Functional hemidecortication by spreading depression or by focal epileptic discharge disrupts spatial memory in rats

Behavioural Processes, Elsevier

10 (1985) 387-398

387

FUNCTIONAL HEMIDECORTICATION BY SPREADING DEPRESSION OR BY FOCAL EPILEPTIC DISCHARGE DISRUPTS SPATIAL MEMORY IN RATS E. PANAKHOVA', 0. BURESOVA and J. BURES Institute of Physiology, Czechoslovak Academy of Sciences, Prague, Czechoslovakia 1 Institute of Physiology, Azerbaidzhanian

Academy of Sciences,

Baku, USSR (Accepted16 April 1984) ABSTRACT Panakhova E., Buregova 0. and Bureg J., 1985. Functional hemidecortication by spreading depression or oy focal epileptic discharge disrupts spatial memory in rats. Behav. Processes10: 387-398. The importance of neocortex for the acquisition and retrieval of the water tank navigation task has been examined in 110 hooded rats. The animals were trained to swim to a small (10 cm in diameter) submersed platform 1 cm below the surface of a large pool (120 cm in diameter) of opaque water. In Experiment 1,naive rats with unilateral cortical spreading depression (SD) elicited by application of a filter paper soaked with 25% KC1 on one hemicortex were unable to find the platform during 1 min in about 50% of the 12 acquisition trials. The performance of functionally hemidecorticated rats did not improve after a single trial with the intact brain, but escape latencies were significantly shortened in rats given 2,6 or 60 pretraining trials with intact brain. Even in the overtrained rats escape latencies under unilateral SD were significantly lon er (14 set) than in intact rats after 12 acquisition trials f 5 set). In Experiment 2, development of an epileptic focus established by local penicillin application onto the exposed cortical surface was electrophysiologically monitored. Regular interictal discharge (0.2 to 0.5 Hz) disrupted acquisition of the navigation task and interfered to a lesser degree with performance in rats given 12 and 36 pretraining trials with the intact brain. The occipital foci were more disruptive than the frontoparietal ones. Performance recovered with cessation of the epileptic discharge. It is concluded that spatial memory mediating navigation in the water tank task requires coordinated activity of both cerebral hemispheres. INTRODUCTION Research into the brain mechanisms of spatial behavior has been recently stimulated by the recognition of the importance of memory mechanisms for foraging behavior. Evolutionary pressures have equipped animals with species specific neural mechanisms which allow them to optimize visits to a multitude a small capacity feeders (Kamil, 1978) or to remember position of hoarded food hidden in thousands of cachets (Shettleworth, 1983). Although laboratory tests based on the above behaviors proved to 0376-6357/85/%03.30 o 1985Elsevier Science Publishers

B.V. (Biomedical

Division)

388

be extremely useful (Olton and Samuelson, 1976; Olton et al., 1979), close relationship of this type of spatial memory with feeding is an important disadvantage. The tests cannot be performed under conditions which decrease the appetitive motivation'of the animal. This may seriously limit usefulness of the radial maze (Olton and Samuelson, 1976) for examination of spatial memory under the effect of drugs, brain lesions or various functional interferences. The experimental possibilities were considerably improved by the introduction of a new water escape task (Morris, 19811, which tests the capability of rats to find a small submersed island in a large pool of opaque water Since immersion into water is a highly arousing stimulus which triggers escape attempts even in severely depressed animals (Ranye and Ungerstedt,

19771, the test is well suited for

examining spatial memory under conditiors incompatible with radial maze performance. The purpose of the present study is to examine the effect of cortical spreading depression

(Leao, 1944;

Bureg et al., 1974) or of a cortical epileptic focus (Woodruff, 1974) on the rats' performance in the water tank task. METHOD Animals, One hundred ten 3-month-old male hooded rats of the Druckras strain were used. The-v were housed 5 oer cage in an animal room with natural light: Food and water-were Freely available in the home cage. The circular swimming pool (120 cm in diameter and 60 Y* cm high was filled 40 cm deep with 25OC water made opaque by adding 1 1 milk per 100 1 water. A clear plastic platform (10 cm in diameter) covered with white cloth and mounted on a heavy supporting column was placed 1 cm below water surface (invisible platform). The visible platform was a brown plastic disk (10 cm in diameter) with a 0.5 cm high rim protruding 1 cm above water surface. The pool was arbitrarily divided into four compass quadrants and the island was always placed at the midpoint of one of the cardinal radii (N,E,S,W) of the basin. The pool was located in a room with a large window, a door and shelves with various pieces of laboratory equipment visible from the water surface. Procedure. The rats were first given a 3 air, swimming test in the pool with the island removed. Training started on the next day with the visible or invisible island in a fixed position (e.g. NJ. The rat was dropped into water by hand so that it faced the plastic wall of the pool at the end point of one of the remaining radii (e.g. E,S or W). When the animal did not find the island during 60 set, it was placed upon it by hand. After 30 set on the island the animal was started agai,. from a different position. Up to 12 trials were given on Day 2 and in some groups training continued for several days. The escape latencies were measured with 0.5 set accuracy. The number of

389

trials given before and after various treatments is indicated in the Result section. Functional decortication. The animal was anesthetised with ether and a trephine opening 4 mm in diameter was made above occipital or frontoparietal cortex of one hemisphere. The cut skin was joined with a single loose suture and the animal was returned to the home cage. Functional decortication was achieved by eliciting repeated waves of spreading depression in one hemisphere. The animal was restrained by hand, the suture was cut and the filter paper (2x2 mm) soaked with 25% KC1 was applied on the exposed dura. The trephine opening was than closed with a low nylon cap (Fig.lA) the serrated elastic rim of which fitted exactly into the hole in the bone and held tightly to its circumference. The cap protected the exposed brain surface against contact with water during the subsequent swimming test. The presence of spreading depression was verified by the absence of cortical postural reflexes (BureS et al., 1976) on the side contralateral to the depressed hemicortex before and after the water ~001 tests. Eoiieptic focus. In the epileptic focus experiments, trephine ooenings were DreDared in the same wav as for eliciting soreading depressio;. bn the following day"the animal was restrained, the suture was cut and about 50 ug of powdered Ka-penicillin was applied onto the exposed cortical surface. The trephine opening was than closed with a plastic cap with two golden plated electrodes (Fig.lB) fixed inside the cap close to its rim. The electrodes were used to Dick UD bioolar EEG across the focus. They were connected through two-pin Amphenol connector and a flexible cable to the input of a conventional polygraph. EEG activity was monitored until a typical interictal discharge of The cable was than disconnected the penicillin focus developed. and the animal was tested in the water tank. Immediately after the test, EEG was sampled again. In some experiments the cap was removed,. the cerebral-surface was washed with saline and after replacement of the cap, EEG recording continued until the epileptic activity completely disappeared. Animals with similar trephine openings to which filter papers with physiological saline were applied instead of KC1 or penicillin, served as operated controls.

Fig.1. The plastic caps used in the spreading depression (A) and epileptic focus (B) experiments to protect the exposed cerebral cortex during swimming.

EESULTS Experiment 1. The effect of functional hemidecortication. Pilot experiments with bilateral spreading depression showed, that functionally decorticated naive animals (n=lO) swim normal&v (Islam and BureSova, 1975) but that they are unable to find the visible island and to escape onto it. They continued to climb the smooth wall: of the plastic tank, swam around its circumference and almost never tried to explore the interior of the pool. When placed after 60 set of swimming on the island, they did not stay on it but fell or jumped into water and did not even attempt to climb back. No improvement of this behavior was seen during twelve successive trials. s

_ pw

A

Wo’

SD

w

C

“: L OJ

r

1

.

1

1

3

1 4

3

l

w

% Lo1

sD

uo :\/

¶o-

0

C

Al

1

, 1

I

.

2

3

1

4

L 1

I

3

l

1 1

I

.

1

3

s l

Fig.2. The effect of unilateral spreading depression on performance in the water tank task. N - naive animals; T(l), T(2), T(6), T(60) - animals receiving 1,2,6 or 60 pretraining trials, respectively. A: average escape latencies 2 SEM in blocks of three trials. B: percentage of trials in which the animals failed to escape during 60 sec. The SD and C curves in sectio:. i; correspond to functionally hemidecorticated and intact animals, respectively. The dotted lines connect the learni!lg curves in the sectio:s T(1) to T(60) of part A with points indicating average escape latency in the last pre-CSD trialis).

391

Whereas

iiltact rats (r=lO) rapidly decreased average escape

laterlcies from 24 s in the first 3 trials to 4 s in trials 10 to 12 (Fig.2A, group N-C), .laive fu:.ctio.!allyhemidecorticated rats (n-10) were also severely impaired. They explored the ir.terior of the pool a::d occasionally

found arid climbed upon the

icvisible platform, but escape latencies did I ot improve during the 12 trials (Fig.&,

significantly

group N-SD). The rats

failed to fi!:d the invisible platform irlmore than half of the trials. A t!vo-way allalysis of variance
indicated that escape latency decreased

faster ir, the ir_tact thai. irA the hemidecorticated rats. In the other 3 groups of 10 animals each, the operated rats were allowed on Day 2 to filld the visible platform 1,2 or 6 times, respectively.

Four to 6 hours later- spreading depression was

elicited in one hemicortex and traini::g continued for 12 trials. krAother group (i.=lO) received 5 sessions of 12 daily trials on Days 2 to 6, was operated after the last session on Day 6 and .tested under unilateral spreading depression. or; Day 7. Fig.2A shows that the experience received with the intact grain has considerably

improved performance

of the hemidecorticate

al:imals;

Whereas orie active trial did l;ot influence the learning curve, two trials caused a clear decrease of average latencies in trials 7 to 12. Similar results were obtaiy,ed with 6 pretrair:ing trials. Long overtrainirlg with intact brain (60 trials) caused only a slight further improvement of the performance of hemidecorticated rats. Two-v;ay a..alysis of variar.ce (pretraining level x blocks) with repeated measures 0~; the second factor indicated sig;.ificant mair: effects of the treatment (F:4,45)=56.9, pqO.001) arld blocks :~(3,135)=114.8, prO.OO1) as well as significant ii:teraction ;F(12,135j=9.1, prO.01). The effect of pretraiiing was still better expressed when only trials i..;:hich the a-imals failed to reach the islarld during GO SCC 'siG2T-C t&kC?Yl iTit ecourt (Fig.2u!. Their i:.cidence was 55% ir. the fut.ctior,allyhemidecorticate naive rats. It was .ot reduced by or,e ;retrai;_ir:gtrial, but dropped to 8% after 2 trials and to zero after still longer 2retrai*_ing with the intact brain. The differe::oe betweeri the groups receivir,g 1 ar.d 2 pretraining

392

trials was significant at the r, 0.01 level

(binomial

compari-

son). The interfering effect of u,,ilateral SD was appare1.t eveIL ir, the long overtrained group. Compariso

of performance or. the

predepression Day 6 with the performa:,ce of ful.ctionally heG.decorticated rats or, Day 7 i:dicates that the escape latt:.cies were significantly larger u.:der the latter corditioul ;t(9)z3.5, prO.01,

paired t-test). The overtreir.ed al,imnls ?:itlnunilateral

spreading depressio!l actively explored the poo; and eventually found platform but apparently did not remember its exact position. Experiment 2.

The effect of ur:ilateral epileptic focus.

Fig. 3 shows examples of the epileptic discharge recorded il, two rats before and after testing i.. the water tarlk. Development of interictal activity irL the occipital or fro:;toparietal penicillin foci was not accompanied by naliifest cofivulsio:.sor jerking movements. Spontaneous locomotio!, was reduced, however a:.d the animals had a tendericy to assume a crouchi::g posture in the corner of the observation box.

Fig. 3. Examples of the il:terictal epileptic activity recorded in overtrained rats from the penicillin focus immediately befor~t: (left recordings) and after (right recordings) the water ta1.k test. F - frontal focus. 0 - occipital focus. The :.umbers i:.dicate average escape latencies (iI. see) before agplicatio!. of penicillir- and durir:g the i!iterictal activity. Rote that the occipital focus impairs the water tar:k task performance more than the frontal focus. Calibration: 2 n;7, 30 sec. Naive animals (!,=lO)T>;ithir.terictal activity ra:?gi:g from 7 to 35 spikes/mil. (median 12 spikes/mir.) were severely i:n;aired in the acquisitio.: of the water tar:k tas:k. They swam !:ormally but were ur,able to find the submersed isla;:d. They stayed close to the wall of the tack, repeated attempts to climb the wall of the pool a:id only exceptionally ventured inside the pool. '//heI.

393

placed or:to the submersed platform by har,d, the rats stayed upor, it but did not stand up and explore their surroundir.gs. i)nly 2 rats were able to find the platform consistently,

although

with rather long latency. None of the remaining 8 rats succeeded to escape at least once. Average escape latencies are shown in Fig. 4. ti,

EF

w

a0 -

T C 20

O-

-

L I

1

1

9

r

I

.

1

2

3

a

1

2

3

N

102)

I *

I 1

I

I

2

3

3 a blocks

0106)

Fig. 4. The effect of unilateral epileptic focus 01: performance in the water tank task. I\I- naive animals; T(12) a:.d r)T(36) traiyLed and overtrained a:limals after 12 and 36 pretraining trials, respectively. Ordinate - average escape latency 2 SEM. Abscissa: blocks of 3 trials. The upper curves (EF) correspond to rats with epileptic foci, the lower curves (C) to a group of similarly traiiled intact rats. There are 10 animals per group. Another group of 10 rats received 12 trials irl the pool in the morrli:Lgand was tested with ur:ilateral epileptic focus in the afterl.oor,.The acquisitio:, with intact brain proceeded normally and tkie escai;c latencies averaged 5.1 2 0.8 s in trials 7 to 12. The >e?licilli,!discharge severely disrupted performance of spatial z,emory. As short i)-1 Fig. 4, the average escape latel.cies :'.c:'e r,ot sigr.ifica..tlydiffere:lt from those of the : aive alLirals. The 2er%ce:.t:17ge of failures to find the submersed Flatfor: ..,lthi.. 60 s d.ro?pc-?, 50 5776, ho.,:cvc-, a:.: o:.ly 2 :-at?

394

failed to find the platform i:. a11 12 trials. The above trend was still better expressed in rats (n=lO) which were trained to find the island ir: the same position on 3 consecutive days (3x12 daily trials). Their escape latezcies averaged 5.5 + 0.8 s ir. the last sessio:.. The trephine o;7er.ir:gs were made after completiorl of the third trair-ing .;essio.;ar,d the epileptic focus was elicited or. the :.ext d&y. As show_. in Fig. 4, the penicillin focus was less disruptive i. the overtrained animals than in naive rats. All animals succeeded to escape ar.d the average escape latencies decreased from 31.0 2 40.0 s itI trials l-3 to 19.3 2 3.5 s i:, trials 10 to 12. The percentagt of trials in which the rats failed to fir.d the submersed platform within the 60 s limit dropped to 15%. The performarLce in the second half of trials was still considerably worse than that of trained intact controls and ever1 of naive colitrols (Fig. 41.

B

II, Irc rlo: IjFlu 4

F

0

F

0

F

0

F

0

F

NW

0 010

F

Fig. 5. Comparison of the effect of frontoparietal :FI a1.d occipital (0) epileptic foci on the performance in the water tank task. N - naive, T(12) - trained and OTC361 - overtrairled rats. A: average escape lateroies 2 SEM. ~3: percerltage of trials in which the animals failed to escape duririg 60 sec. Each column is based on 60 trials in 5 animals. Half of the animals in each group had the focus in the occipital and half in the frorltoparietal cortex. As showr. ir: Fig. 5 the impairment caused by foci localized i. these two areas was not significantly different in the rlaive and trair.ed animals. In the overtrained group, occipital foci impaired performance more than the frontal ones (t(8) = 3.16, p

0.91:.

After completion of testing under the influence of per.icilli::,

395

the trephine opening was washed with physiological

saline in the

overtrained group. When no epileptic activity had been detected in the EEG for at least 10 min, the animals were tested in the pool again. Fig. 6 shows that their performance was almost undistinguishable

from that shown by the same animals on the

preceding day.

bib

40

20 T

0

f 1

I

1

i

t

2

3

4

1

03

.

I

2 3 Db- Pne

t

t

4

1

I

I

2 3 04.kl

i

4 blocks

Fig. 6. The reversibility of the disruptive effect of interictal epileptic activity in overtrained rats (n=lO). Ordinate: average escape latencies + SEM. D3: last day of intact brain pretraining. D4-Pnc: performance during interictal focal discharge. D4-Sal: performance after removal of penicillin and cessation of the epileptic activity several hours later. DISCUSSION The present study shows that unilateral functional decortication or presence of unilateral cortical epileptic focus interferes with acquisition and retrieval of place navigation in the Morris water tank task. This effect can be due to disruption of a neocortical link in the spatial memory mechanisms, to blockade of cortical inputs to non-neocortical

centers (hippo-

campus) or to interfering activity projected from the neocortical focus to remote brain regions. Although lesion studies confirmed

the indispensability

of hippocampus for the water tank task

(Morris et al., 1982; Sutherland

et al., 1982; Sutherland

et al.,

19831, recent evidence indicates important contributions

of other

cortical (Kolb et al., 1981, 1982, 1983) and subcortical

(Milner

and Lines, 1983) sites to spatial orientation. Kolb et a1.(1983) demonstrated complete disruptio:- of the water tank task by total removal of neocortex. The effect was not less expressed in rats highly overtrained before decortication

and the task could not

be relearned by decorticated animals. The effect of bilateral cortical spreading depression in the present study esselitially confirms the above reports. The effects of unilateral lesions were studied less frequent-

ly. Kolb et al.

(1981)

reported that unilateral decortication

interferes with acquisition of the water tank task and that left hemisphere lesions produce a more severe deficit than similar lesion in the right hemicortex. Partially

impaired acquisition

of the water tank task was found in rats with unilateral

colchi-

tine lesions of fascia dentata by Sutherland et al. (1983). The authors suggest that accurate spatial mapping requires availability of inputs from both left and right entorhinal cortices. Functional hemidecortication

induces deafferentation

of the

ipsilateral dentate gyrus and eliminates also the medial frontal cortex, which was shown to play an important role in spatial localization

(Kolb et al., 1983). The more marked effect of uni-

lateral SD as compared with unilateral lesion most probably reflects the time available for compensation: whereas SD effect

was acute, the effect of lesion was tested only 3 weeks after the injection of colchicine. The differential effect of functional hemidecortication

on

naive and trained animals is consonant with the view (Sutherland et al., 1983) that the bilateral input to the dentate gyrus is not necessary for spatial performance well practised prior to the lesion. The present results indicate that already two active trials improve the rats' performance during subsequent hemidecortication. On the other hand, even after prolonged pretrairing (60 trials), escape latencies remained longer in the hemidecorticated than in the intact control rats and approached values typical for experiments in which platform location is randomly changed from trial to trial (Morris, 1981). It seems, therefore, that the overtrained functionally hemidecorticated

rats were

397

able to

Solve

the task by systematically

visiting the possible

platform locations, but that they did not use a true mapping strategy. Since epileptic focus can be viewed as a functional ablation procedure

(Woodruff, 1974),

it

is

not surprisirLg that it affects

place ILavigation similarly as SD. The finding that acquisition in naive a:?imals was disrupted more than performance in overtrair,ed rats is in agreement with the report indicating that focal seizures elicited by application

of aluminum hydroxide

onto the occipital cortex of monkeys interfere with visual discrimination learning more than with the performance of the already learned task (iSraft et al., 1960). Daily injections of penicillin

into the occipital cortex of rats co!lsiderably re-

tarded acquisition of an operant brightness discrimination but caused o_ly a partial impairment of performance

task

in rats

previously overtrained with intact brail, (Woodruff et al., 1974). Bquir;o-Cias et al. (1972) four,d acquisitio:l of a passive avoidance task impaired by focal penicillin discharge in the occipital cortex of oile hemisphere. Even though the discharge spreads far beyond the region of the site of the focus is important.

peinicillin applicatior,

Absence of differences in the effects of from.toparietal and occipital foci was probably due to a ceiling effect both in naive and acutely trained a..imals. In the overtrair,ed rats occipital foci were significantly more disruptive than the froiAtoparietal ones. Place navigation requires complex interpretation of distal visual cues a:.d cannot be mastered by blind rats (Sutherland et al., 1982). It is conceivable that the peLcil1i.r. spikes in the visual cortex disrupt the high level processing of visual input required for competent spatial orientation. The performance of overtrained rats with penicillin foci did not improve in the course of 12 trials, probably because the rats used a non-mapping

strategy based on an effi-

cient search pattern, but were uzable to swim directly to the platform. The interference explav?atior:is also supported by the observatio- that the focal activity leaves rAo lasting aftereffects, but that the impairment disappears shortly after the cessatio!. of the ir:ternictaldischarges.

398

REFERENCES Aquino-Cias, J., Aneirox-Riba, R., Ferliandez-Yero, F. a!:d Hernandez-Mesa, X., 1972. Effects of epileptic foci in the visual cortex of the rat or. passive avoidance learning. Physiol. dehav., 8: 957-961. BureS, J., &re8ova, 0. and,K?iv&ek, J., 1974. The Mechar.ism and Applications of Leao s Spreading Depressioli of Electror encephalographic Activity. Academic Press, New York, 412~~. Islam, S., and BureSova, O., 1975. The effect of cortical spreading depression on motor performance and depth avoidar.ce in rats. Activitas nervosa superior, 17: S-14. Kamil, A.C., 1978. Learning and memory ir, the wild: Systematic foraging by a nectar feedir.g bird, the Amakihi (Loxops wirens). J. Comp. Physiol., 92: 388-396. Kolb, 13., Sutherlarid, R.J. and Whishaw, I.Q., 1981. The relative contribution of cortical site arid side in the cox:trol of praxic and spatial behavior. Sot. Neurosci. tibstr., 7: 521. Kolb, B., Pittman, K., Sutherlar,d, R.J. and Whishaw, I.Q., 1982. Dissociation of the contribution of the prefrontal cortex alli: dorsomedial thalamic nucleus to spatially guided behavior in the rat. Behav. arail, Res., 6: 365-378. Kolb, B., Sutherland, R.J. and Whisha:-, I.Q., 1983. A conparisor of the contributions of the frontal a::d parietal association cortex to spatial localizatio? ir: rats. dehav. Xeurosci., 1: 13-27. Kraft, h?.S., Obrist, W.D. and Pribram, K.H., 196C. The effect of irritative lesions of the striate cortex on learning ana visual discriminations in monkeys. J. camp. physiol. Psychol., 53: 17-22. Leao, A.A.P., 1944. Spreadiu.g depression-!of activity i:. the cerebral cortex. J.Neuro;;hysiol., 7: 359-390. Milner, A.D., and Lii:es, C-R., 1983. Stimulus sampli:&g and the use of distal visual cues in rats with 1esior.s of the superior colliculus. Rehav. &ail! Res., 8: 367-461. Morris, R.G.M., 1981. Spatial localizatior; does r,ot require the presence of local cues. Learning and Motivation, 12: 239-263. Morris, R.G.M., Garrud, P., Rawlirls, J. and O"Xeefe, J., 1932. Place rlavigation impaired iI1 rats rvith hippocampal lesio:.s. Nature (Land.), 297: 681-683. Olton, D.S., aecker, J.T. at:d Her,delmaA,,G.E., 1979. Iiippocar.:)us, space and memory. aehav. &ail? Sci., 2: 313-322. Clton, D.S. and Samuelson, R.J., 1976. Remembrances of places passed: Spatial memory in rats. J. 2xp. Psychol., 2: 97-116. Railye, C. and Uclgerstedt, U., 1977. Lack of acquisition in dopamine denervated animals tested i:, the underwater Y-maze. Brain Res., 134: 95-111. Shettleworth, S.J., 1983. Memory ir, food-h0ardir.g birds. Scie,,tific Amer., 248:102-110. Sutherland, R.J., Kolb, Y. and Whishaw, I.Q., 1982. Spatial by hippocampal or medial mapping: Definitive disruptio frontal cortical damage in the rat. Neurosci. Lett., 31: 271-276. Sutherland, R.J., Whishaw, I.&. a.:dKolb, d., 1983. A beha.jioural al,alysis of spatial localization followi:g electrolytic, kainate- or colchicice-induced damage to the hippocampal formation in the rat. behav. arain Res., 7: 133-153. Woodruff, M.L., 1974. Subcol:vulsive epileptiform discharge a,.d behavioral imoairment. aehav. 3iol.. 11: 431-458. Woodruff, M.L., Kearley, R.C. al?d Isaacson, R.L., 1974. Deficiert bright::ess discrimination acquisition produced by daily i:.tracranial ir:jectiors of per:icillir:in rats. aehav. siol., 12: 445-460.