminals converge on spiny neurons and are in apposition to each other. Brain Res 1990;527:266–79. [36] Setlow B, McGaugh JL. Sulpiride infused into the nucleus accumbens posttraining impairs memory of spatial water maze training. Behav Neurosci 1998;112:603–10. [37] Smith-Roe SL, Sedaghian K, Kelley AE. Spatial learning and performance in the radial arm maze is impaired after N-methyl-d-aspartate (NMDA) receptor blockade in striatal subregions. Behav Neurosci 1999;113:703–17. [38] Steffenach HA, Witter M, Moser MB. Spatial memory in the rat requires the dorsolateral band of the entorhinal cortex. Neuron 2005;45(2):301–13. [39] Sutherland RJ, Rodriguez AJ. The role of the fornix/fimbria and some related subcortical structures in place learning and memory. Behav Brain Res 1989;32:265–77. [40] Tarazi FI, Campbell A, Yeghiayan SK, Baldessarini RJ. Localization of ionotropic glutamate receptors in caudate-putamen and nucleus accumbens septi of rat brain: comparison of NMDA, AMPA, and kainate receptors. Synapse 1998;30:227–35. [41] Upchurch M, Wehner JM. Inheritance of spatial learning ability in inbred mice: a classical genetic analysis. Behav Neurosci 1989;103:1251–8. [42] Wright CI, Groenewegen HJ. Patterns of overlap and segregation between insular cortical, intermediodorsal thalamic and basal amygdaloid afferents in the nucleus accumbens of the rat. Neuroscience 1996;73:359–73.
Research report
Effects of intra-accumbens NMDA and AMPA receptor antagonists on short-term spatial learning in the Morris water maze task Valentina Ferretti a,b,c , Francesca Sargolini a,b,1 , Alberto Oliverio b,c,d , Andrea Mele b,c,d , Pascal Roullet a,∗ a
Centre de Recherches sur la Cognition Animale, Universit´e Paul Sabatier, CNRS-UMR 5169, 118 route de Narbonne, 31062 Toulouse Cedex 9, France b Dipartimento di Genetica e Biologia Molecolare, Universit` a di Roma ‘La Sapienza’, P.le Aldo Moro 5, 00185 Roma, Italy c Centro di Neurobiologia-D. Bovet, Universit` a di Roma ‘La Sapienza’, P.le Aldo Moro 5, 00185 Roma, Italy d Istituto di Neuroscienze, CNR-CERC, Roma, Italy Received 5 September 2006; received in revised form 4 January 2007; accepted 9 January 2007 Available online 13 January 2007
Abstract Glutamatergic transmission within the nucleus accumbens (Nac) is considered to subserve the transfer of different types of information from the cortical and limbic regions. In particular, it has been suggested that glutamatergic afferences from the hippocampus and the prefrontal cortex provide the main source of contextual information to the Nac. Accordingly, several authors have demonstrated that the blockade of glutamate receptors within the Nac impairs various spatial tasks. However, the exact role of the different classes of glutamate receptors within the Nac in short-term spatial memory is still not clear. In this study we investigated the involvement of two major classes of glutamate receptors, NMDA and AMPA receptors, within the Nac in the acquisition of spatial information, using the Morris water maze task. Focal injections of the NMDA antagonist, AP-5 (0.1 and 0.15 g/side), and the AMPA antagonist, DNQX (0.005, 0.01 g/side), were performed before a massed training phase, and mice were tested for retention immediately after. NMDA and AMPA receptor blockade induced no effect during training. On the contrary, injection of the two glutamatergic antagonists impaired spatial localization during the probe test. These data demonstrate an involvement of the Nac in short-term spatial learning. Moreover, they prove that within this structure the short-term processing of spatial information needs the activation of both NMDA and AMPA receptors. © 2007 Elsevier B.V. All rights reserved. Keywords: Glutamate; AP-5; DNQX; Spatial memory; Short-term memory; Mice
1. Introduction In recent years there has been much interest on the possible role of the nucleus accumbens (Nac) in mediating cognitive functions. This structure receives projections from different brain regions such as the hippocampus, the prefrontal cortex and the amygdala [14,17,42], all involved in different kinds of learning [5,10]. In particular the activity of the cortico-hippocampal network (including hippocampal, parahippocampal and prefrontal cortical areas) is essential for spatial
∗
Corresponding author. Tel.: +33 5 61 55 65 69; fax: +33 5 61 55 61 54. E-mail address: [email protected] (P. Roullet). 1 Present address: Laboratoire de Neurobiologie de la Cognition, Universit´ e de Provence, CNRS-UMR 6155, Marseille, France. 0166-4328/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bbr.2007.01.009
navigation [15,26,29,30,38]. Thus, it seems conceivable that the Nac, which is directly connected with those cortical and hippocampal regions might contribute to spatial learning and memory [33,36,39]. This hypothesis is sustained by a certain amount of experimental evidence. For example, cell firing according to location in space, similar to that shown by ‘place cells’ in the hippocampus, has been reported in the Nac [20,24]. Moreover, several studies have provided behavioural evidence that manipulations of the Nac impair performance in spatial learning tasks [2,11,23,31,32,37,39]. However, it is not completely clear which role the Nac might play. Indeed several factors could contribute to spatial navigation: information associated with different sensory modalities, reference systems (i.e. egocentric and allocentric), or motivation. Thus, the impairments observed after lesions or pharmacological manipulation of the Nac could be due in deficits to a variety of
44
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
cognitive abilities (for a discussion, see [2,7,25,37]). A further issue that should be considered is whether the Nac activation may be required during specific stages of spatial information processing—acquisition, memory consolidation and recall. In this regard, microinfusions of selective agonists and antagonists of specific receptor classes are extremely useful tools. In fact, they allow manipulation of the receptors’ activity within a limited time-window, thus permitting precise analysis of the effects induced during the different stages of memory. Furthermore, they enable the investigation of the specific role of the different receptor subtypes. Receptor binding studies have shown high levels of the different subtypes of glutamate receptors on the Nac neurons [1,40]. In particular, both NMDA and AMPA receptors are localized on the medium spiny neurons within this structure [6,35]. In light of the fact that the putative neurotransmitter for most of the inputs impinging on the Nac is glutamate, it has been suggested that NMDA and AMPA receptors could mediate the transmission of information to the Nac [32,37]. Interestingly, recent data point to a functional dissociation between the NMDA and AMPA receptors within the Nac in the modulation of spatial memory. For example, it has been demonstrated that NMDA, but not AMPA receptors are necessary for consolidating spatial information in an open field task, designed to estimate the ability of rodents to detect and react to object displacement [31]. Similar conclusions have been drawn on the basis of the effects observed using the Morris water maze [34], as well as non-spatial tasks, such as object recognition [33] and one-trial inhibitory avoidance [7]. Unfortunately, it is still not clear whether a similar difference between the two receptors of the Nac also exists in the short-term processing of spatial information. Recent reports provide evidence that AMPA but not NMDA receptors are involved in the acquisition of the one-trial inhibitory avoidance and object recognition tasks [7,33], thus suggesting that AMPA but not NMDA receptor activation is needed for the acquisition/encoding of information. On the other hand, a study by Maldonado-Irizarry and Kelley [23] demonstrated that pre-training focal injections within the Nac of both NMDA and AMPA receptor antagonists impaired rats in a spatial food-search task, but NMDA receptor blockade seemed more disruptive. Similarly, intra-accumbens administration of NMDA, but not AMPA antagonists, affect the acquisition of a Pavlovian approach task [8]. The aim of this study was therefore to further investigate the involvement of the two classes of glutamate receptors located in the Nac on short-term processing of spatial information. For this purpose, focal administrations of AP-5 and DNQX (NMDA and AMPA receptor antagonists, respectively) were performed intra-accumbens immediately before training, in the Morris water maze. The behavioural paradigm assesses the ability of mice to navigate in an environment using distal cues and requires the constitution of complex associations among stimuli. In addition, the entire behavioural procedure lasted 45 min, thus excluding the possibility that mice have to rely on long-term consolidated information in order to navigate in the maze.
2. Materials and methods 2.1. Animals A total of 95 CD1 male mice (IFFA CREDO, Lyon, France) were used in this study. Upon arrival, the animals were housed in groups of five in standard breeding cages (21 cm × 21 cm × 12 cm) placed in a rearing room at a constant temperature (22 ± 1 ◦ C) under diurnal conditions (light–dark: 08:00–20:00), with food and water ad libitum. At the time of surgery, they were approximately 8–9-weeks-old and their weights ranged from 35 to 40 g. All experiments were run in the afternoon, between 14:00 and 18:00 h. Every possible effort was made to minimise animal suffering and all procedures were in strict accordance with European community laws and regulations on the use of animals in research and NIH guidelines on animal care.
2.2. Surgery Mice underwent surgery 1 week after their arrival. They were anaesthetised with chloral hydrate (400 mg/kg) and placed in a stereotaxic frame (David Kopf Instruments, Tujunga, CA, USA) with mouse adapter and lateral ear bars. After placing the animals on the stereotaxic apparatus the head skin was cut longitudinally and bilateral guide cannulae (0.5 mm in diameter) were fixed on the calvarium with dental acrylic. The following coordinates with lambda and bregma in the same horizontal plane were used: anterior to bregma, +1.7 mm; lateral to midline, ±1 mm; ventral from the dura, +2.3 mm, according to Franklin and Paxinos [13]. Mice were then left in their home cage for a recovery period of 1 week.
2.3. Apparatus The circular swimming pool (110 cm in diameter and 30 cm in height) was made of ivory-coloured PVC, filled with water (25 ± 1 ◦ C) made opaque with Lytron 631, to 15 cm below the edge of the wall. Four start positions were located equidistantly around the edge of the maze, dividing it into four equal quadrants. During training, a circular goal platform painted white (9 cm of diameter) was laid in the centre of one quadrant, 15 cm from the wall. The platform had a rough (metal grid) surface providing sufficient grip for the animals to climb on top of it. The apparatus was placed in a separate room and surrounded by white curtains containing several extra-maze cues. It was illuminated by a white light (60 W) and had a video camera placed overhead and connected to a video recorder and monitor.
2.4. Procedure The general procedure consisted of three different phases: a familiarization phase, a massed training phase and a probe test [12,34]. Several extra-maze visual cues, approximately 50–100 cm away from the pool, were attached to the walls surrounding the apparatus. The mice were required to navigate to the invisible platform using the spatial cues available in the room. On the first day, the mice were individually submitted to a single familiarization session of three trials, with the platform protruding 0.5 cm above the surface of the water. This familiarization procedure is used to allow the mice to acquire the procedural components of the task (the presence of a platform, the handling), which is especially important when a massed training protocol is used. The session started with the mouse standing on the platform for 60 s. At the beginning of each trial the mice were introduced into the maze facing the wall at one of the four designated starting points and allowed to swim freely or until they reached the platform. Mice failing to find the platform within a fixed period of 60 s were gently guided by hand to the platform and a maximum escape latency of 60 s was recorded. After the animals climbed onto the platform, they were allowed to remain on it for additional 60 s, and subsequently replaced in the maze from a different starting position. The starting positions were determined in a pseudorandom order, such that each was only used once in a single session. The same spatial location of the platform was used for all the animals. Training started the next day. The mice were submitted to four consecutive sessions of three trials. The procedure was the same as for the familiarization phase, except for the platform, which was submerged 0.5 cm beneath the surface of the water. The platform was left in the same position during familiarization and
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49 training. Animals were introduced into the maze from different starting points and the sequence of starting locations randomised. Testing was conducted in sets of two mice, such that following a session the first mouse was placed in a holding cage while the second was run; following that session the first mouse was re-tested. Immediately after the last training session, the mice were submitted to a single probe test. The platform was removed and the mice, starting from the centre of the pool, were allowed 60 s to search for the platform.
2.5. Drugs Selective and reversible NMDA and AMPA receptor blockade was obtained through direct infusions into the Nac of (−)-AP-5 (0.1 and 0.15 g/side) and DNQX (0.005 and 0.01 g/side), competitive antagonists of the NMDA and the AMPA receptors respectively. AP-5 was dissolved in saline solution (0.9% NaCl in distilled water), and DNQX was dissolved in a solution of 2% DMSO in distilled water. Mice were assigned to a drug treatment group and received injections 10 min before the first training trial. Drug injections were performed by lowering into the guide cannula a 9 mm long injector connected with polyethylene tubing to a 1.0 l Hamilton syringe. The volume injected was always 0.2 l. Each injection lasted 2 min and the injector was left in place for an additional minute to allow diffusion. Before the injections mice were manipulated daily in order to habituate them to handling. On the injection day they were hand-held to insert the injector while during the injection they where allowed to move freely in the cage. Drug-treated animals were always compared to control mice injected with the same volume of vehicle solution, under the same conditions.
2.6. Data collection and statistics Data collection was performed using a video camera suspended above the pool and interfaced with a video tracking system (Ethovision® 2.3, Noldus Information Technology, Wageningen, The Netherlands). Several parameters of behavioural performance were recorded. The average swimming speed was calculated during all sessions. During training sessions, path length to mount onto the platform was recorded in each trial. During the probe test the number of annulus crossings was scored, i.e. the number of times a mouse crossed an ideal circle (14 cm of diameter) located around each of the four possible platform positions in the four quadrants. We used this behavioural parameter instead of the percent time spent in the training quadrant during the probe test, which has been reported to overestimate animals’ performance in term of spatial ability [4]. All data were analysed by analysis of variance (ANOVA). For the training phase, the path length for each session (average of three trials) was analysed with a repeated measure ANOVA analysis (between factor: treatment, three levels; within factor: sessions, four levels). When a significant session effect was found, a further repeated measure ANOVA analysis (within factor: sessions, four levels) was conducted for each group. The number of annulus crossings in each of the four quadrants during the probe test was analysed using a two-way ANOVA design (first factor: treatment, three levels; second factor: quadrants, four levels). A one-way (treatment: three levels) ANOVA design was employed to analyse possible differences in swimming speed. Post hoc multiple comparisons analysis was carried out when allowed, using Tukey’s honestly significant difference (HSD) test.
2.7. Histology At the completion of the experiment, mice were sacrificed by an overdose of chloral hydrate, the brains were removed and frozen at −20 ◦ C. Cannula placements were determined by examination of serial coronal sections (25 m) stained with Cresyl Violet.
3. Results 3.1. Cannula placement verification Fig. 1 shows a schematic representation of injector placements for the two experiments. Only animals showing a correct
45
placement within the Nac were included in the statistical analysis. 3.2. Experiment 1: NMDA receptors blockade Fig. 2A shows the path length to reach the platform during the training phase in saline (n = 10), AP-5 0.1 (n = 9) and AP-5 0.15 (n = 9) groups. Both saline and AP-5 groups decreased their path length to find the platform along the sessions. The repeated measures ANOVA analysis showed a significant session effect (F3,75 = 16.98, p < 0.001), no treatment effect (F2,25 = 1.789, p = 0.188) and no interaction between the two factors (F6,75 = 0.307, p = 0.931). A further ANOVA analysis showed a significant effect of session for each of the three groups (saline: F3,27 = 4.495, p = 0.011; AP-5 0.1: F3,24 = 4.880, p = 0.009; AP-5 0.15: F3,24 = 8.534, p = 0.001). During probe test, similarly to that observed during training sessions, the path lengths to reach the annulus in the target quadrant did not differ between saline and AP-5-injected mice (F2,25 = 0.002, p = 0.998) (data not shown). On the other hand, control animals crossed the annulus located in the correct quadrant significantly more than those located in the remaining three quadrants (Fig. 2B). Pre-training AP-5 injections significantly decreased the number of annulus crossings in the correct quadrant. Two-way ANOVA showed a significant quadrant effect (F3,100 = 19.089, p < 0.001), no treatment effect (F2,100 = 0.047, p = 0.954), and a significant interaction between the two factors (F6,100 = 3.050, p = 0.009) indicating a different profile of quadrants’ exploration in the three groups of mice. Post hoc comparison revealed a significant difference between the correct quadrant and the other three quadrants for control animals (p < 0.001). For both groups of AP-5 treated mice, no significant difference among all four quadrants was revealed even though, in the AP-5 0.1 group, there was a marginal effect on the difference between the correct quadrant and the left and opposite quadrants (p = 0.055). Note that post hoc comparison also showed a significant difference for annulus crossing in the correct quadrant between the control animals and the animals treated with the highest dose of AP-5 (p = 0.025). A further comparison between the annulus crossing in the correct quadrant in the first and in the second halves of the probe test revealed a significant decrease over time (F3,150 = 2.35, p = 0.034) but no difference among the experimental groups (F6,150 = 1.595, p = 0.152) (data not shown). Table 1 shows the effects of pre-training AP-5 injections on swimming speed during the probe test. No significant difference was observed between saline and AP-5-injected mice (F2,25 = 1.492, p = 0.244). 3.3. Experiment 2: AMPA receptors blockade Fig. 3A shows the path length to reach the platform during the training phase in saline (n = 9), DNQX 0.005 (n = 9) and DNQX 0.01 (n = 8) groups. Both vehicle and DNQX groups decreased the distance to find the platform across the four sessions. The repeated measure ANOVA analysis showed a significant session effect (F3,69 = 19.93, p < 0.001), no treatment
46
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
Fig. 1. Schematic representation of cannula placements in the two experiments. Each symbol represents the site of injection for one animal. (A) Experiment 1: pre-training AP-5 injections: (*) saline; () AP-5 0.1 g/side; (䊉) AP-5 0.15 g/side. (B) Experiment 2: pre-training DNQX injections. (*) vehicle; () DNQX 0.005 g/side; () DNQX 0.01 g/side.
effect (F2,23 = 1.288, p = 0.295), and no interaction between the two factors (F6,69 = 0.435, p = 0.853). A further ANOVA analysis showed a significant effect of session for each of the three groups (vehicle: F3,24 = 11.526, p < 0.001; DNQX 0.005: F3,24 = 4.778, p = 0.009; DNQX 0.01: F3,21 = 5.331, p = 0.007). Fig. 3B represents the number of annulus crossing in the four quadrants during the probe test, by vehicle- and DNQX-injected mice. The control group as well as the 0.005 DNQX-treated mice crossed the annulus in the correct quadrant significantly Table 1 Mean swimming speed ± S.E.M. in probe test, for the two experiments Group
Speed
Experiment 1 Saline (n = 10) AP-5 0.1 g/side (n = 9) AP-5 0.15 g/side (n = 9)
20.05 ± 2.86 22.64 ± 3.10 21.64 ± 2.87
Experiment 2 Vehicle (n = 9) DNQX 0.005 g/side (n = 9) DNQX 0.01 g/side (n = 8)
20.87 ± 1.59 20.72 ± 1.65 19.93 ± 1.15
more often, compared to the remaining three quadrants. The same did not occur for the mice injected with the highest dose of DNQX. The two-way ANOVA showed a significant quadrant effect (F3,92 = 20.158, p < 0.001) and no treatment effect (F2,92 = 0.030, p = 0.970). The three groups of mice displayed a different profile of exploration as demonstrated by the interaction between the quadrant and treatment factors (F6,92 = 2.365, p = 0.036). Post hoc comparison revealed a significant difference between the correct quadrant and the remaining three quadrants for both control animals (correct versus right and opposite p < 0.001; correct versus left p = 0.028) and mice injected with the lower dose of DNQX (correct versus left and opposite p < 0.001; correct versus right p = 0.028). For the DNQX 0.01 group, we observed a significant difference only between the correct and the opposite quadrant (p = 0.039), but no difference between the correct and the two adjacent quadrants. We performed a further analysis on the annulus crossings during the first and the second halves of the probe test (data not shown). No significant difference was found among the experimental groups (F6,132 = 0.721, p = 0.633). The number of annulus crossings in the target quadrant decreased over time, however this did not reach statistical significance (F3,132 = 2.52, p = 0.061). Sim-
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
Fig. 2. Experiment 1: effect of pre-training intra-accumbens AP-5 administrations on mouse performance during training and probe test, in the Morris water maze. (A) Mean of path length (cm) during training sessions in mice injected with saline, AP-5 0.1 g/side and AP-5 0.15 g/side. (B) Mean number of annulus crossings in the four quadrants, during the probe test for saline, AP-5 0.1 g/side groups and AP-5 0.15 g/side groups. (C) representative paths taken by animals treated with saline, AP-5 0.1 g/side and AP-5 0.15 g/side. * p ≤ 0.05 correct vs. opposite, right, left quadrants, within groups. + p ≤ 0.05, correct quadrant saline vs. AP-5 groups.
ilarly to that observed in the first experiment, an ANOVA on the path length to first entry in annulus revealed no treatment effect (F2,22 = 1.683, p = 0.209). Table 1 reports the effects of pre-training DNQX injections on swimming speed during probe test. No significant differences were observed between saline and DNQX-injected mice (F2,23 = 0.106, p = 0.900). 4. Discussion The present study demonstrates that intra-accumbens focal administrations of the NMDA competitive antagonist, AP-5, and the AMPA antagonist, DNQX, impair the ability of mice to find the hidden platform in a short-term version of the Morris
47
Fig. 3. Experiment 2: effect of pre-training intra-accumbens DNQX administrations on mouse performance during training and the probe test, in the Morris water maze. (A) Mean of path length during training sessions in mice injected with vehicle, DNQX 0.005 g/side groups and DNQX 0.01 g/side groups. (B) Mean number of annulus crossing in the four quadrants for vehicle, DNQX 0.005 g/side groups, and DNQX 0.01 g/side groups. (C) Representative paths taken by animals treated with vehicle, DNQX 0.005 g/side and DNQX 0.01 g/side. (*) p ≤ 0.05 correct vs. opposite, right, left quadrants, within groups.
water maze [12,27,34]. The results demonstrate that the Nac is necessary for short-term place navigation, and this requires the activation of both glutamate receptor subtypes. In both experiments, control and drug-injected mice progressively decreased the path length to reach the platform on the training phase. During the probe test, control mice crossed the annulus more often in the quadrant where the platform was previously located. On the contrary AP-5 and DNQX administrations dose-dependently decreased the number of annulus crossings in the correct quadrant. Interestingly in the two experiments a decrease in path length along the sessions was always observed independently of the effect on annulus crossing. This dissociation between the two measures was also observed within the probe trial, in which the path length to reach the annulus did not differ among experimental groups. Several authors have reported
48
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
a similar dissociation on the two parameters following different CNS manipulations. For example, it has been shown that temporary or permanent inactivation of the hippocampus, does not affect escape latency on training days but induces a significant deficit on the probe test, in both rats [28] and mice [19]. It has been suggested by different authors that a decrease in the path length or in the latencies does not necessarily represent a good assessment of spatial navigation abilities [19]. In fact, path length or latencies could improve with the acquisition of the procedural components of the task. On the contrary, the time spent or the number of annulus crossings in the target quadrant on probe test are considered to be directly linked to the acquisition of the correct spatial localization of the platform and could give a more appropriate measurement of spatial learning [19,21,27,41]. In our experiments, AP-5 and DNQX injected mice decreased the number of correct annulus crossings during the probe test but the learning performances along the training session is the same in the different groups showing that experimental mice certainly used another strategy to find the platform. The use of a different strategy seems also confirmed by the differences in the representative paths of control and of experimental animals during the probe test (Figs. 2C and 3C). A further caveat to be considered is that the number of annulus crossings during the probe test may represent an index of behavioural flexibility. According to this view, it is possible that during the probe test, control mice actively search for the platform in the correct spatial location and persist in their fruitless search inside the target sector. In contrast, mice injected with the two glutamate antagonists first searched for the platform in the correct quadrant and quickly extend their search to the entire pool. However, this does not seem to be the case because we observed a similar decrease in the number of annulus crossings during the first and the second halves of the probe test in all groups. This demonstrates that NMDA and AMPA receptor blockade in the Nac do not directly affect the behavioural flexibility. Rather, the activation of both NMDA and AMPA receptors within this structure is needed to maintain information for short periods of time. The results of the present study confirm but also extend previous findings in the literature. Several reports have investigated the role of the Nac, and in particular the contribution of AMPA and NMDA receptors within this structure in the different steps of memory formation. They provided convincing evidence on the role of both receptor subtypes in memory consolidation and retrieval. In fact, independently of the behavioural test, posttraining administrations of NMDA but not AMPA antagonist impair the ability to perform the task 24 h later [7,31,34], demonstrating that NMDA receptor activation is needed in the early stages of consolidation. On the contrary, AMPA but not NMDA antagonist injections before testing affect performance in tasks based on a large variety of cognitive abilities [7,8,31,37]. This suggests that AMPA but not NMDA receptors are involved in information retrieval. Based on these observations, the current hypothesis states that NMDA receptor activation in the Nac is needed when plastic changes are required in order to maintain and stabilize the memory trace (i.e. consolidation), while AMPA receptors are necessary when this information needs to be utilized.
In the light of the short time lapse between the acquisition phase and the test, the procedure used in the present study did not require any consolidation process for the task to be performed. In fact, even though on the one hand it seems conceivable that the consolidation processes started as soon as the animals began training, on the other hand distinct biochemical and neuroanatomical substrates underlie short and long-term memory processes [16]. It is thus unlikely that long-term processes are engaged in animals that have to provide an efficient response in a short delay. For the same reason, based on the current hypothesis on memory, the involvement of an active process of retrieval can also be excluded. Rather, in order to perform the task, the animals have to construct a spatial map of the environment, keep this information on a short-term buffer and interface it with the motor system. Pre-training administrations do not allow to dissect these processes and in the present study we did not perform any pre-test administration to distinguish encoding and maintenance from the use of information. However, as mentioned before, NMDA receptor activation does not seem to be necessary for using information in order to guide behaviour [3,8,31,37]. On the contrary, pre-test focal administrations of AMPA antagonists both in the accumbens [7,31] and in the hippocampus [3] have been demonstrated to impair performance in different behavioural tasks. Therefore, it seems conceivable to hypothesise that NMDA receptor activation within the accumbens is needed only to acquire and maintain spatial information. On the contrary, AMPA receptors within this structure may be required either to interface the cognitive process with the motor system or for both the acquisition-maintenance and the use of such information. Experimental evidence seems to support this latter hypothesis, but manipulations allowing better time resolution will be needed to address this issue more fully. A wide body of evidence is now available in the literature as regards to the possible biological mechanisms underlying long-term storage of information [9,18,22]. Instead, less is known about the putative changes necessary for short-term maintenance. It is interesting in this regard to compare the effects induced by intra-accumbens pre-training administrations of NMDA and AMPA antagonists in the Morris water maze with those observed in one-trial inhibitory avoidance task [7]. In fact, blockade of AMPA antagonists affected the ability to solve both tasks, while the NMDA antagonist impaired performance only in the water maze. This suggests that within the nucleus accumbens the need for the activation of the NMDA but not the AMPA receptors depends upon the kind of information to be processed. Further neuropharmacological studies based on a larger number of tasks requiring different cognitive processes will help to elucidate this issue. However, on the basis of these two studies solely, it could be speculated that during short-term information processing, the NMDA receptor might come into play only when complex associations among stimuli are required, such as in the Morris water maze. Acknowledgements This study was supported by an I.S.E.R.M. fellowship to FS, a fellowship from the University of Rome ‘La Sapienza’ to VF,
V. Ferretti et al. / Behavioural Brain Research 179 (2007) 43–49
F.I.R.B. and PRIN grants by MIUR to AO and AM, and a grant from the University of Rome ‘La Sapienza’ to AO. We thank Peter Winterton and Anne-Marie Mc Gauran for proofreading this manuscript. References [1] Albin RL, Makowiec RL, Hollingsworth Z, Dure IVLS, Penney JB, Young AB. Excitatory amino acid binding sites in the basal ganglia of the rat: a quantitative autoradiographic study. Neuroscience 1991;46:35–48. [2] Annett LE, McGregor A, Robbins TW. The effects of ibotenic acid lesions of the nucleus accumbens on spatial learning and extinction in the rat. Behav Brain Res 1989;31:231–42. [3] Bast T, da Silva BM, Morris RG. Distinct contributions of hippocampal NMDA and AMPA receptors to encoding and retrieval of one-trial placememory. J Neurosci 2005;25:5845–56. [4] Blokland A, Geraerts E, Been M. A detailed analysis of rats’ spatial memory in a probe trial of a Morris task. Behav Brain Res 2004;154(1):71–5. [5] Burns LH, Annett L, Kelley AE, Everitt BJ, Robbins TW. Effects of lesions to amygdala, ventral subiculum, medial prefrontal cortex, and nucleus accumbens on the reaction to novelty: implication for limbic-striatal interactions. Behav Neurosci 1996;110:60–73. [6] Daw NW, Stein PS, Fox K. The role of NMDA receptors in information processing. Annu Rev Neurosci 1993;16:207–22. [7] De Leonibus E, Costantini VJ, Castellano C, Ferretti V, Oliverio A, Mele A. Distinct roles of the different ionotropic glutamate receptors within the nucleus accumbens in passive-avoidance learning and memory in mice. Eur J Neurosci 2003;18(8):2365–73. [8] Di Ciano P, Cardinal RN, Cowell RA, Little SJ, Everitt BJ. Differential involvement of NMDA, AMPA/kainate, and dopamine receptors in the nucleus accumbens core in the acquisition and performance of pavlovian approach behavior. J Neurosci 2001;21(23):9471–7. [9] Dudai Y. Molecular bases of long-term memories: a question of persistence. Curr Opin Neurobiol 2002;12:211–6. [10] Floresco SB, Seamans JK, Phillips AG. Differential effects of lidocaine infusions into the ventral CA1/subiculum or the nucleus accumbens on the acquisition and retention of spatial information. Behav Brain Res 1996;81:163–71. [11] Floresco SB, Seamans JK, Phillips AG. Selective roles for hippocampal, prefrontal cortical and ventral striatal circuits in radial-arm maze tasks with or without a delay. J Neurosci 1997;17:1880–90. [12] Florian C, Roullet P. Hippocampal CA3-region is crucial for acquisition and memory consolidation in Morris water maze task in mice. Behav Brain Res 2004;154(2):365–74. [13] Franklin BJ, Paxinos G. The mouse brain in stereotaxic coordinates. San Diego: Academic Press; 1997. [14] Friedman DP, Aggleton JP, Saunders RC. Comparison of hippocampal, amygdala, and perirhinal projections to the nucleus accumbens: combined anterograde and retrograde tracing study in the Macaque brain. J Comp Neurol 2002;450:345–65. [15] Hok V, Save E, Lenck-Santini PP, Poucet B. Coding for spatial goals in the prelimbic/infralimbic area of the rat frontal cortex. Proc Natl Acad Sci USA 2005;102(12):4602–7. [16] Izquierdo I, Medina JH, Vianna MR, Izquierdo LA, Barros DM. Separate mechanism for short- and long-term memory. Behav Brain Res 1999;103(1):1–11. [17] Jay TM, Witter MP. Distribution of hippocampal CA1 and subicular efferents in the prefrontal cortex of the rat studied by means of anterograde transport of Phaseolus vulgaris leucoagglutinin. J Comp Neurol 1991;313:574–86. [18] Kandel ER. The molecular biology of memory storage: a dialogue between genes and synapses. Science 2001;294:1030–8. [19] Lassalle JM, Bataille T, Halley H. Reversible inactivation of the hippocampal mossy fiber synapses in mice impairs spatial learning, but neither consolidation nor memory retrieval, in the Morris navigation task. Neurobiol Learn Mem 2000;73:243–57.
49
[20] Lavoie AM, Mizumori SJ. Spatial, movement- and reward-sensitive discharge by medial ventral striatum neurons of rats. Brain Res 1994;638:157–68. [21] Lipp HP, Wolfer DP. Genetically modified mice and cognition. Curr Opin Neurobiol 1998;8:272–80. [22] MacGaugh JL. Memory—a century of consolidation. Science 2000;287: 248–51. [23] Maldonado-Irizarry CS, Kelley AE. Excitatory amino acid receptors within nucleus accumbens subregions differentially mediate spatial learning in the rat. Behav Pharmacol 1995;6:527–39. [24] Martin PD, Ono T. Effects of reward anticipation, reward presentation, and spatial parameters on the firing of single neurons recorded in the subiculum and nucleus accumbens of freely moving rats. Behav Brain Res 2000;116:23–38. [25] Mele A, Avena M, Roullet P, De Leonibus E, Mandillo S, Sargolini F, et al. Nucleus accumbens dopamine receptors in the consolidation of spatial memory. Behav Pharmacol 2004;15(5–6):423–31. [26] Morris RGM, Garrud P, Rawlins JNP, O’Keefe J. Place navigation in rats with hippocampal lesions. Nature 1982;297:681–3. [27] Morris RGM. Developments of a water-maze procedure for studying spatial learning in the rat. J Neurosci Meth 1984;11:47–60. [28] Moser E, Moser MB, Andersen P. Spatial learning impairment parallels the magnitude of dorsal hippocampal lesions, but is hardly present following ventral lesions. J Neurosci 1993;13:3916–25. [29] O’Keefe J, Dostrovski J. The hippocampus as a spatial map: preliminary evidence from unit activity in the freely moving rat. Brain Res 1971;34:171–5. [30] O’Keefe J, Nadel L. The hippocampus as a cognitive map. Oxford: Clarendon Press; 1978. [31] Roullet P, Sargolini F, Oliverio A, Mele A. NMDA and AMPA antagonist infusions into the ventral striatum impair different steps of spatial information processing in a nonassociative task in mice. J Neurosci 2001;21:2143–9. [32] Sargolini F, Roullet P, Oliverio A, Mele A. Effects of lesions to the glutamatergic afferents to the nucleus accumbens in the modulation of reactivity to spatial and non-spatial novelty in mice. Neuroscience 1999;93:855– 67. [33] Sargolini F, Roullet P, Oliverio A, Mele A. Effects of intra-accumbens focal administrations of glutamate antagonists on object recognition memory in mice. Behav Brain Res 2003;138:153–63. [34] Sargolini F, Florian C, Oliverio A, Mele A, Roullet P. Differential involvement of NMDA and AMPA receptors within the nucleus accumbens in consolidation of information necessary for place navigation and guidance strategy, in mice. Learn Mem 2003;10:285–92. [35] Sesack SR, Pickel VM. In the rat medial nucleus accumbens, hippocampal and catecholaminergic terminals converge on spiny neurons and are in apposition to each other. Brain Res 1990;527:266–79. [36] Setlow B, McGaugh JL. Sulpiride infused into the nucleus accumbens posttraining impairs memory of spatial water maze training. Behav Neurosci 1998;112:603–10. [37] Smith-Roe SL, Sedaghian K, Kelley AE. Spatial learning and performance in the radial arm maze is impaired after N-methyl-d-aspartate (NMDA) receptor blockade in striatal subregions. Behav Neurosci 1999;113:703–17. [38] Steffenach HA, Witter M, Moser MB. Spatial memory in the rat requires the dorsolateral band of the entorhinal cortex. Neuron 2005;45(2):301–13. [39] Sutherland RJ, Rodriguez AJ. The role of the fornix/fimbria and some related subcortical structures in place learning and memory. Behav Brain Res 1989;32:265–77. [40] Tarazi FI, Campbell A, Yeghiayan SK, Baldessarini RJ. Localization of ionotropic glutamate receptors in caudate-putamen and nucleus accumbens septi of rat brain: comparison of NMDA, AMPA, and kainate receptors. Synapse 1998;30:227–35. [41] Upchurch M, Wehner JM. Inheritance of spatial learning ability in inbred mice: a classical genetic analysis. Behav Neurosci 1989;103:1251–8. [42] Wright CI, Groenewegen HJ. Patterns of overlap and segregation between insular cortical, intermediodorsal thalamic and basal amygdaloid afferents in the nucleus accumbens of the rat. Neuroscience 1996;73:359–73.