- Email: [email protected]
Selective loss of early suppression in the dentate gyrus precedes kainic acid induced electrographic seizures
Epilepsy Research 31 (1998) 143 – 152
Selective loss of early suppression in the dentate gyrus precedes kainic acid induced electrographic seizures C.J. Ikeda-Douglas a, E. Head a, R.M.D. Holsinger a, L. Tremblay a, R. Racine b, N.W. Milgram a,* a
Life Science Di6ision, Scarborough College, Uni6ersity of Toronto, 1265 Military Trail, Scarborough, Ontario, M1C 1A4, Canada b Department of Psychology, McMaster Uni6ersity, Hamilton, Ontario, L8S 4K1, Canada Received 17 December 1997; received in revised form 1 March 1998; accepted 20 April 1998
Abstract The role of inhibitory and facilitatory processes in the induction of seizures was studied in a kainic acid (KA) model of epilepsy. The dentate gyrus (DG) response to paired-pulse stimulation of the perforant path (PP) was monitored prior to and immediately following the initial KA induced afterdischarge (AD) in rats chronically prepared with stimulation recording electrodes. The subjects received a 1-h program of stimulation consisting of repeated sequences of pulse pairs at a short (20–30 ms), intermediate (45 – 90 ms), and long (200 – 300 ms) interpulse interval (IPIs). The stimulation program was administered both under control conditions and immediately following systemic injection of KA. During the control condition, stable suppression of population spike measures was obtained at the short (early phase) and long (late phase) IPIs, while facilitation was observed at the intermediate IPI. Administration of KA resulted in a progressive loss of suppression prior to the initial AD at the short IPI; neither facilitation nor the late phase of suppression were significantly affected. The early phase decreased further following the initial discharge. Since the early phase most likely reflects recurrent inhibition, these results provide evidence that inhibitory loss precedes the occurrence of KA induced AD, and that this inhibitory loss is increased further following the initial evoked AD. A use-dependent disinhibition is one possible explanation for the change in responsiveness that precedes the AD. This disinhibition could result from a depressed response at GABA-A receptors, an increased responsiveness at GABA-B receptors or possibly both. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Dentate gyrus; Kainic acid; Electrographic seizures
* Corresponding author. Tel.: +1 416 2877402; fax: +1 416 2877642; e-mail: [email protected] 0920-1211/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S0920-1211(98)00028-X
144
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
1. Introduction Generation of epileptic seizures is widely believed to depend, at least in part, on modifications in inhibitory transmission. Two lines of evidence support this view. The first is based on pharmacological studies with agents that affect GABA receptors. Antagonists, such as bicuculline, that block GABA transmission, can have potent epileptogenic effects (Bradford, 1995). Agents that augment GABA transmission, in contrast have anticonvulsant effects. These include agonists such as muscimol (Morrisett et al., 1993), modulators of GABA, such as diazepam (Francis et al., 1994), and reuptake inhibitors, such as tiagabine, that prolong the action of GABA (Faingold et al., 1994). The second line of evidence deals with changes in neural transmission that accompany or follow the induction of seizures. A loss of inhibition has been reported to result from the kindling procedure, which involves the repeated induction of seizures (Tuff et al., 1983). Inhibitory loss has been reported following electrically and chemically induced status epilepticus (Bekenstein et al., 1993). Such seizure related loss of inhibition is particularly notable in epilepsy prone animals (Evans et al., 1994). The research discussed above, however, does not establish disruption of neural inhibition as an essential normal trigger of seizures. Treatments affecting transmitter systems other than GABA, such as glutamate (Joy and Albertson, 1988) or acetylcholine (Turski et al., 1989), can also have epileptogenic effects. Furthermore, decreased inhibition does not always occur in epilepsy models. Increased GABA inhibition has been reported to follow both kindling (Tuff et al., 1983; Milgram et al., 1995) and status epilepticus (Milgram et al., 1991; Haas et al., 1996). Apparently, the effect of seizures on neural inhibition is pathway specific, with some pathways showing an increase and others a decrease (Zhao and Leung, 1992). Another issue relates to the temporal relationship between the occurrence of inhibitory loss and the occurrence of seizures. If there is a link between inhibition and the induction of seizures in the dentate gyrus (DG), then inhibitory transmission should decrease prior to the occurrence of seizures.
Few studies have monitored inhibition preceding seizures, and those that have present contrasting results. Sloviter found that perforant path (PP) stimulation repeated over several hours induced a loss of paired-pulse inhibition, leading to the occurrence of multiple population spikes (Sloviter, 1990). The significance of this finding, however, is unclear. The repeated stimulation paradigm causes multiple seizures, which could be interpreted in either of two ways: (1) multiple seizures are a consequence of stimulation induced inhibitory loss; or (2) the inhibitory loss is a result of the induction of multiple seizures. More direct evidence of a relationship between inhibition and seizure development was obtained by Milgram et al. (1991) who found a loss in early suppression in the DG following systemic administration of kainic acid (KA). Tuff et al. (1983) also reported decreased paired-pulse suppression in the DG prior to afterdischarge (AD) evoked by trains of pulse pairs to the PP. In contrast, Stringer (1993) reported that seizures evoked by repeated trains of electrical stimulation were preceded by an increase in paired-pulse inhibition in the DG, and a decrease in cell field CA1. However, in the DG the repeated stimulation led to progressive increments in the response evoked by the first pulse (C-pulse) of the pulse pair. Paired-pulse suppression is directly linked to the amplitude of the response evoked by the C-pulse: the larger the response the greater the suppression (Milgram et al., 1991; Burdette and Gilbert, 1995). Thus, a stimulation-induced increase in field potential response to the C-pulse can account for the inhibitory increase that precedes the development of seizures. The present investigation had two goals: (1) to establish a relationship between paired-pulse suppression and the development of AD activity in response to systemic KA; and (2) to determine whether changes in DG transmission were restricted to those occurring during the early phase of suppression. This second goal stemmed from the known sensitivity of the paired-pulse procedure to interpulse interval (IPI). At short IPIs (less than about 40 ms) suppression in the response to pulse two (T-pulse) is primarily due to recurrent collateral inhibition caused by chloride conductances (Lømo, 1971; Tuff et al., 1983; Joy and Albertson,
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
1988; Mody et al., 1994; Haas et al., 1996). At longer IPIs (between about 45 and 90 ms) the response to the T-pulse is larger than that evoked by the C-pulse, reflecting a phase of paired-pulse facilitation (Joy and Albertson, 1988). At still longer intervals, the T-pulse response is depressed, which represents a late phase of inhibition that is mediated by potassium conductances (Joy and Albertson, 1988; Scharfman and Schwartzkroin, 1990; Haas et al., 1996). We were particularly interested in the relationship between the late phase of inhibition and seizure development because the late phase of inhibition has been linked to the GABA-B receptor (Raushe et al., 1989; Steffensen and Hendriksen, 1991). Furthermore, several reports suggest a critical role of GABA-B inhibition in epilepsy (Morrisett et al., 1993; Brucato et al., 1995). To accomplish these goals, we monitored the DG response to a program of repeated paired-pulse stimulation to the PP under a control condition and again following systemic administration of KA. KA treatment causes multiple ADs and eventually leads to SE (Ben-Ari, 1985). However, AD in the DG typically develops slowly; the initial discharge generally occurs after a minimum of 20 min, but may take as long as 90 min. The present experiment focuses only on those events that lead to and immediately follow the initial discharge.
2. Methods Subjects were ten male hooded Long Evans rats chronically implanted with DG and PP electrodes. The subjects were selected from a larger population of animals on the basis of having stable evoked field potentials, which showed a topography characteristic of the DG. We also limited the subject pool to animals that showed stable paired-pulse suppression at a short IPI.
2.1. Surgical and electrophysiological test procedures Rats were implanted with two pairs of bipolar electrodes constructed from twisted lengths of Teflon-coated stainless steel wires 250 mm in diame-
145
ter. The tip separation was approximately 500 mm. The stimulating electrodes were placed in the PP and recording electrodes in the granular layer of the DG using coordinates previously described (Lømo, 1971). The surgical procedures were performed under sodium pentobarbital anaesthesia (65 mg/ kg). A stainless steel wire wrapped around jewelers’ screws embedded into the skull served as the ground electrode. The electrodes were connected to microconnectors, which were then inserted into nine pin assemblies and fixed to the animal’s skull using dental cement. For all electrophysiological testing, the animals were placed in shielded cages and connected through leads constructed from low noise cable (Microdot) to both the stimulating and recording units. Stimulation consisted of constant current biphasic pulses 0.10 ms in duration with a 0.05-ms delay between the first and second pulse delivered through a stimulus isolation unit (ISO-Flex, AMPI Instruments). The signals from the animal were taken to a preamplifier (Grass model) and filtered between 1 and 3000 Hz with the gain set individually to maximize the response. The output from the preamplifier was fed to an analog to digital converter and digitized at 12000 Hz. The output of the A/D converter was taken to a computer, and the signals were stored on floppy disks for offline analysis. All data acquisition and analysis was performed under computer control using dedicated programs developed in the ASYST programming language.
2.2. Test procedure Following an initial stage of screening, averaged evoked potentials from six individual sweeps were taken at 13 different intensities, ranging from 100 to 2000 mA, which were used to construct inputoutput (I/O) curves. Two response parameters were measured: the population spike (PS), which provides a field measure of cellular discharge, and the population excitatory post-synaptic potential (pEPSP), which provides a measure of synaptic current. The PS was calculated by projecting a line from the negative peak to a tangent line connecting the spike onset and offset. The EPSP slope was
146
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
calculated by placing a cursor at a fixed point on the rising phase of the response, differentiating a segment of the sweep around the cursor, and taking the mean of the three largest values of the derivative function (Milgram et al., 1991). The baseline test was conducted 24 h prior to KA treatment, and was administered over 50 min. For each animal, the stimulation intensity was individually adjusted at 50 – 70% of the maximum population spike amplitude. Paired pulses were delivered in repeated blocks containing sequences of three interpulse intervals, with a 20-s interval separating each pulse pair. Each sequence contained an IPI of: (1) 20 – 30 ms, the range necessary for producing early phase of suppression; (2) an interval between 45 and 90 ms, the range used to produce facilitation; and (3) an interval between 200 and 300 ms, which was used to evaluate the late phase of suppression. Thus, a total of 50 paired pulses were delivered at each of the three IPIs. For the treatment phase, the animals were first injected with a 12-mg/kg dose of KA (Sigma), which was dissolved in physiological saline and prepared as a 1% solution. The pH was neutralized with 10 N sodium hydroxide. Immediately following intraperitoneal (i.p.) injection, the animals were placed in the test chamber and administered a paired-pulse procedure identical to that used for the baseline test. EEG activity from the DG was continuously monitored to establish the exact time when the first AD occurred.
3. Results
3.1. Beha6ioral response to systemic kainic acid As we have previously reported, an initial period of hypoactivity followed systemic administration of KA (Milgram et al., 1991). This, in turn, was followed by an interval of increased behavioral activity with occasional wet dog shakes. The initial electroencephalographic afterdischarge (AD) in the DG occurred during this second interval, generally between 20 and 40 min after the injection, and was unaccompanied by behavioral seizures. The first AD was short, and was
followed by a transient period of postictal depression. Subsequent ADs increased in duration and showed more obvious behavioral correlates, but these were not of concern for this study.
3.2. Early phase of suppression At the short IPI, the T-pulse response was consistently suppressed relative to the C-pulse during the baseline test (Fig. 1B and Fig. 2A). This suppression decreased progressively following administration of KA (Fig. 1A and Fig. 2B). To test for the significance of this effect, we compared the average response to five pulse pairs at the short IPI immediately following administration of KA with the average response to the corresponding five that immediately preceded the development of the first AD. We also looked at the average of five pulse pairs following recovery from the AD (Fig. 3). The results were evaluated using a repeated measures ANOVA with treatment (KA versus control) and time following treatment as main effects. The time dependent change in suppression caused by KA was indicated by a significant interaction between time and treatment (F(2,18)=4.97; PB0.019). To determine whether these effects were due to changes in the response to pulse 1, we analyzed the response to the C-pulse alone. There was a slight decrease over time, but there was considerable variability; the response decreased prior to the first AD in only three instances. There was a slightly larger response in the other seven animals. However, the overall C-pulse amplitude was not significantly different from the response at the start of the session. Following the initial AD, the response to the C-pulse was markedly suppressed and the PS was typically blocked; these effects were transient. The pEPSP measure was analyzed in the same way as the PS. The results of the repeated measure ANOVA revealed an interaction between time following KA and treatment, (F(2,18)= 12.28; PB 0.0002). As illustrated in Fig. 4, the suppression score for the slope measure did not change prior to the occurrence of the AD; following the AD, however, there was a switch to facilitation.
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
147
Fig. 1. Characteristic pairs of evoked DG field potentials at three different time points are illustrated for the kainic acid (KA) condition (A) and baseline condition (B). T0, beginning of session; T1, immediately preceding the first afterdischarge (AD); T2, following recovery from the AD. Corresponding pulse pairs are taken from the baseline (B).
3.3. Changes in facilitation
3.4. Changes in late phase inhibition
As illustrated in Fig. 3B facilitation of the PS decreased after treatment with KA. However, there were no statistically significant changes in either of the PS measures or the EPSP ratio (Fig. 4B).
As shown in Fig. 3C there was a slight but non-significant loss in the late phase of suppression after KA administration (Fig. 4C).
148
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
Fig. 2. Changes in the early phase of suppression both (A) before and (B) following systemic kainic acid (KA) are shown for a representative subject. Each point represents the average amplitude of the population spike measure for the conditioning and test pulses of three consecutive pulse pairs. The bars represent standard errors.
4. Discussion The present study demonstrates that pairedpulse suppression decreases progressively following the systemic administration of KA, and that this decrease precedes the occurrence of AD activity in the dentate gyrus. Paired-pulse suppression decreased further following the initial AD. The loss in suppression that preceded the occurrence of AD was also selective; only the early phase of paired-pulse suppression was significantly affected. Joy and Albertson (1988) have previously reported a loss in the early phase of suppression following administration of KA. The present results supplement their findings by showing the progressive nature of the change, and also the potential link to the occurrence of seizures. As mentioned previously, the early phase of paired-pulse suppression is largely a result of recurrent collateral inhibition and is mediated by activation of GABA-A receptors (Tuff et al.,
1983; Evans et al., 1994; Mody et al., 1994; During et al., 1996). The present findings therefore, suggest first that a gradual loss in effectiveness of GABA-A mediated inhibition follows systemic treatment with KA and, second that this loss is temporally linked to the development of AD. These results differ from those reported by Stringer (1993), who found a progressive increase in paired-pulse suppression in the DG prior to the occurrence of AD. These discrepant observations can be reconciled by taking into consideration procedural differences. Stringer used repeated trains of electrical stimulation to induce seizures. This procedure lead to progressive increases in DG excitability, which was manifest by increases in the amplitude of the response evoked by the C-pulse. This increased response amplitude probably accounts for the increased paired-pulse suppression, since suppression in the DG is known to vary directly with the amplitude of the response evoked by the C-pulse (Milgram et al., 1991;
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
149
Fig. 3. The effects of KA on the population spike measure of suppression. The suppression score was normalized by dividing the paired pulse ratio by the ratio from the first 5 min. (A) Changes occurring during the early phase of suppression. Changes in facilitation are show in (B) while changes in the late phase are shown in (C). Control and post KA data are compared at three different time points: at the beginning of each session, prior to the first afterdischarge (AD), and following recovery from the first AD.
Burdette and Gilbert, 1995). The present study differed in that the response to the C-pulse did not change significantly during the interval preceding the occurrence of AD. The finding that the C-pulse response did not significantly decrease after treatment with KA, suggests that the KA-induced AD is not caused by a generalized increase in excitability. Joy and Albertson (1988) reached a similar conclusion. They found one class of convulsants, which were GABA-A antagonists, caused an increase in granule cell excitability. A second class of convulsants acted on glutamate receptors and depressed both paired-pulse facilitation and paired-pulse depression; KA was a member of this class. Four possible mechanisms could account for the selective inhibitory loss that preceded the occurrence of seizures: (1) an increase in the phase of facilitation, which could mask recurrent inhibition; (2) a use-dependent loss of transmission at
fast inhibitory synapses; (3) increased activation or responsiveness of presynaptic GABA-B receptors; and (4) inhibitory block caused by kainic acid induced depolarization. The first possibility, enhanced facilitation masking recurrent collateral inhibition can be ruled out. The present results allow us to rule out this possibility. We found no evidence that KA increased facilitation; in fact, there was a small loss of facilitation. These findings are consistent with a report by Joy and Albertson (1988) of a decrease in facilitation following systemic administration of KA. The second possibility, that of a use-dependent loss of GABA-A is suggested by intracellular studies of inhibitory transmission in slices. A single PP stimulation is sufficient to induce a transient use-dependent inhibitory loss (McCarren and Alger, 1985; Mott et al., 1993). Scharfman and Schwartzkroin (1990) observed a more long-
150
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
Fig. 4. The effects of KA on the pEPSP slope measure of paired pulse suppression in the DG. The suppression scores were normalized. Control and post KA data are compared at three time points: the first 5 min of the session, a period prior to the first AD, and following recovery from the AD. Changes in the early phase of suppression are shown in (A), (B) illustrates the effects of facilitation and (C) shows the effects on the late phase.
lasting loss in paired-pulse inhibition when activation of the perforant path was repeated over several hours. Repeated stimulation of the CA3CA1 pathway has also been repeated to produce a long-lasting loss of paired-pulse suppression (Kapur and Lothman, 1989; Kapur et al., 1989). These effects could result from depletion of synaptically available GABA (Kapur and Lothman, 1989; Kapur et al., 1989; Pfeiffer et al., 1996). In the present study, a use-dependent loss of GABA could account for both the progressive loss that precedes AD and the further loss of suppression that follows AD. However, the synaptic depletion hypothesis can not account for the changes we observed in the EPSP slope measure that did not show a loss in suppression prior to AD, but showed significant inhibitory loss following AD. The third possibility relates to GABA-B receptors. Three types of GABA-B receptors are presynaptic to granule cells that can modulate the release of GABA: autoreceptors (Morrisett et al.,
1993; Mott et al., 1993; Mody et al., 1994; Haas et al., 1996), presynaptic receptors on excitatory terminals, and somatic or dendritic receptors that synapse on inhibitory interneurons (Misgeld et al., 1989; Mott et al., 1993). Activation of each of these receptors inhibits the release of GABA, and consequently decreases GABA mediated suppression in the DG (Misgeld et al., 1992; Mott et al., 1993; Mody et al., 1994; Haas et al., 1996). Furthermore, GABA-B receptor agonists such as baclofen can facilitate epileptogenic activity (Steffensen and Hendriksen, 1991; Misgeld et al., 1992). GABA-B receptor blockers by contrast are reported to increase paired-pulse suppression in the DG (Haas et al., 1996) and are frequently known to have antiepileptic effects. Our failure to find any consistent changes in the late phase of suppression argues against a direct effect of presynaptic GABA-B involvement. However, postsynaptically GABA-B receptors may be affected following KA administration. Postsynaptic GABA-B receptors are linked to long-lasting
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
potassium currents which are involved in slow inhibitory processes (McCarren and Alger, 1985; Mott et al., 1993; Mody et al., 1994). More consistent changes in the late phase of paired-pulse suppression would perhaps lean towards this hypothesis. The last possibility, that inhibitory loss is a result of a sustained cellular depolarization, is suggested by intracellular recordings showing that application of KA causes cellular depolarization, which can result in a depolarization block (Fisher and Alger, 1984). Cellular depolarization could counteract an effect of GABA-A receptor activation, which acts primarily on chloride channels (Mott et al., 1993; Mody et al., 1994). According to this hypothesis, the convulsive effect of KA is attributable to ineffective inhibition rather than an effect on inhibition. This suggestion is consistent with our failure to find increased facilitation in KA treated animals. In summary, we have demonstrated that a decrease in the early phase of suppression predicts the onset of AD when using a KA model. Neither facilitation nor the late phase of suppression show a similar relationship. The loss in suppression is probably a reflection of changes in GABAergic transmission. The most likely explanation is that there is a sustained cellular depolarization as a result of the mechanisms of KA. This could lead to a use-dependent change in responsiveness of GABA-A receptors. We are currently trying to distinguish these alternatives pharmacologically.
Acknowledgements This research was supported by a grant to Dr N.W. Milgram from the National Science and Engineering Research Council. Special thanks to Vidya Ann Balkissoon, Jilllian Fecteau, Simone Joseph, and M. Paul Murphy for their technical assistance.
References Bekenstein, J., Rempe, D., Lothman, E., 1993. Decreased heterosynaptic and homosynaptic paired pulse inhibition in
151
the rat hippocampus as a chronic sequela to limbic status epilepticus. Brain Res. 601, 111 – 120. Ben-Ari, Y., 1985. Limbic seizure and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy. Neuroscience 14 (2), 375 – 403. Bradford, H.F., 1995. Glutamate, GABA and epilepsy. Prog. Neurobiol. 47, 477 – 511. Brucato, F.H., Mott, D.D., Lewis, D.V., Swartzwelder, H.S., 1995. GABA-B receptors modulate synaptically-evoked responses in the rat dentate gyrus, in vivo. Brain Res. 677, 326 – 332. Burdette, L.J., Gilbert, M.E., 1995. Stimulus parameters affecting paired-pulse depression of dentate granule cell field potentials: I. Stimulus intensity. Brain Res. 680, 53 – 62. During, M.J., Ryder, K.M., Spencer, D.D., 1996. Hippocampal transporter function in temporal-lobe epilepsy. Nature 376, 174 – 177. Evans, M.S., Viola-McCabe, K.E., Caspary, D.M., Faingold, C.L., 1994. Loss of synaptic inhibition during repetitive stimulation in genetically epilepsy-prone rats. Epilepsy Res. 18, 97 – 105. Faingold, C.L., Randall, M.E., Boersma Anderson, C.A., 1994. Blockade of GABA uptake in tiagabine inhibits audiogenic seizures and reduced neuronal firing in the inferior colliculus of the genetically epilepsy-prone rat. Exp. Neurol. 126, 225 – 232. Fisher, R.S., Alger, B.E., 1984. Electrophysiological mechanisms of kainic acid-induced epileptiform activity in the rat hippocampal slice. J. Neurosci. 4, 1312 – 1323. Francis, J., Mihic, S.J., Sneddon, W.B., Burnham, W.M., 1994. The effect of anticonvulsant drugs on GABA stimulated chloride uptake. Can. J. Neurol. Sci. 21 (1), 3 – 8. Haas, K.Z., Sperber, E.F., Moshe´, S.L., Stanton, P.K., 1996. Kainic acid-induced seizures enhance dentate gyrus inhibition by downregulation of GABA-B receptors. J. Neurosci. 16, 4250 – 4260. Joy, R.M., Albertson, T.E., 1988. Convulsant-induced changes in perforant path-dentate gyrus excitability in urethane anesthetized rats. J. Pharmacol. Exp. Ther. 246 (3), 887 – 895. Kapur, J., Lothman, E.W., 1989. Loss of inhibition precedes delayed spontaneous seizures in the hippocampus after tetanic stimulation. J. Neurophysiol. 61 (2), 427 – 434. Kapur, J., Stringer, J., Lothman, E.W., 1989. Evidence that repetitive seizures in the hippocampus cause a lasting reduction of GABAergic inhibition. J. Neurophysiol. 61 (2), 417 – 426. Lømo, T., 1971. Patterns of activation in a monosynaptic cortical pathway: the perforant path input to the dentate area of the hippocampal formation. Exp. Brain Res. 12, 18 – 45. McCarren, M., Alger, B.E., 1985. Use-dependent depression of IPSPs in rat hippocampal pyramidal cells in vitro. J. Neurophysiol. 53 (2), 557 – 571. Milgram, N.W., Yearwood, T., Khurgel, M., Ivy, G.O., Racine, R., 1991. Changes in inhibitory processes in the hippocampus following recurrent seizures induced by sys-
152
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
temic administration of kainic acid. Brain Res. 551, 236– 246. Milgram, N.W., Michael, M., Cammisuli, S., Head, E., Ferbinteanu, J., Reid, C., Murphy, M.P., Racine, R.J., 1995. Development of spontaneous seizures over extended electrical kindling: II. Persistence of dentate inhibitory suppression. Brain Res. 670, 112–120. Misgeld, U., Muller, W., Brunner, H., 1989. Effects of ( − )baclofen on inhibitory neurons in the guinea pig hippocampal slice. Pflu¨g. Arch. 414, 139–144. Misgeld, U., Bijak, M., Burner, H., 1992. Granule cell inhibition and the activity of hilar neurons. In: Ribak, C.E., Gall, C.M., Mody, I. (Eds.), The Dentate Gyrus and Its Role in Seizures. Epilepsy Research, suppl. 7. Elsevier, Amsterdam, pp. 113 –118 Mody, I., De Koninck, Y., Otis, T.S., Soltesz, I., 1994. Bridging the cleft at GABA synapses in the brain. Trends Neurosci. 17 (12), 517–525. Morrisett, R.A., Lewis, D.V., Swartzwelder, H.S., Wilson, W.A., 1993. Antiepileptic effects of GABA-B receptor activation in area CA3 of rat hippocampus. Brain Res. 600, 235 – 242. Mott, D., Xie, C., Wilson, W.A., Swartzwelder, H.S., Lewis, D.V., 1993. GABA-B autoreceptors mediate activity-dependent disinhibition and enhance signal transmission in the dentate gyrus. J. Neurophysiol. 69 (3), 674–690. Pfeiffer, M., Draguhn, A., Meierkord, H., Heinemann, U., 1996. Effects of GABA agonists and GABA uptake inhibitors on pharmacosensitive and pharmacoresistant epileptiform activity in vitro. Br. J. Pharmacol. 119 (3), 569 – 577.
.
Raushe, G., Sarvey, J.M., Heinemann, U., 1989. Slow synaptic inhibition in relation to frequency habituation in dentate granule cells of the rat hippocampal slices. Exp. Brain Res. 78, 233 – 242. Scharfman, H.E., Schwartzkroin, P.A., 1990. Responses of cells of the rat fascia dentata to prolonged stimulation of the perforant path: sensitivity of hilar cells and changes of granule cell excitability. Neuroscience 35, 491 – 504. Sloviter, R.S., 1990. Feedforward and feedback inhibition of hippocampal principal cell activity evoked by perforant path stimulation: GABA-mediated mechanisms that regulate excitability in vivo. Hippocampus 1, 31 – 40. Steffensen, S.C., Hendriksen, S.J., 1991. Effects of baclofen and bicuculline on inhibition in the fascia dentata and hippocampus regio superior. Brain Res. 538, 46 – 53. Stringer, J.L., 1993. Alterations in both excitation and inhibition occur before the onset of seizure activity in the hippocampus. Epilepsy Res. 16, 99 – 109. Tuff, L.P., Racine, R.J., Adamec, R., 1983. The effect of kindling on GABA-mediated inhibition in the dentate gyrus of the rat. I. Paired pulse depression. Brain Res. 27, 79 – 90. Turski, L., Ikonomidou, C., Turski, W.A., Bortolotto, Z.A., Cavalheiro, E.A., 1989. Review: cholinergic mechanisms and epileptogenesis. The seizures induced by pilocarpine: a novel experimental model of intractable epilepsy. Synapse 3 (2), 154 – 171. Zhao, D., Leung, L.S., 1992. Hippocampal kindling induced paired-pulse depression in the dentate gyrus and pairedpulse facilitation in CA3. Brain Res. 582, 163 – 167.
.
Selective loss of early suppression in the dentate gyrus precedes kainic acid induced electrographic seizures C.J. Ikeda-Douglas a, E. Head a, R.M.D. Holsinger a, L. Tremblay a, R. Racine b, N.W. Milgram a,* a
Life Science Di6ision, Scarborough College, Uni6ersity of Toronto, 1265 Military Trail, Scarborough, Ontario, M1C 1A4, Canada b Department of Psychology, McMaster Uni6ersity, Hamilton, Ontario, L8S 4K1, Canada Received 17 December 1997; received in revised form 1 March 1998; accepted 20 April 1998
Abstract The role of inhibitory and facilitatory processes in the induction of seizures was studied in a kainic acid (KA) model of epilepsy. The dentate gyrus (DG) response to paired-pulse stimulation of the perforant path (PP) was monitored prior to and immediately following the initial KA induced afterdischarge (AD) in rats chronically prepared with stimulation recording electrodes. The subjects received a 1-h program of stimulation consisting of repeated sequences of pulse pairs at a short (20–30 ms), intermediate (45 – 90 ms), and long (200 – 300 ms) interpulse interval (IPIs). The stimulation program was administered both under control conditions and immediately following systemic injection of KA. During the control condition, stable suppression of population spike measures was obtained at the short (early phase) and long (late phase) IPIs, while facilitation was observed at the intermediate IPI. Administration of KA resulted in a progressive loss of suppression prior to the initial AD at the short IPI; neither facilitation nor the late phase of suppression were significantly affected. The early phase decreased further following the initial discharge. Since the early phase most likely reflects recurrent inhibition, these results provide evidence that inhibitory loss precedes the occurrence of KA induced AD, and that this inhibitory loss is increased further following the initial evoked AD. A use-dependent disinhibition is one possible explanation for the change in responsiveness that precedes the AD. This disinhibition could result from a depressed response at GABA-A receptors, an increased responsiveness at GABA-B receptors or possibly both. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Dentate gyrus; Kainic acid; Electrographic seizures
* Corresponding author. Tel.: +1 416 2877402; fax: +1 416 2877642; e-mail: [email protected] 0920-1211/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S0920-1211(98)00028-X
144
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
1. Introduction Generation of epileptic seizures is widely believed to depend, at least in part, on modifications in inhibitory transmission. Two lines of evidence support this view. The first is based on pharmacological studies with agents that affect GABA receptors. Antagonists, such as bicuculline, that block GABA transmission, can have potent epileptogenic effects (Bradford, 1995). Agents that augment GABA transmission, in contrast have anticonvulsant effects. These include agonists such as muscimol (Morrisett et al., 1993), modulators of GABA, such as diazepam (Francis et al., 1994), and reuptake inhibitors, such as tiagabine, that prolong the action of GABA (Faingold et al., 1994). The second line of evidence deals with changes in neural transmission that accompany or follow the induction of seizures. A loss of inhibition has been reported to result from the kindling procedure, which involves the repeated induction of seizures (Tuff et al., 1983). Inhibitory loss has been reported following electrically and chemically induced status epilepticus (Bekenstein et al., 1993). Such seizure related loss of inhibition is particularly notable in epilepsy prone animals (Evans et al., 1994). The research discussed above, however, does not establish disruption of neural inhibition as an essential normal trigger of seizures. Treatments affecting transmitter systems other than GABA, such as glutamate (Joy and Albertson, 1988) or acetylcholine (Turski et al., 1989), can also have epileptogenic effects. Furthermore, decreased inhibition does not always occur in epilepsy models. Increased GABA inhibition has been reported to follow both kindling (Tuff et al., 1983; Milgram et al., 1995) and status epilepticus (Milgram et al., 1991; Haas et al., 1996). Apparently, the effect of seizures on neural inhibition is pathway specific, with some pathways showing an increase and others a decrease (Zhao and Leung, 1992). Another issue relates to the temporal relationship between the occurrence of inhibitory loss and the occurrence of seizures. If there is a link between inhibition and the induction of seizures in the dentate gyrus (DG), then inhibitory transmission should decrease prior to the occurrence of seizures.
Few studies have monitored inhibition preceding seizures, and those that have present contrasting results. Sloviter found that perforant path (PP) stimulation repeated over several hours induced a loss of paired-pulse inhibition, leading to the occurrence of multiple population spikes (Sloviter, 1990). The significance of this finding, however, is unclear. The repeated stimulation paradigm causes multiple seizures, which could be interpreted in either of two ways: (1) multiple seizures are a consequence of stimulation induced inhibitory loss; or (2) the inhibitory loss is a result of the induction of multiple seizures. More direct evidence of a relationship between inhibition and seizure development was obtained by Milgram et al. (1991) who found a loss in early suppression in the DG following systemic administration of kainic acid (KA). Tuff et al. (1983) also reported decreased paired-pulse suppression in the DG prior to afterdischarge (AD) evoked by trains of pulse pairs to the PP. In contrast, Stringer (1993) reported that seizures evoked by repeated trains of electrical stimulation were preceded by an increase in paired-pulse inhibition in the DG, and a decrease in cell field CA1. However, in the DG the repeated stimulation led to progressive increments in the response evoked by the first pulse (C-pulse) of the pulse pair. Paired-pulse suppression is directly linked to the amplitude of the response evoked by the C-pulse: the larger the response the greater the suppression (Milgram et al., 1991; Burdette and Gilbert, 1995). Thus, a stimulation-induced increase in field potential response to the C-pulse can account for the inhibitory increase that precedes the development of seizures. The present investigation had two goals: (1) to establish a relationship between paired-pulse suppression and the development of AD activity in response to systemic KA; and (2) to determine whether changes in DG transmission were restricted to those occurring during the early phase of suppression. This second goal stemmed from the known sensitivity of the paired-pulse procedure to interpulse interval (IPI). At short IPIs (less than about 40 ms) suppression in the response to pulse two (T-pulse) is primarily due to recurrent collateral inhibition caused by chloride conductances (Lømo, 1971; Tuff et al., 1983; Joy and Albertson,
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
1988; Mody et al., 1994; Haas et al., 1996). At longer IPIs (between about 45 and 90 ms) the response to the T-pulse is larger than that evoked by the C-pulse, reflecting a phase of paired-pulse facilitation (Joy and Albertson, 1988). At still longer intervals, the T-pulse response is depressed, which represents a late phase of inhibition that is mediated by potassium conductances (Joy and Albertson, 1988; Scharfman and Schwartzkroin, 1990; Haas et al., 1996). We were particularly interested in the relationship between the late phase of inhibition and seizure development because the late phase of inhibition has been linked to the GABA-B receptor (Raushe et al., 1989; Steffensen and Hendriksen, 1991). Furthermore, several reports suggest a critical role of GABA-B inhibition in epilepsy (Morrisett et al., 1993; Brucato et al., 1995). To accomplish these goals, we monitored the DG response to a program of repeated paired-pulse stimulation to the PP under a control condition and again following systemic administration of KA. KA treatment causes multiple ADs and eventually leads to SE (Ben-Ari, 1985). However, AD in the DG typically develops slowly; the initial discharge generally occurs after a minimum of 20 min, but may take as long as 90 min. The present experiment focuses only on those events that lead to and immediately follow the initial discharge.
2. Methods Subjects were ten male hooded Long Evans rats chronically implanted with DG and PP electrodes. The subjects were selected from a larger population of animals on the basis of having stable evoked field potentials, which showed a topography characteristic of the DG. We also limited the subject pool to animals that showed stable paired-pulse suppression at a short IPI.
2.1. Surgical and electrophysiological test procedures Rats were implanted with two pairs of bipolar electrodes constructed from twisted lengths of Teflon-coated stainless steel wires 250 mm in diame-
145
ter. The tip separation was approximately 500 mm. The stimulating electrodes were placed in the PP and recording electrodes in the granular layer of the DG using coordinates previously described (Lømo, 1971). The surgical procedures were performed under sodium pentobarbital anaesthesia (65 mg/ kg). A stainless steel wire wrapped around jewelers’ screws embedded into the skull served as the ground electrode. The electrodes were connected to microconnectors, which were then inserted into nine pin assemblies and fixed to the animal’s skull using dental cement. For all electrophysiological testing, the animals were placed in shielded cages and connected through leads constructed from low noise cable (Microdot) to both the stimulating and recording units. Stimulation consisted of constant current biphasic pulses 0.10 ms in duration with a 0.05-ms delay between the first and second pulse delivered through a stimulus isolation unit (ISO-Flex, AMPI Instruments). The signals from the animal were taken to a preamplifier (Grass model) and filtered between 1 and 3000 Hz with the gain set individually to maximize the response. The output from the preamplifier was fed to an analog to digital converter and digitized at 12000 Hz. The output of the A/D converter was taken to a computer, and the signals were stored on floppy disks for offline analysis. All data acquisition and analysis was performed under computer control using dedicated programs developed in the ASYST programming language.
2.2. Test procedure Following an initial stage of screening, averaged evoked potentials from six individual sweeps were taken at 13 different intensities, ranging from 100 to 2000 mA, which were used to construct inputoutput (I/O) curves. Two response parameters were measured: the population spike (PS), which provides a field measure of cellular discharge, and the population excitatory post-synaptic potential (pEPSP), which provides a measure of synaptic current. The PS was calculated by projecting a line from the negative peak to a tangent line connecting the spike onset and offset. The EPSP slope was
146
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
calculated by placing a cursor at a fixed point on the rising phase of the response, differentiating a segment of the sweep around the cursor, and taking the mean of the three largest values of the derivative function (Milgram et al., 1991). The baseline test was conducted 24 h prior to KA treatment, and was administered over 50 min. For each animal, the stimulation intensity was individually adjusted at 50 – 70% of the maximum population spike amplitude. Paired pulses were delivered in repeated blocks containing sequences of three interpulse intervals, with a 20-s interval separating each pulse pair. Each sequence contained an IPI of: (1) 20 – 30 ms, the range necessary for producing early phase of suppression; (2) an interval between 45 and 90 ms, the range used to produce facilitation; and (3) an interval between 200 and 300 ms, which was used to evaluate the late phase of suppression. Thus, a total of 50 paired pulses were delivered at each of the three IPIs. For the treatment phase, the animals were first injected with a 12-mg/kg dose of KA (Sigma), which was dissolved in physiological saline and prepared as a 1% solution. The pH was neutralized with 10 N sodium hydroxide. Immediately following intraperitoneal (i.p.) injection, the animals were placed in the test chamber and administered a paired-pulse procedure identical to that used for the baseline test. EEG activity from the DG was continuously monitored to establish the exact time when the first AD occurred.
3. Results
3.1. Beha6ioral response to systemic kainic acid As we have previously reported, an initial period of hypoactivity followed systemic administration of KA (Milgram et al., 1991). This, in turn, was followed by an interval of increased behavioral activity with occasional wet dog shakes. The initial electroencephalographic afterdischarge (AD) in the DG occurred during this second interval, generally between 20 and 40 min after the injection, and was unaccompanied by behavioral seizures. The first AD was short, and was
followed by a transient period of postictal depression. Subsequent ADs increased in duration and showed more obvious behavioral correlates, but these were not of concern for this study.
3.2. Early phase of suppression At the short IPI, the T-pulse response was consistently suppressed relative to the C-pulse during the baseline test (Fig. 1B and Fig. 2A). This suppression decreased progressively following administration of KA (Fig. 1A and Fig. 2B). To test for the significance of this effect, we compared the average response to five pulse pairs at the short IPI immediately following administration of KA with the average response to the corresponding five that immediately preceded the development of the first AD. We also looked at the average of five pulse pairs following recovery from the AD (Fig. 3). The results were evaluated using a repeated measures ANOVA with treatment (KA versus control) and time following treatment as main effects. The time dependent change in suppression caused by KA was indicated by a significant interaction between time and treatment (F(2,18)=4.97; PB0.019). To determine whether these effects were due to changes in the response to pulse 1, we analyzed the response to the C-pulse alone. There was a slight decrease over time, but there was considerable variability; the response decreased prior to the first AD in only three instances. There was a slightly larger response in the other seven animals. However, the overall C-pulse amplitude was not significantly different from the response at the start of the session. Following the initial AD, the response to the C-pulse was markedly suppressed and the PS was typically blocked; these effects were transient. The pEPSP measure was analyzed in the same way as the PS. The results of the repeated measure ANOVA revealed an interaction between time following KA and treatment, (F(2,18)= 12.28; PB 0.0002). As illustrated in Fig. 4, the suppression score for the slope measure did not change prior to the occurrence of the AD; following the AD, however, there was a switch to facilitation.
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
147
Fig. 1. Characteristic pairs of evoked DG field potentials at three different time points are illustrated for the kainic acid (KA) condition (A) and baseline condition (B). T0, beginning of session; T1, immediately preceding the first afterdischarge (AD); T2, following recovery from the AD. Corresponding pulse pairs are taken from the baseline (B).
3.3. Changes in facilitation
3.4. Changes in late phase inhibition
As illustrated in Fig. 3B facilitation of the PS decreased after treatment with KA. However, there were no statistically significant changes in either of the PS measures or the EPSP ratio (Fig. 4B).
As shown in Fig. 3C there was a slight but non-significant loss in the late phase of suppression after KA administration (Fig. 4C).
148
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
Fig. 2. Changes in the early phase of suppression both (A) before and (B) following systemic kainic acid (KA) are shown for a representative subject. Each point represents the average amplitude of the population spike measure for the conditioning and test pulses of three consecutive pulse pairs. The bars represent standard errors.
4. Discussion The present study demonstrates that pairedpulse suppression decreases progressively following the systemic administration of KA, and that this decrease precedes the occurrence of AD activity in the dentate gyrus. Paired-pulse suppression decreased further following the initial AD. The loss in suppression that preceded the occurrence of AD was also selective; only the early phase of paired-pulse suppression was significantly affected. Joy and Albertson (1988) have previously reported a loss in the early phase of suppression following administration of KA. The present results supplement their findings by showing the progressive nature of the change, and also the potential link to the occurrence of seizures. As mentioned previously, the early phase of paired-pulse suppression is largely a result of recurrent collateral inhibition and is mediated by activation of GABA-A receptors (Tuff et al.,
1983; Evans et al., 1994; Mody et al., 1994; During et al., 1996). The present findings therefore, suggest first that a gradual loss in effectiveness of GABA-A mediated inhibition follows systemic treatment with KA and, second that this loss is temporally linked to the development of AD. These results differ from those reported by Stringer (1993), who found a progressive increase in paired-pulse suppression in the DG prior to the occurrence of AD. These discrepant observations can be reconciled by taking into consideration procedural differences. Stringer used repeated trains of electrical stimulation to induce seizures. This procedure lead to progressive increases in DG excitability, which was manifest by increases in the amplitude of the response evoked by the C-pulse. This increased response amplitude probably accounts for the increased paired-pulse suppression, since suppression in the DG is known to vary directly with the amplitude of the response evoked by the C-pulse (Milgram et al., 1991;
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
149
Fig. 3. The effects of KA on the population spike measure of suppression. The suppression score was normalized by dividing the paired pulse ratio by the ratio from the first 5 min. (A) Changes occurring during the early phase of suppression. Changes in facilitation are show in (B) while changes in the late phase are shown in (C). Control and post KA data are compared at three different time points: at the beginning of each session, prior to the first afterdischarge (AD), and following recovery from the first AD.
Burdette and Gilbert, 1995). The present study differed in that the response to the C-pulse did not change significantly during the interval preceding the occurrence of AD. The finding that the C-pulse response did not significantly decrease after treatment with KA, suggests that the KA-induced AD is not caused by a generalized increase in excitability. Joy and Albertson (1988) reached a similar conclusion. They found one class of convulsants, which were GABA-A antagonists, caused an increase in granule cell excitability. A second class of convulsants acted on glutamate receptors and depressed both paired-pulse facilitation and paired-pulse depression; KA was a member of this class. Four possible mechanisms could account for the selective inhibitory loss that preceded the occurrence of seizures: (1) an increase in the phase of facilitation, which could mask recurrent inhibition; (2) a use-dependent loss of transmission at
fast inhibitory synapses; (3) increased activation or responsiveness of presynaptic GABA-B receptors; and (4) inhibitory block caused by kainic acid induced depolarization. The first possibility, enhanced facilitation masking recurrent collateral inhibition can be ruled out. The present results allow us to rule out this possibility. We found no evidence that KA increased facilitation; in fact, there was a small loss of facilitation. These findings are consistent with a report by Joy and Albertson (1988) of a decrease in facilitation following systemic administration of KA. The second possibility, that of a use-dependent loss of GABA-A is suggested by intracellular studies of inhibitory transmission in slices. A single PP stimulation is sufficient to induce a transient use-dependent inhibitory loss (McCarren and Alger, 1985; Mott et al., 1993). Scharfman and Schwartzkroin (1990) observed a more long-
150
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
Fig. 4. The effects of KA on the pEPSP slope measure of paired pulse suppression in the DG. The suppression scores were normalized. Control and post KA data are compared at three time points: the first 5 min of the session, a period prior to the first AD, and following recovery from the AD. Changes in the early phase of suppression are shown in (A), (B) illustrates the effects of facilitation and (C) shows the effects on the late phase.
lasting loss in paired-pulse inhibition when activation of the perforant path was repeated over several hours. Repeated stimulation of the CA3CA1 pathway has also been repeated to produce a long-lasting loss of paired-pulse suppression (Kapur and Lothman, 1989; Kapur et al., 1989). These effects could result from depletion of synaptically available GABA (Kapur and Lothman, 1989; Kapur et al., 1989; Pfeiffer et al., 1996). In the present study, a use-dependent loss of GABA could account for both the progressive loss that precedes AD and the further loss of suppression that follows AD. However, the synaptic depletion hypothesis can not account for the changes we observed in the EPSP slope measure that did not show a loss in suppression prior to AD, but showed significant inhibitory loss following AD. The third possibility relates to GABA-B receptors. Three types of GABA-B receptors are presynaptic to granule cells that can modulate the release of GABA: autoreceptors (Morrisett et al.,
1993; Mott et al., 1993; Mody et al., 1994; Haas et al., 1996), presynaptic receptors on excitatory terminals, and somatic or dendritic receptors that synapse on inhibitory interneurons (Misgeld et al., 1989; Mott et al., 1993). Activation of each of these receptors inhibits the release of GABA, and consequently decreases GABA mediated suppression in the DG (Misgeld et al., 1992; Mott et al., 1993; Mody et al., 1994; Haas et al., 1996). Furthermore, GABA-B receptor agonists such as baclofen can facilitate epileptogenic activity (Steffensen and Hendriksen, 1991; Misgeld et al., 1992). GABA-B receptor blockers by contrast are reported to increase paired-pulse suppression in the DG (Haas et al., 1996) and are frequently known to have antiepileptic effects. Our failure to find any consistent changes in the late phase of suppression argues against a direct effect of presynaptic GABA-B involvement. However, postsynaptically GABA-B receptors may be affected following KA administration. Postsynaptic GABA-B receptors are linked to long-lasting
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
potassium currents which are involved in slow inhibitory processes (McCarren and Alger, 1985; Mott et al., 1993; Mody et al., 1994). More consistent changes in the late phase of paired-pulse suppression would perhaps lean towards this hypothesis. The last possibility, that inhibitory loss is a result of a sustained cellular depolarization, is suggested by intracellular recordings showing that application of KA causes cellular depolarization, which can result in a depolarization block (Fisher and Alger, 1984). Cellular depolarization could counteract an effect of GABA-A receptor activation, which acts primarily on chloride channels (Mott et al., 1993; Mody et al., 1994). According to this hypothesis, the convulsive effect of KA is attributable to ineffective inhibition rather than an effect on inhibition. This suggestion is consistent with our failure to find increased facilitation in KA treated animals. In summary, we have demonstrated that a decrease in the early phase of suppression predicts the onset of AD when using a KA model. Neither facilitation nor the late phase of suppression show a similar relationship. The loss in suppression is probably a reflection of changes in GABAergic transmission. The most likely explanation is that there is a sustained cellular depolarization as a result of the mechanisms of KA. This could lead to a use-dependent change in responsiveness of GABA-A receptors. We are currently trying to distinguish these alternatives pharmacologically.
Acknowledgements This research was supported by a grant to Dr N.W. Milgram from the National Science and Engineering Research Council. Special thanks to Vidya Ann Balkissoon, Jilllian Fecteau, Simone Joseph, and M. Paul Murphy for their technical assistance.
References Bekenstein, J., Rempe, D., Lothman, E., 1993. Decreased heterosynaptic and homosynaptic paired pulse inhibition in
151
the rat hippocampus as a chronic sequela to limbic status epilepticus. Brain Res. 601, 111 – 120. Ben-Ari, Y., 1985. Limbic seizure and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy. Neuroscience 14 (2), 375 – 403. Bradford, H.F., 1995. Glutamate, GABA and epilepsy. Prog. Neurobiol. 47, 477 – 511. Brucato, F.H., Mott, D.D., Lewis, D.V., Swartzwelder, H.S., 1995. GABA-B receptors modulate synaptically-evoked responses in the rat dentate gyrus, in vivo. Brain Res. 677, 326 – 332. Burdette, L.J., Gilbert, M.E., 1995. Stimulus parameters affecting paired-pulse depression of dentate granule cell field potentials: I. Stimulus intensity. Brain Res. 680, 53 – 62. During, M.J., Ryder, K.M., Spencer, D.D., 1996. Hippocampal transporter function in temporal-lobe epilepsy. Nature 376, 174 – 177. Evans, M.S., Viola-McCabe, K.E., Caspary, D.M., Faingold, C.L., 1994. Loss of synaptic inhibition during repetitive stimulation in genetically epilepsy-prone rats. Epilepsy Res. 18, 97 – 105. Faingold, C.L., Randall, M.E., Boersma Anderson, C.A., 1994. Blockade of GABA uptake in tiagabine inhibits audiogenic seizures and reduced neuronal firing in the inferior colliculus of the genetically epilepsy-prone rat. Exp. Neurol. 126, 225 – 232. Fisher, R.S., Alger, B.E., 1984. Electrophysiological mechanisms of kainic acid-induced epileptiform activity in the rat hippocampal slice. J. Neurosci. 4, 1312 – 1323. Francis, J., Mihic, S.J., Sneddon, W.B., Burnham, W.M., 1994. The effect of anticonvulsant drugs on GABA stimulated chloride uptake. Can. J. Neurol. Sci. 21 (1), 3 – 8. Haas, K.Z., Sperber, E.F., Moshe´, S.L., Stanton, P.K., 1996. Kainic acid-induced seizures enhance dentate gyrus inhibition by downregulation of GABA-B receptors. J. Neurosci. 16, 4250 – 4260. Joy, R.M., Albertson, T.E., 1988. Convulsant-induced changes in perforant path-dentate gyrus excitability in urethane anesthetized rats. J. Pharmacol. Exp. Ther. 246 (3), 887 – 895. Kapur, J., Lothman, E.W., 1989. Loss of inhibition precedes delayed spontaneous seizures in the hippocampus after tetanic stimulation. J. Neurophysiol. 61 (2), 427 – 434. Kapur, J., Stringer, J., Lothman, E.W., 1989. Evidence that repetitive seizures in the hippocampus cause a lasting reduction of GABAergic inhibition. J. Neurophysiol. 61 (2), 417 – 426. Lømo, T., 1971. Patterns of activation in a monosynaptic cortical pathway: the perforant path input to the dentate area of the hippocampal formation. Exp. Brain Res. 12, 18 – 45. McCarren, M., Alger, B.E., 1985. Use-dependent depression of IPSPs in rat hippocampal pyramidal cells in vitro. J. Neurophysiol. 53 (2), 557 – 571. Milgram, N.W., Yearwood, T., Khurgel, M., Ivy, G.O., Racine, R., 1991. Changes in inhibitory processes in the hippocampus following recurrent seizures induced by sys-
152
C.J. Ikeda-Douglas et al. / Epilepsy Research 31 (1998) 143–152
temic administration of kainic acid. Brain Res. 551, 236– 246. Milgram, N.W., Michael, M., Cammisuli, S., Head, E., Ferbinteanu, J., Reid, C., Murphy, M.P., Racine, R.J., 1995. Development of spontaneous seizures over extended electrical kindling: II. Persistence of dentate inhibitory suppression. Brain Res. 670, 112–120. Misgeld, U., Muller, W., Brunner, H., 1989. Effects of ( − )baclofen on inhibitory neurons in the guinea pig hippocampal slice. Pflu¨g. Arch. 414, 139–144. Misgeld, U., Bijak, M., Burner, H., 1992. Granule cell inhibition and the activity of hilar neurons. In: Ribak, C.E., Gall, C.M., Mody, I. (Eds.), The Dentate Gyrus and Its Role in Seizures. Epilepsy Research, suppl. 7. Elsevier, Amsterdam, pp. 113 –118 Mody, I., De Koninck, Y., Otis, T.S., Soltesz, I., 1994. Bridging the cleft at GABA synapses in the brain. Trends Neurosci. 17 (12), 517–525. Morrisett, R.A., Lewis, D.V., Swartzwelder, H.S., Wilson, W.A., 1993. Antiepileptic effects of GABA-B receptor activation in area CA3 of rat hippocampus. Brain Res. 600, 235 – 242. Mott, D., Xie, C., Wilson, W.A., Swartzwelder, H.S., Lewis, D.V., 1993. GABA-B autoreceptors mediate activity-dependent disinhibition and enhance signal transmission in the dentate gyrus. J. Neurophysiol. 69 (3), 674–690. Pfeiffer, M., Draguhn, A., Meierkord, H., Heinemann, U., 1996. Effects of GABA agonists and GABA uptake inhibitors on pharmacosensitive and pharmacoresistant epileptiform activity in vitro. Br. J. Pharmacol. 119 (3), 569 – 577.
.
Raushe, G., Sarvey, J.M., Heinemann, U., 1989. Slow synaptic inhibition in relation to frequency habituation in dentate granule cells of the rat hippocampal slices. Exp. Brain Res. 78, 233 – 242. Scharfman, H.E., Schwartzkroin, P.A., 1990. Responses of cells of the rat fascia dentata to prolonged stimulation of the perforant path: sensitivity of hilar cells and changes of granule cell excitability. Neuroscience 35, 491 – 504. Sloviter, R.S., 1990. Feedforward and feedback inhibition of hippocampal principal cell activity evoked by perforant path stimulation: GABA-mediated mechanisms that regulate excitability in vivo. Hippocampus 1, 31 – 40. Steffensen, S.C., Hendriksen, S.J., 1991. Effects of baclofen and bicuculline on inhibition in the fascia dentata and hippocampus regio superior. Brain Res. 538, 46 – 53. Stringer, J.L., 1993. Alterations in both excitation and inhibition occur before the onset of seizure activity in the hippocampus. Epilepsy Res. 16, 99 – 109. Tuff, L.P., Racine, R.J., Adamec, R., 1983. The effect of kindling on GABA-mediated inhibition in the dentate gyrus of the rat. I. Paired pulse depression. Brain Res. 27, 79 – 90. Turski, L., Ikonomidou, C., Turski, W.A., Bortolotto, Z.A., Cavalheiro, E.A., 1989. Review: cholinergic mechanisms and epileptogenesis. The seizures induced by pilocarpine: a novel experimental model of intractable epilepsy. Synapse 3 (2), 154 – 171. Zhao, D., Leung, L.S., 1992. Hippocampal kindling induced paired-pulse depression in the dentate gyrus and pairedpulse facilitation in CA3. Brain Res. 582, 163 – 167.
.