Behavioural Brain Research, 47 (1992) 151-157 9 1992 Elsevier Science Publishers B.V. All rights reserved. 0166-4328/92/S05.00
t5t
BBR 01268
Effects of intraventricular infusion of the N-methyl-D-aspartate ( N M D A ) receptor antagonist AP5 on spatial memory of rats in a radial arm maze Johan J. Bolhuis and Ian C. Reid Department of Pharmacologo', University of Edinburgh Medical School, Edinburgh (UK) (Received 29 July 1991) (Revised version received 4 November 1991) (Accepted 14 January 1992) Key words: N-Methyl-D-aspartic acid receptor; AP5; Spatial memory; Learning; Radial maze; Rat
Rats were trained to asymptotic performance in an 8-arm radial maze. They then received chronic intraventricular infusion ofeither artificial CSF or the N-methyl-D-aspartate (NM DA) receptor antagonist D-2-amino-5-phosphonopentanoic acid (AP5), at a concentration (30 mM)that has been shown previously to prevent the induction of long-term potentiation in the dentate gyrus of the hippocampus in vivo. Subsequently the rats received another 9 trials in the maze in a quasi-random order, 3 uninterrupted trials, and another 6 trials each with mid-trial delays of 5, 20 or 60 min during which the animals were placed in their home cage. The mean number of errors for the AP5 rats did not differ significantly from that of the controls in the uninterrupted trials throughout the experiment, nor did it differ from that of the controls in any of the 3 delayed trials when these were first introduced. However,'the control animals performed better at the longer delays when these were introduced for the second time, whilst there was no such improvement (but rather a deterioration) for the AP5 animals. The impairment of performance in the AP5 rats during the second block ofdelayed trials was significant, and independent of the length of the delay. These results show that NMDA receptor blockade does not impair working memory in the radial maze per se, but that it does prevent an improvement of working memory persistence with further training.
INTRODUCTION There is increasing evidence that a particular class o f excitatory amino acid receptors, the N-methyl-D-aspartate ( N M D A ) receptors, is involved in changes in the efficacy o f synapses. Several studies 6,~ i.~2 have shown that N M D A receptors play a role in a particular form o f artificially induced synaptie plasticity, long-term potentiation ( L T P ) L It has been proposed that the same, or similar mechanisms to those underlying L T P (involving N M D A receptors) might be involved in learning and m e m o r y storage 2.t7"z~ Morris et al. 24 showed that intraventricular infusion o f the competitive N M D A receptor antagonist D,L-2-amino-5-phosphonopentanoic acid (AP5) blocked the induction o f L T P in vivo and impaired place navigation learning in an open field water maze. Visual
Correspondence: J.J. Bolhuis, University of Cambridge, Department of Zoology, Downing Street, Cambridge CB2 3EJ, UK.
discrimination learning in the same apparatus was not impaired. Also, retention o f previously acquired spatial information was not affected by the drug 23. Davis et al. 9 have subsequently demonstrated a parallel d o s e response curve for the action o f AP5 with regard to blockade o f the induction o f L T P and the impairment o f spatial learning. Recent studies have suggested effects o f AP5 in a differential responding to low rates ( D R L ) paradigm 32, in some forms of olfactory learning 3~ and in taste-potentiated o d o u r aversion 7. It has been suggested that the effects o f AP5 and o f certain non-competitive N M D A antagonists on learning and m e m o r y in rats are mainly due to their effects on hippocampal N M D A receptors 18"22-25"28. The reason for this belief is that, on the one hand, there is a particularly high density o f these receptors in this structure 2~, and on the other hand the behavioural effects o f the drugs mirror those o f hippocampal lesions 24"28"32. Furthermore, Morris et al. 25 have shown that local infusion o f AP5 restricted to the hippocampus, is sufficient to obtain an impairment in the
152 open field water maze and to block the induction of LTP. The present study was designed to investigate the effects of intraventricular infusion of AP5 in the radial a r m maze 27, a spatial working memory 26 task that has been shown to be sensitive to hippocampal damage 26. By varying the retention interval 3"4, possible effects of N M D A receptor blockade on retention were studied.
MATERIALS AND METHODS
SllbjecIs Eighteen male hooded Lister rats were used, weighing 240-340 g at the beginning of the experiment. The animals were kept in individual cages in a room with a light/dark regimen of 14/10 (lights on from 08.00 to 22.00 h). Seven days before the start of the experiment, the rats were subjected to a restricted feeding schedule, 1-2 h feeding each day after the end o f the last trial, to maintain the animals at 85-90 ~ of free-feeding weight. Water was available ad libitum throughout the experiment.
Apparatus An 8-arm radial maze was used, similar to that introduced by Olton and Samuelson 27, and used in previous studies by Bolhuis et al. 3-5. It consisted of a central platform (diameter 35 cm) from which 8 arms (96 cm long, 6 cm wide) radiated. Each of the arms was surrounded by a 1-cm high wooden ledge. Five cm from the end of each arm was a hole with a plastic foodcup (diameter 2 cm, I cm deep). Around the edge of the central platform was a 16-cm high wall of wooden segments, with an opening at the beginning of each arm, in which transparent Perspex guillotine doors were placed that could be operated by means of overhead lines. The maze was made of wood and painted matt grey. It was placed 1 m above floor level on a table that could be rotated.
Procedure On the first day of the experiment (day - 15) the rats were placed individually on the central platform, with all eight doors raised. The animal was left in the maze to explore and retrieve raisins that were scattered throughout the apparatus and in the food cups. The next two days, the rats were allowed to explore the maze and retrieve 200 Id of a 23% sucrose solution that was placed in each of the foodcups. After these 2 trials, the guillotine doors were operated during each trial. At the beginning of a trial, the rat was placed on the central platform, with all doors closed. After 10 s, all doors
were raised simultaneously and the rat was allowed to enter an arm. All doors were then lowered, except for the door leading to the arm that the rat was visiting. After the rat had returned to the central platform, this door was lowered too. After 5 s all doors were raised simultaneously and the rat was allowed again to enter an arm. This procedure was repeated until the rats had retrieved all 8 sucrose rewards. The use of doors in this way has been shown to prevent simple turning strategies ofchoosing adjacent arms 3. From the 2nd day until the 9th day of the experiment (day - 14 to - 7 ) the rats received two trials per day, with an interval between trials of at least 2 h, to prevent possible proactive interference 3. Re-entry of an arm with all four paws that had been visited previously in the same trial was considered to be an error. On day - 5, the rats received only one .trial, with a mid-trial interval of 5 min. When the animal had returned to the central platform after the fourth choice, it was removed from the maze and placed back in its home cage. After 5 min it was placed on the central platform again and after 5 s the animal was allowed to complete its trial. This procedure was repeated on day - 4 , but now the maze was rotated 180 ~ during the mid-trial interval. The sucrose rewards were placed in the four arms in the spatial position that the rat had not visited in the first four choices of the trial. Thus, if the animal was guided by extra-maze cues only, its performance would be similar to that in the non-rotation trial. If, on the other hand, intra-maze cues such as odours or visual aspects of the maze would be important in guiding the animal, performance could be expected to deteriorate, compared to that in the non-rotation trial 3. There was no significant difference between the mean number of errors in choices 5 - 8 (random choice level -- 2 errors) in the non-rotation trial (0.50 + 0.13) and that in the rotation trial (0.63 + 0.20), indicating that the animals were not guided by intra-maze c u e s 3. In order to prevent the use ofintra-maze cues, the maze was rotated 180 ~ during the mid-trial intervals in the postoperative delayed trials, replacing the sucrose reward in the same (unvisited) spatial locations. From the 4th day after surgery (see below) onwards, the rats received another 9 trials, one trial per day. The animals received uninterrupted trials on days 4, 8 and 12, trials with a mid-trial delay of 5 min at days 5 and 10, with a 20-rain delay at days 6 and 9, and with a 60-min delay at days 7 and 11. Thus, the order of postoperative trials was: 0-5-20-60-0-20-5-60-0min mid-trial delay. During the mid-trial delays the maze was rotated 180 ~ as described above. The first 3 delayed trials (days 5-7) are called Block 1 and the second 3 delayed trials (days 9-11) are called Block 2.
153
Surgery For 2 days before surgery (days - 2 and - 1) the rats had ad tibitum access to food. On day 0, the rats were anaesthetized with tribromoethanol (Avertin; 29.9 g/l, I0 ml/kg, i.p.) and placed in a stereotaxic apparatus, with the skull surface between bregma and lambda in a horizontal plane. An incision was made in the skin on the head and a single hole was drilled in the skull, 0.9 mm posterior to bregma and 1.3 mm to the right of the midline. A 25-gauge !-shaped cannula was lowered through this hole, to a depth of 4.2 mm below skull surface. The cannula was secured in this position by dental cement and by three stainless steel screws that were screwed into the skull. The cannula was connected via silicone rubber tubing (ESCO rubber; internal diameter 0.25 mm, outside 0.50 mm) to an osmotic minipump (Alzet, model 2002, Alza Co.; pumping rate 0.5 id/h) which was inserted under the dorsal skin. For the vehicle control group, the minipump contained 247/zl of artificial cerebrospinaI fluid (CSF: 8.66g NaCI, 0.224g KCI, 0.206g CaCIz-2H20 , 0.163g MgC12" 6H20, 0.214 g NazHPO4" 7H20 and 0.027 g N a H E P O a ' H 2 0 per 1 sterile HzO). The pumps in the AP5 group contained 247 Ill of a 30 mM solution of D-2-amino-5-phosphonopentanoic acid (d(ssolved in equimolar NaOH, then diluted to 30 mM in artificial CSF). The scalp was sutured and the animals were allowed to recover, with continuous access to food for 2 days after surgery (days I and 2). Two rats died during surgery. On day 3 after surgery the remaining rats (n = 16) were placed on a food-deprivation schedule again to maintain bodyweight at 85-90~o of freefeeding weight. The postoperative experiment lasted for 12 days. The output rate of the minipumps would have started to decline at day 14, according to the manufacturer's f i g u r e s .
Histology and tissue analysis At the end of the experiment, 12 days after surgery, the animals were given an overdose of pentobarbital (i.p.). A slab of brain tissue around the site of cannula insertion was removed and placed in a I0~o formaldehyde solution. Coronal sections of 50 izm were cut and stained with fast cresyl violet Nissl stain. The hippocampi were dissected out of the remaining part of the brain, weighed and stored at - 20 ~ Levels of AP5 in the left and right hippocampus were determined by means of HPLC analysis 23.
CSF) and Trial (lst, 2nd or 3d uninterrupted trial) with repeated measures on the factor Trial 35. The results of the delayed trials were analysed by means of a threefactor ANOVA with factors Treatment, Delay (5, 20 or 60 min) and Block (lst or 2nd block of delayed trials, see Procedure), with repeated measures on the last two factors.
RESULTS
At trial 21 of pre-operative training (day - 2 ) the mean number of errors (re-entry of arms visited previously in the same trial) was 0.05 (_+ 0.05, S.E.M.). Histological examination after the end of the experiment showed that all cannulae ended in the lateral ventricle. The mean concentration of AP5 in ttiedorsal hippocampus was 0.18 nmol/mg tissue (+0.03, S.E.M.) for the left hippocampus and 0.20 nmol/mg (_+ 0.04) for the right hippocampus. The mean number of errors during choices 5-8 in the three uninterrupted trials is shown in Fig. i. There were no significant differences between the two groups ( T r e a t m e n t : Fl,14 = 3.43, P > 0.08). The means for the 2 blocks of delayed trials are shown in Fig. 2. Threefactor analysis of variance with repeated measures revealed significant effects of AP5 (Fz,14 = 11.76, P < 0.005) and Delay (Fz.28 = 4.54, P < 0.05). There was a significant interaction between Treatment and Block (Ft,14 = 21.24, P < 0.001), indicating that the effect of AP5 on the number of errors was dependent on the block of trials. Subsequent ANOVAs performed on the results in the two blocks of delayed trials showed that there was no significant effect of AP5 on the number of errors in block 1 (Treatment: FI,14 = 2.79, P > 0.10), but there was a significant effect of Delay (F2.28 = 3.72, P < 0.05). In contrast, in the second block of trials there was a significant impairment of
CSF AP5 4-1 o
o o
r: 0 4
Statistical atzal.!'sis 9 The results of the three uninterrupted postoperative trials were analysed by means of a two-factor analysis of variance (ANOVA) with factors Treatment (AP5 or
8
12
days after surgery
Fig. 1. glean number of errors ( + S.E.M.) for the CSF and AP5 rats in the three uninterrupted trials run at different days after surgery. n = 8 in both groups.
154
tO +1
2
. l A _ - - - - - - -~- - + ' - -
T
o
APs-2
AP5-1
"6
=o
1
0
0
I 5
...---.------ CSF-2
I 20
I 60
log m i d - t r i a l d e l a y (rain)
Fig. 2. Mean number of errors (+ S.E.M.) in choices 4-8 in two blocks of delayed trials for AP5 rats (block l: filled circles; block 2: filled triangles) and CSF control rats (block l: open circles; block 2: open triangles). Note log scale on abscissa. The 4 straight lines are drawn for clarity and are linear curve fits based on the means in each group, i1 = 8 in both groups.
400
o - - o CSF
AP5
H
8 300
200
o 100
E
I
0
-3
I
I
I
I
I
I
I
I
I
I
-2 t /
4
5
6
7
8
9
10
11
12
surgery
days
Fig. 3. Mean time ( + S.E.M.) to complete the first 8 arm choices, before and after surgery (on day 0) for the AP5 rats (filled circles) and CSF rats (open circles), n = 8 in both groups.
400 O--O CSF H AP5 m
300 o
E 200
I
|
I
I
I
I
I
!
I
I
I
I
-1
0
1
2
3
4
5
6
7
8
9
10
t
surgery
1
I
11 12
days
Fig. 4. Mean bodyweights ( +_S.E.M.) ofthe AP5 (filled circles) and CSF rats (open circles) before, and after surgery. Surgery was performed on day 0. The rats had ad libitum access to food from 2 days before surgery to 2 days after surgery, after which they were again placed on a restricted feeding schedule (see Materials and Methods). n = 8 in both groups.
performance in the AP5 rats (Treatment: F l , 1 4 = 20.78, P < 0.001), which was independent of the length of the delay (Treatment • interaction: F < 1). Also, there was no significant effect of Delay in the second block of trials (F2,28 = 1.42, P > 0.20). The mean time taken to make the first 8 arm choices for the two groups is shown in Fig. 3. There was no significant difference between the two groups (F < 1). The mean bodyweight of the rats in the two groups is shown in Fig. 4. There was no significant difference between the means throughout the experiment (F < 1).
DISCUSSION
The results of the control animals in the present study (Fig. 2) confirm previous findings, showing that (i) spatial working memory in the radial maze has an exponential forgetting curve 3, (ii) memory persistence improves with further training with delayed trials 4'3~. The results of the uninterrupted trials and the first block of delayed trials did not show a significant effect of AP5. However, in contrast to the controls, there was no improvement of performance in delayed trials with further training in the AP5 rats. A possible explanation of the latter effect of AP5 is that the effects of intraventricular infusion of the drug increased with time after surgery. However, there is no support for this suggestion from previous studies using the same technique 24"32. Furthermore, there was no increased effect of the drug on performance of the rats in the uninterrupted trials (Fig. 1). Thus, the mechanism by which rats benefit from training with delayed trials seems to be important to explain the appearance of the impairment of performance by AP5. Either the AP5 rats are unable to improve the persistence of memory storage with further training or they are unable to build up a resistance to interference during delays. Tonkiss et al. 3z reported a significant weight loss in rats with intraventricular infusion of AP5 (30 mM D,L-AP5), compared to vehicle controls. Although the rats in their study also received additional feeding and gained weight after surgery, the difference in weight between AP5 rats and controls remained. In the present study mean bodyweight of the AP5 rats was slightly lower than that of the controls (Fig. 4), but the difference was not significant. It is unlikely therefore that the effects on performance in the maze were a result of an effect of the drug on bodyweight. Further indications that AP5 did not affect the animals' motivation comes from a comparison of the mean times to complete the first 8 arm choices (Fig. 3), which did not differ significantly between the two groups.
155 The present results show that the same dose that significantly impairs acquisition of an open field water maze reference memory task 8"9"22, does not affect performance in a radial maze working memory task. A similar dissociation was reported by McNaughton et al. 19, who found that saturation of hippocampal synapses with stimulation-induced LTP severely impaired acquisition in a spatial reference memory task, but did not affect spatial working memory in a radial arm maze. However, from their report it is not clear whether their rats had actually learned the radial maze task or whether or not the animals were able to solve it without simple response strategies 3. That is, LTP saturation may not have affected performance of the rats in the study by McNaughton et al. because the rats did not make use of spatial cues to solve the task. Of course, the improved performance at longer delays after further training observed in the present study is, by definition, still a manifestation of 'working memory'z6. In that sense, working memory is significantly impaired by AP5 infusion. The results of the present study differ from those in a preliminary report by Danysz et al. t~ These authors found dose-dependent impairments of performance of rats in a radial maze after single intraventricular injections of low doses (8, 16 or 32 nM) of AP5. It is not clear what the level of performance was in these rats prior to drug injections. Further, it is not known whether the rats could (and did) use response strategies to solve the task. Ward et al. 33 studied the effects of systemic injection of the competitive N M D A receptor blocker CPP and the channel blocker MK-801 on performance of rats in a radial maze. MK-801 impaired performance at a dose of 0.5 mg/kg, 2 h after intraperitoneal injection. Previously, it was shown that MK-801 at this dose caused a 35~o reduction in LTP (population spike potential) in the dentate gyrus t. However, at the dose at which there was a complete blockade of the induction of LTP (I.0 mg/kg) there were severe non-specific effects on the rats' behaviour and the animals could not be tested in the maze. Surprisingly, 0.5 mg/kg MK-801, which impaired performance at uninterrupted trials, did not significantly effect performance in a trial with a mid-trial delay of 1 h, when the drug as injected 2 h before the start of the trial. The latter finding suggests that the effect of MK-801 on uninterrupted maze performance may have been nonspecific. A 3-h delay between i.p. injection and performance in the second half of the trial, as occurred in the delayed trials, may have been long enough for the drug t o lose its effect on behaviour. The competitive receptor blocker CPP blocked the induction of LTP and impaired performance in the radial maze at the same
dose (10 mg/kg). These authors also did not control for the possible use by the rats of simple response strategies or intra-maze cues. Shapiro and Caramanos 29 found that 0.06 mg/kg of MK-801 severely impaired acquisition of t h e r a d i a l maze task, but that the drug did not affect reference or working memory in well-trained rats with doses up to 0.1 mg/kg. At higher doses the drug not only impaired performance but also produced ataxia. The mechanisms involved in initial acquisition of the task may be similar to those involved in the improvement of performance with delayed trials seen in the present study. The results of Shapiro and Caramanos are consistent with those of the present study, showing that radial maze performance in well-trained rats is not affected by N M D A blockers at doses at which an improvement with training is impaired. Given that the impairment of performance in the present study was not delay-dependent, and that it increased with further training, it is possible that the effects of AP5 were due to a performance deficit. The degree of training received prior to AP5 infusion is clearly an important variable, and may explain some of the discrepant findings discussed above. Subtle interference with sensory or motor (rather than cognitive) function by AP5 may have a more disruptive effect on novel tasks than on those which have been well learnt for this reason alone. An important difference between the open field watermaze and the radial maze, for example, is that while the latter has a restricted spatial layout in which all locations are visited repeatedly, the former is an open field in which many possible locations may or may not be visited in the course of learning to find a particular location. As a consequence, learning to approach a new (further) location in the watermaze may require the animal to pass through relatively less familiar pool sub-areas than those encountered in initial training. In the radial maze, however, each test requires the animal to visit the same, well-learned locations, while still requiring new learning of which of these locations have been visited each day. Subtle sensory impairment may therefore have more impact on performance in the water maze when a new location is learned, than on the new learning required each day in the radial maze task (e.g. ref. 23 vs. the present findings). N M D A receptors may play a role in synaptic plasticity underlying memory storage that is different from their role in LTP. For instance, the number of N M D A receptors may change with learning. Examples of this were reported by Wenk et al. 34 for spatial memory in rats and McCabe and Horn 16'17 for the learning process of imprinting in chicks. The significant changes in receptor number found by McCabe and Horn 16"~7
156 were limited to a restricted region of the brain (left IMHV) that has been shown to be intimately involved in the learning process and that is likely to be a site of long-term memory storage ~3. Whether the N M D A receptor-related mechanisms involved in learning and memory are related to those underlying LTP 14 remains to be investigated.
10
11
12 ACKNOWLEDGEMENTS
We are grateful to Richard Morris for advice and to Gabriel Horn for valuable comments on the manuscript. Thanks to Lisa Kendall and Steve Butcher for performing the tissue analysis, to Elma Forrest for preparing the histology, and to Sabrina Davis, Caroline Stewart and Grant MeLennan for their help. Financial support was provided by the Wellcome Trust to I.C.R., by the Netherlands Organization for Scientific Research (NWO) and the Dr. J.L. Dobberke Foundation for Comparative Psychology (Amsterdam) to J.J.B. and by an MRC Programme Grant to R.G.M. Morris and D. Willshaw.
REFERENCES 1 Abraham, W.C. and Mason, S.E., Effects of the NMDA receptor/channel antagonists CPP and MKS01 on hippocampal field potentials and long-term potentiation in anesthetized rats, Brain Res., 462 (1988) 40-46. 2 Bliss, T.V.P. and Lomo, T., Long-lasting potentiation ofsynaptic transmission in the dentate area of the anaesthetised rabbit following stimulation of the perforant path, J. PhysioL, 232 (1973) 331-356. 3 Bolhuis, J.J., Bijlsma, S. and Ansmink, P., Exponential decay of spatial memory of rats in a radial maze, Behav. Neural BioL, 46 (1986) 115-122. 4 Bolbuis, J.J., Strijkstra, A.M. and Kramers, R.J.K., Effects of scopolamine on performance of rats in a delayed-response radial maze task, PhysioL Behav., 43 (1988) 403-409. 5 Bolhuis, J.J. and Van Kampen, H.S., Serial position curves in spatial memory of rats: primacy and recency effects, Quart. J. Exp. Psychol., 40B (1988) 135-149. 6 Collingridge, G.L., Kehl, S.J. and McLennan, H., Excitatory amino acids in synaptic transmission in the Schaffer collateralcommissural pathway of the rat hippocampus, J. PhysioL, 334 (1983) 33-46. 7 Crooks Jr., G.B., Robinson Jr., G.S., Hatfield, T.J., Graham, P.W. and Gallagher, M., Intraventricular administration of the NMDA antagonist APV disrupts learning of an odor aversion that is potentiated by taste, Soc. NeuroscL Abstr., 15 (1989) 464. 8 Davis, S., Butcher, S.P. and Morris, R.G.M., The role of NMDA receptors in certain kinds of learning. In E. Cavalheiro, J. Lehman and L. Turski (Eds.), Recent Advances in Excitatory Amh~o Acid Research, Alan R. Liss, New York, 1988, pp. 385-392. 9 Davis, S., Butcher, S.P. and Morris, R.G.M., The N-methylD-aspartate receptor antagonist 2-amino-5-phosphonopentano-
13 14
15 16
17
18
19
20
21
22
23
24
25
26 27
28
ate (AP5) impairs spatial learning and LTP in vivo at comparable intracerebral concentrations to those that block LTP in vitro, J. Neurosci., 12 (1992) 21-34. Danysz, W., Wroblewski, J.T. and Costa, E., Learning impairment in rats by N-methyl-D-aspartate receptor antagonists, Neuropharmacology, 27 (1988) 653-656. Errington, M.L, Lynch, M.A. and Bliss, T.V.P., Long-term potentiation in the dentate gyrus; induction and increased glutamate release are blocked by D(-)aminophosphonovalerate, Neuroscience, 20 (1987) 279-284. Harris, E.W., Ganong, A.H. and Cotman, C.W., Long-term potentiation in the hippocampus involves activation of N-methyl-D-aspartate receptors, Brain Res., 323 (1984) 132-137. Horn, G., 3Iemory, hnprinthlg, attd the Brahe, Clarendon Press, Oxford, 1985. Horn, G. and McCabe, B.J., Learning by seeing: N-methylD-aspartate receptors and recognition memory. In Y. Ben-Ari (Ed.), Excitatory Amhlo Acids and Neuronal Plasticity, Plenum Press, New York, 1990. Lynch, G. and Baudry, M., The biochemistry of memory: a new and specific hypothesis, Science, 224 (1984) 1057-1063. McCabe, B.J. and Horn, G., Learning and me,mory: regional changes in N-methyl-D-aspartate receptors in the chick brain after imprinting, Proc. NatL Acad. Sci. U.S.A., 85 (1988) 2849-2853. McCabe, B.J. and Horn, G., Synaptic transmission and recognition memory: time course of changes in N-methyl-D-aspartate receptors after imprinting, Behav. Neurosci., 105 (1991) 289-294. McLamb, R.L., Williams, L.R., Nanry, K.P., Wilson, W.A. and Tilson, H.A., MK-801 impedes the acquisition of a spatial memory task in rats, Pharmacol. Biochen~. Behav., 37 (1990) 41-45. McNaughton, B.L., Barnes, C.A., Rao, G., Baldwin, J. and Rasmussen, M., Long-term enhancement of hippocampal synaptic transmission and the acquisition of spatial information, J. NeuroscL, 6 (1986) 563-571. McNaughton, B.L. and Morris, R.G.M., Hippocampal synaptic enhancement and information storage within a distributed memory system, Trends NeuroscL, 10 (1987) 408-415. Monaghan, D.T. and Cotman, C.W., Distribution of N-methylo-aspartate-sensitive L-[3H]-glutamate-binding sites in rat brain, J. Neurosci., 5 (1985) 2909-2919. Morris, R.G.M., Does synaptic plasticity play a role in information storage in the vertebrate brain? In R.G.M. Morris (Ed.), Parallel Distributed Processhzg: hnplications for Psychology and Neurobiology, Oxford University Press, Oxford, 1989. Morris, R.G.M., Synaptic plasticity and learning: selective impairment &learning in rats and blockade &long-term potentiation in vivo by the N-methyl-D-aspartate receptor antagonist AP5, J. Neurosci., 9 (1989) 3040-3057. Morris, R.G.M., Anderson, E., Lynch, G.S. and Baudry, M., Selective impairment of learning and blockade of long-term potentiation by an N-methyl-D-aspartate receptor antagonist, AP5, Nature, 319 (1986) 774-776. Morris, R.G.M., Halliwell, R.F. and Bowery, N., Synaptic plasticity and learning II: Do different kinds of plasticity underlie different kinds of learning? Neuropsychologia, 27 (1989) 41-59. Ohon, D.S., Becker, J.T. and Handelmann, G.E., Hippocampus, space, and memory, Behav. Brahz Sci., 2 (1979) 313-366. Olton, D.S. and Samuelson, R.J., Remembrance of places passed: spatial memory in rats, .I. Exp. PsychoL : Anita. Behav. Proc., 2 (1976) 97-116. Robinson Jr., G.S., Crooks Jr., G.B., Shinkman, P.G. and
157
29
30
31
32
Gallagher, M., Behavioral effects of MK-801 mimic deficits associated with hippocampal damage, Psychobiology, 17 (1989) 156-164. Shapiro, M.L. and Caramanos, Z., NMDA antagonist MK-801 impairs acquisition but not performance of spatial working and reference memory, Psychobiology, 18 (1990) 231-243. St~iubli, U., Thibault, O., DiLorenzo, M. and Lynch, G., Antagonism of NMDA receptors impairs acquisition but not retention of olfactory memory, Behav. Neurosci., 103 (1989) 54-60. Strijkstra, A.M. and Bolhuis, J.J., Memory persistence ofrats in a radial maze varies with training procedure, Behav. Ne,tral Biol., 47 (1987) 158-166. Tonkiss, J., Morris, R.G.M. and Rawlins, J.N.P., Intra-ventricular infusion of the NMDA antagonist AP5 impairs performance
on a non-spatial operant DRL task in the rat, Exp. Braht Res., 73 (1988) 181-188. 33 Ward, L., Mason, S.E. and Abraham, W.C., Effects of the NMDA antagonists CPP and MK-801 on radial arm maze performance in rats, PharmacoL Biochem. Behav., 35 (1990) 785-790. 34 Wenk, G.L., Grey, C.M., Ingram, D.K., Spangler, E.L. and Olton, D.S., Retention of maze performance inversely correlates with N-methyl-D-aspartate receptor number in hippocampus and frontal neocortex in the rat, Behav. NeuroscL, 103 (1989) 688-690. 35 Winer, B.J., Statistical Principles in Experimental Design, 2nd edn., McGraw-Hill, New York, 1972.
t5t
BBR 01268
Effects of intraventricular infusion of the N-methyl-D-aspartate ( N M D A ) receptor antagonist AP5 on spatial memory of rats in a radial arm maze Johan J. Bolhuis and Ian C. Reid Department of Pharmacologo', University of Edinburgh Medical School, Edinburgh (UK) (Received 29 July 1991) (Revised version received 4 November 1991) (Accepted 14 January 1992) Key words: N-Methyl-D-aspartic acid receptor; AP5; Spatial memory; Learning; Radial maze; Rat
Rats were trained to asymptotic performance in an 8-arm radial maze. They then received chronic intraventricular infusion ofeither artificial CSF or the N-methyl-D-aspartate (NM DA) receptor antagonist D-2-amino-5-phosphonopentanoic acid (AP5), at a concentration (30 mM)that has been shown previously to prevent the induction of long-term potentiation in the dentate gyrus of the hippocampus in vivo. Subsequently the rats received another 9 trials in the maze in a quasi-random order, 3 uninterrupted trials, and another 6 trials each with mid-trial delays of 5, 20 or 60 min during which the animals were placed in their home cage. The mean number of errors for the AP5 rats did not differ significantly from that of the controls in the uninterrupted trials throughout the experiment, nor did it differ from that of the controls in any of the 3 delayed trials when these were first introduced. However,'the control animals performed better at the longer delays when these were introduced for the second time, whilst there was no such improvement (but rather a deterioration) for the AP5 animals. The impairment of performance in the AP5 rats during the second block ofdelayed trials was significant, and independent of the length of the delay. These results show that NMDA receptor blockade does not impair working memory in the radial maze per se, but that it does prevent an improvement of working memory persistence with further training.
INTRODUCTION There is increasing evidence that a particular class o f excitatory amino acid receptors, the N-methyl-D-aspartate ( N M D A ) receptors, is involved in changes in the efficacy o f synapses. Several studies 6,~ i.~2 have shown that N M D A receptors play a role in a particular form o f artificially induced synaptie plasticity, long-term potentiation ( L T P ) L It has been proposed that the same, or similar mechanisms to those underlying L T P (involving N M D A receptors) might be involved in learning and m e m o r y storage 2.t7"z~ Morris et al. 24 showed that intraventricular infusion o f the competitive N M D A receptor antagonist D,L-2-amino-5-phosphonopentanoic acid (AP5) blocked the induction o f L T P in vivo and impaired place navigation learning in an open field water maze. Visual
Correspondence: J.J. Bolhuis, University of Cambridge, Department of Zoology, Downing Street, Cambridge CB2 3EJ, UK.
discrimination learning in the same apparatus was not impaired. Also, retention o f previously acquired spatial information was not affected by the drug 23. Davis et al. 9 have subsequently demonstrated a parallel d o s e response curve for the action o f AP5 with regard to blockade o f the induction o f L T P and the impairment o f spatial learning. Recent studies have suggested effects o f AP5 in a differential responding to low rates ( D R L ) paradigm 32, in some forms of olfactory learning 3~ and in taste-potentiated o d o u r aversion 7. It has been suggested that the effects o f AP5 and o f certain non-competitive N M D A antagonists on learning and m e m o r y in rats are mainly due to their effects on hippocampal N M D A receptors 18"22-25"28. The reason for this belief is that, on the one hand, there is a particularly high density o f these receptors in this structure 2~, and on the other hand the behavioural effects o f the drugs mirror those o f hippocampal lesions 24"28"32. Furthermore, Morris et al. 25 have shown that local infusion o f AP5 restricted to the hippocampus, is sufficient to obtain an impairment in the
152 open field water maze and to block the induction of LTP. The present study was designed to investigate the effects of intraventricular infusion of AP5 in the radial a r m maze 27, a spatial working memory 26 task that has been shown to be sensitive to hippocampal damage 26. By varying the retention interval 3"4, possible effects of N M D A receptor blockade on retention were studied.
MATERIALS AND METHODS
SllbjecIs Eighteen male hooded Lister rats were used, weighing 240-340 g at the beginning of the experiment. The animals were kept in individual cages in a room with a light/dark regimen of 14/10 (lights on from 08.00 to 22.00 h). Seven days before the start of the experiment, the rats were subjected to a restricted feeding schedule, 1-2 h feeding each day after the end o f the last trial, to maintain the animals at 85-90 ~ of free-feeding weight. Water was available ad libitum throughout the experiment.
Apparatus An 8-arm radial maze was used, similar to that introduced by Olton and Samuelson 27, and used in previous studies by Bolhuis et al. 3-5. It consisted of a central platform (diameter 35 cm) from which 8 arms (96 cm long, 6 cm wide) radiated. Each of the arms was surrounded by a 1-cm high wooden ledge. Five cm from the end of each arm was a hole with a plastic foodcup (diameter 2 cm, I cm deep). Around the edge of the central platform was a 16-cm high wall of wooden segments, with an opening at the beginning of each arm, in which transparent Perspex guillotine doors were placed that could be operated by means of overhead lines. The maze was made of wood and painted matt grey. It was placed 1 m above floor level on a table that could be rotated.
Procedure On the first day of the experiment (day - 15) the rats were placed individually on the central platform, with all eight doors raised. The animal was left in the maze to explore and retrieve raisins that were scattered throughout the apparatus and in the food cups. The next two days, the rats were allowed to explore the maze and retrieve 200 Id of a 23% sucrose solution that was placed in each of the foodcups. After these 2 trials, the guillotine doors were operated during each trial. At the beginning of a trial, the rat was placed on the central platform, with all doors closed. After 10 s, all doors
were raised simultaneously and the rat was allowed to enter an arm. All doors were then lowered, except for the door leading to the arm that the rat was visiting. After the rat had returned to the central platform, this door was lowered too. After 5 s all doors were raised simultaneously and the rat was allowed again to enter an arm. This procedure was repeated until the rats had retrieved all 8 sucrose rewards. The use of doors in this way has been shown to prevent simple turning strategies ofchoosing adjacent arms 3. From the 2nd day until the 9th day of the experiment (day - 14 to - 7 ) the rats received two trials per day, with an interval between trials of at least 2 h, to prevent possible proactive interference 3. Re-entry of an arm with all four paws that had been visited previously in the same trial was considered to be an error. On day - 5, the rats received only one .trial, with a mid-trial interval of 5 min. When the animal had returned to the central platform after the fourth choice, it was removed from the maze and placed back in its home cage. After 5 min it was placed on the central platform again and after 5 s the animal was allowed to complete its trial. This procedure was repeated on day - 4 , but now the maze was rotated 180 ~ during the mid-trial interval. The sucrose rewards were placed in the four arms in the spatial position that the rat had not visited in the first four choices of the trial. Thus, if the animal was guided by extra-maze cues only, its performance would be similar to that in the non-rotation trial. If, on the other hand, intra-maze cues such as odours or visual aspects of the maze would be important in guiding the animal, performance could be expected to deteriorate, compared to that in the non-rotation trial 3. There was no significant difference between the mean number of errors in choices 5 - 8 (random choice level -- 2 errors) in the non-rotation trial (0.50 + 0.13) and that in the rotation trial (0.63 + 0.20), indicating that the animals were not guided by intra-maze c u e s 3. In order to prevent the use ofintra-maze cues, the maze was rotated 180 ~ during the mid-trial intervals in the postoperative delayed trials, replacing the sucrose reward in the same (unvisited) spatial locations. From the 4th day after surgery (see below) onwards, the rats received another 9 trials, one trial per day. The animals received uninterrupted trials on days 4, 8 and 12, trials with a mid-trial delay of 5 min at days 5 and 10, with a 20-rain delay at days 6 and 9, and with a 60-min delay at days 7 and 11. Thus, the order of postoperative trials was: 0-5-20-60-0-20-5-60-0min mid-trial delay. During the mid-trial delays the maze was rotated 180 ~ as described above. The first 3 delayed trials (days 5-7) are called Block 1 and the second 3 delayed trials (days 9-11) are called Block 2.
153
Surgery For 2 days before surgery (days - 2 and - 1) the rats had ad tibitum access to food. On day 0, the rats were anaesthetized with tribromoethanol (Avertin; 29.9 g/l, I0 ml/kg, i.p.) and placed in a stereotaxic apparatus, with the skull surface between bregma and lambda in a horizontal plane. An incision was made in the skin on the head and a single hole was drilled in the skull, 0.9 mm posterior to bregma and 1.3 mm to the right of the midline. A 25-gauge !-shaped cannula was lowered through this hole, to a depth of 4.2 mm below skull surface. The cannula was secured in this position by dental cement and by three stainless steel screws that were screwed into the skull. The cannula was connected via silicone rubber tubing (ESCO rubber; internal diameter 0.25 mm, outside 0.50 mm) to an osmotic minipump (Alzet, model 2002, Alza Co.; pumping rate 0.5 id/h) which was inserted under the dorsal skin. For the vehicle control group, the minipump contained 247/zl of artificial cerebrospinaI fluid (CSF: 8.66g NaCI, 0.224g KCI, 0.206g CaCIz-2H20 , 0.163g MgC12" 6H20, 0.214 g NazHPO4" 7H20 and 0.027 g N a H E P O a ' H 2 0 per 1 sterile HzO). The pumps in the AP5 group contained 247 Ill of a 30 mM solution of D-2-amino-5-phosphonopentanoic acid (d(ssolved in equimolar NaOH, then diluted to 30 mM in artificial CSF). The scalp was sutured and the animals were allowed to recover, with continuous access to food for 2 days after surgery (days I and 2). Two rats died during surgery. On day 3 after surgery the remaining rats (n = 16) were placed on a food-deprivation schedule again to maintain bodyweight at 85-90~o of freefeeding weight. The postoperative experiment lasted for 12 days. The output rate of the minipumps would have started to decline at day 14, according to the manufacturer's f i g u r e s .
Histology and tissue analysis At the end of the experiment, 12 days after surgery, the animals were given an overdose of pentobarbital (i.p.). A slab of brain tissue around the site of cannula insertion was removed and placed in a I0~o formaldehyde solution. Coronal sections of 50 izm were cut and stained with fast cresyl violet Nissl stain. The hippocampi were dissected out of the remaining part of the brain, weighed and stored at - 20 ~ Levels of AP5 in the left and right hippocampus were determined by means of HPLC analysis 23.
CSF) and Trial (lst, 2nd or 3d uninterrupted trial) with repeated measures on the factor Trial 35. The results of the delayed trials were analysed by means of a threefactor ANOVA with factors Treatment, Delay (5, 20 or 60 min) and Block (lst or 2nd block of delayed trials, see Procedure), with repeated measures on the last two factors.
RESULTS
At trial 21 of pre-operative training (day - 2 ) the mean number of errors (re-entry of arms visited previously in the same trial) was 0.05 (_+ 0.05, S.E.M.). Histological examination after the end of the experiment showed that all cannulae ended in the lateral ventricle. The mean concentration of AP5 in ttiedorsal hippocampus was 0.18 nmol/mg tissue (+0.03, S.E.M.) for the left hippocampus and 0.20 nmol/mg (_+ 0.04) for the right hippocampus. The mean number of errors during choices 5-8 in the three uninterrupted trials is shown in Fig. i. There were no significant differences between the two groups ( T r e a t m e n t : Fl,14 = 3.43, P > 0.08). The means for the 2 blocks of delayed trials are shown in Fig. 2. Threefactor analysis of variance with repeated measures revealed significant effects of AP5 (Fz,14 = 11.76, P < 0.005) and Delay (Fz.28 = 4.54, P < 0.05). There was a significant interaction between Treatment and Block (Ft,14 = 21.24, P < 0.001), indicating that the effect of AP5 on the number of errors was dependent on the block of trials. Subsequent ANOVAs performed on the results in the two blocks of delayed trials showed that there was no significant effect of AP5 on the number of errors in block 1 (Treatment: FI,14 = 2.79, P > 0.10), but there was a significant effect of Delay (F2.28 = 3.72, P < 0.05). In contrast, in the second block of trials there was a significant impairment of
CSF AP5 4-1 o
o o
r: 0 4
Statistical atzal.!'sis 9 The results of the three uninterrupted postoperative trials were analysed by means of a two-factor analysis of variance (ANOVA) with factors Treatment (AP5 or
8
12
days after surgery
Fig. 1. glean number of errors ( + S.E.M.) for the CSF and AP5 rats in the three uninterrupted trials run at different days after surgery. n = 8 in both groups.
154
tO +1
2
. l A _ - - - - - - -~- - + ' - -
T
o
APs-2
AP5-1
"6
=o
1
0
0
I 5
...---.------ CSF-2
I 20
I 60
log m i d - t r i a l d e l a y (rain)
Fig. 2. Mean number of errors (+ S.E.M.) in choices 4-8 in two blocks of delayed trials for AP5 rats (block l: filled circles; block 2: filled triangles) and CSF control rats (block l: open circles; block 2: open triangles). Note log scale on abscissa. The 4 straight lines are drawn for clarity and are linear curve fits based on the means in each group, i1 = 8 in both groups.
400
o - - o CSF
AP5
H
8 300
200
o 100
E
I
0
-3
I
I
I
I
I
I
I
I
I
I
-2 t /
4
5
6
7
8
9
10
11
12
surgery
days
Fig. 3. Mean time ( + S.E.M.) to complete the first 8 arm choices, before and after surgery (on day 0) for the AP5 rats (filled circles) and CSF rats (open circles), n = 8 in both groups.
400 O--O CSF H AP5 m
300 o
E 200
I
|
I
I
I
I
I
!
I
I
I
I
-1
0
1
2
3
4
5
6
7
8
9
10
t
surgery
1
I
11 12
days
Fig. 4. Mean bodyweights ( +_S.E.M.) ofthe AP5 (filled circles) and CSF rats (open circles) before, and after surgery. Surgery was performed on day 0. The rats had ad libitum access to food from 2 days before surgery to 2 days after surgery, after which they were again placed on a restricted feeding schedule (see Materials and Methods). n = 8 in both groups.
performance in the AP5 rats (Treatment: F l , 1 4 = 20.78, P < 0.001), which was independent of the length of the delay (Treatment • interaction: F < 1). Also, there was no significant effect of Delay in the second block of trials (F2,28 = 1.42, P > 0.20). The mean time taken to make the first 8 arm choices for the two groups is shown in Fig. 3. There was no significant difference between the two groups (F < 1). The mean bodyweight of the rats in the two groups is shown in Fig. 4. There was no significant difference between the means throughout the experiment (F < 1).
DISCUSSION
The results of the control animals in the present study (Fig. 2) confirm previous findings, showing that (i) spatial working memory in the radial maze has an exponential forgetting curve 3, (ii) memory persistence improves with further training with delayed trials 4'3~. The results of the uninterrupted trials and the first block of delayed trials did not show a significant effect of AP5. However, in contrast to the controls, there was no improvement of performance in delayed trials with further training in the AP5 rats. A possible explanation of the latter effect of AP5 is that the effects of intraventricular infusion of the drug increased with time after surgery. However, there is no support for this suggestion from previous studies using the same technique 24"32. Furthermore, there was no increased effect of the drug on performance of the rats in the uninterrupted trials (Fig. 1). Thus, the mechanism by which rats benefit from training with delayed trials seems to be important to explain the appearance of the impairment of performance by AP5. Either the AP5 rats are unable to improve the persistence of memory storage with further training or they are unable to build up a resistance to interference during delays. Tonkiss et al. 3z reported a significant weight loss in rats with intraventricular infusion of AP5 (30 mM D,L-AP5), compared to vehicle controls. Although the rats in their study also received additional feeding and gained weight after surgery, the difference in weight between AP5 rats and controls remained. In the present study mean bodyweight of the AP5 rats was slightly lower than that of the controls (Fig. 4), but the difference was not significant. It is unlikely therefore that the effects on performance in the maze were a result of an effect of the drug on bodyweight. Further indications that AP5 did not affect the animals' motivation comes from a comparison of the mean times to complete the first 8 arm choices (Fig. 3), which did not differ significantly between the two groups.
155 The present results show that the same dose that significantly impairs acquisition of an open field water maze reference memory task 8"9"22, does not affect performance in a radial maze working memory task. A similar dissociation was reported by McNaughton et al. 19, who found that saturation of hippocampal synapses with stimulation-induced LTP severely impaired acquisition in a spatial reference memory task, but did not affect spatial working memory in a radial arm maze. However, from their report it is not clear whether their rats had actually learned the radial maze task or whether or not the animals were able to solve it without simple response strategies 3. That is, LTP saturation may not have affected performance of the rats in the study by McNaughton et al. because the rats did not make use of spatial cues to solve the task. Of course, the improved performance at longer delays after further training observed in the present study is, by definition, still a manifestation of 'working memory'z6. In that sense, working memory is significantly impaired by AP5 infusion. The results of the present study differ from those in a preliminary report by Danysz et al. t~ These authors found dose-dependent impairments of performance of rats in a radial maze after single intraventricular injections of low doses (8, 16 or 32 nM) of AP5. It is not clear what the level of performance was in these rats prior to drug injections. Further, it is not known whether the rats could (and did) use response strategies to solve the task. Ward et al. 33 studied the effects of systemic injection of the competitive N M D A receptor blocker CPP and the channel blocker MK-801 on performance of rats in a radial maze. MK-801 impaired performance at a dose of 0.5 mg/kg, 2 h after intraperitoneal injection. Previously, it was shown that MK-801 at this dose caused a 35~o reduction in LTP (population spike potential) in the dentate gyrus t. However, at the dose at which there was a complete blockade of the induction of LTP (I.0 mg/kg) there were severe non-specific effects on the rats' behaviour and the animals could not be tested in the maze. Surprisingly, 0.5 mg/kg MK-801, which impaired performance at uninterrupted trials, did not significantly effect performance in a trial with a mid-trial delay of 1 h, when the drug as injected 2 h before the start of the trial. The latter finding suggests that the effect of MK-801 on uninterrupted maze performance may have been nonspecific. A 3-h delay between i.p. injection and performance in the second half of the trial, as occurred in the delayed trials, may have been long enough for the drug t o lose its effect on behaviour. The competitive receptor blocker CPP blocked the induction of LTP and impaired performance in the radial maze at the same
dose (10 mg/kg). These authors also did not control for the possible use by the rats of simple response strategies or intra-maze cues. Shapiro and Caramanos 29 found that 0.06 mg/kg of MK-801 severely impaired acquisition of t h e r a d i a l maze task, but that the drug did not affect reference or working memory in well-trained rats with doses up to 0.1 mg/kg. At higher doses the drug not only impaired performance but also produced ataxia. The mechanisms involved in initial acquisition of the task may be similar to those involved in the improvement of performance with delayed trials seen in the present study. The results of Shapiro and Caramanos are consistent with those of the present study, showing that radial maze performance in well-trained rats is not affected by N M D A blockers at doses at which an improvement with training is impaired. Given that the impairment of performance in the present study was not delay-dependent, and that it increased with further training, it is possible that the effects of AP5 were due to a performance deficit. The degree of training received prior to AP5 infusion is clearly an important variable, and may explain some of the discrepant findings discussed above. Subtle interference with sensory or motor (rather than cognitive) function by AP5 may have a more disruptive effect on novel tasks than on those which have been well learnt for this reason alone. An important difference between the open field watermaze and the radial maze, for example, is that while the latter has a restricted spatial layout in which all locations are visited repeatedly, the former is an open field in which many possible locations may or may not be visited in the course of learning to find a particular location. As a consequence, learning to approach a new (further) location in the watermaze may require the animal to pass through relatively less familiar pool sub-areas than those encountered in initial training. In the radial maze, however, each test requires the animal to visit the same, well-learned locations, while still requiring new learning of which of these locations have been visited each day. Subtle sensory impairment may therefore have more impact on performance in the water maze when a new location is learned, than on the new learning required each day in the radial maze task (e.g. ref. 23 vs. the present findings). N M D A receptors may play a role in synaptic plasticity underlying memory storage that is different from their role in LTP. For instance, the number of N M D A receptors may change with learning. Examples of this were reported by Wenk et al. 34 for spatial memory in rats and McCabe and Horn 16'17 for the learning process of imprinting in chicks. The significant changes in receptor number found by McCabe and Horn 16"~7
156 were limited to a restricted region of the brain (left IMHV) that has been shown to be intimately involved in the learning process and that is likely to be a site of long-term memory storage ~3. Whether the N M D A receptor-related mechanisms involved in learning and memory are related to those underlying LTP 14 remains to be investigated.
10
11
12 ACKNOWLEDGEMENTS
We are grateful to Richard Morris for advice and to Gabriel Horn for valuable comments on the manuscript. Thanks to Lisa Kendall and Steve Butcher for performing the tissue analysis, to Elma Forrest for preparing the histology, and to Sabrina Davis, Caroline Stewart and Grant MeLennan for their help. Financial support was provided by the Wellcome Trust to I.C.R., by the Netherlands Organization for Scientific Research (NWO) and the Dr. J.L. Dobberke Foundation for Comparative Psychology (Amsterdam) to J.J.B. and by an MRC Programme Grant to R.G.M. Morris and D. Willshaw.
REFERENCES 1 Abraham, W.C. and Mason, S.E., Effects of the NMDA receptor/channel antagonists CPP and MKS01 on hippocampal field potentials and long-term potentiation in anesthetized rats, Brain Res., 462 (1988) 40-46. 2 Bliss, T.V.P. and Lomo, T., Long-lasting potentiation ofsynaptic transmission in the dentate area of the anaesthetised rabbit following stimulation of the perforant path, J. PhysioL, 232 (1973) 331-356. 3 Bolhuis, J.J., Bijlsma, S. and Ansmink, P., Exponential decay of spatial memory of rats in a radial maze, Behav. Neural BioL, 46 (1986) 115-122. 4 Bolbuis, J.J., Strijkstra, A.M. and Kramers, R.J.K., Effects of scopolamine on performance of rats in a delayed-response radial maze task, PhysioL Behav., 43 (1988) 403-409. 5 Bolhuis, J.J. and Van Kampen, H.S., Serial position curves in spatial memory of rats: primacy and recency effects, Quart. J. Exp. Psychol., 40B (1988) 135-149. 6 Collingridge, G.L., Kehl, S.J. and McLennan, H., Excitatory amino acids in synaptic transmission in the Schaffer collateralcommissural pathway of the rat hippocampus, J. PhysioL, 334 (1983) 33-46. 7 Crooks Jr., G.B., Robinson Jr., G.S., Hatfield, T.J., Graham, P.W. and Gallagher, M., Intraventricular administration of the NMDA antagonist APV disrupts learning of an odor aversion that is potentiated by taste, Soc. NeuroscL Abstr., 15 (1989) 464. 8 Davis, S., Butcher, S.P. and Morris, R.G.M., The role of NMDA receptors in certain kinds of learning. In E. Cavalheiro, J. Lehman and L. Turski (Eds.), Recent Advances in Excitatory Amh~o Acid Research, Alan R. Liss, New York, 1988, pp. 385-392. 9 Davis, S., Butcher, S.P. and Morris, R.G.M., The N-methylD-aspartate receptor antagonist 2-amino-5-phosphonopentano-
13 14
15 16
17
18
19
20
21
22
23
24
25
26 27
28
ate (AP5) impairs spatial learning and LTP in vivo at comparable intracerebral concentrations to those that block LTP in vitro, J. Neurosci., 12 (1992) 21-34. Danysz, W., Wroblewski, J.T. and Costa, E., Learning impairment in rats by N-methyl-D-aspartate receptor antagonists, Neuropharmacology, 27 (1988) 653-656. Errington, M.L, Lynch, M.A. and Bliss, T.V.P., Long-term potentiation in the dentate gyrus; induction and increased glutamate release are blocked by D(-)aminophosphonovalerate, Neuroscience, 20 (1987) 279-284. Harris, E.W., Ganong, A.H. and Cotman, C.W., Long-term potentiation in the hippocampus involves activation of N-methyl-D-aspartate receptors, Brain Res., 323 (1984) 132-137. Horn, G., 3Iemory, hnprinthlg, attd the Brahe, Clarendon Press, Oxford, 1985. Horn, G. and McCabe, B.J., Learning by seeing: N-methylD-aspartate receptors and recognition memory. In Y. Ben-Ari (Ed.), Excitatory Amhlo Acids and Neuronal Plasticity, Plenum Press, New York, 1990. Lynch, G. and Baudry, M., The biochemistry of memory: a new and specific hypothesis, Science, 224 (1984) 1057-1063. McCabe, B.J. and Horn, G., Learning and me,mory: regional changes in N-methyl-D-aspartate receptors in the chick brain after imprinting, Proc. NatL Acad. Sci. U.S.A., 85 (1988) 2849-2853. McCabe, B.J. and Horn, G., Synaptic transmission and recognition memory: time course of changes in N-methyl-D-aspartate receptors after imprinting, Behav. Neurosci., 105 (1991) 289-294. McLamb, R.L., Williams, L.R., Nanry, K.P., Wilson, W.A. and Tilson, H.A., MK-801 impedes the acquisition of a spatial memory task in rats, Pharmacol. Biochen~. Behav., 37 (1990) 41-45. McNaughton, B.L., Barnes, C.A., Rao, G., Baldwin, J. and Rasmussen, M., Long-term enhancement of hippocampal synaptic transmission and the acquisition of spatial information, J. NeuroscL, 6 (1986) 563-571. McNaughton, B.L. and Morris, R.G.M., Hippocampal synaptic enhancement and information storage within a distributed memory system, Trends NeuroscL, 10 (1987) 408-415. Monaghan, D.T. and Cotman, C.W., Distribution of N-methylo-aspartate-sensitive L-[3H]-glutamate-binding sites in rat brain, J. Neurosci., 5 (1985) 2909-2919. Morris, R.G.M., Does synaptic plasticity play a role in information storage in the vertebrate brain? In R.G.M. Morris (Ed.), Parallel Distributed Processhzg: hnplications for Psychology and Neurobiology, Oxford University Press, Oxford, 1989. Morris, R.G.M., Synaptic plasticity and learning: selective impairment &learning in rats and blockade &long-term potentiation in vivo by the N-methyl-D-aspartate receptor antagonist AP5, J. Neurosci., 9 (1989) 3040-3057. Morris, R.G.M., Anderson, E., Lynch, G.S. and Baudry, M., Selective impairment of learning and blockade of long-term potentiation by an N-methyl-D-aspartate receptor antagonist, AP5, Nature, 319 (1986) 774-776. Morris, R.G.M., Halliwell, R.F. and Bowery, N., Synaptic plasticity and learning II: Do different kinds of plasticity underlie different kinds of learning? Neuropsychologia, 27 (1989) 41-59. Ohon, D.S., Becker, J.T. and Handelmann, G.E., Hippocampus, space, and memory, Behav. Brahz Sci., 2 (1979) 313-366. Olton, D.S. and Samuelson, R.J., Remembrance of places passed: spatial memory in rats, .I. Exp. PsychoL : Anita. Behav. Proc., 2 (1976) 97-116. Robinson Jr., G.S., Crooks Jr., G.B., Shinkman, P.G. and
157
29
30
31
32
Gallagher, M., Behavioral effects of MK-801 mimic deficits associated with hippocampal damage, Psychobiology, 17 (1989) 156-164. Shapiro, M.L. and Caramanos, Z., NMDA antagonist MK-801 impairs acquisition but not performance of spatial working and reference memory, Psychobiology, 18 (1990) 231-243. St~iubli, U., Thibault, O., DiLorenzo, M. and Lynch, G., Antagonism of NMDA receptors impairs acquisition but not retention of olfactory memory, Behav. Neurosci., 103 (1989) 54-60. Strijkstra, A.M. and Bolhuis, J.J., Memory persistence ofrats in a radial maze varies with training procedure, Behav. Ne,tral Biol., 47 (1987) 158-166. Tonkiss, J., Morris, R.G.M. and Rawlins, J.N.P., Intra-ventricular infusion of the NMDA antagonist AP5 impairs performance
on a non-spatial operant DRL task in the rat, Exp. Braht Res., 73 (1988) 181-188. 33 Ward, L., Mason, S.E. and Abraham, W.C., Effects of the NMDA antagonists CPP and MK-801 on radial arm maze performance in rats, PharmacoL Biochem. Behav., 35 (1990) 785-790. 34 Wenk, G.L., Grey, C.M., Ingram, D.K., Spangler, E.L. and Olton, D.S., Retention of maze performance inversely correlates with N-methyl-D-aspartate receptor number in hippocampus and frontal neocortex in the rat, Behav. NeuroscL, 103 (1989) 688-690. 35 Winer, B.J., Statistical Principles in Experimental Design, 2nd edn., McGraw-Hill, New York, 1972.
