- Email: [email protected]
Low-dose challenge by the NMDA receptor antagonist dizocilpine exacerbates the spatial learning deficit in entorhinal cortex-lesioned rats
ELSEVIER
BEHAVIOURAL BRAIN RESEARCH Behavioural Brain Research 67 (1995) 255-261
Research report
Low-dose challenge by the NMDA receptor antagonist dizocilpine exacerbates the spatial learning deficit in entorhinal cortex-lesioned rats Ute Keseberg*, Werner J. Schmidt Department of Neuropharmacology, Zoological Institute, Mohlstr. 54/l, D-72074 Tfibingen, Germany Received 25 May 1994; revised 6 October 1994; accepted 6 October 1994
Abstract
We investigated the effects of a bilateral quinolinic acid lesion of the medial entorhinal cortex (EC) on acquisition of a spatial learning task. During reversal of the same task, we challenged the animals by the N-methyl-D-aspartate (NMDA) receptor antagonist dizocilpine (MK-801). Training took postoperatively place in an eight-arm radial maze in which four of eight arms were baited. In the acquisition phase (ten blocks of five trials) of the test, EC-lesioned animals showed a working (WM) and a reference memory (RM) deficit. The WM deficit was prominent at the beginning and fully compensated at the end of the acquisition phase. The RM deficit became more evident during the course of the experiment. In the reversal learning phase (seven blocks of five trials), the formerly unbaited arms were baited and half of the control and lesioned animals were challenged by a low dose of dizocilpine (0.04 mg/kg i.p.) before training. Only lesioned and additionally dizocilpine-treated animals showed a WM deficit that was again compensated and a RM deficit that was stronger at the end of the test. In summary, quinolinic acid lesion of the medial EC induces both WM and RM deficits in rats. The WM deficit is rapidly compensated. Enhancement of these deficits by challenge with dizocilpine in the reversal learning phase suggests that the NMDA receptor system was rendered more sensitive by this type of lesion. Key words: Entorhinal cortex; Quinolinic acid; Excitotoxic lesion; Spatial learning; Eight-arm radial maze; MK-801; Challenge
1. Introduction
The entorhinal cortex (EC) is a major relay station for converging inputs from sensory and association cortices to the hippocampus and projections back from the hippocampal formation to the neocortex. Neurons from layer II of the EC project via the perforant path to the granule cell layer of the dentate gyrus, the first station of the trisynaptic hippocampal loop, and to the subiculum. Projections from layer III of the EC terminate in areas CA3 and CA1. Hippocampal output from CA1 is directed through the subiculum to deeper layers of the EC [11]. Thus the EC has anatomically a key position between the neocortex and the hippocampus. While the role of the hippocampus in spatial learning has been well documented in the last years, the contribution of the EC remains less clear [9,32]. According to a recent review of Jarrard [9] the hippocampus is stronger involved in working memory (WM) than reference me* Corresponding author. Fax: (49) (7071) 29-5017. E-mail: [email protected] 0166-4328/95/$9.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0 1 6 6 - 4 3 2 8 ( 9 4 ) 0 0 1 5 6 - 1
mory (RM) which is especially evident in the acquisition of the spatial but not of the intramaze-cue task in an eight-arm radial maze. In contrast, as Jarrard et al. [ 10] found in earlier electrolytic lesion studies, the EC seems to play a stronger role in RM than WM, at least in the place version of the task. Recent data following aspiration lesions of the EC [9], however, revealed no deficit at all in the place task and only a RM deficit in the intramaze cue task. Thus, there is considerable uncertainty of the functional role of the EC in acquisition of spatial learning due to different lesion methods. Our approach to this problem was therefore (1) to use a neurotoxin that specifically lesions somata leaving passing axons intact and (2) to lesion only a part of the EC instead of damaging the whole structure. We lesioned the more medial part of the EC by quinolinic acid. This excitotoxin is an endogenous metabolite of the tryptophan metabolism and destroys as an agonist of the N-methyl-D-aspartate ( N M D A ) receptor only somata without interrupting fibers of passage [29]. In addition, it is a possible candidate to contribute to excitotoxic damage in vivo [6]. In a previous study of our group [7] le-
256
Behavioural Brain Research 67 (1995) 255-26l
sion of the medial EC induced a deficit in allocentric (i.e., spatial) but not in egocentric (i.e., body movement) learning in the eight-arm radial maze. In the present study we prolonged the acquisition phase and additionally challenged the animals in the reversal phase of the task with a low dose (0.04 mg/kg) of the non-competitive N M D A antagonist dizocilpine (MK-801). The low dizocilpine dose has neither an effect on locomotion [14] nor on learning [23,26] and affects in higher doses only acquisition but not retention of learning tasks [3,17,30]. The challenge was undertaken in order to investigate whether lesioned animals become more sensitive to this treatment. The present results might be of relevance for human neurodegenerative diseases. EC-lesion studies have been often discussed in relation to Alzheimer's disease (AD) (see for example [8,15,19,20,28]), since layer II of the human EC is the first brain area to show neurofibrillary tangles and develops the greatest number of tangles per unit area in the brain of AD patients [2,8]. Damage in cerebral ischemia and traumatic brain injury is induced via similar excitotoxic mechanisms as quinolinic acid lesions [22,31 ]. N M D A receptor antagonists such as dizocilpine are able to protect from excitotoxicity but have at the same time unfavorable effects on cognition [16,18]. These aspects will be considered in the discussion of the present paper. 2. Materials and Methods 2.1. Subjects
Fourty male Sprague-Dawley rats (Interfauna, Tuttlingen), weighing 210-260 g, were used as experimental animals. They were housed in groups of 4 - 6 and were maintained on a 12/12 h light/dark cycle, lights being on from 06.00 h to 18.00 h. Water was freely available but food was restricted to a daily diet of 12 g per animal so that the animals held approximately 85~/o of their free feeding weight. They were fed after completion of the experiment. 2.2. Drugs
Quinolinic acid (Sigma, Deisenhofen) was dissolved in 0.1 M phosphate buffered saline and adjusted to pH 7.4 by 1 M NaOH. Chloralhydrate (Fluka, Neu-Ulm), atropine (Sigma, Deisenhofen) and dizocilpine (( + )-MK-801, RBI, KOln) were dissolved in 0.9~o saline. Atropine (s.c) and dizocilpine (i.p.) were injected in a volume of 1 ml/kg. 2.3. Surgery
Animals were anesthetized with 4 ~o choralhydrate (400 mg/kg i.p.) and treated with atropine sulfate (0.2 mg/animal s.c.) before and after surgery. Surgery was per-
formed in a stereotaxic Stoelting instrument. Stereotaxic coordinates [24] were calculated from bregma: anterioposterior -8.0, mediolateral + 3.5. Injections of 0.5 #1 0.1 M PBS (control, N = 20) or 60 nmol quinolinic acid (lesion, N = 20) were bilaterally given under a mediolateral angle of 15 ° at three different dorsoventral levels: 0.2 #1 at -7.5, 0.2/~1 at -7.0 and 0.1/~1 at -6.0 [7]. The injection syringe was connected by a tube to a hamilton syringe that contained either the neurotoxin or vehicle. 2.4. Spatial learning
We used the same eight-arm radial maze as in previous studies [ 7,13 ]. Preoperatively, animals were made familiar with the maze for 15 min together with cage-mates and two times alone for 10 min on the following days. Postoperatively, they were given access to the eight-arm radial maze for 5 min with food pellets being scattered on the floor. The learning experiment started 10-12 days after surgery. Four of the eight arms were baited by a 45 mg food pellet (Noyes, New York) in a cup. The unbaited arms contained an empty food cup. Several extramaze cues were present in the well-lighted room and remained in a constant position during the course of the experiment. At the beginning of a trial the rat was placed at the end of a randomly chosen unbaited arm. The trial was finished when all four baits had been found. Blocks of five trials were given per day with an interval of approximately 30 s between the trials during which the animal was held in a separate cage. Arm visits were documented. It was distinguished between total, WM and RM errors. WM involves information that is constant only over a short period of time and is represented in a maze by the rule to learn "that each arm should only be visited once in a trial". A WM error is counted when a formerly visited arm is revisited. RM concerns information that remains constant over a longer period of time and means in a maze "to visit only the baited arms". A R M error is counted when an unbaited arm is entered for the first time but it is only noted from the second block of trials because the position of baited arms in relation to extramaze cues is at first unknown to the animals. The experiment consisted of an acquisition and a reversal phase. During acquisition 20 control and 20 lesioned animals were tested. The same arms were baited for one animal during ten blocks of trials (5 days a week). During reversal, half of the controls and half of the lesioned animals were challenged with dizocilpine (0.04 mg/kg i.p.) half an hour before testing. The formerly unbaited arms were changed to become baited arms and vice versa. The position of baited arms remained constant during seven blocks of trials.
Behavioural Brain Research 67 (1995) 255-261
2.5. Histology After completion of the learning experiment animals were deeply anesthetized by Nembutal and intracardially perfused by 4~o buffered paraformaldehyde. The brains were removed, postfixed, stored in 30~o sucrose at 4 °C and sectioned horizontally at 15/~m on a cryostat. Sections were Nissl-stained by Cresyl violet and microscopally analyzed.
257
bilaterally affected. Pre- and parasubiculum were in two cases unilaterally and in three cases bilaterally lesioned. Postoperative dizocilpine treatment had no effect on the lesion size. Fig. 2 shows photographs of Cresyl violetstained sections of control and lesioned animals. In controis (Fig. 2, top and bottom left), all layers of pyramidal neurons were intact. In lesioned animals (Fig. 2, top and bottom right), mainly neurons in the outer layers of the medial EC were missing and the lesioned area was characterized by a high amount of glia.
2.6. Statistics 3.2. Spatial learning Data were analyzed by using a two-way analysis of variance (ANOVA) followed, if necessary, by Tukey's t-test. Results were considered statistically significant at P < 0.05. Diagrams Fig. 3-5 show the accumulated number of errors in five trials expressed as mean + standard error of the mean (S.E.M.).
3. Results
3.1. Histology Microscopic examination of Cresyl violet-stained sections revealed that lesions were confined to the medial part of the EC and neither affected the hippocampus proper nor the subiculum. No fiber degeneration was observed. The EC was in no case entirely lesioned. Fig. 1 summarizes the smallest (crossed areas) and the largest extent (hhtched areas) of the lesion in different animals. The presubiculum was in two cases unilaterally and in one case
During the acquisition phase of the spatial learning experiment, EC-lesioned animals showed a learning deficit (Fig. 3) as indicated by a higher number of total errors (F= 28.02, df 9,399, P<0.0001). There was a significant effect of blocks (F=42.26, df 9,399, P<0.0001) but no significant interaction between treatment and blocks (F= 1.13, df 9,399), since the learning curves of the two groups were parallel. The learning deficit was due to a WM deficit that was prominent at the beginning but compensated at the end of the experiment (Fig. 4) and a RM deficit that became more evident at the end (Fig. 5). The WM deficit was indicated by an effect of treatment (F= 14.60, df 9,399, P < 0.01) and an interaction between treatment and blocks (F= 2.65, df 9,399, P < 0.01). There was also a significant effect of blocks (F= 21.60, df 9,399, P < 0.0001). The RM deficit was indicated by an effect of treatment (F= 7.72, df 8,359, P<0.01). There was a significant effect of blocks (F= 29.70, df 8,359, P<0.0001) but no interaction between blocks and treatment.
-6.6111111 -7.6 him
Fig. 1. Schematicillustration of quinolinic acid lesions of the medial entorhinalcortex. Summaryof smallest (crossed areas) and largest(hatched areas) lesion extents. DG, dentate gyms; EC, entorhinal cortex; PrS, presubiculum; PaS, parasubiculum; S, subiculum. Bregmalevels [24] are indicated on the bottom.
Behavioural Brain Research 67 (1995) 255-261
258
~:
~
Fig. 2. Cresyl-violet stained horizontal sections through the ventral hippocampus and parahippocampal cortices. Scale bar 500/~m. Top left: section of a control animal. All layers of pyramidal neurons in the entorhinal cortex and adjoining areas are intact. Bottom left: enlargement of the injection site in a control animal. Top right: section of a representative quinolinic acid lesioned animal. Pyramidal neurons in the medial part of the entorhinal cortex are missing. Neurons in the subiculum, pre- and parasubiculum and adjoining hippocampal areas are not affected. Bottom right: enlargement of the injection site in a lesioned animal. Glial cells are accumulated at the lesion site.
• L~
Acquisition ~
30
0
I
Reversal
: Les+Diz 30
Los
Con
',,
-~
A
Con
o
,.,,,,
+I}iz
,~ I.~
~
~0
20
,.~ 10
10
O I~
,~ O [.-,
"5 r
2
l
l
l
4
l
l
l
6
l
l
i
8 Blocks
i
i
1?.
10 of
T
i
~
14
16
~
~
i
18
5 trials
Fig. 3. Total errors in five trials (mean + S.E.M.) in the eight-arm radial maze. Acquisition phase: ten blocks of five trials of control (Con, N = 20) and lesioned animals (Les, N = 20), ANOVA P < 0.0001. Reversal phase: seven blocks of five trials of control (Con, N = 10), control plus dizocilpine-treated (Con + Diz, N = 10), lesioned (Les, N = 10) and lesioned plus dizocilpine-treated animals (Les + Diz, N = 10), ANOVA P < 0.001.
During the reversal phase, half of the animals of the
ance according
to their pretreatment of
treatment
as indicated
(F=21.28,
df
6,
by a
control group and of the lesioned group were pretreated
significant
effect
with dizocilpine. The groups showed a different perform-
P<0.001).
Tukey's t-test revealed no difference between
279,
259
Behavioural Brain Research 67 (1995) 255-261
2O
20
~
Acquisition
• L,,~
T
,~
0
~!..
(IOll
i
.~ 15
• L,.~+I~,~
Reversal
•
L~
~,
( ¢)n + [ ) l z
o
(',,,I
15
u~
~
10
I0
.
0
.
.
.
.
2
.
.
4
.
.
6
.
.
.
8
.
.
.
.
I0
Blocks
.
12
~
14
16
,
~
18
of 5 trials
F~g. 4. Working memory (WM) errors in five trials (mean ~ S . E M . ) in th~ c~ght-~m radial m ~ c . Abbreviations see Fig. 3. Acquisition phase: ANOVA P~0.01. Reversal phase: ANOVA ~ 0 . 0 0 0 1 .
Acquisition
15
• 0
u~
Reversal
• l,e~+l)iz
Con
•
I~'~
15
A ('ol~ ~ [ l i z o ('o1~
z. - -~,,,
"~,,,L~---i-..i
lO
i0
0
~
2
1
~
4
1
~
6
1
i
8
i
i
12
I0
Blocks of .5
i
14
i
i
16
i
~
i
18
trials
Fig. 5. Reference memory (RM) errors in five trials (mean _+S.E.M) in the eight-arm radial maze. Abbreviations see Fig. 3. Acquisition phase: ANOVA P<0.01. Reversal phase: ANOVA P<0.0001.
control, lesioned or control plus dizocilpine-treated animals but the lesioned plus dizocilpine-treated group had a strong learning deficit as indicated by a higher number of total errors (Fig. 3). There was an effect of blocks (F=74,70, df 6,279, P < 0 . 0 0 0 1 ) but no interaction between blocks and treatment ( F = 0.80, df 6,279). Correspondingly to the acquisition phase, the reversal deficit was due to a WM deficit at the beginning (Fig. 4) and a RM deficit at the end of the reversal phase (Fig. 5). All four groups made a high number of W M errors in the first block due to perseveration in those arms that had been baited in the acquisition phase. There was a significant effect of treatment (F= 9.52, df 6,279, P < 0 . 0 0 0 1 ) and blocks (F= 35.51, df 6,279, P<0.0001) on W M errors but not of treatment × blocks (F = 0.75, df 6,279). Although RM errors are only counted on the second
block, we also looked at this type of error in the first block of trials. ' R M type errors' were on a higher level in the first block of the reversal phase than in the acquisition phase, i.e., 14 instead of 11 ' R M type errors'. There was no difference between the four groups in the first block of trials (data not presented). Concerning the whole reversal phase, RM errors differed significantly in treatment (F = 24.49, df 5,239, P<0.0001) and blocks (F--29.44, df 5,239, P_<0.0001) but not in treatment × blocks (F= 0.75, df 5,239). The data pool was too little to calculate correlations between lesion extent and learning deficit. However, thorough examination of data revealed no influence of missing versus unilateral or bilateral lesion of the pre- and parasubiculum on acquisition or reversal of spatial learning.
260
Behavioural Brain Research 67 (1995) 255-261
4. Discussion In the present study, lesion of the medial EC led to a small deficit in acquisition of a spatial learning paradigm. Challenge by dizocilpine (0.04 mg/kg) of lesioned animals but not of controls in the reversal learning phase induced a strong deficit. In the acquisition phase, learning curves of both control and lesioned animals were parallel, approximated on the last block of trials and demonstrate a better compensation of WM than RM errrors. The RM deficit usually develops only later in the experiment because, on the first day of training, all animals have to learn about baited and non-baited arms and their positions. In contrast, animals are able to perform the correct WM procedure "to explore every arm only once in a trial" already on the first day of testing without knowing the specific task. According to Shapiro and Caramanos [30], the RM deficit is a true spatial learning deficit whereas WM can be also acquired by using intramaze-cues or a strategy such as to visit adjacent arms. This might have been the case for both lesioned animals in the acquisition phase and lesioned plus challenged animals in the reversal phase. The WM and RM memory deficits in the acquisition phase qualitatively match earlier studies [ 10]. Several reasons may account for the rather small memory deficit that we have observed. (1) All earlier studies showing essential spatial memory deficits in EC-lesioned animals used either electrolytic [10], radiofrequency [27] or aspiration lesion techniques [25]. In contrast, studies in which neurotoxins were applied did not reveal any deficit in spatial learning tasks. Bouffard and Jarrard [ 1] used ibotenate and demonstrated no deficit in the eight-arm radial maze when baited arms were changed every day. In a study of Rothblat et al. [25] NMDA-lesioned animals did not show a deficit in black-white discrimination in the Y-maze in contrast to aspiration-lesioned animals tested in the same task. We conclude that the more pronounced learning deficits in animals which had been lesioned with conventional techniques were mostly due to the damage of passing fibers. These fibers might be projections from the perirhinal cortex that are presumably more important for relaying spatial information to the hippocampus than neurons of the medial EC itself projecting to the hippocampus [20]. (2) There is some evidence for a functional dissociation of the medial and lateral EC. Myhrer [19] showed that only disruption of perforant path fibers originating in the lateral but not in the medial EC induced a deficit in a preoperatively acquired brightness discrimination task. The lateral EC receives strong projections from the perirhinal cortex which is said to be involved in recognition memory [28]. (3) Spatial tasks are claimed to be especially suited to reveal declarative memory deficits that correspond to the type of human amnesia following tem-
poral cortex lesions [27,31 ]. However, some studies provide evidence that these tasks are not the primary choice to investigate the nature of the EC in the rat [ 10,20,25]. Jarrard [ 10] and Rothblat et al. [25], for example, demonstrated that animals with either an aspiration or an NMDA-lesion display strong deficits in concurrent discrimination learning. The present study combined for the first time, to our knowledge, an EC lesion with a pharmacological challenge by a glutamate antagonist. In the reversal phase, dizocilpine definitely impaired lesioned animals whereas controls, challenged controls and non-challenged lesioned animals rapidly improved their perfomance. Lesioned animals showed a comparable RM deficit to the acquisition phase. Dizocilpine exacerbated particularly the lesioninduced RM deficit, i.e., the spatial component of the task, since WM errors continously decreased and RM errors remained on a higher level. These results are in accordance with studies of Shapiro and Caramanos [3,30]. They interpreted the dizocilpine-induced deficits in the way that acquisition of RM and learning of new contexts is especially dependent on N M D A receptors. Numerous articles report on learning deficits in acquisition of spatial [3,12,17,30,33] and non-spatial tasks [4,23] induced by dizocilpine. However, 0.04 mg/kg dizocilpine never induced any hyperlocomotion [14] and even higher doses did not impair performance in animals that were previously trained [3,30]. This is a strong argument against any interference of reasonable dizocilpine doses with motivational or attentional variables. The effectiveness of the low dizocilpine dose in lesioned animals strongly supports the hypothesis that the glutamatergic system was rendered more sensitive by the EC-lesion. What may cause hypersensitivity of the system? It is well known that lesion of the EC causes secondary compensatory changes in the dentate gyrus of the hippocampus. Sprouting of remaining afferents leads to a 70 ~o reinnervation of the perforant path termination zone and new synaptic contacts are formed within 30 days following deafferentiation [21]. Accordingly, compensation of the learning deficit in the present study may express reorganisation of the system in functional terms. In a recent study, traumatic brain-injured rats showed a corresponding sensitivity to dizocilpine treatment (0.1 mg/kg) as seen by an increased retention deficit in passive avoidance [5]. This enhancement of the amnesic dizocilpine effect was prevented by preoperative treatment with a higher dizocilpine dose (0.3 mg/kg). Thus, both quinolinic acid lesions of the EC and traumatic brain injury seem to produce a sensitization of the N M D A receptor system pointing to a common excitotoxic process. We conclude that, in contrast to earlier studies using conventional lesion techniques, a quinolinic acid lesion of
Behavioural Brain Research 67 (l 995) 255-26l
the medial EC induces only temporary and rather confined learning deficits. The WM deficit is more rapidly compensated whereas the RM develops during the course of the learning task. Since the learning deficit as such is rather small, this approach alone does not seem to provide a good animal model of AD. Challenge by the glutamate antagonist dizocilpine, however, resulted in an outstanding increase of the learning deficit which is presumably due to sensitization of the NMDA receptor system. Glutamate antagonists have been widely discussed for their beneficial effects in neuroprotection [16,18]. However, the present study provides more evidence for the rather detrimental effects of NMDA receptor antagonists, even in very low doses, when given late after brain injury.
References [I] Bouffard, J.P. and Jarrard, L.E., Acquisition of a complex place task in rats with selective ibotenate lesions of hippocampal formation: combined lesions of subiculum and entorhinal cortex versus hippocampus, Behav. Neurosci., 102 (1988) 828-834. [2] Braak, H. and Braak, E., Entorhinal-hippocampal interaction in mnestic disorders, Hippocampus, 3 (1993) 239-246. [3] Caramanos, Z. and Shapiro, M.L., Spatial memory and N-methylD-aspartate receptor antagonists APV and MK-801: memory impairments depend on familiarity with the environment, drug dose, and training duration, Behav. Neurosci., 108 (1994) 30-43. [4] Danysz, W. and Wroblewski, J.J.T., Amnesic properties of glutamate receptor antagonists, Neurosci. Res. Com., 5 (1989) 9-18. [5] Hamm, R.J., Pyke, B.R., O'Dell, M.S., Lyeth, B.G., Traumatic brain injury enhances the amnesic effect of an NMDA antagonist in rats, J. Neurosurg., 81 (1994) 267-271. [6] Heyes, M.P., Saito, K., Crowley, J.S., Davis, L.E., Demitrack, M.A., Der, M., Dilling, L.A., Elia, J., Kruesi, M.J.P., Lackner, A., Larsen, S.A., Lee, K., Leonard, H.L., Markey, S.P., Martin, A., Milstein, S., Mouradian, M.M., Pranzatelli, M.R., Quearry, B.J., Salazar, A., Smith, M., Strauss, S.E., Sunderland,T., Swedo, S.W., Tourtellotte, W.W., Quinolinic acid and kynurenine pathway metabolism in inflammatory and non-inflammatory neurological disease, Brain, 115 (1992) 1249-1273. [7] HOlscher, C. and Schmidt, W.J., Quinolinic acid lesion of the rat entorhinal cortex pars medialis produces selective amnesia in allocentric working memory (WM), but not in egocentric WM, Behav. Brain Res., 63 (1994) 187-194. [8] Hyman, B.T., Van Hoesen, G.W. and Damasio, A.R. (1990), Memory-related neural systems in Alzheimer's disease: an anatomic study, Neurology, 40 (1990) 1721-1730. [9] Jarrard, L.E., On the role of the hippocampus in learning and memory in the rat, Behav. Neural Biol., 60 (1993) 9-26. [10] Jarrard, L.E., Okaichi, H., Steward, O. and Goldschmidt, R.B., On the role of hippocampal connections in the performance of place and cue tasks: comparisons with damage to hippocampus, Behav. Neurosci., 6 (1984) 946-654. [11] Jones, R.S.G., Entorhinal-hippocampal connections: a speculative view of their function, Trends Neurosci., 16 (1993) 58-64. [12] Kant, G.J., Wright, W.L., Robinson III, T. and D'Angelo, P., Effects of MKo801 on learning and memory as assessed using a novel water maze, Pharmacol. Biochem. Behav., 39 (1991) 479-485. [ 13] Keseberg, U. and Schmidt, W.J., Tetrahydroaminoacridine attenuates dizocilpine-induced behavioural changes, Behav. Brain Res., 53 (1993) 113-118.
261
[14] Kretschmer, B.D., Zadow, B., Volz, T.L., Volz, L. & Schmidt, W.J., The contribution of the different binding sites of the N-methyl-D-aspartate (NMDA) receptor to the expression of behavior, J. Neural Transm., 87 (1992) 23-35. [ 15] Levisohn, L.F. and Isacson, O., Excitotoxic lesions of the rat entorhinal cortex. Effects of selective neuronal damage on acquisition of a non-spatial reference memory task, Brain Res., 564 (1991) 230-244. [16] Lipton, S.A., Rosenberg, P.A., Excitatory amino acids as a final common pathway for neurologic disorders, New Engl. J. Med., 330 (1994) 613-622. [17] McLamb, R.L., Williams, L.R., Nanry, K.P., Wilson, W.A. and Tilson, H.A., MK-801 impedes the aquisition of a spatial memory task in rats, Pharmacol. Biochem. Behav., 17 (1990) 41-45. [ 18] Meldrum, B. and Garthwaite, J., Excitatory amino acid neurotoxicity and neurodegenerative disease, Trends Pharmacol. Sci., 11 (1990) 379-387. [ 19] Myhrer, T., The role of medial and lateral hippocampal perforant path lesions and object distinctiveness in rats' reaction to novelty, Physiol. Behav., 42 (1988) 371-377. [20] Myhrer, T., Selective lesions in the temporal-hippocampal region of the rat: effects on acquisition and retention of a visual discrimination task, Behav. Neural. Biol., 58 (1992) 8-15. [21] Nitsch, R. and Frotscher, M., Transneuronal changes in dendrites of GABAergic parvalbumin-containing neurons of the rat fascia dentata following entorhinal lesion, Hippocampus, 3 (1993) 481490. [22] Olney, J.W., Excitatory amino acids and neuropsychiatric disorders, Biol. Psychiatry, 26 (1989) 505-525. [23] Parada-Turska, J. and Turski, W.A., Excitatory amino acid antagonists and memory: effect of drugs acting at N-methyl-Daspartate receptors in learning and memory, Neuropharrnacology, 29 (1990) 1111-1116. [24] Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic Press, San diego, 1986. [25] Rothblat, A.R., Vnek, N., Gleason, T.C. and Kromer, L.F., Role of the parahippocampal region in spatial and non-spatial memory: effects of parahippocampal lesions on rewarded alternation and concurrent object discrimination learning in the rat, Behav. Brain Res., 55 (1993) 93-100. [26] Sanger, D.J. and Joly, D. The effects of NMDA antagonists on punished exploration in mice, Behav. Pharmacol., 2 (1991) 57-63. [27] Schenk, F. and Morris, R.G.M., Dissociation between components of spatial memory in rats after recovery from the effects of retrohippocampal lesions, Exp. Brain Res., 58 (1985) 11-28. [28] Schouesboe, A., Bachevalier, J., Braak, H., Heinemann, U., Nitsch, R., Schneider, H. and Wetmore, C., Structural correlates and cellular mechanisms in entorhinal-hippocampal dysfunction, Hippocarnpus, 3 (1993) 293-302. [29] Schwarcz, R., Whetsee Jr., W.O., Maugano, R.M., Quinolinic acid: an endogenous metabolite that produces axon-sparing lesions in rat brain, Science, 219 (1983) 316-318. [30] 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. [31] Siesjr, B.K., Memezawa, H., Smith, M.L., Neurotoxicity: pharmacological implications, Fundam. Clin. Pharmacol., 5 (1991) 755767. [32] Squire, L.R., Declarative and nondeclarative memory: multiple brain systems supporting learning and memory, J. Cogn. Neurosci., 14 (1992) 232-243. [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.
BEHAVIOURAL BRAIN RESEARCH Behavioural Brain Research 67 (1995) 255-261
Research report
Low-dose challenge by the NMDA receptor antagonist dizocilpine exacerbates the spatial learning deficit in entorhinal cortex-lesioned rats Ute Keseberg*, Werner J. Schmidt Department of Neuropharmacology, Zoological Institute, Mohlstr. 54/l, D-72074 Tfibingen, Germany Received 25 May 1994; revised 6 October 1994; accepted 6 October 1994
Abstract
We investigated the effects of a bilateral quinolinic acid lesion of the medial entorhinal cortex (EC) on acquisition of a spatial learning task. During reversal of the same task, we challenged the animals by the N-methyl-D-aspartate (NMDA) receptor antagonist dizocilpine (MK-801). Training took postoperatively place in an eight-arm radial maze in which four of eight arms were baited. In the acquisition phase (ten blocks of five trials) of the test, EC-lesioned animals showed a working (WM) and a reference memory (RM) deficit. The WM deficit was prominent at the beginning and fully compensated at the end of the acquisition phase. The RM deficit became more evident during the course of the experiment. In the reversal learning phase (seven blocks of five trials), the formerly unbaited arms were baited and half of the control and lesioned animals were challenged by a low dose of dizocilpine (0.04 mg/kg i.p.) before training. Only lesioned and additionally dizocilpine-treated animals showed a WM deficit that was again compensated and a RM deficit that was stronger at the end of the test. In summary, quinolinic acid lesion of the medial EC induces both WM and RM deficits in rats. The WM deficit is rapidly compensated. Enhancement of these deficits by challenge with dizocilpine in the reversal learning phase suggests that the NMDA receptor system was rendered more sensitive by this type of lesion. Key words: Entorhinal cortex; Quinolinic acid; Excitotoxic lesion; Spatial learning; Eight-arm radial maze; MK-801; Challenge
1. Introduction
The entorhinal cortex (EC) is a major relay station for converging inputs from sensory and association cortices to the hippocampus and projections back from the hippocampal formation to the neocortex. Neurons from layer II of the EC project via the perforant path to the granule cell layer of the dentate gyrus, the first station of the trisynaptic hippocampal loop, and to the subiculum. Projections from layer III of the EC terminate in areas CA3 and CA1. Hippocampal output from CA1 is directed through the subiculum to deeper layers of the EC [11]. Thus the EC has anatomically a key position between the neocortex and the hippocampus. While the role of the hippocampus in spatial learning has been well documented in the last years, the contribution of the EC remains less clear [9,32]. According to a recent review of Jarrard [9] the hippocampus is stronger involved in working memory (WM) than reference me* Corresponding author. Fax: (49) (7071) 29-5017. E-mail: [email protected] 0166-4328/95/$9.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0 1 6 6 - 4 3 2 8 ( 9 4 ) 0 0 1 5 6 - 1
mory (RM) which is especially evident in the acquisition of the spatial but not of the intramaze-cue task in an eight-arm radial maze. In contrast, as Jarrard et al. [ 10] found in earlier electrolytic lesion studies, the EC seems to play a stronger role in RM than WM, at least in the place version of the task. Recent data following aspiration lesions of the EC [9], however, revealed no deficit at all in the place task and only a RM deficit in the intramaze cue task. Thus, there is considerable uncertainty of the functional role of the EC in acquisition of spatial learning due to different lesion methods. Our approach to this problem was therefore (1) to use a neurotoxin that specifically lesions somata leaving passing axons intact and (2) to lesion only a part of the EC instead of damaging the whole structure. We lesioned the more medial part of the EC by quinolinic acid. This excitotoxin is an endogenous metabolite of the tryptophan metabolism and destroys as an agonist of the N-methyl-D-aspartate ( N M D A ) receptor only somata without interrupting fibers of passage [29]. In addition, it is a possible candidate to contribute to excitotoxic damage in vivo [6]. In a previous study of our group [7] le-
256
Behavioural Brain Research 67 (1995) 255-26l
sion of the medial EC induced a deficit in allocentric (i.e., spatial) but not in egocentric (i.e., body movement) learning in the eight-arm radial maze. In the present study we prolonged the acquisition phase and additionally challenged the animals in the reversal phase of the task with a low dose (0.04 mg/kg) of the non-competitive N M D A antagonist dizocilpine (MK-801). The low dizocilpine dose has neither an effect on locomotion [14] nor on learning [23,26] and affects in higher doses only acquisition but not retention of learning tasks [3,17,30]. The challenge was undertaken in order to investigate whether lesioned animals become more sensitive to this treatment. The present results might be of relevance for human neurodegenerative diseases. EC-lesion studies have been often discussed in relation to Alzheimer's disease (AD) (see for example [8,15,19,20,28]), since layer II of the human EC is the first brain area to show neurofibrillary tangles and develops the greatest number of tangles per unit area in the brain of AD patients [2,8]. Damage in cerebral ischemia and traumatic brain injury is induced via similar excitotoxic mechanisms as quinolinic acid lesions [22,31 ]. N M D A receptor antagonists such as dizocilpine are able to protect from excitotoxicity but have at the same time unfavorable effects on cognition [16,18]. These aspects will be considered in the discussion of the present paper. 2. Materials and Methods 2.1. Subjects
Fourty male Sprague-Dawley rats (Interfauna, Tuttlingen), weighing 210-260 g, were used as experimental animals. They were housed in groups of 4 - 6 and were maintained on a 12/12 h light/dark cycle, lights being on from 06.00 h to 18.00 h. Water was freely available but food was restricted to a daily diet of 12 g per animal so that the animals held approximately 85~/o of their free feeding weight. They were fed after completion of the experiment. 2.2. Drugs
Quinolinic acid (Sigma, Deisenhofen) was dissolved in 0.1 M phosphate buffered saline and adjusted to pH 7.4 by 1 M NaOH. Chloralhydrate (Fluka, Neu-Ulm), atropine (Sigma, Deisenhofen) and dizocilpine (( + )-MK-801, RBI, KOln) were dissolved in 0.9~o saline. Atropine (s.c) and dizocilpine (i.p.) were injected in a volume of 1 ml/kg. 2.3. Surgery
Animals were anesthetized with 4 ~o choralhydrate (400 mg/kg i.p.) and treated with atropine sulfate (0.2 mg/animal s.c.) before and after surgery. Surgery was per-
formed in a stereotaxic Stoelting instrument. Stereotaxic coordinates [24] were calculated from bregma: anterioposterior -8.0, mediolateral + 3.5. Injections of 0.5 #1 0.1 M PBS (control, N = 20) or 60 nmol quinolinic acid (lesion, N = 20) were bilaterally given under a mediolateral angle of 15 ° at three different dorsoventral levels: 0.2 #1 at -7.5, 0.2/~1 at -7.0 and 0.1/~1 at -6.0 [7]. The injection syringe was connected by a tube to a hamilton syringe that contained either the neurotoxin or vehicle. 2.4. Spatial learning
We used the same eight-arm radial maze as in previous studies [ 7,13 ]. Preoperatively, animals were made familiar with the maze for 15 min together with cage-mates and two times alone for 10 min on the following days. Postoperatively, they were given access to the eight-arm radial maze for 5 min with food pellets being scattered on the floor. The learning experiment started 10-12 days after surgery. Four of the eight arms were baited by a 45 mg food pellet (Noyes, New York) in a cup. The unbaited arms contained an empty food cup. Several extramaze cues were present in the well-lighted room and remained in a constant position during the course of the experiment. At the beginning of a trial the rat was placed at the end of a randomly chosen unbaited arm. The trial was finished when all four baits had been found. Blocks of five trials were given per day with an interval of approximately 30 s between the trials during which the animal was held in a separate cage. Arm visits were documented. It was distinguished between total, WM and RM errors. WM involves information that is constant only over a short period of time and is represented in a maze by the rule to learn "that each arm should only be visited once in a trial". A WM error is counted when a formerly visited arm is revisited. RM concerns information that remains constant over a longer period of time and means in a maze "to visit only the baited arms". A R M error is counted when an unbaited arm is entered for the first time but it is only noted from the second block of trials because the position of baited arms in relation to extramaze cues is at first unknown to the animals. The experiment consisted of an acquisition and a reversal phase. During acquisition 20 control and 20 lesioned animals were tested. The same arms were baited for one animal during ten blocks of trials (5 days a week). During reversal, half of the controls and half of the lesioned animals were challenged with dizocilpine (0.04 mg/kg i.p.) half an hour before testing. The formerly unbaited arms were changed to become baited arms and vice versa. The position of baited arms remained constant during seven blocks of trials.
Behavioural Brain Research 67 (1995) 255-261
2.5. Histology After completion of the learning experiment animals were deeply anesthetized by Nembutal and intracardially perfused by 4~o buffered paraformaldehyde. The brains were removed, postfixed, stored in 30~o sucrose at 4 °C and sectioned horizontally at 15/~m on a cryostat. Sections were Nissl-stained by Cresyl violet and microscopally analyzed.
257
bilaterally affected. Pre- and parasubiculum were in two cases unilaterally and in three cases bilaterally lesioned. Postoperative dizocilpine treatment had no effect on the lesion size. Fig. 2 shows photographs of Cresyl violetstained sections of control and lesioned animals. In controis (Fig. 2, top and bottom left), all layers of pyramidal neurons were intact. In lesioned animals (Fig. 2, top and bottom right), mainly neurons in the outer layers of the medial EC were missing and the lesioned area was characterized by a high amount of glia.
2.6. Statistics 3.2. Spatial learning Data were analyzed by using a two-way analysis of variance (ANOVA) followed, if necessary, by Tukey's t-test. Results were considered statistically significant at P < 0.05. Diagrams Fig. 3-5 show the accumulated number of errors in five trials expressed as mean + standard error of the mean (S.E.M.).
3. Results
3.1. Histology Microscopic examination of Cresyl violet-stained sections revealed that lesions were confined to the medial part of the EC and neither affected the hippocampus proper nor the subiculum. No fiber degeneration was observed. The EC was in no case entirely lesioned. Fig. 1 summarizes the smallest (crossed areas) and the largest extent (hhtched areas) of the lesion in different animals. The presubiculum was in two cases unilaterally and in one case
During the acquisition phase of the spatial learning experiment, EC-lesioned animals showed a learning deficit (Fig. 3) as indicated by a higher number of total errors (F= 28.02, df 9,399, P<0.0001). There was a significant effect of blocks (F=42.26, df 9,399, P<0.0001) but no significant interaction between treatment and blocks (F= 1.13, df 9,399), since the learning curves of the two groups were parallel. The learning deficit was due to a WM deficit that was prominent at the beginning but compensated at the end of the experiment (Fig. 4) and a RM deficit that became more evident at the end (Fig. 5). The WM deficit was indicated by an effect of treatment (F= 14.60, df 9,399, P < 0.01) and an interaction between treatment and blocks (F= 2.65, df 9,399, P < 0.01). There was also a significant effect of blocks (F= 21.60, df 9,399, P < 0.0001). The RM deficit was indicated by an effect of treatment (F= 7.72, df 8,359, P<0.01). There was a significant effect of blocks (F= 29.70, df 8,359, P<0.0001) but no interaction between blocks and treatment.
-6.6111111 -7.6 him
Fig. 1. Schematicillustration of quinolinic acid lesions of the medial entorhinalcortex. Summaryof smallest (crossed areas) and largest(hatched areas) lesion extents. DG, dentate gyms; EC, entorhinal cortex; PrS, presubiculum; PaS, parasubiculum; S, subiculum. Bregmalevels [24] are indicated on the bottom.
Behavioural Brain Research 67 (1995) 255-261
258
~:
~
Fig. 2. Cresyl-violet stained horizontal sections through the ventral hippocampus and parahippocampal cortices. Scale bar 500/~m. Top left: section of a control animal. All layers of pyramidal neurons in the entorhinal cortex and adjoining areas are intact. Bottom left: enlargement of the injection site in a control animal. Top right: section of a representative quinolinic acid lesioned animal. Pyramidal neurons in the medial part of the entorhinal cortex are missing. Neurons in the subiculum, pre- and parasubiculum and adjoining hippocampal areas are not affected. Bottom right: enlargement of the injection site in a lesioned animal. Glial cells are accumulated at the lesion site.
• L~
Acquisition ~
30
0
I
Reversal
: Les+Diz 30
Los
Con
',,
-~
A
Con
o
,.,,,,
+I}iz
,~ I.~
~
~0
20
,.~ 10
10
O I~
,~ O [.-,
"5 r
2
l
l
l
4
l
l
l
6
l
l
i
8 Blocks
i
i
1?.
10 of
T
i
~
14
16
~
~
i
18
5 trials
Fig. 3. Total errors in five trials (mean + S.E.M.) in the eight-arm radial maze. Acquisition phase: ten blocks of five trials of control (Con, N = 20) and lesioned animals (Les, N = 20), ANOVA P < 0.0001. Reversal phase: seven blocks of five trials of control (Con, N = 10), control plus dizocilpine-treated (Con + Diz, N = 10), lesioned (Les, N = 10) and lesioned plus dizocilpine-treated animals (Les + Diz, N = 10), ANOVA P < 0.001.
During the reversal phase, half of the animals of the
ance according
to their pretreatment of
treatment
as indicated
(F=21.28,
df
6,
by a
control group and of the lesioned group were pretreated
significant
effect
with dizocilpine. The groups showed a different perform-
P<0.001).
Tukey's t-test revealed no difference between
279,
259
Behavioural Brain Research 67 (1995) 255-261
2O
20
~
Acquisition
• L,,~
T
,~
0
~!..
(IOll
i
.~ 15
• L,.~+I~,~
Reversal
•
L~
~,
( ¢)n + [ ) l z
o
(',,,I
15
u~
~
10
I0
.
0
.
.
.
.
2
.
.
4
.
.
6
.
.
.
8
.
.
.
.
I0
Blocks
.
12
~
14
16
,
~
18
of 5 trials
F~g. 4. Working memory (WM) errors in five trials (mean ~ S . E M . ) in th~ c~ght-~m radial m ~ c . Abbreviations see Fig. 3. Acquisition phase: ANOVA P~0.01. Reversal phase: ANOVA ~ 0 . 0 0 0 1 .
Acquisition
15
• 0
u~
Reversal
• l,e~+l)iz
Con
•
I~'~
15
A ('ol~ ~ [ l i z o ('o1~
z. - -~,,,
"~,,,L~---i-..i
lO
i0
0
~
2
1
~
4
1
~
6
1
i
8
i
i
12
I0
Blocks of .5
i
14
i
i
16
i
~
i
18
trials
Fig. 5. Reference memory (RM) errors in five trials (mean _+S.E.M) in the eight-arm radial maze. Abbreviations see Fig. 3. Acquisition phase: ANOVA P<0.01. Reversal phase: ANOVA P<0.0001.
control, lesioned or control plus dizocilpine-treated animals but the lesioned plus dizocilpine-treated group had a strong learning deficit as indicated by a higher number of total errors (Fig. 3). There was an effect of blocks (F=74,70, df 6,279, P < 0 . 0 0 0 1 ) but no interaction between blocks and treatment ( F = 0.80, df 6,279). Correspondingly to the acquisition phase, the reversal deficit was due to a WM deficit at the beginning (Fig. 4) and a RM deficit at the end of the reversal phase (Fig. 5). All four groups made a high number of W M errors in the first block due to perseveration in those arms that had been baited in the acquisition phase. There was a significant effect of treatment (F= 9.52, df 6,279, P < 0 . 0 0 0 1 ) and blocks (F= 35.51, df 6,279, P<0.0001) on W M errors but not of treatment × blocks (F = 0.75, df 6,279). Although RM errors are only counted on the second
block, we also looked at this type of error in the first block of trials. ' R M type errors' were on a higher level in the first block of the reversal phase than in the acquisition phase, i.e., 14 instead of 11 ' R M type errors'. There was no difference between the four groups in the first block of trials (data not presented). Concerning the whole reversal phase, RM errors differed significantly in treatment (F = 24.49, df 5,239, P<0.0001) and blocks (F--29.44, df 5,239, P_<0.0001) but not in treatment × blocks (F= 0.75, df 5,239). The data pool was too little to calculate correlations between lesion extent and learning deficit. However, thorough examination of data revealed no influence of missing versus unilateral or bilateral lesion of the pre- and parasubiculum on acquisition or reversal of spatial learning.
260
Behavioural Brain Research 67 (1995) 255-261
4. Discussion In the present study, lesion of the medial EC led to a small deficit in acquisition of a spatial learning paradigm. Challenge by dizocilpine (0.04 mg/kg) of lesioned animals but not of controls in the reversal learning phase induced a strong deficit. In the acquisition phase, learning curves of both control and lesioned animals were parallel, approximated on the last block of trials and demonstrate a better compensation of WM than RM errrors. The RM deficit usually develops only later in the experiment because, on the first day of training, all animals have to learn about baited and non-baited arms and their positions. In contrast, animals are able to perform the correct WM procedure "to explore every arm only once in a trial" already on the first day of testing without knowing the specific task. According to Shapiro and Caramanos [30], the RM deficit is a true spatial learning deficit whereas WM can be also acquired by using intramaze-cues or a strategy such as to visit adjacent arms. This might have been the case for both lesioned animals in the acquisition phase and lesioned plus challenged animals in the reversal phase. The WM and RM memory deficits in the acquisition phase qualitatively match earlier studies [ 10]. Several reasons may account for the rather small memory deficit that we have observed. (1) All earlier studies showing essential spatial memory deficits in EC-lesioned animals used either electrolytic [10], radiofrequency [27] or aspiration lesion techniques [25]. In contrast, studies in which neurotoxins were applied did not reveal any deficit in spatial learning tasks. Bouffard and Jarrard [ 1] used ibotenate and demonstrated no deficit in the eight-arm radial maze when baited arms were changed every day. In a study of Rothblat et al. [25] NMDA-lesioned animals did not show a deficit in black-white discrimination in the Y-maze in contrast to aspiration-lesioned animals tested in the same task. We conclude that the more pronounced learning deficits in animals which had been lesioned with conventional techniques were mostly due to the damage of passing fibers. These fibers might be projections from the perirhinal cortex that are presumably more important for relaying spatial information to the hippocampus than neurons of the medial EC itself projecting to the hippocampus [20]. (2) There is some evidence for a functional dissociation of the medial and lateral EC. Myhrer [19] showed that only disruption of perforant path fibers originating in the lateral but not in the medial EC induced a deficit in a preoperatively acquired brightness discrimination task. The lateral EC receives strong projections from the perirhinal cortex which is said to be involved in recognition memory [28]. (3) Spatial tasks are claimed to be especially suited to reveal declarative memory deficits that correspond to the type of human amnesia following tem-
poral cortex lesions [27,31 ]. However, some studies provide evidence that these tasks are not the primary choice to investigate the nature of the EC in the rat [ 10,20,25]. Jarrard [ 10] and Rothblat et al. [25], for example, demonstrated that animals with either an aspiration or an NMDA-lesion display strong deficits in concurrent discrimination learning. The present study combined for the first time, to our knowledge, an EC lesion with a pharmacological challenge by a glutamate antagonist. In the reversal phase, dizocilpine definitely impaired lesioned animals whereas controls, challenged controls and non-challenged lesioned animals rapidly improved their perfomance. Lesioned animals showed a comparable RM deficit to the acquisition phase. Dizocilpine exacerbated particularly the lesioninduced RM deficit, i.e., the spatial component of the task, since WM errors continously decreased and RM errors remained on a higher level. These results are in accordance with studies of Shapiro and Caramanos [3,30]. They interpreted the dizocilpine-induced deficits in the way that acquisition of RM and learning of new contexts is especially dependent on N M D A receptors. Numerous articles report on learning deficits in acquisition of spatial [3,12,17,30,33] and non-spatial tasks [4,23] induced by dizocilpine. However, 0.04 mg/kg dizocilpine never induced any hyperlocomotion [14] and even higher doses did not impair performance in animals that were previously trained [3,30]. This is a strong argument against any interference of reasonable dizocilpine doses with motivational or attentional variables. The effectiveness of the low dizocilpine dose in lesioned animals strongly supports the hypothesis that the glutamatergic system was rendered more sensitive by the EC-lesion. What may cause hypersensitivity of the system? It is well known that lesion of the EC causes secondary compensatory changes in the dentate gyrus of the hippocampus. Sprouting of remaining afferents leads to a 70 ~o reinnervation of the perforant path termination zone and new synaptic contacts are formed within 30 days following deafferentiation [21]. Accordingly, compensation of the learning deficit in the present study may express reorganisation of the system in functional terms. In a recent study, traumatic brain-injured rats showed a corresponding sensitivity to dizocilpine treatment (0.1 mg/kg) as seen by an increased retention deficit in passive avoidance [5]. This enhancement of the amnesic dizocilpine effect was prevented by preoperative treatment with a higher dizocilpine dose (0.3 mg/kg). Thus, both quinolinic acid lesions of the EC and traumatic brain injury seem to produce a sensitization of the N M D A receptor system pointing to a common excitotoxic process. We conclude that, in contrast to earlier studies using conventional lesion techniques, a quinolinic acid lesion of
Behavioural Brain Research 67 (l 995) 255-26l
the medial EC induces only temporary and rather confined learning deficits. The WM deficit is more rapidly compensated whereas the RM develops during the course of the learning task. Since the learning deficit as such is rather small, this approach alone does not seem to provide a good animal model of AD. Challenge by the glutamate antagonist dizocilpine, however, resulted in an outstanding increase of the learning deficit which is presumably due to sensitization of the NMDA receptor system. Glutamate antagonists have been widely discussed for their beneficial effects in neuroprotection [16,18]. However, the present study provides more evidence for the rather detrimental effects of NMDA receptor antagonists, even in very low doses, when given late after brain injury.
References [I] Bouffard, J.P. and Jarrard, L.E., Acquisition of a complex place task in rats with selective ibotenate lesions of hippocampal formation: combined lesions of subiculum and entorhinal cortex versus hippocampus, Behav. Neurosci., 102 (1988) 828-834. [2] Braak, H. and Braak, E., Entorhinal-hippocampal interaction in mnestic disorders, Hippocampus, 3 (1993) 239-246. [3] Caramanos, Z. and Shapiro, M.L., Spatial memory and N-methylD-aspartate receptor antagonists APV and MK-801: memory impairments depend on familiarity with the environment, drug dose, and training duration, Behav. Neurosci., 108 (1994) 30-43. [4] Danysz, W. and Wroblewski, J.J.T., Amnesic properties of glutamate receptor antagonists, Neurosci. Res. Com., 5 (1989) 9-18. [5] Hamm, R.J., Pyke, B.R., O'Dell, M.S., Lyeth, B.G., Traumatic brain injury enhances the amnesic effect of an NMDA antagonist in rats, J. Neurosurg., 81 (1994) 267-271. [6] Heyes, M.P., Saito, K., Crowley, J.S., Davis, L.E., Demitrack, M.A., Der, M., Dilling, L.A., Elia, J., Kruesi, M.J.P., Lackner, A., Larsen, S.A., Lee, K., Leonard, H.L., Markey, S.P., Martin, A., Milstein, S., Mouradian, M.M., Pranzatelli, M.R., Quearry, B.J., Salazar, A., Smith, M., Strauss, S.E., Sunderland,T., Swedo, S.W., Tourtellotte, W.W., Quinolinic acid and kynurenine pathway metabolism in inflammatory and non-inflammatory neurological disease, Brain, 115 (1992) 1249-1273. [7] HOlscher, C. and Schmidt, W.J., Quinolinic acid lesion of the rat entorhinal cortex pars medialis produces selective amnesia in allocentric working memory (WM), but not in egocentric WM, Behav. Brain Res., 63 (1994) 187-194. [8] Hyman, B.T., Van Hoesen, G.W. and Damasio, A.R. (1990), Memory-related neural systems in Alzheimer's disease: an anatomic study, Neurology, 40 (1990) 1721-1730. [9] Jarrard, L.E., On the role of the hippocampus in learning and memory in the rat, Behav. Neural Biol., 60 (1993) 9-26. [10] Jarrard, L.E., Okaichi, H., Steward, O. and Goldschmidt, R.B., On the role of hippocampal connections in the performance of place and cue tasks: comparisons with damage to hippocampus, Behav. Neurosci., 6 (1984) 946-654. [11] Jones, R.S.G., Entorhinal-hippocampal connections: a speculative view of their function, Trends Neurosci., 16 (1993) 58-64. [12] Kant, G.J., Wright, W.L., Robinson III, T. and D'Angelo, P., Effects of MKo801 on learning and memory as assessed using a novel water maze, Pharmacol. Biochem. Behav., 39 (1991) 479-485. [ 13] Keseberg, U. and Schmidt, W.J., Tetrahydroaminoacridine attenuates dizocilpine-induced behavioural changes, Behav. Brain Res., 53 (1993) 113-118.
261
[14] Kretschmer, B.D., Zadow, B., Volz, T.L., Volz, L. & Schmidt, W.J., The contribution of the different binding sites of the N-methyl-D-aspartate (NMDA) receptor to the expression of behavior, J. Neural Transm., 87 (1992) 23-35. [ 15] Levisohn, L.F. and Isacson, O., Excitotoxic lesions of the rat entorhinal cortex. Effects of selective neuronal damage on acquisition of a non-spatial reference memory task, Brain Res., 564 (1991) 230-244. [16] Lipton, S.A., Rosenberg, P.A., Excitatory amino acids as a final common pathway for neurologic disorders, New Engl. J. Med., 330 (1994) 613-622. [17] McLamb, R.L., Williams, L.R., Nanry, K.P., Wilson, W.A. and Tilson, H.A., MK-801 impedes the aquisition of a spatial memory task in rats, Pharmacol. Biochem. Behav., 17 (1990) 41-45. [ 18] Meldrum, B. and Garthwaite, J., Excitatory amino acid neurotoxicity and neurodegenerative disease, Trends Pharmacol. Sci., 11 (1990) 379-387. [ 19] Myhrer, T., The role of medial and lateral hippocampal perforant path lesions and object distinctiveness in rats' reaction to novelty, Physiol. Behav., 42 (1988) 371-377. [20] Myhrer, T., Selective lesions in the temporal-hippocampal region of the rat: effects on acquisition and retention of a visual discrimination task, Behav. Neural. Biol., 58 (1992) 8-15. [21] Nitsch, R. and Frotscher, M., Transneuronal changes in dendrites of GABAergic parvalbumin-containing neurons of the rat fascia dentata following entorhinal lesion, Hippocampus, 3 (1993) 481490. [22] Olney, J.W., Excitatory amino acids and neuropsychiatric disorders, Biol. Psychiatry, 26 (1989) 505-525. [23] Parada-Turska, J. and Turski, W.A., Excitatory amino acid antagonists and memory: effect of drugs acting at N-methyl-Daspartate receptors in learning and memory, Neuropharrnacology, 29 (1990) 1111-1116. [24] Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic Press, San diego, 1986. [25] Rothblat, A.R., Vnek, N., Gleason, T.C. and Kromer, L.F., Role of the parahippocampal region in spatial and non-spatial memory: effects of parahippocampal lesions on rewarded alternation and concurrent object discrimination learning in the rat, Behav. Brain Res., 55 (1993) 93-100. [26] Sanger, D.J. and Joly, D. The effects of NMDA antagonists on punished exploration in mice, Behav. Pharmacol., 2 (1991) 57-63. [27] Schenk, F. and Morris, R.G.M., Dissociation between components of spatial memory in rats after recovery from the effects of retrohippocampal lesions, Exp. Brain Res., 58 (1985) 11-28. [28] Schouesboe, A., Bachevalier, J., Braak, H., Heinemann, U., Nitsch, R., Schneider, H. and Wetmore, C., Structural correlates and cellular mechanisms in entorhinal-hippocampal dysfunction, Hippocarnpus, 3 (1993) 293-302. [29] Schwarcz, R., Whetsee Jr., W.O., Maugano, R.M., Quinolinic acid: an endogenous metabolite that produces axon-sparing lesions in rat brain, Science, 219 (1983) 316-318. [30] 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. [31] Siesjr, B.K., Memezawa, H., Smith, M.L., Neurotoxicity: pharmacological implications, Fundam. Clin. Pharmacol., 5 (1991) 755767. [32] Squire, L.R., Declarative and nondeclarative memory: multiple brain systems supporting learning and memory, J. Cogn. Neurosci., 14 (1992) 232-243. [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.