Induction of long-term depression and potentiation by ... - Science Direct

Brain Res'earch, 4(15(19S7i 10!! 107 Elscvic~

l[l(I BRE 12389

Induction of long-term depression and potentiation by low- and high-frequency stimulation in the dentate area of the anesthetized rat: magnitude, time course and EEG* Clive R. Bramham and Bolek Srebro Department of Physiology, Universityof Bergen, Bergen (Norway) (Accepted 15 July 1986)

Key words. Long-term depression: Long-term potentiation; Perforant path; Dentate area; Tetanization frequency: Spreading depression

We investigated the possible importance of stimulus train frequency for the induction and magnitude of long-term synaptic plasticity in the perforant path-granule cell pathway. Under the same experimental conditions, low- (15 Hz) or high-frequency (400 Hz) stimulation could elicit a profound long-term depression (LTD), or typical long-term potentiation (LTP), of the population spike amplitude, excitatory postsynaptic potential (EPSP) amplitude and spike onset latency. In addition, changes in the relationship between the EPSP and population spike amplitude indicated that granule cell excitability was enhanced during LTP and reduced during LTD. LTD occurred primarily after low-frequency stimulation (5 of 6 cases), and was always accompanied by striking changes in the EEG. most notably a biphasic slow potential. While the EEG changes were confined to the first 5 min after the tetanus. LTD lasted from 1 to 4 h, The nature of the EEG events is still unclear, it is suggested that they may represent a spreading depression-like episode. Finally, we found that LTP evoked by high-frequency stimulation was larger and generally reached peak magnitude faster than when it followed low-frequency stimulation. A possible mechanism and role for hippocampal LTD is proposed.

INTRODUCTION Brief trains of electrical stimuli applied to the hippocampal pathways can generate a long-term potentiation (LTP) of evoked potentials which lasts for hours or more. No consensus has been reached on the mechanism of LTP for any specific pathway despite the abundant interest in LTP as a model of synaptic plasticity or neuronal ' m e m o r y ' (see refs. 6, 37 and 38 for review). So far, there have been only a few studies on LTP in relation to the parameters of afferent stimulation14'2°'24'31"~°; a circumstance perhaps best reflected by the diversity of stimulus frequencies used. Indeed, train frequencies ranging from 0.2 to 500 Hz have been reported to evoke LTP in the perforant path-granule cell pathway 7"12"13'27'33"36. The possibility arises, therefore,

that the descriptive

properties of LTP, and by implication the mechanisms, may at least partly depend on the frequency of afferent firing. In the Schaffer c o l l a t e r a l - C A t pathway low-frequency trains (20 Hz) can evoke a remarkable longterm depression (LTD) of synaptic transmission 29'~, although others have used similar stimulation condition to produce LTP 14,3t,36. The reason for this apparent discrepancy is unclear, and the mechanism and possible significance of hippocampal L T D is unknown. Recently, however, neuronal calcium accumulation and enhanced L-glutamate binding, two events central to a hypothesis of a postsynaptic origin of LTP 5'22, were reported to occur during L T D 29'3°. It has also been suggested that some form of spreading cortical depression might be involved in L T D 4, but this question needs to be examined electrographical-

* A preliminary account of this work has been reported elsewhere 9. Correspondence: C.R. Bramham, Department of Physiology, University of Bergen, N-5000 Bergen, Norway. 0006-8993/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)

101

ly in intact animals. In the present study we investigated the possible importance of stimulus train frequency for the induction and magnitude of two types of long-term synaptic plasticity, LTP and LTD, in the hippocampus. Low- (15 Hz) and high-frequency (400 Hz) trains were delivered to the perforant pathway and field potentials recorded from the dentate granule cell layer. The E E G was continuously monitored on a lowbandpass filter to allow detection of the slow, sustained potentials which characterize spreading depression. MATERIALS AND METHODS Thirty-five male Sprague-Dawley rats weighing 250-350 g were anesthetized with chloral hydrate/ pentobarbital (17 mg/kg and 3.9 mg/kg i.p., respectively). Animals were closely monitored and given supplemental injections to ensure a deep level of surgical anesthesia. Body temperature was kept between 36.5 and 37.5 °C by an electrical heating pad. Bipolar stimulating electrodes (SNEX-100, Rhodes Medical Instruments) were positioned in the angular bundle (coordinates: 7.9 mm posterior to bregma and 4.1 mm from the midline). Stainless-steel, insulated insect pin (size 000) recording electrodes were then slowly lowered into the hippocampus (coordinates: 3.9 mm posterior to bregma and 2.4 mm from the midline) until stimulation elicited a characteristic field potential, with a maximal population spike in the granule cell body layer or just below in the dentate hilus. The depth of the stimulating electrode was then adjusted to give a minimal current threshold for eliciting the population spike. Two stainless-steel screws fixed to the frontal bones served as a reference recording electrode and electrical ground. The hippocampal electrode and screws were connected to a source-follower amplifier with field effect transistor (FET) input stages. The signal from the hippocampal electrode was fed into an AC preamplifier, passed through a low-bandpass filter (0.15-75 Hz; Grass 7P5), and recorded on polygraph paper. The low-pass filter was used to detect slow, sustained potentials in the EEG. The amplifier output was then fed into a Tektronix 468 Digital Storage Oscilloscope where the evoked field potentials were averaged, measured and photographed. The compo-

nents of the field potential were measured as follows: the amplitude of the population spike from peak negativity to midway between the first and second positive peaks; the excitatory postsynaptic potential (EPSP) at a fixed latency near the middle of the first positive wave; and the spike latency from the stimulus artifact to the onset of the population spike. Test stimuli consisted of biphasic, constant-current pulses with a half-width of 150 ~s. A stable baseline was established by measuring the components of an average field potential (8 sweeps at 0.05 Hz) every 5 min until a 45-min period of steady potentials was observed. The mean of the values obtained during the stable period was taken as a measure of baseline amplitude. Animals whose response did not stabilize or fluctuated beyond 10% of baseline were discarded from testing. After baseline measurements, animals received a stimulus train of 100 pulses delivered at either 15 Hz (low-frequency stimulation; LFS) or 400 Hz (high-frequency stimulation; HFS). The type of train given to each animal was selected at random. Post-train response changes were followed by collecting averages (4 sweeps) at fixed time intervals for at least 1 h and up to 5 h. The current intensity (for trains and test stimuli) was selected, in each animal, from the middle of the stimulus strength-response function (input-output curves) in order to evoke a submaximal population spike. Intensities used ranged from 200 to 500/~A and were about 150 ~A above spike threshold. Data for input-output curves

TABLE I Effects of low- and high-frequency tetanization of the perforant path on the direction of long-term changes in dentate population spike amplitude

The effectsof 15- and 400-Hzstimulation were significantlydifferent on the basis of the ~2 test (;(2= 6.8; P < 0.05). The magnitude of the change refers to the mean of the population spike amplitudes recorded at 6 fixed times during the first hour after the tetanus. Train frequency (lOOpulses)

15 Hz 400 Hz

n

Number of animals and (range of the magnitude of change) Long-term potentiation

16 ' 5 (36 to 103) 15 9 (44 to 249)

Long-term depression

No change

5 6 (-28 to-75) (-11 to 12) 1 5 (-70) (-13 to 5)

102

were collected during pretrain baseline and about 30 rain after the tetanus. Plots of the relation between EPSP and population spike amplitude served to measure changes in granule cell excitability (after Andersen et al.3). The position of the recording electrode was verified histologically after testing. Changes in dentate field potentials measured following tetanization were classified as long-term potentiation (LTP), long-term depression (LTD), or no change. The amplitude criterion for LTP and LTD was set for each type of response measure in percent of baseline, as follows: population spike _+20%, EPSP __+15%, and spike onset latency -+5%. The criterion for a long-term change was met when a response measure exceeded the amplitude criterion for 15 min, starting at least 10 min after the train. Response values obtained at 5, 10, 15, 30, 45 and 60 min post-train were averaged as an estimate of the magnitude of LTP in each animal, and expresed in percent of baseline. The following statistical tests were used: Student's t-test, Z2 and Kendall Rank Correlation r.

RESULTS The effects of low- and high-frequency stimulation (LFS and HFS, respectively) on the induction of long-term synaptic plasticity are summarized in Table I. Under carefully controlled experimental conditions, both types of stimulation could produce a long-term depression (LTD) or characteristic longterm potentiation (LTP) of evoked field potentials. Furthermore, train frequency seemed to be an important parameter for determining the direction of the response change (2'2 = 6.8, df = 2, P < 0.05). Thus, while HFS primarily evoked LTP, LFS produced LTD and LTP in an equal number of cases. About 35% of the animals showed no long-term response change. The effects of train frequency on the magnitude of long-term response changes in population spike amplitude are illustrated by representative graphs in Fig. 1A, along with traces of the corresponding dentate field potentials (Fig. 1B). Analysis of the posttrain time courses revealed that the magnitude of

A

B PRE-TRAIN

&5 MIN POST-TRAIN

400 Hz LTP ¢-

0 J:l

---..e 15 Hz LTP 100 ~o

~s

&00 Hz LTD 15 Hz LTD

,If ~.00 Hz

LTP ~

5o 15Hz

E 0 ~

2! v

0

1 Ores

C

9

Q

&00 Hz .~

-5o

LTD

C

TRAIN

g C

2

u

-ioo

TRAIN~o

1 20

I 40

1 ~.x 60

I 90

l 100

; 115

Tlrrm ( r n i n )

Fig. 1. Long-term potentiation and depression of the dentate area population spike followinglow- (15 Hz) and high- (400 Hz) frequency stimulation of the perforant path. A: sample post-train time courses illustrating LTD, and the effect of train frequency on the magnitude of LTP. Note also the abrupt 'reversal' of LTD by a 400-Hz train delivered to the perforant path 90 rain after the first train. B: traces of averaged field potentials recorded pre- (8 sweeps) and post-train (4 sweeps) from each animal plotted in A.

103 LTP was significantly greater in the group receiving HFS (Student's t-test; P < 0.05). The average magnitude of LTP was 127% in the HFS group compared to only 52% after LFS (the ranges for each group are indicated in Table I). In addition, the rise-time to the maximum LTP value was often faster after HFS; 4 out of 9 animals reached peak LTP within 1 min after HFS, whereas none did after LFS. While,any possible differential effects of train frequency on the magnitude of LTD could not be assessed with only one case of LTD in the HFS group, several observations concerning the time course and properties of LTD were made. The depression of the field potential always started abruptly, within the first minute after the train, and lasted between 1 and 4 h. After recovery from depression, the response measures remained near the baseline amplitude for the rest of the experiment (i.e. up to 2 h); no second phase of depression or potentiation was detected. Furthermore, it should be pointed out here, that a 400-Hz train given to the perforant path during established LTD (two cases) resulted in a rapid and sustained recovery of the potential, to the original baseline level or above (see Fig. 1A). It was also possible to elicit typical LTP after both spontaneous and tetanus-induced recovery of the potential. In another experiment, we examined the role of the number of stimuli per train for the induction of LTP and LTD. Trains in which the total number of pulses was reduced from 100 to 8 produced LTP but not LTD (n -6), thus implying that prolonged or excessive stimulation might be involved in LTD (unpublished observation). The results described above for the population spike pertain equally to the EPSP amplitude and spike latency, as both of these measures significantly correlated with changes in population spike amplitude (Kendall Rank Correlation r, P < 0.05). Despite this, there was a consistent disproportion between the strength of the synaptic drive (EPSP) and the size of the sub sequeht population spike. The population spike was small in relation to the EPSP during LTD, but large during LTP. Thus, in addition to the changes in synaptic efficacy, it appears that the excitability of granule cells is reduced during LTD and enhanced in LTP. Fig. 2 illustrates this point. It has recently been reported that selective tetanization of the lateral component of the perforant path

• [.5 MIN POST-TRAIN

-6

• PRE'TRAIN

A

v

i. *'-'~

B

•& . ~ - "

06

,j/LTD

P c O

l

I

I

I

02

I

I

06

I

I

10

I

I

02

I

[

06

10

EPSP emplitude (normolized)

Fig. 2. Plots of the relationship between EPSP and population spike amplitude taken before and after a 15-Hz train. The curve is shifted to the left during LTP (case A) and to the right during LTD (case B). Each point is the average of 8 responses. Maximum spike heights were 6.6 and 7.2 mV for A and B, respectively.

can evoke an LTD of the EPSP in the adjacent (heterosynaptic) medial perforant pathway 1.2. An interesting property of this heterosynaptic LTD was that it could not be induced during LTP of the medial pathway 2. Fig. 3 indicates that this was not the case for the EPSP depression described here, LTD could still be evoked during established LTP. Inspection of the E E G records revealed characteristic E E G changes in all 6 cases of LTD (Fig. 4A). Prominent among these was a sustained potential (SP) shift starting between 5 and 45 s post-train and lasting for 10-15 s. This shift was typically biphasic; consisting of a 5-15-mV negative wave followed by a similar positive wave. A single 10-15-s burst of paroxysmal afterdischarges also occurred. However, AO0 HZ

15 Hz

TRAIN

TRAIN

i

oo o o

E

0

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

-40

w

c

-80 o

L)

L

I

L

L

I

20

40

60

80

lOO

Time (rain)

Fig. 3. Induction of long-term depression of the EPSP amplitude by a 15-Hz train given during established LTP. Two cases (Q and O) are presented. Each point represents the average of 4 sweeps. The EPSP amplitude was measured at a fixed latency on the rising phase.

1()4 there was no strict temporal relationship between these afterdischarges and the SP shift. In fact afterdischarges could start before or after the shift. Finally, a short-lasting depression of the E E G amplitude and frequency was identified in all cases of LTD. This E E G depression began during the SP shift but always returned to normal within 5 rain. No E E G abnormalities could be detected after this time. With regard to a possible causal relationship between the early E E G events and LTD, it is important to note that L T D did not occur when afterdischarges and E E G depression appeared alone (5 cases), without the typical SP shift (see Fig. 4B). Thus it seems unlikely that the combination of epileptiform discharges and E E G depression per se can precipitate LTD. Although we could not precisely determine the onset of L T D in relation to the various E E G events, responses to single test-pulses indicated that field potentials could be suppressed before any overt af-

terdischarge activity began, and almost completely disappeared at the time of the SP shift (Fig, 4C). In contrast to the above, LTP was generally not ass()ciated with changes in the EEG. Afterdischarges and E E G depression occurred in only two of 14 cases of LTP, one of which is illustrated in Fig. 4B. DISCUSSION The results of this study suggest that the train frequency parameter is important in determining the direction of long-term changes in synaptic transmission and neuronal excitability in the perforant p a t h - g r a nule cell pathway. Long-term depression (LTD) occurred primarily after low-frequency stimulation while high-frequency stimulation usually produced LTP, in agreement with studies from the Schaffer collateral CA] pathway 14'29. However, the fact that the same train frequency could evoke both L T D and

C

A

PRE-TRAIN SP%,

/,oo

t Hz TRAIN Afterdischarges ,

Aft_erdischcl_.._.~es

.SP~.~,.

POST-TRAIN 30-35s 5mm

5 MIN POST.TRAINIomv~ ~ j

EEG-depression 4'

8mY ~ ~ .

5ms

4~

'

15HzTRAIN B

Afterdischarges

EEG-depression

D 5mY L

15HzTRAIN

•

,,

45

Fig. 4. A and B: low-pass EEG records from the granule cell body layer of the dentate area before and after tetanization of the perforant path. A continuous trace from the first minute post-train is followed by a 10-s segment recorded 5 rain later. A: typical traces from animals which demonstrated LTD. Slow potentials (SP) occurred only in cases of LTD. Examples of paroxysmal afterdischarges starting before (lower trace) or after the slow potential (upper trace). Note the recovery of the EEG at 5 rain post-train. B: trace from an animal which developed LTP illustrating aflerdischarges and EEG depression, withouLan SP shift. C and D: changes in evoked dentate field potentials recorded from animals in A and B. Test pulses delivered at 30-35 s (single pulse) and 5 rain (4 sweeps) after the train are indicated by an asterisk in A and B. The potential was almost flat during the slow potential. In the upper EEG trace, the potential was suppressed before overt afterdischarges occurred.

105 LTP, under seemingly identical experimental conditions, clearly suggests that factors other than the specific train parameters are involved in the induction of synaptic plasticity in the hippocampus. The LTD described here seems to be different from heterosynaptic LTD in the dent.ate area 1'2, since it: (1) occurs in the tetanized pathway; (2) involves a reduction in granule cell excitability; and (3) can be induced during established LTP. Perhaps the most significant finding with regard to LTD was its consistent association with changes in the E E G during the first 5 min after the train. Inspection of the E E G records revealed that the polarity and amplitude of the sustained potential shift, the duration of the EEG depression and the pattern of the epileptiform spiking are similar to changes reported in 'spreading' cortical depression (SD) of neocortica115A6'21'23 and hippocampal areas 18'26'34. The only apparent difference in our experiment was the brevity of the sustained potential shift, which lasted 10-15 s rather than 1-2 min. There are several possible explanations for this discrepancy. It could simply be because we used AC rather than DC signal amplification. However, the fact that SD is a graded reaction of a neuronal population may be important. Typical stimuli used to provoke SD experimentally, such as repetitive tetanization and direct cortical application of KCI, all involve intense neuronal excitation 16'21'26'34. Consequently, the abbreviated shift recorded in our experiments may well reflect our use of relatively moderate stimulation conditions. Alternatively, the observed E E G changes may depict seizure activity and postictal depression, instead of SD. Although we cannot dismiss this possibility, our results do not favor such an interpretation, since LTD did not occur when afterdischarge activity and EEG depression appeared without the slow potential shift. In fact, LTP could be observed under these conditions. We therefore suggest that the E E G changes associated with LTD may represent a brief, spreading depression-like event. If LTD and SD are somehow connected, this might help to clarify some of the peculiar properties of hippocampal LTD. The tendency for LTD to follow lowfrequency stimulation, for instance, may be explained by a report that SD in the dentate area is optimally induced by low-frequency stimulation of the perforant path. Using a wide range of frequencies

(4-100 Hz) Somjen et al. 34 found that SD occurred only after 10-, 15- and 20-Hz trains. Similarly, an earlier report that low-frequency trains do not induce LTD in the dentate of the rat hippocampal slice 36 may be understood in light of the recent observation that SD cannot be induced in the dentate area of this preparation 34. Furthermore, the variable effects of tetanization observed in the present study could reflect differences between the individual animals in factors which predispose to SD, and are not readily controlled, such as the extent of mechanical damage or anoxia around the recording electrode 16"23. Very little is known about the cellular mechanisms underlying LTD in the hippocampus. Sastry and colleagues 29'3° found that the calcium channel antagonist, verapamil, blocks LTD (but not LTP) in the Schaffer collateral CAj pathway, and suggested that neuronal calcium accumulation was involved. Interestingly, a link between SD and LTD seems to support this idea. SD in the dentate area is accompanied by a tremendous drop in extracellular calcium concentration in both the soma and dendritic layers 35. This change in calcium concentration is about 5 times greater than during a seizure 35, and occurs mainly during the slow-potential phase of SD, when neurons are almost completely depolarized 1°,35. Similarly, Krjnevi619 reported an equally large rise of intracellular calcium in CA 3 pyramidal cells in association with a large, depolarizing slow potential which occurred between 5 and 30 s after 10-Hz fimbrial stimulation. Thus, in a working model of hippocampal LTD, it would seem necessary to consider the possible role of postsynaptic neuronal calcium accumulation during an SD-iike episode. What role might LTD have in hippocampal physiology? The present results indicate that LTD does not reflect permanent cell damage, since: (1) highfrequency stimulation abolishes the depression within seconds; (2) LTD is transient; and (3) the capacity for LTP is retained after LTD. Instead, LTD appears to be a sustained, compensatory reaction to an episode of extreme neuronal excitation, which in our case may be an SD-like event. Others have suggested that the dentate may dampen the propagation of kindled seizures through the hippocampus tt,17, possibly by enhancing synaptic inhibition 17. We would like to add, that by temporarily impairing neurotransmission, a phenomenon like LTD might serve to protect

106 against the neurotoxic effects of excitatory amino acids in conditions like hippocampal epilepsy :5, isch• 3-~", hypoglycemia 39 and anoxia 2s. emta Finally, we found that the magnitude and develo p m e n t of LTP in the dentate area depends on the frequency of perforant path stimulation• A f t e r highfrequency stimulation, LTP of the population spike and EPSP was larger and generally reached maximum sooner than after low-frequency trains. This confirms and extends an earlier report by Douglas and G o d d a r d 13 in which frequencies ranging from 10 to 60 Hz gave larger LTP than 0.2 and 3 Hz. A t present, the mechanism(s) underlying the differential effects of train frequency on the magnitude of LTP is unknown. O n e possibility is that LTP is partially m a s k e d by a concurrent L T D p r o c e s s1"2 ~0; f•f t h e o p posing depression is smaller after high-frequency stimulation, then the net effect should be a larger LTP. However, this hypothesis is w e a k e n e d by our result that the depressed response spontaneously recovered only to the original baseline, instead of proREFERENCES l Abraham, W.C., Bliss, T.V.P. and Goddard, G.V., Heterosynaptic changes accompany long-term potentiation but not short-term potentiation of the perforant path in the anaesthetized rat, J. Physiol. (London), 363 (1985) 335-349. 2 Abraham, W.C. and Goddard, G.V. Asymmetric relationships between homosynaptie long-term potentiation and heterosynaptie long-term depression, Nature (London), 305 (1983) 717-719. 3 Andersen, P., Sundberg, S.H., Sveen, O., Swann, J.W. and Wigstrom, H., Possible mechanisms for long-lasting potentiation of synaptie transmission in hippocampal slices from guinea pigs, J. Physiol. (London), 302 (1980) 463-482. 4 Barrionuevo, G., Schottler, F. and Lynch, G., The effects of repetitive low frequency stimulation on control and 'potentiated' synaptic responses in the hippocampus, Life Sci., 27 (1980) 2385-2391. 5 Baudry, M. and Lynch, G., Hypothesis regarding the cellular mechanisms responsible for long-term synaptic potentiation in the hippocampus, Exp, Neurol., 68 (1980) 202-204. 6 Bliss, T.V.P. and Dolphin, A.C., What is the mechanism of long-term potentiation in the hippocampus?, Trends Neurosci., 5 (1982) 289-290. 7 Bliss, T.V.P. and LOmo, T., Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path, J. Physiol. (London), 232 (1973) 331-356. 8 Bramham, C.R. and Bliss, T.V.P., Naloxone blocks the induction of long-term potentiation in the lateral perforant pathway in vivo, Acta Physiol. Scand.. 128 (1986) A-16.

ceeding to demonstrate LTP. A more likely explanation perhaps, is that high frequencies may augment some process in LTP induction, such as N M D A rec e p t o r - m e d i a t e d calcium influx. These mechanisms do not exclude each other or the possible involvement of novel high-frequency-dependent events, for instance the action of a n e u r o p e p t i d e s. F u r t h e r study of long-term synaptic plasticity in relation to the frequency and pattern of afferent activation may help to resolve these issues. In conclusion, we suggest that both L T D and LTP may be useful models of long-term synaptic plasticity in the hippocampus; however, their relevance to physiological processes remains to be demonstrated. ACKNOWLEDGEMENTS We thank Drs. T.V.P. Bliss and N.W. Milgram for carefully reading an earlier version of the manuscript. This work was s u p p o r t e d by the Norwegian Research Council for Science and the Humanities. 9 Bramham, C.R. and Srebro, B., A comparison of longterm potentiation and depression induced by low (15 Hz) and high (400 Hz) frequency trains in the dentate area of the anesthetized rat: magnitude, time course, and DC shifts, Neurosci. Lett., Suppl. 22 (1985) S 5t. 10 Collewijn, H. and Van Harreveld, A., Membrane potential of cerebral cortical cells during spreading depression and asphyxia, Exp. Neurol.. 15 (19661 425-436. ll Collins, R.C., Tearse, R.G. and Lothman, E.W.. Functional anatomy of limbic seizures: focal discharges from medial entorhinal cortex in rat, Brain Research, 281) (1983) 25-40. 12 Douglas, R.M., Long lasting potentiation in the rat dentate gyrus following brief high frequency stimulation, Brain Research, 126 (19771 361-365. 13 Douglas, R.M. and Goddard, G.V., Long-term potentiation of the perforant path-granule cell synapse in the rat hippocampus, Brain Research, 86 (1975) 2(15- 215. 14 Dunwiddie, T. and Lynch, G., Long-term potentiation and depression of synaptic responses in the rat hippocampus: localization and frequency dependency, J. Physiol. (London), 276 (1978) 353-367. 15 Futamachi, K.J., Mutani, R. and Prince, D.A., Potassium activity in rabbit cortex, Brain Research. 75 (1974) 5-25. 16 Grafstein, B., Mechanism of spreading cortical depression, J. Neurophysiol., 19 (1956) 154-171. 17 King, G.L., Dingledine, R., Giacchino, J.L. and McNamara, J.O., Abnormal neuronal excitability in hippocampal slices from kindled rats, J. Neurophysiol.. 54 (19851 1295-13(/4. 18 KrnjcviC K., Morris, M.E. and Reiffenstein, R.J., Changes in extracellular Ca ++ and K ÷ activity accompanying hippocampal discharges, Can. J. Physiol, Pharmacol., 58 (1980) 579-583.

107 19 Krnjevi~, K., Morris, M.E. and Ropert, N., Changes in free calcium ion concentration recorded inside hippocampal pyramidal cells in situ, Brain Research, 374 (1986) 1-11. 20 Larson, J., Wong, D. and Lynch, G., Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation, Brain Research, 368 (1986) 347-350. 21 Leao, A.A.P., Spreading depression of activity in the cerebral cortex, J. Neurochem., 7 (1944) 359-390. 22 Lynch, G. and Baudry, M., The biochemistry of memory: a new and specific hypothesis, Science, 224 (1984) 1057-1063. 23 Marshall, W.H., Spreading cortical depression of Leao, Physiol. Rev., 39 (1959) 239-279. 24 McNaughton, B.L., Douglas, R.M. and Goddard, G.V., Synaptic enhancement in fascia dentata: cooperativity among coactive afferents, Brain Research, 157 (1978) 277-293. 25 Meldrum, B.S., Amino acid neurotransmitters and new approaches to anticonvulsant drug action, Epilepsia, 25 (1984) S140-S149. 26 Monakhov, K.K., Fifkova, E. and Bures, J., Steady potential field of hippocampal spreading depression, J. Cell. Comp. Physiol., 59 (1962) 155-161. 27 Racine, R.J., Milgram, N.W. and Hafner, S., Long-term potentiation phenomena in the rat limbic forebrain, Brain Research, 260 (1983) 217-231. 28 Rothman, S., Synaptic release of excitatory amino acid neurotransmitter mediates anoxic neuronal death, J. Neurosci., 4 (1984) 1884-1891. 29 Sastry, B.R., Chirwa, S.S., Goh, J.W., Maretic, H. and Pandanaboina, M.M., Verapamil counteracts depression not but long-lasting potentiation of the hippocampal population spike, Life Sci., 34 (1984) 1075-1086. 30 Sastry, B.R. and Goh, J.W., Long-lasting potentiation in hippocampus is not due to an increase in glutamate recep-

tors, Life Sci., 34 (1984) 1497-1601. 31 Schwartzkroin, P.A. and Wester, K., Long-lasting facilitation of a synaptic potential following tetanization in the in vitro hippocampal slice, Brain Research, 89 (1975) 107-119. 32 Simon, R.P., Swan, J.H., Griffiths, T. and Meldrum, B.S., Blockade of N-methyl-D-aspartate receptors may protect against ischemic damage in the brain, Science, 226 (1984) 850-852. 33 Skelton, R.W., Miller, J.J. and Phillips, A.G., Low-frequency stimulation of the perforant path produces longterm potentiation in the dentate gyrus of unanesthetized rats, Can. J. Physiol. Pharmacol., 61 (1983) 1156-1161. 34 Somjen, G.G., Aitken, P.G., Giacchino, J.L. and McNamara, J.O., Sustained potential shifts and paroxysmal discharges in hippocampal formation, J. Neurophysiol., 53 (1985) 1079-1097. 35 Somjen, G.G. and Giacchino, J.L., Potassium and calcium concentrations in interstitial fluid of hippocampal formation during paroxysmal responses, J. NeurophysioL, 53 (1985) 1098-1108. 36 Teyler, T.J., Alger, B.E., Bergman, T. and Livingston, K., A comparison of long-term potentiation in the in vitro and in vivo hippocampal preparations, Behav. Biol., 19 (1977) 24-34. 37 Teyler, T.J. and Discenna, P., Long-term potentiation as a candidate mnemonic device, Brain Res. Rev., 7 (1984) 15-28. 38 Voronin, L.L., Long-term potentiation in the hippocampus, Neurosci., 10 (1983) 1051-1069. 39 Wieloch, T., Hypoglycemia-induced neuronal damage prevented by an N-methyl-D-aspartate antagonist, Science, 230 (1985) 681-683. 40 Yamamoto, C. and Sawada, S., Important factors in induction of long-term potentiation in thin hippocampal sections, Exp. Neurol., 74 (1981) 122-130.