- Email: [email protected]
Changes in paired-pulse depression during the triggering of seizures by 2 Hz dentate gyrus stimulation: effect of the kindling
Brain Research 776 Ž1997. 250–254
Short communication
Changes in paired-pulse depression during the triggering of seizures by 2 Hz dentate gyrus stimulation: effect of the kindling Kenji Emori, Hiroshi Katsumori, Yoshio Minabe
)
DiÕision of Cortical Function Disorder, National Institute of Neuroscience, NCNP, 4-1-1 Ogawahigashi, Kodaira, Tokyo 187, Japan Accepted 2 September 1997
Abstract In this study, seizures in the dentate gyrus were triggered by 2 Hz electrical stimulation while, at the same time, serial changes in paired-pulse depression was measured in free-moving rats. The perforant path was stimulated by paired pulses 25 ms apart and recordings were made in the dentate gyrus. We also observed the effect of kindling procedure on this change of evoked field potential. Paired-pulse depression started to fail following the onset of epileptic afterdischarge, both before and after kindling. As kindling progressed, paired-pulse depression increased in the initial part of the stimulus train, and also the time delay from the start of the stimulation to when paired-pulse depression started to fail increased significantly. These findings suggest that the collapse of local early, presumably GABA A-mediated, inhibition may occur after the dentate gyrus seizure onset and development of kindling epileptogenesis does not weaken but, on the contrary, enhances this local inhibitory function. Finally we propose that seizures triggered by 2 Hz electrical stimulation constitutes a useful model for examining physiological or biochemical changes during seizure initiation of awake, free-moving animals since it provides long Žusually several seconds or more. latency from the start of stimulation to seizure onset. q 1997 Elsevier Science B.V. Keywords: Dentate gyrus; Hippocampus; Low-frequency stimulation; Paired-pulse depression; Seizure; Kindling; Perforant path
Although research on epilepsy has been going on for many years, very little is known how epileptic seizures begin. Since a loss or reduction of GABA-mediated inhibition causes seizure-like activity in vitro and convulsive seizure in vivo studies w3x, a decrease of GABA activity has long been considered to play a major role in seizure initiation and epileptogenesis. In addition it is known that GABA A -mediated inhibition fades during repetitive electrical stimulation of the brain, referred to as activity-dependent depression w5x, and this use-dependent modulation of early inhibition may provide an important gating function of neuronal excitation until inhibitory strength is no longer sufficient to prevent the triggering of epileptic afterdischarge ŽAD. w1x. However, to our knowledge, there have been no studies examining specifically whether activity-dependent disinhibition or failure of local inhibition occurs prior to seizure onset by recording serial changes of evoked field potentials in the seizure triggering site of awake free-moving animals.
) Corresponding author. Fax: q81 Ž423. 461748; E-mail: [email protected]
0006-8993r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 7 . 0 1 1 1 8 - 9
We have previously reported that seizures can be triggered and kindled with 2 Hz electrical stimulation and AD threshold can be measured from the number of stimulating pulses required for AD onset, referred to as pulse-number threshold ŽPNT. w2,4,6x. This procedure has the advantage that seizure threshold is evaluated through a single stimulus train and physiological or biochemical changes may be observed in the interval between the start of 2 Hz stimulus train and the onset of AD that is usually over several seconds. Therefore in this study seizures in the dentate gyrus ŽDG. were triggered by 2 Hz electrical stimulation. At the same time the perforant path ŽPP. was stimulated by paired pulses 25 ms apart, and paired-pulse depression measured in the DG. Male Wistar rats ŽClea Japan, Tokyo., weighing 230– 250 g at surgery, were used. The rats were housed in plastic cages with wood chip bedding under controlled conditions Žtemperature: 23–258C; humidity: 50–60%; 12-h lightrdark cycle, lights on at 08:00 h. and permitted to free access to food and water. The rats were anesthetized with pentobarbital Ž50 mgrkg i.p.. and mounted in a stereotaxic instrument. A tripolar electrode Žtwisted
K. Emori et al.r Brain Research 776 (1997) 250–254
251
Fig. 1. Experimental design and EEG recording. A: epileptic seizure was triggered by 2 Hz kindling stimulation in DG. During the delivery of kindling stimulus pulses, evoked field potentials of DG by paired-pulse stimulation of PP were also recorded. EC, entorhinal cortex. B: 2 Hz DG kindling stimulating pulses Ž0.5 mA, 1 ms biphasic. were delivered 25 ms after paired PP stimulating pulses Ž0.2–0.8 mA, 0.1 ms, 25 ms apart..
0.2 mm diameter, polyurethane-coated stainless steel lines. was implanted in the right side DG and a bipolar one in the same side PP. Stainless steel screws served as anchors
and the reference electrodes. The electrodes were attached to the skull by acrylic dental cement and connected with a socket. Two weeks later, test pulses Ž1r20 s., set at the
Fig. 2. Intracranial EEG recording from the stimulating site ŽDG. during 2 Hz kindling stimulation. Consecutive tracings are shown from top to bottom. 1: start of stimulation; 2: onset of AD; 3: end of stimulus train. In this case PNT, number of stimulating pulses required for AD triggering, is 14 Žfrom 1 to 2..
252
K. Emori et al.r Brain Research 776 (1997) 250–254
intensity triggering half of the maximal population spike height, were given to the PP for 10 min, and a control value was determined by averaging. Then the rats were kindled by low-frequency DG stimulation. 2 Hz DG kindling stimulating pulses for 17.5 s Ž500 mA base to peak, 1 ms duration biphasic square-wave pulses. were delivered 25 ms after 2 Hz paired PP stimulating pulses Ž25 ms apart, 0.2–0.7 mA, 0.1 ms duration monophasic squarewave pulses., using computer-assisted automatic stimulating systems ŽNihon Koden, SEN-7103; National Instrument, Labview 2. and constant-current units. The rats were stimulated every 3 h until kindling was completed. As shown in Figs. 1 and 2, we recorded EEG each time to determine the onset of AD. The EEG recording was digitized at a sampling rate of 10 kHz by an analog-digital converter ŽNational Instrument, NB-MIO-16X., acquired into a personal computer system and saved to MO-disk to be analyzed. The rat’s behavior was videotaped by an EEG-VTR system ŽNihon Koden, VY-440A.. A scaled-up EEG sample during the DG stimulation is shown in Fig. 3. In this case AD was triggered after the 14th stimulating pulse, and after the onset of AD the height of population spike decreased and then increased
with failure of paired-pulse depression. At the time of failure of paired-pulse depression, the shape of AD changed with multiple-population spike-like components. Fig. 4 shows the serial changes of % population spike heights, compared to the pre-kindling control value Žmean: 3.08 " 0.68 mV; n s 9., in the kindling process. As kindling progressed, paired-pulse depression increased in the initial part of the stimulus train, and also the time delay from the start of the stimulation to when the paired-pulse depression began to fail, as measured by the appearance of multiplespike AD, increased from 29.6 " 1.6 to 38.4 " 2.4 s Ž P 0.05, Wilcoxon test.. To our knowledge, this is the first report showing that failure of the early, presumably GABA A mediated w9x, inhibition of the seizure triggering site ŽDG. occurred after the onset of AD, and this inhibition failure was associated with the change of AD shape, containing a multiple-population spike-like component which appeared suddenly. Another major finding is that following kindling, the time delay from the start of the stimulation to when the pairedpulse depression started to fail increased significantly, indicating an enhanced resistance to inhibition failure. In the pioneering study of 1983 w9x, Tuff et al. showed that
Fig. 3. Scaled-up EEG recording of Fig. 2. A: epileptic AD onset in DG after the 14th stimulating pulses. Note that PP paired stimulation was delivered 25 ms before each stimulating pulse of DG. B: changes of paired-pulse response. During 2 Hz kindling stimulus train the height of evoked population spike decreased after AD onset, then reappeared with failure of paired-pulse depression. C: AD shape changed with multiple population spike-like component at the time of failure of paired-pulse depression.
K. Emori et al.r Brain Research 776 (1997) 250–254
Fig. 4. Effect of kindling development on the change of % population spike height Ž% of pre-kindling value, 3.08"0.68 mV, ns9.. P1, P2; 1st and 2nd response by paired-pulse stimulation, respectively, PNT, pulsenumber threshold meaning the onset of AD. Values represent mean" SEM.
paired-pulse depression in DG by PP stimulation significantly increased following amygdala kindling and 5 Hz stimulation applied to PP caused failure of paired-pulse depression immediately preceding the appearance of AD in DG. However, they did not stimulate the DG in order to trigger AD and they identified the onset of AD using the appearance of a multiple spike-like AD, due to the lack of continuous EEG recording and analysis during the stimulation. Indeed a distinct change in the shape of the induced AD in the DG, consisting of a sudden appearance of bursts of large amplitude population spikes, has been reported in anesthetized w7x and unanesthetized w8x animals, referred to as maximal dentate activation. Associated with this phenomenon is an abrupt secondary rise in extracellular potassium and a sudden sustained negative shift of the DC potential w8x. Our data suggest that maximal dentate activa-
253
tion appears with local inhibition failure of the DG. Recently Burdette et al. w1x have examined changes in the strength of early Ž20 ms interpulse interval. and late Ž200 ms interpulse interval. paired-pulse depression in the DG of unanesthetized rats during the initiation of AD triggered by 5 Hz PP stimulation. They concluded that failure of late, but not of early, paired-pulse depression may be a precipitating event in AD initiation. However, as reported by Tuff et al. w9x, they also identified the onset of AD using the presence of multiple spike-like AD and did not test the effect of DG stimulation on the paired-pulse depression of the same site. In any case, it will be necessary to examine the change of late paired-pulse depression using our experimental design in the future. We do not know the reason why the height of the first population spike deceased transiently Žf 2 s. before the failure of paired-pulse depression, but it is possible that inhibitory volley in DG reached the maximum state preceding its failure. In conclusion we discovered for the first time that epileptic afterdischarge can be triggered by 2 Hz dentate gyrus stimulation immediately preceding the failure of early local inhibition, indicated by the collapse of 25 ms apart paired-pulse depression by perforant path stimulation. Furthermore the shape of triggered afterdischarge suddenly changed with multiple-population spike-like components when the inhibition failure occurred, meaning that the change of afterdischarge shape is a good indicator of any failure of local inhibition in the seizure triggering site. Kindling procedure seems to enhance the early inhibition by increasing the time delay from the start of the stimulation to the collapse of paired-pulse depression. These results suggest that failure of GABA A -mediated inhibition of the dentate gyrus may occur after seizure initiation of the same site, meaning that failure of GABA A -mediated inhibition may not be a precipitating event in seizure initiation in the epileptic focus. Finally we propose that dentate gyrus seizure triggered by 2 Hz electrical stimulation is a useful model of epilepsy using awake free-moving animals, particularly when examining events during seizure initiation.
Acknowledgements This work was supported by the grant from Japanese Ministry of Education, Science and Culture, and from Japanese Ministry of Health and Welfare to Y.M.
References w1x L.J. Burdette, G.J. Hart, L.M. Masukawa, Changes in dentate granule cell field potentials during afterdischarge initiation triggered by 5 Hz perforant path stimulation, Brain Res. 722 Ž1996. 39–49.
254
K. Emori et al.r Brain Research 776 (1997) 250–254
w2x K. Emori, Y. Minabe, Effects of anticonvulsants on hippocampusgenerating seizure, Brain Res. 511 Ž1990. 217–221. w3x J.G.R. Jefferys, Experimental neurobiology of epilepsies, Curr. Opin. Neurol. 7 Ž1994. 113–122. w4x H. Katsumori, Y. Ito, H. Higashida, M. Hashii, Y. Minabe, Anti- and proconvulsive actions of levcromakalim, an opener of ATP-sensitive K channel, in the model of hippocampus-generating partial seizure in rats, Eur. J. Pharmacol. 311 Ž1996. 37–44. w5x D.S.F. Lin, L.S. Bernardo, Activity-dependent depression of monosynaptic fast IPSC in hippocampus: contribution from reduction in chloride driving force and conductance, Brain Res. 670 Ž1995. 142– 146.
w6x Y. Minabe, K. Emori, Two types of neuroplasticities in the kindling phenomenon, effects of chronic MK-801 and methamphetamine, Brain Res. 585 Ž1992. 237–241. w7x J.L. Stringer, E.W. Lothman, Maximal dentate gyrus activation: characteristics and alteration after repeated seizures, J. Neurophysiol. 62 Ž1988. 136–143. w8x J.L. Stringer, J.M. Williamson, E.W. Lothman, Maximal dentate activation is produced by amygdala stimulation in unanesthetized rats, Brain Res. 542 Ž1991. 336–342. w9x L.P. Tuff, J. Racine, R. Adamec, The effect of kindling on GABAmediated inhibition in the dentate gyrus of the rat. 1. Paired-pulse depression, Brain Res. 277 Ž1983. 79–90.
Short communication
Changes in paired-pulse depression during the triggering of seizures by 2 Hz dentate gyrus stimulation: effect of the kindling Kenji Emori, Hiroshi Katsumori, Yoshio Minabe
)
DiÕision of Cortical Function Disorder, National Institute of Neuroscience, NCNP, 4-1-1 Ogawahigashi, Kodaira, Tokyo 187, Japan Accepted 2 September 1997
Abstract In this study, seizures in the dentate gyrus were triggered by 2 Hz electrical stimulation while, at the same time, serial changes in paired-pulse depression was measured in free-moving rats. The perforant path was stimulated by paired pulses 25 ms apart and recordings were made in the dentate gyrus. We also observed the effect of kindling procedure on this change of evoked field potential. Paired-pulse depression started to fail following the onset of epileptic afterdischarge, both before and after kindling. As kindling progressed, paired-pulse depression increased in the initial part of the stimulus train, and also the time delay from the start of the stimulation to when paired-pulse depression started to fail increased significantly. These findings suggest that the collapse of local early, presumably GABA A-mediated, inhibition may occur after the dentate gyrus seizure onset and development of kindling epileptogenesis does not weaken but, on the contrary, enhances this local inhibitory function. Finally we propose that seizures triggered by 2 Hz electrical stimulation constitutes a useful model for examining physiological or biochemical changes during seizure initiation of awake, free-moving animals since it provides long Žusually several seconds or more. latency from the start of stimulation to seizure onset. q 1997 Elsevier Science B.V. Keywords: Dentate gyrus; Hippocampus; Low-frequency stimulation; Paired-pulse depression; Seizure; Kindling; Perforant path
Although research on epilepsy has been going on for many years, very little is known how epileptic seizures begin. Since a loss or reduction of GABA-mediated inhibition causes seizure-like activity in vitro and convulsive seizure in vivo studies w3x, a decrease of GABA activity has long been considered to play a major role in seizure initiation and epileptogenesis. In addition it is known that GABA A -mediated inhibition fades during repetitive electrical stimulation of the brain, referred to as activity-dependent depression w5x, and this use-dependent modulation of early inhibition may provide an important gating function of neuronal excitation until inhibitory strength is no longer sufficient to prevent the triggering of epileptic afterdischarge ŽAD. w1x. However, to our knowledge, there have been no studies examining specifically whether activity-dependent disinhibition or failure of local inhibition occurs prior to seizure onset by recording serial changes of evoked field potentials in the seizure triggering site of awake free-moving animals.
) Corresponding author. Fax: q81 Ž423. 461748; E-mail: [email protected]
0006-8993r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 7 . 0 1 1 1 8 - 9
We have previously reported that seizures can be triggered and kindled with 2 Hz electrical stimulation and AD threshold can be measured from the number of stimulating pulses required for AD onset, referred to as pulse-number threshold ŽPNT. w2,4,6x. This procedure has the advantage that seizure threshold is evaluated through a single stimulus train and physiological or biochemical changes may be observed in the interval between the start of 2 Hz stimulus train and the onset of AD that is usually over several seconds. Therefore in this study seizures in the dentate gyrus ŽDG. were triggered by 2 Hz electrical stimulation. At the same time the perforant path ŽPP. was stimulated by paired pulses 25 ms apart, and paired-pulse depression measured in the DG. Male Wistar rats ŽClea Japan, Tokyo., weighing 230– 250 g at surgery, were used. The rats were housed in plastic cages with wood chip bedding under controlled conditions Žtemperature: 23–258C; humidity: 50–60%; 12-h lightrdark cycle, lights on at 08:00 h. and permitted to free access to food and water. The rats were anesthetized with pentobarbital Ž50 mgrkg i.p.. and mounted in a stereotaxic instrument. A tripolar electrode Žtwisted
K. Emori et al.r Brain Research 776 (1997) 250–254
251
Fig. 1. Experimental design and EEG recording. A: epileptic seizure was triggered by 2 Hz kindling stimulation in DG. During the delivery of kindling stimulus pulses, evoked field potentials of DG by paired-pulse stimulation of PP were also recorded. EC, entorhinal cortex. B: 2 Hz DG kindling stimulating pulses Ž0.5 mA, 1 ms biphasic. were delivered 25 ms after paired PP stimulating pulses Ž0.2–0.8 mA, 0.1 ms, 25 ms apart..
0.2 mm diameter, polyurethane-coated stainless steel lines. was implanted in the right side DG and a bipolar one in the same side PP. Stainless steel screws served as anchors
and the reference electrodes. The electrodes were attached to the skull by acrylic dental cement and connected with a socket. Two weeks later, test pulses Ž1r20 s., set at the
Fig. 2. Intracranial EEG recording from the stimulating site ŽDG. during 2 Hz kindling stimulation. Consecutive tracings are shown from top to bottom. 1: start of stimulation; 2: onset of AD; 3: end of stimulus train. In this case PNT, number of stimulating pulses required for AD triggering, is 14 Žfrom 1 to 2..
252
K. Emori et al.r Brain Research 776 (1997) 250–254
intensity triggering half of the maximal population spike height, were given to the PP for 10 min, and a control value was determined by averaging. Then the rats were kindled by low-frequency DG stimulation. 2 Hz DG kindling stimulating pulses for 17.5 s Ž500 mA base to peak, 1 ms duration biphasic square-wave pulses. were delivered 25 ms after 2 Hz paired PP stimulating pulses Ž25 ms apart, 0.2–0.7 mA, 0.1 ms duration monophasic squarewave pulses., using computer-assisted automatic stimulating systems ŽNihon Koden, SEN-7103; National Instrument, Labview 2. and constant-current units. The rats were stimulated every 3 h until kindling was completed. As shown in Figs. 1 and 2, we recorded EEG each time to determine the onset of AD. The EEG recording was digitized at a sampling rate of 10 kHz by an analog-digital converter ŽNational Instrument, NB-MIO-16X., acquired into a personal computer system and saved to MO-disk to be analyzed. The rat’s behavior was videotaped by an EEG-VTR system ŽNihon Koden, VY-440A.. A scaled-up EEG sample during the DG stimulation is shown in Fig. 3. In this case AD was triggered after the 14th stimulating pulse, and after the onset of AD the height of population spike decreased and then increased
with failure of paired-pulse depression. At the time of failure of paired-pulse depression, the shape of AD changed with multiple-population spike-like components. Fig. 4 shows the serial changes of % population spike heights, compared to the pre-kindling control value Žmean: 3.08 " 0.68 mV; n s 9., in the kindling process. As kindling progressed, paired-pulse depression increased in the initial part of the stimulus train, and also the time delay from the start of the stimulation to when the paired-pulse depression began to fail, as measured by the appearance of multiplespike AD, increased from 29.6 " 1.6 to 38.4 " 2.4 s Ž P 0.05, Wilcoxon test.. To our knowledge, this is the first report showing that failure of the early, presumably GABA A mediated w9x, inhibition of the seizure triggering site ŽDG. occurred after the onset of AD, and this inhibition failure was associated with the change of AD shape, containing a multiple-population spike-like component which appeared suddenly. Another major finding is that following kindling, the time delay from the start of the stimulation to when the pairedpulse depression started to fail increased significantly, indicating an enhanced resistance to inhibition failure. In the pioneering study of 1983 w9x, Tuff et al. showed that
Fig. 3. Scaled-up EEG recording of Fig. 2. A: epileptic AD onset in DG after the 14th stimulating pulses. Note that PP paired stimulation was delivered 25 ms before each stimulating pulse of DG. B: changes of paired-pulse response. During 2 Hz kindling stimulus train the height of evoked population spike decreased after AD onset, then reappeared with failure of paired-pulse depression. C: AD shape changed with multiple population spike-like component at the time of failure of paired-pulse depression.
K. Emori et al.r Brain Research 776 (1997) 250–254
Fig. 4. Effect of kindling development on the change of % population spike height Ž% of pre-kindling value, 3.08"0.68 mV, ns9.. P1, P2; 1st and 2nd response by paired-pulse stimulation, respectively, PNT, pulsenumber threshold meaning the onset of AD. Values represent mean" SEM.
paired-pulse depression in DG by PP stimulation significantly increased following amygdala kindling and 5 Hz stimulation applied to PP caused failure of paired-pulse depression immediately preceding the appearance of AD in DG. However, they did not stimulate the DG in order to trigger AD and they identified the onset of AD using the appearance of a multiple spike-like AD, due to the lack of continuous EEG recording and analysis during the stimulation. Indeed a distinct change in the shape of the induced AD in the DG, consisting of a sudden appearance of bursts of large amplitude population spikes, has been reported in anesthetized w7x and unanesthetized w8x animals, referred to as maximal dentate activation. Associated with this phenomenon is an abrupt secondary rise in extracellular potassium and a sudden sustained negative shift of the DC potential w8x. Our data suggest that maximal dentate activa-
253
tion appears with local inhibition failure of the DG. Recently Burdette et al. w1x have examined changes in the strength of early Ž20 ms interpulse interval. and late Ž200 ms interpulse interval. paired-pulse depression in the DG of unanesthetized rats during the initiation of AD triggered by 5 Hz PP stimulation. They concluded that failure of late, but not of early, paired-pulse depression may be a precipitating event in AD initiation. However, as reported by Tuff et al. w9x, they also identified the onset of AD using the presence of multiple spike-like AD and did not test the effect of DG stimulation on the paired-pulse depression of the same site. In any case, it will be necessary to examine the change of late paired-pulse depression using our experimental design in the future. We do not know the reason why the height of the first population spike deceased transiently Žf 2 s. before the failure of paired-pulse depression, but it is possible that inhibitory volley in DG reached the maximum state preceding its failure. In conclusion we discovered for the first time that epileptic afterdischarge can be triggered by 2 Hz dentate gyrus stimulation immediately preceding the failure of early local inhibition, indicated by the collapse of 25 ms apart paired-pulse depression by perforant path stimulation. Furthermore the shape of triggered afterdischarge suddenly changed with multiple-population spike-like components when the inhibition failure occurred, meaning that the change of afterdischarge shape is a good indicator of any failure of local inhibition in the seizure triggering site. Kindling procedure seems to enhance the early inhibition by increasing the time delay from the start of the stimulation to the collapse of paired-pulse depression. These results suggest that failure of GABA A -mediated inhibition of the dentate gyrus may occur after seizure initiation of the same site, meaning that failure of GABA A -mediated inhibition may not be a precipitating event in seizure initiation in the epileptic focus. Finally we propose that dentate gyrus seizure triggered by 2 Hz electrical stimulation is a useful model of epilepsy using awake free-moving animals, particularly when examining events during seizure initiation.
Acknowledgements This work was supported by the grant from Japanese Ministry of Education, Science and Culture, and from Japanese Ministry of Health and Welfare to Y.M.
References w1x L.J. Burdette, G.J. Hart, L.M. Masukawa, Changes in dentate granule cell field potentials during afterdischarge initiation triggered by 5 Hz perforant path stimulation, Brain Res. 722 Ž1996. 39–49.
254
K. Emori et al.r Brain Research 776 (1997) 250–254
w2x K. Emori, Y. Minabe, Effects of anticonvulsants on hippocampusgenerating seizure, Brain Res. 511 Ž1990. 217–221. w3x J.G.R. Jefferys, Experimental neurobiology of epilepsies, Curr. Opin. Neurol. 7 Ž1994. 113–122. w4x H. Katsumori, Y. Ito, H. Higashida, M. Hashii, Y. Minabe, Anti- and proconvulsive actions of levcromakalim, an opener of ATP-sensitive K channel, in the model of hippocampus-generating partial seizure in rats, Eur. J. Pharmacol. 311 Ž1996. 37–44. w5x D.S.F. Lin, L.S. Bernardo, Activity-dependent depression of monosynaptic fast IPSC in hippocampus: contribution from reduction in chloride driving force and conductance, Brain Res. 670 Ž1995. 142– 146.
w6x Y. Minabe, K. Emori, Two types of neuroplasticities in the kindling phenomenon, effects of chronic MK-801 and methamphetamine, Brain Res. 585 Ž1992. 237–241. w7x J.L. Stringer, E.W. Lothman, Maximal dentate gyrus activation: characteristics and alteration after repeated seizures, J. Neurophysiol. 62 Ž1988. 136–143. w8x J.L. Stringer, J.M. Williamson, E.W. Lothman, Maximal dentate activation is produced by amygdala stimulation in unanesthetized rats, Brain Res. 542 Ž1991. 336–342. w9x L.P. Tuff, J. Racine, R. Adamec, The effect of kindling on GABAmediated inhibition in the dentate gyrus of the rat. 1. Paired-pulse depression, Brain Res. 277 Ž1983. 79–90.