- Email: [email protected]
Cellular mechanisms underlying spontaneous interictal spikes in an acute model of focal cortical epileptogenesis
Pergamon
PII:
Neuroscience Vol. 88, No. 1, pp. 107–117, 1999 Copyright 1998 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/99 $19.00+0.00 S0306-4522(98)00201-2
CELLULAR MECHANISMS UNDERLYING SPONTANEOUS INTERICTAL SPIKES IN AN ACUTE MODEL OF FOCAL CORTICAL EPILEPTOGENESIS M. de CURTIS,* C. RADICI and M. FORTI Department of Experimental Neurophysiology, Istituto Nazionale Neurologico, via Celoria 11, 20133 Milan, Italy Abstract––The cellular mechanisms involved in the generation of spontaneous epileptiform potentials were investigated in the pirifom cortex of the in vitro isolated guinea-pig brain. A single, unilateral injection of bicuculline (150–200 nmol) in the anterior piriform cortex induced locally spontaneous interictal spikes that recurred with a period of 8.814.47 s and propagated caudally to the ipsi- and contralateral hemispheres. Simultaneous extra- and intracellular recordings from layer II and III principal cells showed that the spontaneous interictal spike correlates to a burst of action potentials followed by a large afterdepolarization. Intracellular application of the sodium conductance blocker, QX-314 (80 mM), abolished bursting activity and unmasked a high-threshold slow spike enhanced by the calcium chelator EGTA (50 mM). The slow spike was abolished by membrane hyperpolarization and by local perfusion with 2 mM cadmium. The depolarizing potential that followed the primary burst was reduced by arterial perfusion with the N-methyl--aspartate receptor antagonist, -2-amino-5-phosphonopentanoic acid (100–200 µM). The non-N-methyl--aspartate glutamate receptor antagonist, 6-cyano-7-nitroquinoxaline2,3-dione (20 µM), completely and reversibly blocked the spontaneous spikes. The interictal spikes were terminated by a large afterpotential blocked either by intracellular QX-314 (80 mM) or by extracellular application of phaclofen and 2-hydroxysaclofen (10 and 4 mM, respectively). The present study demonstrates that, in an acute model of epileptogenesis, spontaneous interictal spikes are fostered by a primary burst of fast action potentials that ride on a regenerative high-threshold, possibly calcium-mediated spike, which activates a recurrent, glutamate-mediated potential responsible for the entrainment of adjacent and remote cortical regions. The bursting activity is controlled by a GABAB receptor-mediated inhibitory synaptic potential. 1998 IBRO. Published by Elsevier Science Ltd. Key words: epileptogenesis, interictal spikes, intracellular recordings, isolated brain preparation, piriform cortex.
Spontaneous interictal spikes (sISs) represent the electrographic mark of an active epileptic focus in several epileptic human conditions and can be reproduced in experimental models of focal epilepsy.1,15 Interictal spikes are fast transients clearly distinguished from the baseline, followed by a slow potential of variable amplitude and duration.2 Experimental studies demonstrated that sISs occur synchronously in a large population of cortical neurons and correlate at the single cell level to a depolarization of the membrane potential, designated as paroxysmal depolarization shift.23,28,34 In vitro studies on stimulus-evoked epileptiform discharges acutely established by various means in different cortical *To whom correspondence should be addressed. Abbreviations: AHP, afterhyperpolarizing potential; AP-5, -2-amino-5-phosphonopentanoic acid; APC, anterior piriform cortex; CNQX, 6-cyano-7-nitroquinoxaline2,3-dione; EGTA, ethyleneglycolbis(aminoethyl ether) tetra-acetate; HEPES, N-2-hydroxyethylpiperazine-N -2ethanesulphonic acid; LOT, lateral olfactory tract; NMDA, N-methyl--aspartate; sIS, spontaneous interictal spike. 107
regions demonstrated that the paroxysmal depolarization shift is initiated by a primary voltagedependent intrinsic burst of action potentials, which gives rise to a large excitatory postsynaptic potential mediated by the recurrent activation of local glutamatergic synapses.12,13,15,31,39 Although it has never been demonstrated directly, such a sequence of events is expected to occur during the generation of spontaneous interictal potentials. The piriform cortex has been shown to be involved in limbic epileptogenesis.5,11,13,16,26,27,30 sISs in the piriform cortex were observed early during the development of kindling epileptogenesis11,16 and were found to propagate from the piriform lobe to the hippocampus, where the ability to generate epileptic potentials developed later, when the kindling process is completed.30 We demonstrated that a single, brief intracerebral injection of bicuculline in the piriform cortex of the isolated guinea-pig brain activates persistent spontaneous interictal potentials that show the characteristics of sISs.5,7 The same model of acute focal epileptogenesis is utilized here to characterize the sequence of intrinsic and synaptic events
108
M. de Curtis et al.
Fig. 1. sISs and subthreshold spontaneous potentials induced by a single 10-s application of a 2-mM bicuculline solution in the APC. Immediately after bicuculline, the frequency of repetition of sIS was lower (middle trace), and increased within 10 min (bottom trace recorded 2 h after bicuculline ejection). As illustrated in the upper control trace, sporadic, low-amplitude potentials were observed before bicuculline application in this experiment. The recordings were performed at 800 µm depth.
responsible for the generation and for the termination of sISs. The utilization of the isolated guineapig brain preparation allows (i) study of the sISs in an intact brain, in the absence of any mechanical alteration of the cerebral tissue, (ii) correlation of events recorded intracellularly during the sIS discharge with simultaneous extracellular responses obtained at various cortical depths/locations in the piriform cortex, and (iii) pharmacological manipulations of sISs. Preliminary results have been published in abstract form.6 EXPERIMENTAL PROCEDURES
Young adult Hartley guinea-pigs (150–250 g; Charles River, Italy) anaesthetized with penthotal sodium (20 mg/ kg) were perfused for 2 min through the aorta with an oxygenated, slightly acidic, cold saline solution (10C, pH 7.1). After craniotomy performed in hypothermia, the brain was removed and transferred to a perfusion chamber, according to the standard procedure.3,4,21,24 All efforts were made to minimize animal suffering and to reduce the number of animals used for the study. The isolated brain was perfused through the basilar artery with an artificial plasma solution (composition, mM: NaCl 126, KCl 2.3, NaH2CO3 26, MgSO4 1.3, CaCl2 2.4, KH2PO4 1.2, glucose 15, HEPES 5, thiourea 0.4 and 3% dextran 70000) oxygenated with a 5% CO2/95% O2 gas mixture (pH 7.3). The perfusion rate was 5.5–6 ml/min. The temperature in the perfusion chamber was gradually raised to 32C after completing the isolation manoeuvre. Bicuculline methiodide (2 mM; RBI) dissolved in Ringer’s solution was injected for 10 s at 800 µm depth in the anterior piriform cortex (APC) through an 8- to 10-µm tip diameter glass pipette connected to a graduated syringe. The pipette was removed after the bicuculline ejection. As described in previous studies,5,7 the total volume ejected was c. 150–200 nl (150–200 nmol of bicuculline). The effect of bicuculline was followed by studying the lateral olfactory tract (LOT)-evoked field responses. Bicuculline was washed out within 60–90 min.7 A bipolar stimulating silver-wire electrode positioned on the LOT was utilized to evoke responses in the APC. Extracellular potentials were recorded with glass micropipettes filled with 0.9% NaCl (3–10 MÙ resistance). Intracellular recordings were performed by means of micropipettes filled with either 3 M or 2 M potassium acetate and 1% biocytine (40–80 MÙ resistance) from principal neurons in layers II and III at 300–700 µm depth. Extra- and intracellular activity was recorded with a Biomedical
Engeneering Amplifier (Thornwood, NY, U.S.A.) and with a Neurodata Amplifier (New York, NY, U.S.A.). Signals were stored via a DTR 2602 Digital Tape Recorder (Biologic, France) for off-line analysis with a Microvax 3400 computer system (Digital, Italy) and a 486 PC. The N-methyl--aspartate (NMDA) and non-NMDA receptor blockers -2-amino-5-phosphonopentanoic acid (AP-5; 100–200 µM; RBI) and 6-cyano-7-nitroquinoxaline2,3-dione (CNQX; 20 µM; RBI) were applied by arterial perfusion. QX-314 (80 mM; RBI) and EGTA (50 mM; Sigma) were dissolved in 1 M potassium acetate and delivered intracellularly through the recording pipette (pH 7.2). Phaclofen (10 mM; Sigma), 5-hydroxysaclofen (4 mM; RBI) and cadmium (Cd2+, 2 mM; Sigma) diluted in standard solution without dextran were locally injected in the tissue, close to the intracellular recording electrode. The effect of the locally applied drugs was rapid (peak activity within 1 min) and completely reversible. When micropipettes filled with biocytin were used for intracellular staining, the isolated brains were fixed overnight with 4% paraformaldehyde. Coronal sections (100 µm thick) were then processed for avidin–horseradish peroxidase visualization and were counterstained with Neutral Red to locate the cortical depth of the cells.7 RESULTS
In all experiments (n=28), sISs were activated by a single 10-s injection of bicuculline and persisted for 4–6 h after bicuculline washout.7 In 21 of 28 tests, bicuculline induced transient activation of spontaneous afterdischarges.7 sISs recurred with a period of 12.763.32 s immediately after bicuculline injection (Fig. 1, middle trace) and gradually stabilized within 10 min to one event every 8.814.47 s (Fig. 1, lower trace). In 28% of the experiments, smallamplitude potentials subthreshold for spike generation were observed in isolation after bicuculline. Non-periodic, sporadic, low-amplitude potentials, subthreshold for population spike generation, were detected in 11% of the experiments before bicuculline was applied (as shown in the upper trace in Fig. 1). In order to identify the site of generation of sISs, simultaneous extracellular recordings were carried out in different regions in the olfactory lobes, ispiand contralateral to the site of bicuculline injection (Fig. 2; n=5). Recordings were performed with the
Spontaneous interictal spikes in focal epileptogenesis
109
Fig. 2. Characteristics of propagation of the sIS in the olfactory and limbic cortices. The sIS recorded at the site of bicuculline injection in the APC is marked by an asterisk (site 2). Recordings were performed in both hemispheres, ipsi- and contralateral to the bicuculline injection site. See text for specifications of the position of the recording electrodes, indicated here on a picture of the ventral view of the isolated brain.
micropipette utilized to inject bicuculline as reference electrode (trace 2 in Fig. 2, marked by an asterisk), moving a second electrode to record simultaneous activity in the olfactory tubercle (site 12), the olfactory bulb (site 10), the LOT (site 11), the APC rostral and caudal to the drug application site (sites 1 and 3), the endopiriform nucleus (site 4), the posterior piriform cortex (sites 5 and 7), the periamygdaloid cortex (sites 6 and 8), the lateral entorhinal cortex (site 9) and the contralateral APC, posterior piriform cortex and entorhinal cortex (sites 13, 14 and 15, respectively). The sIS features were highly reproducible and the delay of the potentials recorded at different sites was correlated in time with the sIS simultaneously recorded with electrode 2. The results demonstrate that sISs originate at the site of bicuculline application in the APC. From here, they propagate throughout the olfactory cortices and into the entorhinal cortex of both hemispheres. Simultaneous extra- and intracellular recordings were performed in the APC at the site of bicuculline injection. Intracellular recordings were performed from 71 APC neurons recorded in 28 isolated guineapig brains. Twenty of 71 neurons were injected with biocytin and were identified morphologically as pyramidal cells (not shown). Twenty-two cells were recorded before, during and after bicuculline injection. Forty-one of 71 neurons were recorded 1 h after the injection of bicuculline. Averaged resting membrane potential and input resistance were 74.184.65 mV (n=55) and 35.535.84 MÙ (n= 24). As reported previously,7 neurons in layers II and III showed similar membrane properties. No differences were observed in sISs recorded intracellularly from layer II and III neurons. The similarity between
the LOT-evoked epileptiform potential and the sIS is illustrated in Fig. 3. The onset of the extracellular population spike in the sIS correlated to a burst of high-frequency action potentials (Fig. 3B). The spontaneous burst/population spike was preceded by a small-amplitude potential (arrowhead in Fig. 3A and B and the following figures). The intracellular burst was followed by a large after-depolarizing potential, which correlated with a surface-negative/depthpositive extracellular potential (asterisk in Fig. 3B). The depth location of the sinks responsible for the components of the LOT-evoked epileptiform potential have been described in detail previously.5,7 Surface-negative/depth-positive potentials correspond to synaptic events located in the superficial layer Ib generated by corticocortical associative synapses. Depth-negative/surface-positive potentials are due to non-synaptic sinks (intrinsic bursting activity) located in the superficial portion of layer II. The burst discharge in the sIS is generated by fast sodium action potentials superimposed on a slow regenerative spike, probably mediated by calcium. The intracellular diffusion through the recording electrode of the sodium channel blocker QX-314 (80 mM) progressively abolished the sodium spikes (n=15) and revealed a large potential enhanced by the co-application through the recording pipette of the calcium chelator EGTA (50 mM), as illustrated in Fig. 4A (from left to right: sISs at cell impalement and after 3 and 8 min of intracellular recording). The calcium dependence of this plateau potential was demonstrated when the inorganic antagonist of voltage-gated calcium conductances, cadmium (2 mM; n=3), was briefly perfused in the extracellular space for 10 s through a 10-µm tip pipette adjacent to
110
M. de Curtis et al.
Fig. 3. Comparison between LOT-evoked epileptiform potential and sIS recorded simultaneously intraand extracellularly 40 min after bicuculline injection. (A) Slow time scale of two sISs, intercalated by an LOT-evoked response (a). (B) Expanded LOT-evoked potential (a) and sIS (b) are shown. Simultaneous extracellular recordings at 100, 500 and 850µm depth from the cortical surface (top, middle and bottom traces in the lower part of B, respectively) close to the intracellular recording (upper traces) are illustrated. Stimulus artifact is marked by an arrow. The arrowheads indicate the small-amplitude potential that precedes the burst discharge. Such a potential can be detected in both the intra- and extracellular recordings. Resting membrane potential: 78 mV.
the intracellular recording electrode. Cadmium abolished the early component of the sIS associated with the depth-negative extracellular potential. The calcium plateau potential was restored when the drug was washed out (Fig. 4C, D). The transient abolition of the calcium plateau isolated a depolarizing potential temporally correlated to the depth-positive synaptic potential defined previously as post-burst after depolarization. The field potential was reduced slightly during the application of cadmium. The amplitude and duration of the post-burst after depolarization and the associated extracellular potential were reduced by arterial perfusion with the NMDA receptor antagonist AP-5 (100–200 µM; n=6). The AP-5-sensitive component of the burst after depolarization was enhanced when the membrane potential was depolarized to increase the NMDA receptor-mediated response (insets in Fig. 5A). sIS periodicity was slowed to 6.482.21 s during AP-5 perfusion. Arterial perfusion with the nonNMDA receptor antagonist CNQX (20 µM) in addition to AP-5 completely abolished the sISs (n=4). The sIS was preceded by a small-amplitude
potential correlated with a tiny deflection in the extracellular response, which was easily detected in surface recordings (Fig. 5A, arrowhead in the right panel; also see Figs 3 and 6). The pre-sIS potential showed a depth reversal around 400µm depth (not shown); it was preserved during perfusion with 100– 200 µM AP-5 solutions, but was abolished by 20 µM CNQX. The post-burst afterdepolarizing potential could be discriminated as a deflection after the early slow spike when the membrane potential was artificially depolarized during sodium conductance blockade with intracellular QX-314 (arrows in Fig. 5B and C). Such a late potential and its extracellular correlate were reduced during the early phase of a perfusion with 20 µM CNQX and 100 µM AP-5 (Fig. 5B, right traces, and C; n=4), before the complete blockade of the sIS occurred. The reduction of the glutamate receptor-mediated synaptic event isolated a slow potential correlated with the extracellular population spike (Fig. 5C). The results described above suggest that the early component of the sIS (i) can be isolated by glutamate
Spontaneous interictal spikes in focal epileptogenesis
111
Fig. 4. Fast sodium conductance blocker unmasked a plateau potential that subtends the sIS. (A) Gradual effect of intracellular diffusion of QX-314 (80 mM) and EGTA (50 mM). Traces were recorded 30 s, 3 and 8 min after cell impalement (from left to right). (B) The first and third traces shown in A are superimposed. The simultaneous extracellular recording recorded at 800 µm depth was not modified by the intracellular QX-314 and EGTA diffusion. Note that the plateau potential onset correlates to the depth-negative extracellular potential. (C, D) Local microperfusion with cadmium (Cd2+, 2 µM) reversibly abolished the early shoulder in the intracellular potential. Washout was complete 10 min after cadmium ejection. In D, three pairs of traces from another experiment show the correlation between the intra- and extracellular potentials before, during perfusion and after cadmium washout. Note that cadmium perfusion reduced the amplitude of the field response. Resting membrane potential was 77 mV for A and B, 74 mV for C and 78 mV for D.
receptor blockade and (ii) is associated with a cadmium-sensitive potential. To further differentiate this potential from the late excitatory synaptic component of the sIS, large hyperpolarizing currents were injected intracellularly during the fast sodium spike blockade by QX-314 (n=4). As illustrated in Fig. 6, membrane hyperpolarization to values negative to 110 mV (resting membrane potential marked by the horizontal dotted line: 72 mV) abolished the early component of the sIS coupled with the depth-negative population spike and, as is predictable for an excitatory synaptic potential, increased the amplitude of the recurrent burst after depolarization correlated with the depth-positive potential. The burst/afterdepolarization was consistently followed by a late potential associated with an extracellular surface-positive/depth-negative deflection (asterisk in Fig. 7B). The late potential was identified as an afterhyperpolarizing potential (AHP) when the membrane potential was depolarized by a steady intracellular current injection (Fig. 7A and B, right trace). The AHP lasted 507173.33 ms (n=16), inverted close to the potassium equilibrium potential (79.45.5 mV; n=13; Fig. 7A) and was reversibly abolished by saclofen (4 mM) and phaclofen
(10 mM) applied together in the superficial cortical layers, close to the intracellular recording electrode (Fig. 8B; n=4). In control conditions, before bicuculline injection, saclofen and phaclofen completely abolished the late inhibitory postsynaptic potential mediated by GABAB receptors (Fig. 8A). Either one of the two compounds alone showed an incomplete blocking effect. The blockade of the AHP prolonged the sIS and induced secondary bursting activity (middle traces in Fig. 8B). The late AHP was similarly abolished by the lidocaine derivative QX-314 (80 mM) applied intracellularly through the recording micropipette (Fig. 8C; n=12). DISCUSSION
The sISs described in the present study show similarities with those observed in focal human epilepsy,1 in chronic animal models of focal epilepsy11,20,30 and in acute epileptogenic conditions.13,26,29,35 The sequence of events that sustain spontaneous epileptiform interictal spikes in the piriform cortex of the isolated guinea-pig brain can elucidate the mechanisms that control sIS generation in other cortical areas and, in general, might help to understand interictal focal epileptogenesis. Four
112
M. de Curtis et al.
Fig. 5. (A) The duration of the sIS burst (upper traces) and the corresponding depth-positive extracellular potential (asterisk in lower traces) are reduced by arterial perfusion with AP-5 (150 µM). The AP-5 effect is more evident when the late burst afterdepolarization is enhanced by membrane potential depolarization (inset; membrane potential depolarized to 48 and 50 mV in control conditions and after AP-5, respectively). The NMDA receptor antagonist did not affect the small-amplitude potential that precedes the sIS. Resting membrane potential: 81 mV. The extracellular potential was recorded at 850 µm depth. (B) sISs during depolarizations of the membrane potential induced by a steady intracellular current injection. QX-314 (80 mM) and EGTA (50 mM) were applied intracellularly through the recording electrode. A late depolarizing potential (arrows in the left panel) was activated in the sIS when the membrane potential was depolarized. On the right, the effect of the early phase of an arterial perfusion with 20 µM CNQX and 100 µM AP-5 is displayed. The traces were acquired within 2 min after perfusion onset, before the complete abolition of the sIS occurred (after 5 min perfusion). (C) Traces from B obtained before (arrows) and after glutamate receptor antagonist perfusion superimposed. The extracellular responses recorded simultaneously at 850 µm depth (lower traces in each panel) show that the depth-positive potential is gradually reduced during blockade of glutamate receptors. This blockade enhanced the amplitude of the population spike component of the sIS. Resting membrane potential (dotted line) was 71 mV.
major original findings are described here: (i) sISs generated at the site of bicuculline injection in the piriform cortex propagate to the limbic cortices bilaterally along anatomical pathways; (ii) the sISs are initiated by a synaptic small-amplitude potential mediated by non-NMDA-type glutamate receptors; (iii) the sISs are associated with a primary burst, possibly sustained by a calcium-dependent regenerative potential; (iv) the sIS duration is terminated by an AHP mediated primarily by a GABAB receptormediated inhibitory postsynaptic potential. We also demonstrate that, as for other cortical areas, the component of the sIS that follows the primary burst is mediated by the recurrent activation of associative, glutamatergic synapses. The similarity between the late part of the LOTevoked epileptiform potential and the sIS shown in Fig. 3B suggests that the same sequence of events underlies both the activity evoked artificially by elec-
trical stimulation and the spontaneous epileptiform potentials. LOT-evoked discharges in the isolated brain piriform cortex have been analysed in previous studies, in which current source density analysis of field potential laminar profiles was utilized to locate the depth of the extracellular current sinks and sources responsible for sub-components of the epileptiform potential, and to define their temporal correlation with the potentials recorded intracellularly.5,7 These studies demonstrated that the primary bursting activity is associated with a surface-positive, depthnegative potential generated by a current sink located in layer II. Since no synaptic inputs are recognized at such a cortical depth,10 we suggested that primary bursting activity could be due to the activation of a non-synaptic intrinsic regenerative potential.7 The same studies demonstrated that burst afterdepolarization correlated with a surface-negative, depthpositive potential generated by a current sink located
Spontaneous interictal spikes in focal epileptogenesis
Fig. 6. Effect of membrane potential hyperpolarization on the early and late components of the sIS associated with the depth-negative and depth-positive extracellular potentials, respectively (correspondence marked by the vertical dotted lines). QX-314 was added to the intracellular solution to block fast sodium firing. The extracellular potentials recorded simultaneously at resting membrane potential (72 mV) and after membrane hyperpolarization at 110 mV are superimposed in the bottom part of the panel.
in layer Ib, where the corticocortical associative fibres that sustain a recurrent excitatory feedback on the dendrites of layer II and III neurons are
113
located.9,10,18 Since current source density analysis of averaged sISs is technically difficult to achieve, the correlation between extracellular potentials and the intracellular events that make up the sIS was deduced from a comparison with the LOT-evoked potentials. Therefore, we assumed that surface-negative potentials inverting in polarity at 250–400 µm depth represent synaptic associative potentials, whereas surface-positive potentials with a depth reversal at 400–600 µm are due to non-synaptic events generated at the border between layers Ib and II. The possible contribution of a synaptic recurrent excitation in the basal dendrites of layer II neurons cannot be excluded; to verify this possibility, a detailed current source density analysis of sISs is needed. It is already known from studies on hippocampal and neocortical neurons that sISs correlate with a primary burst of fast spikes, superimposed on a regenerative potential31,36,37,43 possibly mediated by a calcium current.42 In our experiments, a large plateau potential similar to that observed in neocortical pyramidal neurons bathed in tetrodotoxin and tetraethylammonium8 is activated at high threshold during the sIS when the sodium conductance is blocked in the presence of a calcium chelator. Such a plateau potential was (i) reversibly abolished by cadmium, a selective blocker of the calcium conductances, (ii) blocked by membrane potential
Fig. 7. The duration of the sIS burst is controlled by an AHP unmasked by membrane potential depolarization. (A) sISs during steady hyperpolarizing and depolarizing current injections are shown (two traces on the left and three traces on the right, respectively). (B) The first (a), third (b) and sixth (c) traces from A reported with a faster time-scale. The AHP is associated with a depth-negative wave in the extracellular traces marked by the asterisk (recorded at 800 µm depth). The pre-sIS small-amplitude potential (arrowheads in the intracellular trace in c) generated an action potential when the membrane potential was depolarized (right trace). Such an action potential precedes the sIS burst, which correlates with the depth-negative extracellular component of the field potential. Resting membrane potential: 81 mV (dotted line).
114
M. de Curtis et al.
Fig. 8. The AHP is abolished by the GABAB receptor antagonists saclofen and phaclofen applied locally together (4 and 10 mM, respectively). (A) Effect of the GABAB receptor antagonists on the late inhibitory postsynaptic potential (asterisk) evoked by LOT stimulation before bicuculline application. The extracellular response in shown in the lower traces. (B) In the same neuron, the post-burst AHP in the bicuculline-induced sIS is reversibly abolished by saclofen and phaclofen. Secondary bursts (and relative extracellular potentials) were induced by the AHP blockade. The membrane potential was depolarized 10 mV to show the inhibitory postsynaptic potential and the AHP in both A and B, respectively. (C) Intracellular diffusion of QX-314 (80 mM) from the recording electrode abolished both the fast sodium spikes and the AHP. The extracellular response was not modified (lower traces). Resting membrane potentials, marked by the dotted lines, were 78 mV (cell in A and B) and 72 mV (cell in C).
hyperpolarization and (iii) associated with the extracellular potential generated in layer II. The exact threshold of activation of this calcium potential could not be determined by analysing the sIS. Recent studies suggest that the involvement of a lowthreshold calcium conductance could be excluded, since layer II neurons lack a T-type low-voltage calcium current.22 The calcium-dependent burst is followed by an afterdepolarization generated by a recurrent synaptic excitation sustained by the primary bursting activity
itself.12,39 As mentioned above, the highly synchronized recurrent synaptic input responsible for the burst afterdepolarization is mediated by corticocortical associative synapses. The extracellular potential associated with the burst after depolarization, indeed, inverts within 400 µm depth, where the associative fibres contact the apical portion of the pyramidal cell dendrites.5,7,9,18 The duration of the sIS is increased when the membrane potential is artificially depolarized to values positive to 50 mV, as demonstrated in control condition and after blockade of the
Spontaneous interictal spikes in focal epileptogenesis
intracellular burst with QX-314, suggesting that the depolarization-activated NMDA subtype of glutamate receptors contributes to the burst afterdepolarization. As for other cortical regions,14,19,40 the burst afterdepolarization was found to be mediated by both ionotropic postsynaptic glutamate receptors, since it was reduced by pharmacological blockade with the specific NMDA receptor antagonist AP-5, and was further diminished by the non-NMDA receptor antagonist CNQX. Prolonged perfusion with CNQX finally induced a complete abolition of sISs. Previous reports demonstrated that prominent GABAB receptor-mediated inhibitory synaptic potentials are activated either during epileptiform activity induced by GABAA receptor blockade32,41 or when sustained epileptic discharges are determined by the kindling procedure.11 Large GABAB receptormediated inhibitory synaptic potentials are able to dampen cortical excitability even in the absence of fast GABAA-mediated inhibition, probably because of the enhanced GABA release mediated by a diffuse increase in excitability that promotes intense firing in inhibitory interneurons.32 Accordingly, in our study large shunting potentials that inverted close to the potassium equilibrium potential were observed after bicuculline injection. Such AHPs were abolished (i) during local application of the GABAB receptor antagonists, saclofen and phaclofen, and (ii) by intracellular application of the lidocaine derivative QX314, a drug that has been demonstrated to block GABAB receptors in CA1 pyramidal neurons25 and in piriform cortex pyramidal cells.17 Fig. 7B showed that the extracellular blockade of GABAB receptors determined secondary bursting activity after the sIS. The prolongation of the sIS and the tendency to transform a single sIS in a short interictal repetitive discharge suggest that GABAB receptors may play a role in the control of the transition from a state of sporadic interictal spiking to a situation in which prolonged ictal discharges are activated.33 Our observations suggest that a single cortical injection of bicuculline induces sISs for several hours after bicuculline washout. This phenomenon could be due to a persistence in the tissue of concentrations of bicuculline lower then the concentration threshold of the method utilized to detect bicuculline in the tissue in a previous study (0.9 µM).7 This explanation is questionable, since we observed that local application of bicuculline at a concentration lower then 200 µM did not change GABAA-mediated responses in our
115
preparation. Alternatively, as discussed elsewhere,7 persistent excitability changes that sustain a spontaneous interictal spiking may occur in the APC as a consequence of the transient afterdischarges observed just after bicuculline injection. Further experiments are necessary to explain this phenomenon. Even though most of the events involved in sIS expression are established, no definite understanding of the mechanisms that initiate an sIS discharge is evinced from the present study. Our results indicate that the sIS is initiated by a population event that occurs in the dendrites of layer II and III neurons, identified as a small-amplitude potential in both intracellular and surface extracellular recordings. The remote dendritic location of the pre-sIS potential is suggested by (i) its depth reversal at the border between layers I and II, (ii) its ability to generate a fast sodium spike upon membrane depolarization (see Fig. 7B, trace c), and (iii) its insensitivity to current injected intracellularly at the soma. The pre-sIS potential (and therefore the sIS) was not abolished by blocking the NMDA receptors in our experimental conditions, even when high concentrations of AP-5 (up to 200 µM) were applied by arterial perfusion, but was eliminated by the nonNMDA receptor antagonist. These results are in agreement with previous reports that demonstrated the role of non-NMDA receptor-mediated excitatory postsynaptic potentials in the generation of bursting activity in CA3 neurons38 and in piriform cortex neurons exposed to low magnesium in slices.13 In our study, sISs were generated in superficial layers at the site of bicuculline injection and persisted for hours, even after the washout of the drug. The leading role of the endopiriform nucleus hypothesized previously in different experimental epileptic conditions2,13 has therefore to be ruled out in the model we utilized. The spontaneous activity generated at the site of bicuculline injection propagates diffusely to cortical and subcortical structures synaptically connected to the site of sIS origin, suggesting that propagation of interictal epileptic activity broadly follows anatomical pathways and does not spread via non-synaptic mechanisms. Acknowledgements—Partial support of this work was provided by the Italian Health Ministry and by the PNRMURST, through Pharmacia-Upjohn, Italy. M.F. was sponsored by the Italian Research Council (C.N.R., grant 9502877-CT14).
REFERENCES
1.
Ajmone Marsan C. (1978) Clinical electrographic correlations of partial seizures. In Modern Perspectives in Epilepsy (ed. Wada J. A.), pp. 76–98. Eden, Montreal. 2. Chatrian G. E., Bergamini L., Dondey M., Klass E. W., Lennox-Buchtal M. and Petersen I. (1974) A glossary of terms most commonly used by clinical electroencephalographers. Electroenceph. clin. Neurophysiol. 37, 538–548. 3. de Curtis M., Pare` D. and Llinas R. (1991) The electrophysiology of the olfactory–hippocampal circuit in the isolated and perfused adult mammalian brain in vitro. Hippocampus 1, 341–354.
116 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.
M. de Curtis et al. de Curtis M., Arcelli P., Debiasi S., Spreafico R. and Avanzini G. (1994) Ultrastructural examination of the isolated guinea pig brain perfused in vitro. Neuroscience 59, 775–788. de Curtis M., Biella G., Forti M. and Panzica F. (1994) Multifocal epileptiform activity induced by restricted convulsing ejections in the piriform cortex of the isolated guinea pig brain. J. Neurophysiol. 71, 2463–2476. de Curtis M., Forti M. and Radici C. (1996) Spontaneous interictal epileptiform potentials in the piriform cortex of the isolated guinea pig brain. Soc. Neurosci. Abstr. 822, 6. Forti M., Biella G., Caccia S. and de Curtis M. (1997) Persistent excitability changes in the piriform cortex of the isolated guinea pig brain after transient exposure to bicuculline. Eur. J. Neurosci. 9, 435–451. Friedman A. and Gutnick M. J. (1989) Intracellular calcium and control of burst generation in neurons of guinea pig neocortex in vitro. Eur. J. Neurosci. 1, 374–381. Haberly L. B. and Price J. L. (1978) Association and commissural fiber system of the olfactory cortex of the rat. J. comp. Neurol. 178, 711–740. Haberly L. B. and Bower J. M. (1989) Olfactory cortex: model circuit for study of associative memory? Trends Neurosci. 12, 258–264. Haberly H. B. and Sutula T. P. (1992) Neuronal processes that underlie expression of kindled epileptiform events in the piriform cortex in vivo. J. Neurosci. 12, 2211–2224. Hablitz J. J. (1984) Picrotoxin-induced epileptiform activity in hippocampus: role of endogenous versus synaptic factors. J. Neurophysiol. 51, 1011–1027. Hoffman W. H. and Haberly L. B. (1993) Role of synaptic excitation in the generation of bursting-induced epileptiform potentials in the endopiriform nucleus and piriform cortex. J. Neurophysiol. 70, 2550–2561. Hwa G. G. C. and Avoli M. (1991) The involvement of excitatory amino acids in neocortical epileptogenesis: NMDA and non-NMDA receptors. Expl Brain Res. 86, 248–256. Jefferys J. G. R. (1990) Basic mechanisms of focal epilepsy. Expl Physiol. 75, 127–162. Kairiss E. W., Racine R. J. and Smith G. K. (1984) The development of the interictal spike during kindling in the rat. Brain Res. 322, 101–110. Kanter E. D., Kapur A. and Haberly L. B. (1996) A dendritic GABAA mediated IPSP regulates facilitation of NMDA-mediated responses to burst stimulation of afferent fibers in piriform cortex. J. Neurosci. 16, 307–312. Ketchum K. L. and Haberly L. B. (1993) Membrane currents evoked by afferent fiber stimulation in the rat piriform cortex. I. Current source density analysis. J. Neurophysiol. 69, 248–260. Lee W. L. and Hablitz J. J. (1991) Initiation of epileptiform activity by excitatory amino acid receptors in the disinhibited rat neocortex. J. Neurophysiol. 65, 87–95. Leung L. S. (1991) Spontaneous hippocampal interictal spikes following local kindling: time-course of change and relation to behavioral seizures. Brain Res. 513, 308–314. Llinas R., Yarom Y. and Sugimori M. (1981) Isolated mammalian brain in vitro: new technique for analysis of electrical activity of neuronal circuit function. Fedn Proc. 40, 2240–2245. Magistretti J. and de Curtis M. (1998) Low and high voltage activated calcium currents in guinea pig piriform cortex neurons. J. Neurophysiol. 79, 808–816. Matsumoto H. and Ajmone-Marsan C. (1964) Cortical cellular phenomena in experimental epilepsy: interictal manifestations. Expl Neurol. 9, 286–304. Muhlethaler M., de Curtis M., Walton K. and Llinas R. (1993) The isolated and perfused brain of the guinea pig in vitro. Eur. J. Neurosci. 5, 915–926. Nathan T., Jensen M. S. and Lambert J. D. (1990) The slow IPSP in rat hippocampal CA1 neurons is blocked by intracellular injection of QX314. Neurosci. Lett. 110, 309–313. Pelletier M. R. and Carlen P. L. (1996) Repeated tetanic stimulation in piriform cortex in vitro: epileptogenesis and pharmacology. J. Neurophysiol. 75, 4069–4078. Piredda S. and Gale K. (1985) A crucial epileptogenic site in the deep pirifrom cortex. Nature 317, 623–625. Prince D. A. (1968) The depolarization shift in epileptic neurons. Expl Neurol. 21, 467–485. Prince D. A. (1971) Cortical cellular activities during cyclically occurring inter-ictal epileptiform discharges. Electroenceph. clin. Neurophysiol. 31, 469–484. Racine R. J., Mosher M. and Kairiss E. W. (1988) The role of pyriform cortex in the generation of interictal spikes in the kindled preparation. Brain Res. 454, 251–263. Rutecki P. A., Lebeda F. J. and Johnston D. (1985) Epileptiform activity induced by changes in extracellular potassium in hippocampus. J. Neurophysiol. 54, 1363–1374. Scanziani M., Ga¨hwiler B. H. and Thompson S. M. (1991) Paroxysmal inhibitory potentials mediated by GABAB receptors in partially disinhibited rat hippocampal slice cultures. J. Physiol. 444, 375–396. Scanziani M., Debanne D., Muller M., Ga¨hwiler B. H. and Thompson S. M. (1994) Role of excitatory amino acid and GABAB receptors in the generation of epileptiform activity in disinhibited hippocampal slices cultures. Neuroscience 61, 823–832. Schwartzkroin P. A. and Wyler A. R. (1980) Mechanisms underlying epileptiform burst discharge. Ann. Neurol. 7, 95–107. Sherwin I. (1984) Ictal–interictal unit firing pattern differences in penicillin-induced primary and secondary epileptogenic foci. Expl Neurol. 84, 463–477. Silva-Barrat C., Araneda S., Menini Ch., Champagnat J. and Naquet R. (1992) Burst generation in neocortical neurons after GABA withdrawal in the rat. J. Neurophysiol. 67, 715–727. Traub R. D. and Llinas R. (1979) Hippocampal pyramidal cells: significance of dendritic conductances for neuronal function and epileptogenesis. J. Neurophysiol. 42, 476–496. Traub R. D. and Dingledine R. (1990) Model of synchronized epileptiform bursts induced by high potassium in CA3 region of rat hippocampal slice. Role of spontaneous EPSPs in initiation. J. Neurophysiol. 64, 1009–1018. Traub R. D., Miles R. and Jefferys J. G. R. (1993) Synaptic and intrinsic conductances shape picrotoxin-induced synchronized afterdischarges in the guinea pig hippocampal slice. J. Physiol. 461, 525–547. Traub R. D., Jefferys J. G. R. and Whittington M. A. (1994) Enhanced NMDA conductance can account for epileptiform activity induced by low Mg2+ in the rat hippocampal slices. J. Physiol. 478, 379–393.
Spontaneous interictal spikes in focal epileptogenesis 41.
117
Witte O. (1994) Afterpotentials of penicillin-induced epileptiform neuronal discharges in the motor cortex of the rat in vivo. Epil. Res. 18, 43–55. 42. Wong R. K. S. and Prince D. A. (1978) Participation of calcium spike during intrinsic burst firing in hippocampal neurons. Brain Res. 159, 365–390. 43. Wong R. K. S. and Prince D. A. (1979) Dendritic mechanisms underlying penicillin induced epileptiform activity. Science 204, 1228–1231. (Accepted 1 April 1998)
PII:
Neuroscience Vol. 88, No. 1, pp. 107–117, 1999 Copyright 1998 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/99 $19.00+0.00 S0306-4522(98)00201-2
CELLULAR MECHANISMS UNDERLYING SPONTANEOUS INTERICTAL SPIKES IN AN ACUTE MODEL OF FOCAL CORTICAL EPILEPTOGENESIS M. de CURTIS,* C. RADICI and M. FORTI Department of Experimental Neurophysiology, Istituto Nazionale Neurologico, via Celoria 11, 20133 Milan, Italy Abstract––The cellular mechanisms involved in the generation of spontaneous epileptiform potentials were investigated in the pirifom cortex of the in vitro isolated guinea-pig brain. A single, unilateral injection of bicuculline (150–200 nmol) in the anterior piriform cortex induced locally spontaneous interictal spikes that recurred with a period of 8.814.47 s and propagated caudally to the ipsi- and contralateral hemispheres. Simultaneous extra- and intracellular recordings from layer II and III principal cells showed that the spontaneous interictal spike correlates to a burst of action potentials followed by a large afterdepolarization. Intracellular application of the sodium conductance blocker, QX-314 (80 mM), abolished bursting activity and unmasked a high-threshold slow spike enhanced by the calcium chelator EGTA (50 mM). The slow spike was abolished by membrane hyperpolarization and by local perfusion with 2 mM cadmium. The depolarizing potential that followed the primary burst was reduced by arterial perfusion with the N-methyl--aspartate receptor antagonist, -2-amino-5-phosphonopentanoic acid (100–200 µM). The non-N-methyl--aspartate glutamate receptor antagonist, 6-cyano-7-nitroquinoxaline2,3-dione (20 µM), completely and reversibly blocked the spontaneous spikes. The interictal spikes were terminated by a large afterpotential blocked either by intracellular QX-314 (80 mM) or by extracellular application of phaclofen and 2-hydroxysaclofen (10 and 4 mM, respectively). The present study demonstrates that, in an acute model of epileptogenesis, spontaneous interictal spikes are fostered by a primary burst of fast action potentials that ride on a regenerative high-threshold, possibly calcium-mediated spike, which activates a recurrent, glutamate-mediated potential responsible for the entrainment of adjacent and remote cortical regions. The bursting activity is controlled by a GABAB receptor-mediated inhibitory synaptic potential. 1998 IBRO. Published by Elsevier Science Ltd. Key words: epileptogenesis, interictal spikes, intracellular recordings, isolated brain preparation, piriform cortex.
Spontaneous interictal spikes (sISs) represent the electrographic mark of an active epileptic focus in several epileptic human conditions and can be reproduced in experimental models of focal epilepsy.1,15 Interictal spikes are fast transients clearly distinguished from the baseline, followed by a slow potential of variable amplitude and duration.2 Experimental studies demonstrated that sISs occur synchronously in a large population of cortical neurons and correlate at the single cell level to a depolarization of the membrane potential, designated as paroxysmal depolarization shift.23,28,34 In vitro studies on stimulus-evoked epileptiform discharges acutely established by various means in different cortical *To whom correspondence should be addressed. Abbreviations: AHP, afterhyperpolarizing potential; AP-5, -2-amino-5-phosphonopentanoic acid; APC, anterior piriform cortex; CNQX, 6-cyano-7-nitroquinoxaline2,3-dione; EGTA, ethyleneglycolbis(aminoethyl ether) tetra-acetate; HEPES, N-2-hydroxyethylpiperazine-N -2ethanesulphonic acid; LOT, lateral olfactory tract; NMDA, N-methyl--aspartate; sIS, spontaneous interictal spike. 107
regions demonstrated that the paroxysmal depolarization shift is initiated by a primary voltagedependent intrinsic burst of action potentials, which gives rise to a large excitatory postsynaptic potential mediated by the recurrent activation of local glutamatergic synapses.12,13,15,31,39 Although it has never been demonstrated directly, such a sequence of events is expected to occur during the generation of spontaneous interictal potentials. The piriform cortex has been shown to be involved in limbic epileptogenesis.5,11,13,16,26,27,30 sISs in the piriform cortex were observed early during the development of kindling epileptogenesis11,16 and were found to propagate from the piriform lobe to the hippocampus, where the ability to generate epileptic potentials developed later, when the kindling process is completed.30 We demonstrated that a single, brief intracerebral injection of bicuculline in the piriform cortex of the isolated guinea-pig brain activates persistent spontaneous interictal potentials that show the characteristics of sISs.5,7 The same model of acute focal epileptogenesis is utilized here to characterize the sequence of intrinsic and synaptic events
108
M. de Curtis et al.
Fig. 1. sISs and subthreshold spontaneous potentials induced by a single 10-s application of a 2-mM bicuculline solution in the APC. Immediately after bicuculline, the frequency of repetition of sIS was lower (middle trace), and increased within 10 min (bottom trace recorded 2 h after bicuculline ejection). As illustrated in the upper control trace, sporadic, low-amplitude potentials were observed before bicuculline application in this experiment. The recordings were performed at 800 µm depth.
responsible for the generation and for the termination of sISs. The utilization of the isolated guineapig brain preparation allows (i) study of the sISs in an intact brain, in the absence of any mechanical alteration of the cerebral tissue, (ii) correlation of events recorded intracellularly during the sIS discharge with simultaneous extracellular responses obtained at various cortical depths/locations in the piriform cortex, and (iii) pharmacological manipulations of sISs. Preliminary results have been published in abstract form.6 EXPERIMENTAL PROCEDURES
Young adult Hartley guinea-pigs (150–250 g; Charles River, Italy) anaesthetized with penthotal sodium (20 mg/ kg) were perfused for 2 min through the aorta with an oxygenated, slightly acidic, cold saline solution (10C, pH 7.1). After craniotomy performed in hypothermia, the brain was removed and transferred to a perfusion chamber, according to the standard procedure.3,4,21,24 All efforts were made to minimize animal suffering and to reduce the number of animals used for the study. The isolated brain was perfused through the basilar artery with an artificial plasma solution (composition, mM: NaCl 126, KCl 2.3, NaH2CO3 26, MgSO4 1.3, CaCl2 2.4, KH2PO4 1.2, glucose 15, HEPES 5, thiourea 0.4 and 3% dextran 70000) oxygenated with a 5% CO2/95% O2 gas mixture (pH 7.3). The perfusion rate was 5.5–6 ml/min. The temperature in the perfusion chamber was gradually raised to 32C after completing the isolation manoeuvre. Bicuculline methiodide (2 mM; RBI) dissolved in Ringer’s solution was injected for 10 s at 800 µm depth in the anterior piriform cortex (APC) through an 8- to 10-µm tip diameter glass pipette connected to a graduated syringe. The pipette was removed after the bicuculline ejection. As described in previous studies,5,7 the total volume ejected was c. 150–200 nl (150–200 nmol of bicuculline). The effect of bicuculline was followed by studying the lateral olfactory tract (LOT)-evoked field responses. Bicuculline was washed out within 60–90 min.7 A bipolar stimulating silver-wire electrode positioned on the LOT was utilized to evoke responses in the APC. Extracellular potentials were recorded with glass micropipettes filled with 0.9% NaCl (3–10 MÙ resistance). Intracellular recordings were performed by means of micropipettes filled with either 3 M or 2 M potassium acetate and 1% biocytine (40–80 MÙ resistance) from principal neurons in layers II and III at 300–700 µm depth. Extra- and intracellular activity was recorded with a Biomedical
Engeneering Amplifier (Thornwood, NY, U.S.A.) and with a Neurodata Amplifier (New York, NY, U.S.A.). Signals were stored via a DTR 2602 Digital Tape Recorder (Biologic, France) for off-line analysis with a Microvax 3400 computer system (Digital, Italy) and a 486 PC. The N-methyl--aspartate (NMDA) and non-NMDA receptor blockers -2-amino-5-phosphonopentanoic acid (AP-5; 100–200 µM; RBI) and 6-cyano-7-nitroquinoxaline2,3-dione (CNQX; 20 µM; RBI) were applied by arterial perfusion. QX-314 (80 mM; RBI) and EGTA (50 mM; Sigma) were dissolved in 1 M potassium acetate and delivered intracellularly through the recording pipette (pH 7.2). Phaclofen (10 mM; Sigma), 5-hydroxysaclofen (4 mM; RBI) and cadmium (Cd2+, 2 mM; Sigma) diluted in standard solution without dextran were locally injected in the tissue, close to the intracellular recording electrode. The effect of the locally applied drugs was rapid (peak activity within 1 min) and completely reversible. When micropipettes filled with biocytin were used for intracellular staining, the isolated brains were fixed overnight with 4% paraformaldehyde. Coronal sections (100 µm thick) were then processed for avidin–horseradish peroxidase visualization and were counterstained with Neutral Red to locate the cortical depth of the cells.7 RESULTS
In all experiments (n=28), sISs were activated by a single 10-s injection of bicuculline and persisted for 4–6 h after bicuculline washout.7 In 21 of 28 tests, bicuculline induced transient activation of spontaneous afterdischarges.7 sISs recurred with a period of 12.763.32 s immediately after bicuculline injection (Fig. 1, middle trace) and gradually stabilized within 10 min to one event every 8.814.47 s (Fig. 1, lower trace). In 28% of the experiments, smallamplitude potentials subthreshold for spike generation were observed in isolation after bicuculline. Non-periodic, sporadic, low-amplitude potentials, subthreshold for population spike generation, were detected in 11% of the experiments before bicuculline was applied (as shown in the upper trace in Fig. 1). In order to identify the site of generation of sISs, simultaneous extracellular recordings were carried out in different regions in the olfactory lobes, ispiand contralateral to the site of bicuculline injection (Fig. 2; n=5). Recordings were performed with the
Spontaneous interictal spikes in focal epileptogenesis
109
Fig. 2. Characteristics of propagation of the sIS in the olfactory and limbic cortices. The sIS recorded at the site of bicuculline injection in the APC is marked by an asterisk (site 2). Recordings were performed in both hemispheres, ipsi- and contralateral to the bicuculline injection site. See text for specifications of the position of the recording electrodes, indicated here on a picture of the ventral view of the isolated brain.
micropipette utilized to inject bicuculline as reference electrode (trace 2 in Fig. 2, marked by an asterisk), moving a second electrode to record simultaneous activity in the olfactory tubercle (site 12), the olfactory bulb (site 10), the LOT (site 11), the APC rostral and caudal to the drug application site (sites 1 and 3), the endopiriform nucleus (site 4), the posterior piriform cortex (sites 5 and 7), the periamygdaloid cortex (sites 6 and 8), the lateral entorhinal cortex (site 9) and the contralateral APC, posterior piriform cortex and entorhinal cortex (sites 13, 14 and 15, respectively). The sIS features were highly reproducible and the delay of the potentials recorded at different sites was correlated in time with the sIS simultaneously recorded with electrode 2. The results demonstrate that sISs originate at the site of bicuculline application in the APC. From here, they propagate throughout the olfactory cortices and into the entorhinal cortex of both hemispheres. Simultaneous extra- and intracellular recordings were performed in the APC at the site of bicuculline injection. Intracellular recordings were performed from 71 APC neurons recorded in 28 isolated guineapig brains. Twenty of 71 neurons were injected with biocytin and were identified morphologically as pyramidal cells (not shown). Twenty-two cells were recorded before, during and after bicuculline injection. Forty-one of 71 neurons were recorded 1 h after the injection of bicuculline. Averaged resting membrane potential and input resistance were 74.184.65 mV (n=55) and 35.535.84 MÙ (n= 24). As reported previously,7 neurons in layers II and III showed similar membrane properties. No differences were observed in sISs recorded intracellularly from layer II and III neurons. The similarity between
the LOT-evoked epileptiform potential and the sIS is illustrated in Fig. 3. The onset of the extracellular population spike in the sIS correlated to a burst of high-frequency action potentials (Fig. 3B). The spontaneous burst/population spike was preceded by a small-amplitude potential (arrowhead in Fig. 3A and B and the following figures). The intracellular burst was followed by a large after-depolarizing potential, which correlated with a surface-negative/depthpositive extracellular potential (asterisk in Fig. 3B). The depth location of the sinks responsible for the components of the LOT-evoked epileptiform potential have been described in detail previously.5,7 Surface-negative/depth-positive potentials correspond to synaptic events located in the superficial layer Ib generated by corticocortical associative synapses. Depth-negative/surface-positive potentials are due to non-synaptic sinks (intrinsic bursting activity) located in the superficial portion of layer II. The burst discharge in the sIS is generated by fast sodium action potentials superimposed on a slow regenerative spike, probably mediated by calcium. The intracellular diffusion through the recording electrode of the sodium channel blocker QX-314 (80 mM) progressively abolished the sodium spikes (n=15) and revealed a large potential enhanced by the co-application through the recording pipette of the calcium chelator EGTA (50 mM), as illustrated in Fig. 4A (from left to right: sISs at cell impalement and after 3 and 8 min of intracellular recording). The calcium dependence of this plateau potential was demonstrated when the inorganic antagonist of voltage-gated calcium conductances, cadmium (2 mM; n=3), was briefly perfused in the extracellular space for 10 s through a 10-µm tip pipette adjacent to
110
M. de Curtis et al.
Fig. 3. Comparison between LOT-evoked epileptiform potential and sIS recorded simultaneously intraand extracellularly 40 min after bicuculline injection. (A) Slow time scale of two sISs, intercalated by an LOT-evoked response (a). (B) Expanded LOT-evoked potential (a) and sIS (b) are shown. Simultaneous extracellular recordings at 100, 500 and 850µm depth from the cortical surface (top, middle and bottom traces in the lower part of B, respectively) close to the intracellular recording (upper traces) are illustrated. Stimulus artifact is marked by an arrow. The arrowheads indicate the small-amplitude potential that precedes the burst discharge. Such a potential can be detected in both the intra- and extracellular recordings. Resting membrane potential: 78 mV.
the intracellular recording electrode. Cadmium abolished the early component of the sIS associated with the depth-negative extracellular potential. The calcium plateau potential was restored when the drug was washed out (Fig. 4C, D). The transient abolition of the calcium plateau isolated a depolarizing potential temporally correlated to the depth-positive synaptic potential defined previously as post-burst after depolarization. The field potential was reduced slightly during the application of cadmium. The amplitude and duration of the post-burst after depolarization and the associated extracellular potential were reduced by arterial perfusion with the NMDA receptor antagonist AP-5 (100–200 µM; n=6). The AP-5-sensitive component of the burst after depolarization was enhanced when the membrane potential was depolarized to increase the NMDA receptor-mediated response (insets in Fig. 5A). sIS periodicity was slowed to 6.482.21 s during AP-5 perfusion. Arterial perfusion with the nonNMDA receptor antagonist CNQX (20 µM) in addition to AP-5 completely abolished the sISs (n=4). The sIS was preceded by a small-amplitude
potential correlated with a tiny deflection in the extracellular response, which was easily detected in surface recordings (Fig. 5A, arrowhead in the right panel; also see Figs 3 and 6). The pre-sIS potential showed a depth reversal around 400µm depth (not shown); it was preserved during perfusion with 100– 200 µM AP-5 solutions, but was abolished by 20 µM CNQX. The post-burst afterdepolarizing potential could be discriminated as a deflection after the early slow spike when the membrane potential was artificially depolarized during sodium conductance blockade with intracellular QX-314 (arrows in Fig. 5B and C). Such a late potential and its extracellular correlate were reduced during the early phase of a perfusion with 20 µM CNQX and 100 µM AP-5 (Fig. 5B, right traces, and C; n=4), before the complete blockade of the sIS occurred. The reduction of the glutamate receptor-mediated synaptic event isolated a slow potential correlated with the extracellular population spike (Fig. 5C). The results described above suggest that the early component of the sIS (i) can be isolated by glutamate
Spontaneous interictal spikes in focal epileptogenesis
111
Fig. 4. Fast sodium conductance blocker unmasked a plateau potential that subtends the sIS. (A) Gradual effect of intracellular diffusion of QX-314 (80 mM) and EGTA (50 mM). Traces were recorded 30 s, 3 and 8 min after cell impalement (from left to right). (B) The first and third traces shown in A are superimposed. The simultaneous extracellular recording recorded at 800 µm depth was not modified by the intracellular QX-314 and EGTA diffusion. Note that the plateau potential onset correlates to the depth-negative extracellular potential. (C, D) Local microperfusion with cadmium (Cd2+, 2 µM) reversibly abolished the early shoulder in the intracellular potential. Washout was complete 10 min after cadmium ejection. In D, three pairs of traces from another experiment show the correlation between the intra- and extracellular potentials before, during perfusion and after cadmium washout. Note that cadmium perfusion reduced the amplitude of the field response. Resting membrane potential was 77 mV for A and B, 74 mV for C and 78 mV for D.
receptor blockade and (ii) is associated with a cadmium-sensitive potential. To further differentiate this potential from the late excitatory synaptic component of the sIS, large hyperpolarizing currents were injected intracellularly during the fast sodium spike blockade by QX-314 (n=4). As illustrated in Fig. 6, membrane hyperpolarization to values negative to 110 mV (resting membrane potential marked by the horizontal dotted line: 72 mV) abolished the early component of the sIS coupled with the depth-negative population spike and, as is predictable for an excitatory synaptic potential, increased the amplitude of the recurrent burst after depolarization correlated with the depth-positive potential. The burst/afterdepolarization was consistently followed by a late potential associated with an extracellular surface-positive/depth-negative deflection (asterisk in Fig. 7B). The late potential was identified as an afterhyperpolarizing potential (AHP) when the membrane potential was depolarized by a steady intracellular current injection (Fig. 7A and B, right trace). The AHP lasted 507173.33 ms (n=16), inverted close to the potassium equilibrium potential (79.45.5 mV; n=13; Fig. 7A) and was reversibly abolished by saclofen (4 mM) and phaclofen
(10 mM) applied together in the superficial cortical layers, close to the intracellular recording electrode (Fig. 8B; n=4). In control conditions, before bicuculline injection, saclofen and phaclofen completely abolished the late inhibitory postsynaptic potential mediated by GABAB receptors (Fig. 8A). Either one of the two compounds alone showed an incomplete blocking effect. The blockade of the AHP prolonged the sIS and induced secondary bursting activity (middle traces in Fig. 8B). The late AHP was similarly abolished by the lidocaine derivative QX-314 (80 mM) applied intracellularly through the recording micropipette (Fig. 8C; n=12). DISCUSSION
The sISs described in the present study show similarities with those observed in focal human epilepsy,1 in chronic animal models of focal epilepsy11,20,30 and in acute epileptogenic conditions.13,26,29,35 The sequence of events that sustain spontaneous epileptiform interictal spikes in the piriform cortex of the isolated guinea-pig brain can elucidate the mechanisms that control sIS generation in other cortical areas and, in general, might help to understand interictal focal epileptogenesis. Four
112
M. de Curtis et al.
Fig. 5. (A) The duration of the sIS burst (upper traces) and the corresponding depth-positive extracellular potential (asterisk in lower traces) are reduced by arterial perfusion with AP-5 (150 µM). The AP-5 effect is more evident when the late burst afterdepolarization is enhanced by membrane potential depolarization (inset; membrane potential depolarized to 48 and 50 mV in control conditions and after AP-5, respectively). The NMDA receptor antagonist did not affect the small-amplitude potential that precedes the sIS. Resting membrane potential: 81 mV. The extracellular potential was recorded at 850 µm depth. (B) sISs during depolarizations of the membrane potential induced by a steady intracellular current injection. QX-314 (80 mM) and EGTA (50 mM) were applied intracellularly through the recording electrode. A late depolarizing potential (arrows in the left panel) was activated in the sIS when the membrane potential was depolarized. On the right, the effect of the early phase of an arterial perfusion with 20 µM CNQX and 100 µM AP-5 is displayed. The traces were acquired within 2 min after perfusion onset, before the complete abolition of the sIS occurred (after 5 min perfusion). (C) Traces from B obtained before (arrows) and after glutamate receptor antagonist perfusion superimposed. The extracellular responses recorded simultaneously at 850 µm depth (lower traces in each panel) show that the depth-positive potential is gradually reduced during blockade of glutamate receptors. This blockade enhanced the amplitude of the population spike component of the sIS. Resting membrane potential (dotted line) was 71 mV.
major original findings are described here: (i) sISs generated at the site of bicuculline injection in the piriform cortex propagate to the limbic cortices bilaterally along anatomical pathways; (ii) the sISs are initiated by a synaptic small-amplitude potential mediated by non-NMDA-type glutamate receptors; (iii) the sISs are associated with a primary burst, possibly sustained by a calcium-dependent regenerative potential; (iv) the sIS duration is terminated by an AHP mediated primarily by a GABAB receptormediated inhibitory postsynaptic potential. We also demonstrate that, as for other cortical areas, the component of the sIS that follows the primary burst is mediated by the recurrent activation of associative, glutamatergic synapses. The similarity between the late part of the LOTevoked epileptiform potential and the sIS shown in Fig. 3B suggests that the same sequence of events underlies both the activity evoked artificially by elec-
trical stimulation and the spontaneous epileptiform potentials. LOT-evoked discharges in the isolated brain piriform cortex have been analysed in previous studies, in which current source density analysis of field potential laminar profiles was utilized to locate the depth of the extracellular current sinks and sources responsible for sub-components of the epileptiform potential, and to define their temporal correlation with the potentials recorded intracellularly.5,7 These studies demonstrated that the primary bursting activity is associated with a surface-positive, depthnegative potential generated by a current sink located in layer II. Since no synaptic inputs are recognized at such a cortical depth,10 we suggested that primary bursting activity could be due to the activation of a non-synaptic intrinsic regenerative potential.7 The same studies demonstrated that burst afterdepolarization correlated with a surface-negative, depthpositive potential generated by a current sink located
Spontaneous interictal spikes in focal epileptogenesis
Fig. 6. Effect of membrane potential hyperpolarization on the early and late components of the sIS associated with the depth-negative and depth-positive extracellular potentials, respectively (correspondence marked by the vertical dotted lines). QX-314 was added to the intracellular solution to block fast sodium firing. The extracellular potentials recorded simultaneously at resting membrane potential (72 mV) and after membrane hyperpolarization at 110 mV are superimposed in the bottom part of the panel.
in layer Ib, where the corticocortical associative fibres that sustain a recurrent excitatory feedback on the dendrites of layer II and III neurons are
113
located.9,10,18 Since current source density analysis of averaged sISs is technically difficult to achieve, the correlation between extracellular potentials and the intracellular events that make up the sIS was deduced from a comparison with the LOT-evoked potentials. Therefore, we assumed that surface-negative potentials inverting in polarity at 250–400 µm depth represent synaptic associative potentials, whereas surface-positive potentials with a depth reversal at 400–600 µm are due to non-synaptic events generated at the border between layers Ib and II. The possible contribution of a synaptic recurrent excitation in the basal dendrites of layer II neurons cannot be excluded; to verify this possibility, a detailed current source density analysis of sISs is needed. It is already known from studies on hippocampal and neocortical neurons that sISs correlate with a primary burst of fast spikes, superimposed on a regenerative potential31,36,37,43 possibly mediated by a calcium current.42 In our experiments, a large plateau potential similar to that observed in neocortical pyramidal neurons bathed in tetrodotoxin and tetraethylammonium8 is activated at high threshold during the sIS when the sodium conductance is blocked in the presence of a calcium chelator. Such a plateau potential was (i) reversibly abolished by cadmium, a selective blocker of the calcium conductances, (ii) blocked by membrane potential
Fig. 7. The duration of the sIS burst is controlled by an AHP unmasked by membrane potential depolarization. (A) sISs during steady hyperpolarizing and depolarizing current injections are shown (two traces on the left and three traces on the right, respectively). (B) The first (a), third (b) and sixth (c) traces from A reported with a faster time-scale. The AHP is associated with a depth-negative wave in the extracellular traces marked by the asterisk (recorded at 800 µm depth). The pre-sIS small-amplitude potential (arrowheads in the intracellular trace in c) generated an action potential when the membrane potential was depolarized (right trace). Such an action potential precedes the sIS burst, which correlates with the depth-negative extracellular component of the field potential. Resting membrane potential: 81 mV (dotted line).
114
M. de Curtis et al.
Fig. 8. The AHP is abolished by the GABAB receptor antagonists saclofen and phaclofen applied locally together (4 and 10 mM, respectively). (A) Effect of the GABAB receptor antagonists on the late inhibitory postsynaptic potential (asterisk) evoked by LOT stimulation before bicuculline application. The extracellular response in shown in the lower traces. (B) In the same neuron, the post-burst AHP in the bicuculline-induced sIS is reversibly abolished by saclofen and phaclofen. Secondary bursts (and relative extracellular potentials) were induced by the AHP blockade. The membrane potential was depolarized 10 mV to show the inhibitory postsynaptic potential and the AHP in both A and B, respectively. (C) Intracellular diffusion of QX-314 (80 mM) from the recording electrode abolished both the fast sodium spikes and the AHP. The extracellular response was not modified (lower traces). Resting membrane potentials, marked by the dotted lines, were 78 mV (cell in A and B) and 72 mV (cell in C).
hyperpolarization and (iii) associated with the extracellular potential generated in layer II. The exact threshold of activation of this calcium potential could not be determined by analysing the sIS. Recent studies suggest that the involvement of a lowthreshold calcium conductance could be excluded, since layer II neurons lack a T-type low-voltage calcium current.22 The calcium-dependent burst is followed by an afterdepolarization generated by a recurrent synaptic excitation sustained by the primary bursting activity
itself.12,39 As mentioned above, the highly synchronized recurrent synaptic input responsible for the burst afterdepolarization is mediated by corticocortical associative synapses. The extracellular potential associated with the burst after depolarization, indeed, inverts within 400 µm depth, where the associative fibres contact the apical portion of the pyramidal cell dendrites.5,7,9,18 The duration of the sIS is increased when the membrane potential is artificially depolarized to values positive to 50 mV, as demonstrated in control condition and after blockade of the
Spontaneous interictal spikes in focal epileptogenesis
intracellular burst with QX-314, suggesting that the depolarization-activated NMDA subtype of glutamate receptors contributes to the burst afterdepolarization. As for other cortical regions,14,19,40 the burst afterdepolarization was found to be mediated by both ionotropic postsynaptic glutamate receptors, since it was reduced by pharmacological blockade with the specific NMDA receptor antagonist AP-5, and was further diminished by the non-NMDA receptor antagonist CNQX. Prolonged perfusion with CNQX finally induced a complete abolition of sISs. Previous reports demonstrated that prominent GABAB receptor-mediated inhibitory synaptic potentials are activated either during epileptiform activity induced by GABAA receptor blockade32,41 or when sustained epileptic discharges are determined by the kindling procedure.11 Large GABAB receptormediated inhibitory synaptic potentials are able to dampen cortical excitability even in the absence of fast GABAA-mediated inhibition, probably because of the enhanced GABA release mediated by a diffuse increase in excitability that promotes intense firing in inhibitory interneurons.32 Accordingly, in our study large shunting potentials that inverted close to the potassium equilibrium potential were observed after bicuculline injection. Such AHPs were abolished (i) during local application of the GABAB receptor antagonists, saclofen and phaclofen, and (ii) by intracellular application of the lidocaine derivative QX314, a drug that has been demonstrated to block GABAB receptors in CA1 pyramidal neurons25 and in piriform cortex pyramidal cells.17 Fig. 7B showed that the extracellular blockade of GABAB receptors determined secondary bursting activity after the sIS. The prolongation of the sIS and the tendency to transform a single sIS in a short interictal repetitive discharge suggest that GABAB receptors may play a role in the control of the transition from a state of sporadic interictal spiking to a situation in which prolonged ictal discharges are activated.33 Our observations suggest that a single cortical injection of bicuculline induces sISs for several hours after bicuculline washout. This phenomenon could be due to a persistence in the tissue of concentrations of bicuculline lower then the concentration threshold of the method utilized to detect bicuculline in the tissue in a previous study (0.9 µM).7 This explanation is questionable, since we observed that local application of bicuculline at a concentration lower then 200 µM did not change GABAA-mediated responses in our
115
preparation. Alternatively, as discussed elsewhere,7 persistent excitability changes that sustain a spontaneous interictal spiking may occur in the APC as a consequence of the transient afterdischarges observed just after bicuculline injection. Further experiments are necessary to explain this phenomenon. Even though most of the events involved in sIS expression are established, no definite understanding of the mechanisms that initiate an sIS discharge is evinced from the present study. Our results indicate that the sIS is initiated by a population event that occurs in the dendrites of layer II and III neurons, identified as a small-amplitude potential in both intracellular and surface extracellular recordings. The remote dendritic location of the pre-sIS potential is suggested by (i) its depth reversal at the border between layers I and II, (ii) its ability to generate a fast sodium spike upon membrane depolarization (see Fig. 7B, trace c), and (iii) its insensitivity to current injected intracellularly at the soma. The pre-sIS potential (and therefore the sIS) was not abolished by blocking the NMDA receptors in our experimental conditions, even when high concentrations of AP-5 (up to 200 µM) were applied by arterial perfusion, but was eliminated by the nonNMDA receptor antagonist. These results are in agreement with previous reports that demonstrated the role of non-NMDA receptor-mediated excitatory postsynaptic potentials in the generation of bursting activity in CA3 neurons38 and in piriform cortex neurons exposed to low magnesium in slices.13 In our study, sISs were generated in superficial layers at the site of bicuculline injection and persisted for hours, even after the washout of the drug. The leading role of the endopiriform nucleus hypothesized previously in different experimental epileptic conditions2,13 has therefore to be ruled out in the model we utilized. The spontaneous activity generated at the site of bicuculline injection propagates diffusely to cortical and subcortical structures synaptically connected to the site of sIS origin, suggesting that propagation of interictal epileptic activity broadly follows anatomical pathways and does not spread via non-synaptic mechanisms. Acknowledgements—Partial support of this work was provided by the Italian Health Ministry and by the PNRMURST, through Pharmacia-Upjohn, Italy. M.F. was sponsored by the Italian Research Council (C.N.R., grant 9502877-CT14).
REFERENCES
1.
Ajmone Marsan C. (1978) Clinical electrographic correlations of partial seizures. In Modern Perspectives in Epilepsy (ed. Wada J. A.), pp. 76–98. Eden, Montreal. 2. Chatrian G. E., Bergamini L., Dondey M., Klass E. W., Lennox-Buchtal M. and Petersen I. (1974) A glossary of terms most commonly used by clinical electroencephalographers. Electroenceph. clin. Neurophysiol. 37, 538–548. 3. de Curtis M., Pare` D. and Llinas R. (1991) The electrophysiology of the olfactory–hippocampal circuit in the isolated and perfused adult mammalian brain in vitro. Hippocampus 1, 341–354.
116 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.
M. de Curtis et al. de Curtis M., Arcelli P., Debiasi S., Spreafico R. and Avanzini G. (1994) Ultrastructural examination of the isolated guinea pig brain perfused in vitro. Neuroscience 59, 775–788. de Curtis M., Biella G., Forti M. and Panzica F. (1994) Multifocal epileptiform activity induced by restricted convulsing ejections in the piriform cortex of the isolated guinea pig brain. J. Neurophysiol. 71, 2463–2476. de Curtis M., Forti M. and Radici C. (1996) Spontaneous interictal epileptiform potentials in the piriform cortex of the isolated guinea pig brain. Soc. Neurosci. Abstr. 822, 6. Forti M., Biella G., Caccia S. and de Curtis M. (1997) Persistent excitability changes in the piriform cortex of the isolated guinea pig brain after transient exposure to bicuculline. Eur. J. Neurosci. 9, 435–451. Friedman A. and Gutnick M. J. (1989) Intracellular calcium and control of burst generation in neurons of guinea pig neocortex in vitro. Eur. J. Neurosci. 1, 374–381. Haberly L. B. and Price J. L. (1978) Association and commissural fiber system of the olfactory cortex of the rat. J. comp. Neurol. 178, 711–740. Haberly L. B. and Bower J. M. (1989) Olfactory cortex: model circuit for study of associative memory? Trends Neurosci. 12, 258–264. Haberly H. B. and Sutula T. P. (1992) Neuronal processes that underlie expression of kindled epileptiform events in the piriform cortex in vivo. J. Neurosci. 12, 2211–2224. Hablitz J. J. (1984) Picrotoxin-induced epileptiform activity in hippocampus: role of endogenous versus synaptic factors. J. Neurophysiol. 51, 1011–1027. Hoffman W. H. and Haberly L. B. (1993) Role of synaptic excitation in the generation of bursting-induced epileptiform potentials in the endopiriform nucleus and piriform cortex. J. Neurophysiol. 70, 2550–2561. Hwa G. G. C. and Avoli M. (1991) The involvement of excitatory amino acids in neocortical epileptogenesis: NMDA and non-NMDA receptors. Expl Brain Res. 86, 248–256. Jefferys J. G. R. (1990) Basic mechanisms of focal epilepsy. Expl Physiol. 75, 127–162. Kairiss E. W., Racine R. J. and Smith G. K. (1984) The development of the interictal spike during kindling in the rat. Brain Res. 322, 101–110. Kanter E. D., Kapur A. and Haberly L. B. (1996) A dendritic GABAA mediated IPSP regulates facilitation of NMDA-mediated responses to burst stimulation of afferent fibers in piriform cortex. J. Neurosci. 16, 307–312. Ketchum K. L. and Haberly L. B. (1993) Membrane currents evoked by afferent fiber stimulation in the rat piriform cortex. I. Current source density analysis. J. Neurophysiol. 69, 248–260. Lee W. L. and Hablitz J. J. (1991) Initiation of epileptiform activity by excitatory amino acid receptors in the disinhibited rat neocortex. J. Neurophysiol. 65, 87–95. Leung L. S. (1991) Spontaneous hippocampal interictal spikes following local kindling: time-course of change and relation to behavioral seizures. Brain Res. 513, 308–314. Llinas R., Yarom Y. and Sugimori M. (1981) Isolated mammalian brain in vitro: new technique for analysis of electrical activity of neuronal circuit function. Fedn Proc. 40, 2240–2245. Magistretti J. and de Curtis M. (1998) Low and high voltage activated calcium currents in guinea pig piriform cortex neurons. J. Neurophysiol. 79, 808–816. Matsumoto H. and Ajmone-Marsan C. (1964) Cortical cellular phenomena in experimental epilepsy: interictal manifestations. Expl Neurol. 9, 286–304. Muhlethaler M., de Curtis M., Walton K. and Llinas R. (1993) The isolated and perfused brain of the guinea pig in vitro. Eur. J. Neurosci. 5, 915–926. Nathan T., Jensen M. S. and Lambert J. D. (1990) The slow IPSP in rat hippocampal CA1 neurons is blocked by intracellular injection of QX314. Neurosci. Lett. 110, 309–313. Pelletier M. R. and Carlen P. L. (1996) Repeated tetanic stimulation in piriform cortex in vitro: epileptogenesis and pharmacology. J. Neurophysiol. 75, 4069–4078. Piredda S. and Gale K. (1985) A crucial epileptogenic site in the deep pirifrom cortex. Nature 317, 623–625. Prince D. A. (1968) The depolarization shift in epileptic neurons. Expl Neurol. 21, 467–485. Prince D. A. (1971) Cortical cellular activities during cyclically occurring inter-ictal epileptiform discharges. Electroenceph. clin. Neurophysiol. 31, 469–484. Racine R. J., Mosher M. and Kairiss E. W. (1988) The role of pyriform cortex in the generation of interictal spikes in the kindled preparation. Brain Res. 454, 251–263. Rutecki P. A., Lebeda F. J. and Johnston D. (1985) Epileptiform activity induced by changes in extracellular potassium in hippocampus. J. Neurophysiol. 54, 1363–1374. Scanziani M., Ga¨hwiler B. H. and Thompson S. M. (1991) Paroxysmal inhibitory potentials mediated by GABAB receptors in partially disinhibited rat hippocampal slice cultures. J. Physiol. 444, 375–396. Scanziani M., Debanne D., Muller M., Ga¨hwiler B. H. and Thompson S. M. (1994) Role of excitatory amino acid and GABAB receptors in the generation of epileptiform activity in disinhibited hippocampal slices cultures. Neuroscience 61, 823–832. Schwartzkroin P. A. and Wyler A. R. (1980) Mechanisms underlying epileptiform burst discharge. Ann. Neurol. 7, 95–107. Sherwin I. (1984) Ictal–interictal unit firing pattern differences in penicillin-induced primary and secondary epileptogenic foci. Expl Neurol. 84, 463–477. Silva-Barrat C., Araneda S., Menini Ch., Champagnat J. and Naquet R. (1992) Burst generation in neocortical neurons after GABA withdrawal in the rat. J. Neurophysiol. 67, 715–727. Traub R. D. and Llinas R. (1979) Hippocampal pyramidal cells: significance of dendritic conductances for neuronal function and epileptogenesis. J. Neurophysiol. 42, 476–496. Traub R. D. and Dingledine R. (1990) Model of synchronized epileptiform bursts induced by high potassium in CA3 region of rat hippocampal slice. Role of spontaneous EPSPs in initiation. J. Neurophysiol. 64, 1009–1018. Traub R. D., Miles R. and Jefferys J. G. R. (1993) Synaptic and intrinsic conductances shape picrotoxin-induced synchronized afterdischarges in the guinea pig hippocampal slice. J. Physiol. 461, 525–547. Traub R. D., Jefferys J. G. R. and Whittington M. A. (1994) Enhanced NMDA conductance can account for epileptiform activity induced by low Mg2+ in the rat hippocampal slices. J. Physiol. 478, 379–393.
Spontaneous interictal spikes in focal epileptogenesis 41.
117
Witte O. (1994) Afterpotentials of penicillin-induced epileptiform neuronal discharges in the motor cortex of the rat in vivo. Epil. Res. 18, 43–55. 42. Wong R. K. S. and Prince D. A. (1978) Participation of calcium spike during intrinsic burst firing in hippocampal neurons. Brain Res. 159, 365–390. 43. Wong R. K. S. and Prince D. A. (1979) Dendritic mechanisms underlying penicillin induced epileptiform activity. Science 204, 1228–1231. (Accepted 1 April 1998)