HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 96/6/3028    most recent 00434.2006v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by de Sevilla, D. F. Articles by Buño, W. Search for Related Content PubMed PubMed Citation Articles by de Sevilla, D. F. Articles by Buño, W.

Calcium-Activated Afterhyperpolarizations Regulate Synchronization and Timing of Epileptiform Bursts in Hippocampal CA3 Pyramidal Neurons

Instituto Cajal, Consejo Superior de Investigaciones Científicas, Madrid, Spain

Submitted 25 April 2006; accepted in final form 30 August 2006

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Calcium-activated potassium conductances regulate neuronal excitability, but their role in epileptogenesis remains elusive. We investigated in rat CA3 pyramidal neurons the contribution of the Ca²⁺-activated K⁺-mediated afterhyperpolarizations (AHPs) in the genesis and regulation of epileptiform activity induced in vitro by 4-aminopyridine (4-AP) in Mg²⁺-free Ringer. Recurring spike bursts terminated by prolonged AHPs were generated. Burst synchronization between CA3 pyramidal neurons in paired recordings typified this interictal-like activity. A downregulation of the medium afterhyperpolarization (mAHP) paralleled the emergence of the interictal-like activity. When the mAHP was reduced or enhanced by apamin and EBIO bursts induced by 4-AP were increased or blocked, respectively. Inhibition of the slow afterhyperpolarization (sAHP) with carbachol, t-ACPD, or isoproterenol increased bursting frequency and disrupted burst regularity and synchronization between pyramidal neuron pairs. In contrast, enhancing the sAHP by intracellular dialysis with KMeSO₄ reduced burst frequency. Block of GABA_A–B inhibitions did not modify the abnormal activity. We describe novel cellular mechanisms where 1) the inhibition of the mAHP plays an essential role in the genesis and regulation of the bursting activity by reducing negative feedback, 2) the sAHP sets the interburst interval by decreasing excitability, and 3) bursting was synchronized by excitatory synaptic interactions that increased in advance and during bursts and decreased throughout the subsequent sAHP. These cellular mechanisms are active in the CA3 region, where epileptiform activity is initiated, and cooperatively regulate the timing of the synchronized rhythmic interictal-like network activity.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Changes in neuronal excitability and synaptic efficacy contribute to induce the coordinated network activity needed to generate abnormal epileptiform bursting. In hippocampal pyramidal neurons a key role in the control of excitability is carried out by Ca²⁺-activated K⁺ currents with medium (mI_AHP) and slow (sI_AHP) deactivation kinetics that mediate the medium AHP (mAHP) and the slow AHP (sAHP), respectively (Borde et al. 1995, 2000; Carrer et al. 2003; Sah and Bekkers 1996; reviewed in Sah and Faber 2002; Stocker 2004; Storm 1987; Vogalis et al. 2003). It has been recognized that an abnormal regulation of the sI_AHP/sAHP may contribute to epileptogenesis (Alger and Nicoll 1980; Alger and Williamson 1988; Martín et al. 2001; Matsumoto and Ajmone-Marsan 1964; Traub et al. 1993; reviewed in de Curtis and Avanzini 2001; McCormick and Contreras 2001). However, important aspects of the contribution of the sAHP to abnormal hyperexcitable states remain under debate. In addition, little attention has been paid to the contribution of the mI_AHP/mAHP to epileptogenesis (Alger and Williamson 1988; Empson and Jefferys 2001; Garduño et al. 2005; McCown and Breese 1990; Verma-Ahuja et al. 1998).

We centered our analysis on the contribution of both I_AHPs/AHPs to epileptogenesis in CA3 pyramidal neurons, the region where hippocampal epileptiform activity is initiated (Colom and Saggau 1994; Luhmann et al. 2000; MacVicar and Dudek 1982; Miles and Wong 1983; Schwartzkroin and Prince 1978). We show that 1) a downregulation of the mI_AHP/mAHP paralleled the emergence of epileptiform bursting; 2) when the mI_AHP/mAHP was reduced or enhanced by pharmacological manipulations, bursts were increased or blocked, respectively; 3) manipulations that decreased the sI_AHP/sAHP increased bursting frequency and decreased network synchronization; 4) in contrast, increasing the sI_AHP/sAHP reduced bursting frequency; and 5) bursting in pyramidal neuron pairs was synchronized by excitatory synaptic interaction that increased shortly in advance and during bursts and decreased throughout the subsequent sAHP. The rhythmic bursting network activity that characterizes CA3 epileptogenesis is regulated by intrinsic cellular mechanisms where the mAHP and the sAHP play different roles, but nevertheless act cooperatively to regulate the synchronized bursting that characterizes the interictal-like network activity in the CA3 region. These cellular mechanisms may also be an integral part in the normal function of hippocampal region by regulating networks dynamics, such as the theta rhythm, where bursts of synchronous population activity occur (e.g., Buño et al. 1978) and are reset by interictal spikes in vivo (Lerma et al. 1984).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Slice preparation

Procedures of animal care, surgery, and slice preparation were in accordance with the guidelines laid down by the European Communities Council. Juvenile Wistar rats (12–15 days) were decapitated and their brains were quickly removed and placed in ice-cold control artificial cerebrospinal fluid (ACSF). The composition of the ACSF was (in mM): 124 NaCl, 2.69 KCl, 1.25 KH₂PO₄, 2 MgSO₄, 26 NaHCO₃, 2 CaCl₂, and 10 glucose. The ACSF was continuously gassed with a 95% O₂-5% CO₂ mixture to attain a pH of 7.3–7.4. Transverse hippocampal slices (400 µm thick) were prepared using a Vibratome (Pelco 101, Series 1000, St. Louis, MO), incubated >1 h at room temperature (20–22°C), and were transferred to a recording chamber (about 1 ml) placed on an inverted (Nikon TMS, Tokyo, Japan) or an upright microscope (Olympus BX51WI, Tokyo, Japan) equipped with infrared differential interference contrast video microscopy and a x40 water-immersion objective. Slices were superfused with gassed ACSF at a rate that completely exchanged the solution in the chamber within about 3 min and maintained at room temperature and in some cases at 32–34°C.

Electrophysiology

Whole cell recordings from pyramidal cells placed in the ventral branch of the CA3 region (Fig. 1A) were both in the current- and voltage-clamp modes with (4–8 M) patch-pipettes connected to an Axoclamp-2A amplifier (Axon Instruments, Foster City, CA). Pipettes were filled with a solution that contained (in mM): 135 K-gluconate, 1 EGTA, 10 KCl, 10 HEPES, 2 ATP, 0.4 GTP, and 1 MgCl₂, buffered to pH 7.2–7.3 with KOH.

View larger version (27K): [in this window] [in a new window]   FIG. 1. A slow afterhyperpolarization (sAHP) and outward current follow epileptiform bursts. A: schematic diagram showing the experimental recording and stimulation setup; pyramidal neurons in gray indicate the recording site. B: representative current-clamp recordings showing examples of the 2 types of responses evoked by de- and hyperpolarizing pulses in 2 CA3 pyramidal cells. C, left: current-clamp recording in control solution. D, left: autocorrelation functions calculated with data (2 min long) from same cell. Cell was "silent" and the autocorrelation function tended to be flat (except for the expected narrow peak at brief delays) in control artificial cerebrospinal fluid (ACSF). E, left and F, left: same as C and D, but >20 min after the onset of superfusion with 4-aminopyridine (4-AP) + Mg2+-free ACSF. Repetitive spike bursts (truncated, as in most other cases) followed by sAHP (interrupted arrows) were induced during epileptiform activity [continuous arrows indicate the absence of the medium afterhyperpolarization (mAHP)]. Autocorrelation function shows periodic peaks separated by slow waves that reach negative estimate values. C–F, right: same as C–F, left, but in voltage-clamp conditions. Note the increased synaptic activity indicated between open arrows and its absence during the slow Ca2+-activated K+ current (sIAHP). Bursts were truncated, as in most other figures.

Stimulation

Synaptic responses were evoked by mossy fiber (MF) stimulation through a pair of nichrome wires ( 60 µm) separated about 100 µm, insulated except at the tips, and placed in the stratum lucidum about 500 µm away from the recorded neuron (Fig. 1A). Electrodes were connected to a stimulator unit (Cibertec, Madrid, Spain) driven by the Clampex program (Axon Instruments). Stimulation was with brief barrages of three current pulses (barrage 110 ms; pulse 0.25 ms; repetition rate 0.1 s^–1).

Induction of epileptiform activity

Epileptiform activity was induced with 50–100 µM 4-AP added to a modified Mg²⁺-free ACSF that contained (in mM): 124 NaCl, 2.69 KCl, 1.25 KH₂PO₄, 26 NaHCO₃, 2 CaCl₂, and 10 glucose. 4-AP increases excitability and presynaptic glutamate release by block of the transient A-type K⁺-mediated current (I_A). Responses mediated by released glutamate by N-methyl-D-aspartate (NMDA)–receptor activation are enhanced in Mg²⁺-free solutions by relieving the voltage-dependent block by extracellular Mg²⁺. We previously showed that 4-AP + Mg²⁺-free and 4-AP in control ACSF had identical epileptogenic effects in the CA1 region (Martín et al. 2001). We used the Mg²⁺-free ACSF because the induction of epileptiform activity was faster and stable for a longer period of time than with 4-AP per se (as tested in six cells not included in this study).

Pharmacology

All the following drugs were added to the solutions and superfused in some experiments: Picrotoxin (PTX; 40 µM), to block -aminobutyric acid (GABA_A)–mediated synaptic inhibition, and saclofen (100 µM), to block GABA_B inhibition. Bicuculline (50 µM), to block GABA_A inhibition; the drug also inhibits the mAHP/mI_AHP (Debarbieux et al. 1998; Stocker et al. 1999). Apamin (100 nM), which specifically blocks small conductance (SK) Ca²⁺-activated K⁺-mediated channels and the mI_AHP/mAHP (see DISCUSSION). EBIO (1-ethyl-2-benzimidazolinone), which enhances channel activity and the Ca²⁺-dependent AHPs in neurons (Pedarzani et al. 2001; reviewed in Stocker 2004). EBIO was the first benzimidazolinone described as an activator of both SK channels and Cl^– secretion (Devor et al. 1996). EBIO was prepared as a stock solution in DMSO, stored at –18°C, diluted before use, and added at concentrations between 200 µM and 1 mM. The DMSO at the concentrations used had no effect on membrane properties or synaptic potentials (n = 4; Garduño et al. 2005). t-ACPD [(±)-1-aminocyclopentane-trans-1,3-dicarboxylic acid, 20 µM], a nonselective metabotropic glutamate receptor (mGluR) agonist; carbachol (CCh, 10 µM), a wide-spectrum nonhydrolysable cholinergic agonist; or isoproterenol (5–10 µM), a -adrenergic agonist—all three are unspecific blockers of the sI_AHP/sAHP. CNQX (6-cyano-7-nitroquinoxaline-2,3-dione, 20 µM), which specifically blocks non-NMDA glutamate receptors. In addition, atropine (10 µM) that inhibits muscarinic receptors, MCPG [(S)--methyl-4-carboxyphenylglycine, 0.5–1.0 mM], or LY341495 (20 µM), group I and group II mGluR antagonists, were superfused throughout the experiment and starting 10 min before switching to the 4-AP solution. Chemicals were purchased from Sigma (St. Louis, MO), Tocris Cookson (Bristol, UK), and Alomone Labs (Jerusalem, Israel).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Results are based on 137 pyramidal neurons from the ventral branch of the CA3 region (Fig. 1A) that were recorded at room temperature and exhibited a stable mean V_m of –69.3 ± 1.9 mV and an input resistance (R_in) of 195.5 ± 33.3 M. In addition, experiments (n = 18) were also performed at 32–34°C and the mean V_m was –62.3 ± 2.1 mV and the R_in 175.5 ± 45.8 M, respectively. Neurons were silent in control conditions and either an initial burst followed by a silence (n = 48 or roughly 35%) or a sustained response with little frequency adaptation (n = 89 or roughly 65%) were evoked by suprathreshold depolarizing pulses. The V_m and R_in were not statistically different in both neuron types and responses evoked by hyperpolarizing pulses always displayed repolarizing sag that was essentially identical in bursting and slowly adapting neurons (Fig. 1A).

Evolution of epileptiform activity

The AHP terminates epileptiform bursts. In current-clamp conditions pyramidal neurons did not show spontaneous spike activity in control ACSF (i.e., were "silent"), did not reveal V_m oscillation, and spontaneous synaptic activity was scarce (Current-clamp, Fig. 1C). Superfusion with 4-AP (50–100 µM) induced epileptiform activity in all the CA3 pyramidal neurons analyzed (n = 104). The abnormal activity started after nearly 10–15 min of superfusion with the 4-AP solution and was initially characterized by repetitive single spikes and spike bursts. This mixed activity rapidly changed to continuous spike bursts (3.3 ± 0.1 spikes, 258.2 ± 30.9 ms duration; at 0.23 ± 0.03 s^–1; n = 35, selected at random from the sample) and stabilized in about 10–20 min (Luhmann et al. 2000; Perreault and Avoli 1992). Each burst rode on a large and prolonged depolarizing wave termed paroxysmal depolarization shift (PDS) that was always terminated by an AHP (Fig. 1E) (Garduño et al. 2005; Goldensohn and Purpura 1963; Martín et al. 2001; Matsumoto and Ajmone-Marsan 1964; reviewed in Avoli et al. 2002; de Curtis and Avanzini 2001; McCormick and Contreras 2001). Brief volleys of synaptic activity could precede or follow spike bursts but were absent or markedly reduced during the sAHP (Voltage-clamp, Fig. 1E). The postburst AHP could display an initial faster and a subsequent slower component (Current-clamp, Fig. 1E). The AHP component with faster decay kinetics (30–90 ms; 5.1 ± 1.8 mV; n = 35; in control conditions, measured at –50 mV) could be reduced in amplitude (by 72. 2 ± 9.3% from control values; 10/35 cells or roughly 29% measured at –50 mV) or disappear altogether (25/35 neurons or roughly 71%) during the evolution of the epileptiform activity (continuous arrows, Current-clamp, Fig. 1E). In contrast, the AHP with slower decay kinetics was always present, it decayed to baseline in 3.7 ± 0.2 s, had a peak amplitude of 10.1 ± 0.3 mV (n = 35), and did not change (97 ± 5.7% of control; P > 0.05; same cells, measured at –50 mV) with time during the epileptiform activity (interrupted arrows, Current-clamp, Fig. 1E). The epileptogenic effects of the 4-AP reverted after a prolonged washout (>20 min; n = 12).

We centered our analysis on this type of bursting that has been termed interictal-like activity (Avoli et al. 1993; reviewed in Avoli et al. 2002; de Curtis and Avanzini 2001; McCormick and Contreras 2001). In some cases (10/35 cells or roughly 29%), ictal-like activity was also generated, characterized by bursts at higher frequency, riding on a sustained depolarization (Schiller 2004; Traub et al. 1993; reviewed in Avoli et al. 2002; de Curtis and Avanzini 2001; McCormick and Contreras 2001), and not followed by AHPs (see following text).

It is noteworthy that both during the mixed initial bursting–single-spike activity and the subsequent interictal-like activity bursts were essentially identical and always synchronized in paired recordings, indicating that similar population activity was occurring in the network during both periods (see following text).

We also calculated autocorrelation functions that provide an estimation of the membrane potential oscillations during the control and abnormal interictal-like activity (n = 10). In control conditions autocorrelations were flat (Current-clamp, Fig. 1D), consistent with the absence of oscillations. During the interictal-like activity autocorrelations revealed periodic peaks separated by slow waves (Current-clamp, Fig. 1F), in harmony with the rhythmic repetitive PDS topped by bursts followed by AHPs.

There were no significant differences between the above-described activity recorded at room temperature and the one induced by 4-AP at 32–34°C (3.8 ± 0.6 spikes, 300.4 ± 42.8-ms duration; at 0.33 ± 0.06 s^–1; n = 10) and the faster (30–90 ms; 4.5 ± 1.5 mV; n = 10) and slower (2.8 ± 0.4 s; 5.6 ± 0.9 mV) AHPs. In addition, both the mAHP and sAHP showed similar behaviors, with the former decreasing or disappearing with the evolution of the abnormal activity and the latter remaining unchanged, respectively.

Under voltage-clamp mode cells were silent in control conditions (Voltage-clamp, Fig. 1D) and the abnormal activity evoked in all cells (n = 33) by the 4-AP challenge (100 µM) was initially (about 10–15 min) typified by single or bursts of "unclamped" action currents riding on an inward current wave followed by a long-lasting outward "tail" current. The inward and outward currents correspond to the PDS and the sAHP recorded under current-clamp mode, respectively. The activity then stabilized at a frequency of 0.18 ± 0.02 s^–1 after about 20 min of superfusion with the 4-AP solution. Burst duration was 170.7 ± 23.9 ms and the number of action currents per burst was 2.7 ± 0.2, respectively (n = 33). Differences in the characteristics of the abnormal activity under current- and voltage-clamp modes may be explained by the distinct recording methods.

The large prolonged outward current that followed bursts could show (12/33 cells or roughly 36%) an early, briefer higher-amplitude (50–150 ms, 39.8 ± 9.8 pA, measured at –50 mV in control conditions) component that usually disappeared with the evolution of the epileptiform activity (9/12 or 75%) or was substantially reduced in amplitude (by 65 ± 9.3% from control values; 3/12 cells or 25%, measured at –50 mV) (filled arrows in Voltage-clamp, Fig. 1E). A slower current (25.7 ± 1.3 pA, decay 5.6 ± 2.6 s; n = 33, measured at –50 mV) was always present and did not change (102 ± 9.1%; P > 0.05; same ells) during epileptiform activity (Voltage-clamp, Fig. 1E). Accordingly, the corresponding autocorrelation functions were flat in control conditions (Voltage-clamp, Fig. 1D) and showed periodic peaks at the bursting frequency during epileptiform activity (Voltage-clamp, Fig. 1F). The spontaneous synaptic activity was clearly reduced during the slow outward currents (open arrows, Voltage-clamp, Fig. 1E).

Synaptic inhibition does not contribute to the AHPs

Voltage- and Ca²⁺-activated K⁺ conductances or synaptic inhibition could contribute to the afterhyperpolarization. Block of GABA_A inhibition with PTX (40 µM) did not change the frequency of epileptiform burst (0.16 ± 0.01 s^–1; P > 0.05; n = 8), burst duration (179.4 ± 22.1 ms; P > 0.05; same cells), number of action currents per burst (3.0 ± 0.1 P > 0.05; same cells), nor the amplitude and duration (26.5 ± 0.3 pA; 5.2 ± 0.9 s; P > 0.05, same neurons) of the outward current that follows epileptiform bursts (Fig. 2, A–C). Moreover, block of GABA_B inhibition with saclofen (100 µM) did not modify the epileptiform activity, the peak amplitude (89.6 ± 10.7% of control; P > 0.05; n = 5) of the outward current, or the area under the outward current (91.3 ± 13% of control; P > 0.05; n = 5; Fig. 2D). The values shown correspond to measurements performed when the interictal-like activity had stabilized >20 min after the first spike bursts. These results challenge the notion of an important contribution of postburst GABAergic inhibition to the postburst AHPs in our experimental conditions. Indeed, the duration of the postburst GABA-mediated inhibition, when present (Fig. 2E and F), is much shorter than the postburst AHPs (reviewed in de Curtis and Avanzini 2001), suggesting that additional mechanisms, such as the sI_AHP/sAHP, must be active in our experimental conditions even if GABAergic inhibition is present.

View larger version (26K): [in this window] [in a new window]   FIG. 2. -Aminobutyric acid types A and B (GABAA and GABAB, respectively)–mediated currents do not contribute to the slow outward currents. A: epileptiform action current burst and subsequent slow outward current were induced after (20 min) superfusion with 4-AP + Mg2+-free ACSF under voltage clamp. B: adding picrotoxin (PTX, 40 µM) did not modify the slow outward current. Note the absence of medium Ca2+-activated K+ current (mIAHP, arrows). C: superimposed records A and B. D: summary data showing the mean peak amplitude of the slow outward current in control conditions and 20 min after the onset of superfusion with 4-AP + Mg2+-free ACSF and when PTX (40 µM; n = 8), bicuculline (50 µM; n = 5), and saclofen (100 µM; n = 6) were added. E: repetitive outward currents corresponding to inhibitory postsynaptic potentials (IPSPs) evoked at the onset (about 5 min) of 4-AP superfusion. F: brief inward repetitive action current burst followed by the sIAHPs, characteristic of the interictal-like activity recorded later (>10 min) in the same cell. All recordings were at a –50 mV holding potential (Vh).

We first tested the effects of an intracellular pipette solution containing Cs-gluconate, which blocks all K⁺-mediated currents in the recorded cell (n = 11) without affecting other neurons in the bursting network. Intracellular Cs⁺ blocked the outward currents and the hyperpolarizations that followed epileptiform bursts in all cells (Fig. 3A), consistent with the AHPs being K⁺-mediated conductances without contribution of Cl^–-mediated GABA_A inhibition. An inward current (IC) (19.8 ± 7.9 pA, measured 50 ms after burst termination; IC, Fig. 3A) or an afterdepolarization (ADP) that followed bursts (15.9 ± 8.3 mV, measured 50 ms after burst termination) and that had decay kinetics similar to that of the mI_AHP/mAHP was usually unmasked (7/11 or roughly 64%) with intracellular Cs⁺ (see following text).

View larger version (27K): [in this window] [in a new window]   FIG. 3. Two components of the AHP that follow epileptiform bursts are mediated by K+ conductances. A: representative voltage-clamp trace obtained with a Cs-gluconate–filled electrode 15 min after the onset of 4-AP superfusion. Note epileptiform burst of inward unclamped action currents (truncated) followed by prolonged inward current (IC) and total absence of outward currents. B: representative voltage-clamp trace showing the mIAHP and sIAHP evoked by depolarizing command pulses (shown above) in control ACSF (top) and after block of the mIAHP with apamin (bottom). C: representative voltage-clamp traces of outward "tail currents" evoked by brief depolarizing command pulses (shown above). C, left: in control ACSF the currents are the early mIAHP and late sIAHP components, respectively. C, middle: mIAHP is isolated after blockade of the sIAHP with (±)-1-aminocyclopentane-trans-1,3-dicarboxylic acid (t-ACPD, 20 µM). C, right: with t-ACPD, during epileptiform activity induced by 4-AP, the isolated mIAHP is also inhibited. D: summary data showing amplitude modifications of the mIAHP and sIAHP (% of control values) when apamin (100 nM; n = 9) and carbachol (CCh, 10 µM; n = 6) were added to the control solution and 20 min after the onset of the epileptiform activity when the 4-AP + Mg2+-free (n = 10) and the 4-AP + Mg2+-free + CCh solutions (n = 6) were superfused. All recordings were at a –50 mV Vh.

The molecular identity of the Ca²⁺-activated K⁺ channels mediating the mI_AHP is known (Bond et al. 2004; Sailer et al. 2002; Stocker 2004; Stocker et al. 1999; Villalobos et al. 2004; however, see DISCUSSION). In addition, the Kv7/KCNQ M-current and the hyperpolarization-activated I_h were previosuly shown to contribute to the mI_AHP/mAHP (Gu et al. 2005; Storm 1987; Young et al. 2004). In contrast, the channels mediating the sI_AHP are different and their nature has not been clarified (Bond et al. 2004; Sah and Faber 2002; Villalobos et al. 2004). There is no known specific blocking agent for the sI_AHP that is insensitive to tetraethylammonium (TEA) and micromolar concentrations of 4-AP (Alger and Williamson 1988; Martín et al. 2001; reviewed in Sah and Faber 2002; Stocker 2004). The clotrimazole analogue UCL2027-2 (PZ323) has been shown to induce a relatively selective inhibition of the sAHP in cultured hippocampal neurons (Shah et al. 2001), but bath application of PZ323 (5–10 µM) had no significant effects on the sAHP in CA3 pyramidal neurons in our experimental conditions (data not shown, n = 2), even when GABAergic inhibition was blocked with PTX (40 µm) and saclofen (100 µM; n = 3). However, the sI_AHP is strongly inhibited in a nonspecific manner by muscarinic, adrenergic, and mGluR agonists (e.g., Borde et al. 2000; Madison and Nicoll 1986; Martín et al. 2001; Melyan et al. 2002; Pedarzani and Storm 1993; reviewed in Sah and Faber 2002; Stocker 2004). We investigated the action of t-ACPD (20 µM; n = 6) added to the 4-AP solution (t-ACPD + 4-AP, Fig. 3C) and in a few cases (n = 5) to the control ACSF (+t-ACPD, Fig. 3C). The t-ACPD challenge (Fig. 3C), CCh (10 µM; n = 6; Fig. 3D), and isoproterenol (10 µM; n = 5; see following text) inhibited the slow outward component, consistent with this current being the sI_AHP. Therefore the I_AHPs/AHPs that follow epileptiform burst are most likely the faster initial mI_AHP that mediates the mAHP and the late slower sI_AHP that underlies the sAHP.

A reduction of the mI_AHP/mAHP parallels the induction of epileptiform bursts

When the mI_AHP was isolated after blocking the sI_AHP with t-ACPD the mI_AHP was not modified, but after the induction of epileptiform activity there was a gradual and marked reduction of the mI_AHP (to 40.8 ± 4.3% of control values; P < 0.001; n = 6) (t-ACPD + 4-AP, Fig. 3C).

To further analyze the possible modifications of the I_AHPs/AHPs that follow epileptiform bursts we compared them with the I_AHPs/AHPs induced by identical depolarizing current pulses both in control conditions and during interictal-like activity (n = 11). With this methodology the changes of the I_AHPs/AHPs during the epileptiform activity could be compared with the I_AHPs/AHPs in control conditions in the same cells. In addition, the contributions of variations in the burst characteristics to the I_AHPs/AHPs during the interictal-like activity were minimized. Moreover, manipulations that modified the I_AHPs/AHPs that followed interictal-like bursts induced parallel changes of the pulse-evoked I_AHPs/AHPs, suggesting that they were mediated buy the same conductances.

In current-clamp conditions the pulse-evoked mAHP was reduced in amplitude (^bx40.1 ± 6.2% of control values; P < 0.001; n = 6) or even disappeared altogether (n = 5) when measured 10–20 min after the establishment of the abnormal activity (Fig. 4, A–D). The mAHP did not recover after a prolonged nearly 45-min washout. In contrast the pulse-evoked sAHP did not change during the abnormal bursting activity (96.9 ± 4.3% of control; P > 0.05; n = 11) (Fig. 4, A–D). Under voltage-clamp conditions the control pulse-evoked mI_AHP was reduced in amplitude (by 45.9 ± 5.7% from control values; P < 0.001; 8/11 cells or roughly 73%) and could even disappear (3/11 cells or roughly 27%) during the epileptiform activity (Fig. 4, E–G), whereas the sI_AHP did not change (98.2 ± 5.9% of control; P > 0.05; n = 9).

View larger version (31K): [in this window] [in a new window]   FIG. 4. A reduction of the mIAHP/mAHP parallels the induction of epileptiform activity, whereas the sIAHP/sAHP is not modified. A: superimposed current-clamp records of pulse-evoked AHPs in: control ACSF (black continuous record), showing the mAHP (arrow) and the sAHP; during epileptiform activity induced by 4-AP (dashed record), with bursts before and after the AHPs (open arrows); and during a washout 20 min after the withdrawal of the 4-AP (gray record). B: superimposed representative current-clamp traces (n = 3) obtained in control ACSF showing the spike burst and the sAHP evoked by depolarizing current pulses (shown below). C: same as B, but during epileptiform activity after 30 min of superfusion with 4-AP + Mg2+-free ACSF. sAHP that follows epileptiform bursts is similar to the one evoked by depolarizing pulses. Note the lack of abnormal bursting during the sAHPs (double-headed arrows). B and C: same neuron. D: superimposed traces B and C showing the reduction of the mAHP (arrow) but not of the sAHP during epileptiform activity. E and F: same as B and C, but under voltage-clamp, showing the epileptiform bursts of inward unclamped action currents followed by small IAHPs and the larger pulse-evoked IAHPs. Note the absence of bursts (double-headed arrows) during the pulse-evoked IAHPs with epileptiform activity. G: same as D, but showing superimposed IAHPs; note reduction of the mIAHP (arrow) and unchanged sIAHP during epileptiform activity.

View larger version (59K): [in this window] [in a new window]   FIG. 5. Manipulations that inhibit and enhance the mIAHP strengthen and weaken bursting, respectively. A: current-clamp record showing epileptiform bursting under 4-AP. Note the increased paroxysmal depolarization shift (PDS) amplitude and burst duration (arrows) when apamin (100 nM) was added, and the recovery of the abnormal bursting 20 min after the removal of apamin; without important changes in the interburst intervals. B: expanded versions of the recordings enclosed by the dotted rectangles in A. Note the prolonged PDSs and bursts followed by afterdepolarization (ADP, arrows). C: current-clamp record showing epileptiform bursting under 4-AP. Note the increased mAHP (arrows) and the gradual reduction of burst and prolonged silences (i.e., block of epileptiform activity) when 1-ethyl-2-benzimidazolinone (EBIO, 400 µM) was added. Abnormal bursting recovered 15 min after the removal of EBIO.

In some cases apamin was effective in enhancing bursts even in cells where the mAHP was not evident (4/14 or roughly 29%), probably indicating that the mAHP was nevertheless present but masked by the ADP. Moreover, in those cases the ADP could be revealed under apamin (Fig. 5B). The ADP is also exposed in CA3 pyramidal cells when the sAHP is blocked under muscarinic receptor activation (e.g., McQuiston and Madison 1999). Therefore the increased ADP could result from an increased acetylcholine (ACh) release arising from the activation of cholinergic afferents during the abnormal activity. However, superfusion of atropine (10 µM) starting 10 min before applying the 4-AP challenge did not modify bursting and did not change the ADP nor the subsequent sAHP (n = 4; data not shown), suggesting that an increased release of ACh was not functional in our conditions and that other mechanisms were active (Schiller 2004). It is noteworthy that apamin added to the control ACSF at similar concentration did not generate abnormal activity (n = 4; data not shown), suggesting that other processes besides the block of the mI_AHP/mAHP were needed to induce the abnormal activity. The added process could be the increased ADP induced by the epileptiform activity, a possibility that waits to be investigated.

Agonists of mGluRs may reduce both the sAHP and mAHP in CA3 pyramidal neurons (Young et al. 2004), an action that could be caused by activation of mGluR induced by the increased glutamate release that parallels the epileptiform activity as occurs in CA1 pyramidal cells (Martín et al. 2001). In addition, activation of mGluRs may also induce a reduction of the background conductance and the activation of a voltage-gated inward current that contributes to the ADP (Chuang et al. 2002; Young et al. 2004). However, neither MCPG (0.5–1.0 mM) nor LY341495 (20 µM), which block type I and type II mGluRs, prevented the mI_AHP/mAHP reduction that paralleled the epileptiform activity (n = 4; data not shown), suggesting that other mechanisms were active to depress the conductance. This view is consistent with the action of the mGluR agonist t-ACPD that inhibited the sI_AHP but not the mI_AHP (Fig. 3C).

The sI_AHP/sAHP regulates the interburst interval and rhythmicity

Another important issue that remains to be clarified is what factors determine the timing of the periodic network interictal-like activity. Therefore we tested the possible contribution of the sI_AHP/sAHP to the timing of the bursting activity. The constancy of epileptiform bursts characteristics in different cells suggests that a uniform influx of Ca²⁺ was induced in all cells in the bursting network. Therefore we hypothesized that: 1) bursts were supported by the increased excitability caused by the reduction of the mI_AHP/mAHP and 2) the synchronized bursting and the resulting massive influx of Ca²⁺ activated the sI_AHP/sAHP that had similar characteristics and tended to terminate at fixed intervals after a network burst. This view is in accordance with the absence of epileptiform bursts and the reduction of the synaptic activity during the sI_AHP/sAHP. It also agrees with the occurrence of the subsequent synchronized bursts in the network when sI_AHP/sAHP had terminated, as observed with paired recordings (see following text). This notion is consistent with the synchronized bursting being caused by a recovery of the V_m and excitability that brought cells in the network to fire in close synchrony to a level of population firing that depolarized neurons and triggered the burst (Menéndez de la Prida et al. 2006).

We further tested the above assumptions in two ways. First, by estimating the probability of occurrence of epileptiform bursts before and after pulse-evoked sI_AHP/sAHPs, we found that there was a dramatic reduction of bursting both in current-clamp (by 77.9 ± 0.8% from control values; P < 0.01; n = 10) and voltage-clamp conditions (by 60.2 ± 0.3% from control values; P < 0.01; n = 10) during the pulse-evoked sI_AHP/sAHP (Fig. 6, A–C). In addition, under current-clamp mode the pulse-evoked burst–sAHP sequence induced a reset of the abnormal bursting interictal-like activity that was characterized by repeated bursts that tended to occur at specified times after the pulse. The successive bursts were synchronized by the consecutive pulse-evoked burst–sAHP sequences (Fig. 6A). Second, we tested the effects of blocking the sI_AHP/sAHP during the epileptiform activity with t-ACPD (20 µM) that disrupted rhythmicity, reduced the silent interval that followed pulse-evoked bursts, and increased the frequency of epileptiform bursts by 175.5 ± 22.3% (P < 0.005, n = 6). Therefore under block of the sI_AHP with t-ACPD epileptiform bursts occurred at irregular intervals and with similar probability through the record (Fig. 6D), suggesting that synchronization was not a direct consequence of the bursts but resulted from the combined pulse-evoked burst–sAHP sequence.

View larger version (60K): [in this window] [in a new window]   FIG. 6. Epileptiform bursts are inhibited during the sIAHP/sAHP and preceded by increased synaptic activity. A: superimposed representative current-clamp traces showing pulse-evoked sAHPs (n = 6, as in the rest of the traces) after 20 min of superfusion with 4-AP. Note the synchronization of bursts after the pulse-evoked sAHP; i.e., the sAHP "resets" the ongoing interictal-like activity. B, top: superimposed current-clamp traces showing pulse-evoked sAHPs 20 min after superfusion with 4-AP. Note the absence of bursts during the sAHP (double-headed arrow). B, bottom: histogram showing epileptiform burst occurrences (%) before and after spike bursts and sAHPs obtained from the same cell as in B, top. C, top and bottom: same as B, top and bottom, but in voltage-clamp conditions in another cell. D, top and bottom: same as C, top and bottom, but 30 min after onset of superfusion with 4-AP and added t-ACPD (20 µM) in another neuron. Note the absence of sIAHP and of burst synchronization. Histograms in each experimental group (n = 10 cells) were constructed with responses evoked by 20 successive stimuli in each case. E: current-clamp recordings, showing examples of the increased synaptic activity (dotted squares) immediately preceding (about 1.5 s) epileptiform bursts (i.e., "follower cell"). F: same as E, but the increased synaptic activity (dotted squares) occurred after (about 1.0 s) bursts in another neuron (i.e., "leader cell"). G: same as E, showing complete bursting cycle and interburst interval (IBI). E and G: same cell. H: histogram showing postsynaptic potential occurrences % after (n = 6 cells) and before (n = 5 neurons) bursts. Data were averaged from synaptic potentials paralleling 10 successive bursts in each cell. IBI was normalized by scaling to the briefest one.

These results are consistent with the regulation of network interactions by the sAHP that probably acts by decreasing population activity by reducing excitability in the cells that compose the bursting network. Many neurons showed synaptic potentials during bursts (21/35 or 60.0%), suggesting that they fired in close synchrony with other cells in the network (see following text). The above determinations were made in the first 20 min after the onset of the interictal-like activity. However, synaptic activity preceding and following burst was scarce (4/35 or 11.4%) later during the evolution of epileptiform activity, suggesting that burst synchronization improved in the network with the evolution of the abnormal activity (see following text). The above results are consistent with the burst–sAHP sequence being a key factor in the regulation of the frequency and synchronization of the bursting rhythm, whereas the burst per se was not because desynchronization was observed in conditions where the burst persisted and the sAHP was absent. The abnormal activity was inhibited by 20 µM CNQX (n = 3; data not shown), indicating that excitatory synaptic interactions were of major importance in the genesis of the interictal-like activity.

We verified the above assumptions by recording pairs of CA3 pyramidal neurons (about 50–100 µm apart; n = 22 pairs). We could not detect synaptic interactions between pairs, but their absence is not surprising in view of the low probability of functional excitatory interconnections in acute slices (Miles and Wong 1986).

First, during epileptiform activity we blocked the sAHP with 10 µm CCh (n = 5 pairs) that desynchronized bursting between neurons and increased the bursting rate (by 148.3 ± 18.3%, 3/10 cells or 30%). The ACh challenge also usually increased synaptic activity (4/10 cells or 40%). Both effects were probably caused by the depolarization and the increased excitability induced by CCh in the cells composing the network. Indeed, CCh inhibits the sI_AHP/sAHP and also increases excitability by blocking several K⁺-mediated conductances (reviewed in Storm 1987). However, it was previously reported that at higher temperature (about 32°C) and under GABA_A blockade with bicuculine, instead of desynchronization, CCh per se induces synchronized population activity in the CA3 region in vitro (Psarropoulou and Dallaire 1998). The synchronized activity is blocked by muscarinic antagonists and is thought to be mediated by local excitatory circuits enhanced by muscarinic activity in the absence of inhibition. We tested the effects of CCh applied at 32–34°C on the behavior of pyramidal neuron pairs (separated by about 100 µM; n = 5 pairs) in experiments in which GABA_A inhibition was blocked with PTX (40 µM) in replacement of the bicuculline used by Psarropoulou and Dallaire (1998) because this drug also blocks the mAHP (Debarbieux et al. 1998; Stocker et al. 1999), thus favoring epileptogenesis. Synchronized bursts (5.1 ± 1.1 spikes, 340 ± 20.8 ms duration, at 0.38 ± 0.1 s^–1, n = 5) were induced by CCh (10 µM) (CCh, Fig. 7A), thus confirming the results of Psarropoulou and Dallaire (1998). In addition, synchronized bursts at a higher more irregular rate (8.3 ± 1.8 spikes, 450 ± 30.2 ms duration, at 0. 75 ± 0.1 s^–1, n = 5 pairs) continued when 4-AP (100 µM) was added to the CCh solution (CCh + 4-AP, Fig. 7A). In contrast, the synchronized interictal-like bursting activity induced by 100 µM 4-AP (4-AP, Fig. 7B) was totally desynchronized by adding CCh (10 µM), confirming our results obtained at room temperature (n = 4 pairs). The above differences between the effects of superfusion with CCh before and after the induction of the interictal-like activity by 4-AP at high temperatures might be caused by divergence in the sequences of the blockade of different potassium conductances by both agents, especially because the muscarinic metabotropic activity may be long lived, may induce potentiation (Fernández de Sevilla et al. 2005), and may disclose an ADP (McQuiston and Madison 1999). In any case the differences are highly interesting and should be analyzed in detail in future studies.

View larger version (60K): [in this window] [in a new window]   FIG. 7. Paired recordings: effects of blocking or enhancing the sAHP in one of the cells. A, left: representative current-clamp recordings in a pair of pyramidal neurons (1 and 2; about 50-µm separation) showing rhythmic bursting activity induced by CCh at 32–24°C (10 µM). A, right: after adding 4-AP (50 µM) the bursting frequency increased and bursts remained synchronized in both neurons. B, left: records showing synchronized bursting interictal-like activity induced by 4-AP (50 µM). B, right: when 50 µm CCh was added bursts were at a higher frequency in both cells and totally desynchronized. C1: representative traces obtained with an electrode filled with the KMeSO4 solution that enhances the sAHP. D2: simultaneous recording with the normal K-gluconate electrode solution. Note the absence of burst synchronization and the more prolonged and larger sAHP in D1 during the interictal-like activity. Cells synchronized at the onset of the ictal episode. E: digitized differential interference contrast (DIC) image of the pyramidal neuron pair corresponding to the recordings shown in D.

Second, we tested the contribution of the sAHP to the bursting rhythm by enhancing the sAHP in one cell with an intracellular KMeSO₄ solution (Zhang et al. 1994), whereas the other cell was recorded with normal K-gluconate pipette solution (n = 4 pairs). The cells initially (<20 min after the start of the abnormal bursts) did not burst in synchrony. The neuron dialyzed with KMeSO₄ showed larger sAHPs of longer duration and tended to burst at a lower frequency than the other neuron (by 30.9 ± 5.2%, P < 0.005, n = 4 pairs) (Interictal, Fig. 7C). However, both cells fired in synchrony during the ictal-like episodes when the sAHPs were substantially reduced in both cells and a prolonged depolarization was evoked (Ictal, Fig. 7C). In addition, both cells also tended to burst in synchrony later (>20 min) with the evolution of the epileptiform activity (data not shown).

Third, we recorded pairs with one cell loaded with Cs-gluconate to test the effects blocking all K⁺ conductances in the recorded cell (n = 5 pairs) without affecting other neurons in the bursting network or the cell recorded with the normal intracellular solution. With intracellular Cs⁺ the AHPs were completely blocked and the Cs⁺-loaded cells initially (<20 min after the onset of bursts) tended to burst irregularly at a higher rate (by 192.3 ± 7.2%, P < 0.001, n = 5 pairs) and were not synchronized. However, later (>20 min) both cells tended to burst in synchrony, although bursts were much longer and were followed by ADPs in the neuron dialyzed with Cs⁺ (n = 6) (Fig. 8A). Pulse-evoked bursts were followed by AHPs in control conditions but ADPs that could generate plateaulike depolarization topped by prolonged spike bursts were evoked under Cs⁺ (ACSF, Fig. 8A).

View larger version (57K): [in this window] [in a new window]   FIG. 8. Paired recordings: abnormal bursting in cells dialyzed with Cs+ and 2-bis(2-aminophenoxy) ethane-N,N,N',N'-tetraacetic acid (BAPTA) and with control intracellular solution. A, left: representative current-clamp records obtained in control conditions showing pulse evoked responses (the 200-ms pulse shown below) of a neuron with control intracellular solution (1) and another dialyzed with Cs+ (2). Note sAHP in (1) is absent in (2) and a plateau-like ADP is evoked in (2). A, center: abnormal bursting activity induced by 4-AP (during the first 20 min) showing the irregular high-frequency bursting of the neuron dialyzed with Cs+ (2) and the much lower rate bursting of the cell with control intracellular solution (1). A, right: records obtained 30 min after, showing partial synchronization with occasional out-of-phase bursts. Note that more and much longer bursts followed by ADPs are evoked under Cs+ (2). B, left and middle: same as A, left, but one neuron loaded with control solution (1) and the other with BAPTA (2). Effects were comparable to those of Cs+, but the bursting rate in the BAPTA-loaded cell was lower. B, right: superfusion with isoprotereonol (10 µM) completely desynchronized bursting that was at a higher rate in the BAPTA-loaded cell (2). Note also the increased spontaneous synaptic activity.

The above results offer additional support to the notion that the sI_AHP/sAHP plays a key role in determining the interburst cycle and regulating the frequency and synchronization of the bursting network. The results also explain why, when the sAHP has different kinetics in different cells, bursts may not be entirely synchronized. However, bursts became synchronized later even if the AHPs and other K⁺-mediated currents were inhibited in one of the recorded cells, indicating that other neurons in the network with active AHPs and K⁺-mediated conductances were driving the recorded neurons.

Synaptic excitation synchronizes CA3 pyramidal neuron ensembles

As described above, in many cells there was an increase in synaptic activity preceding or following the interictal burst in the recorded cell (Fig. 10A). The synaptic activity increased in amplitude with imposed hyperpolarization and decreased with depolarization, consistent with excitatory postsynaptic potentials (EPSPs, not shown). Therefore the bursting network was driven by mutual excitatory interconnections that depolarized and drove the recorded cell to the bursting threshold (Menédez de la Prida et al. 2006).

View larger version (39K): [in this window] [in a new window]   FIG. 9. Paired recordings: burst synchronization increases with the evolution of epileptiform activity. A, left: current-clamp records showing irregular and partially synchronized bursting early (10 min) during superfusion with 4-AP. B, left: time-expanded version of framed record in A, left. A, right: cross-correlation of data in A, obtained from a 2-min-long recording. Gray graph is a time-expanded version of the black record. C and D, left and right: same as A and B, but later (20 min) when the abnormal activity had stabilized. Note higher bursting rate, the increased regularity and synchronization of the rhythmic bursting, and the reduction of delay between bursts in both cells from about 20 to about 2 ms with the evolution of the abnormal activity.

View larger version (53K): [in this window] [in a new window]   FIG. 10. Paired recordings: firing synchronization during epileptiform activity; activation of a single neuron can drive the interictal-like activity. A, left: change from synchronous excitatory postsynaptic potentials (EPSPs, dotted arrows), to simultaneous bursts and bursts and EPSPs (open arrow) in 2 neurons (1 and 2). A, right: EPSPs could precede bursts and occur simultaneously with burst in the other neuron (dotted arrows) later during the establishment of the abnormal activity in two neurons (1 and 2). B: superimposed traces (n = 10) showing that the activation of one neuron (1) with a brief depolarizing pulse could "drive" a simultaneously recorded cell (2). Dotted arrows indicate the silent period during the sAHPs.

Additional support to the increased excitatory synaptic interactions was provided by paired recordings, which revealed that initially during the interictal-like activity (5–10 min) EPSPs could occur simultaneously in both cells (18/22 pairs or 86%), consistent with a common excitatory input from other neurons in the network (arrows, left records, Fig. 10A). Later with the stabilization of the abnormal activity (20–30 min) simultaneous EPSPs disappeared and coincident spikes and EPSPs could occur during bursts (20/22 or 90%) (arrows, right records, Fig. 10A). These changes are also consistent with an increase in excitability and excitatory synaptic interactions within the network.

The above results suggest that the activation of a single cell could eventually, as a consequence of the increase synaptic interactions, control the interictal-like network activity. Therefore we used paired recordings and direct stimulation of one of the recorded cells with a brief high-intensity current pulse that evoked the burst–sAHP sequence to test the possible evolution of network interactions. Activation of either of the neurons of the pair never reveled direct excitatory connections in control conditions in our sample (n = 22). However, when the interictal-like activity had stabilized (>20 min) a brief current pulse in one cell, which induced the usual spike burst followed by a silent period, was paralleled by a brief period of increased bursting activity followed by a silent interval in the other neuron (Fig. 10B). This synchronization of activity suggests that the pulse-evoked burst and subsequent sAHP in the recorded cell induced changes in the excitability of the other neuron (4/22 pairs or roughly 18%), implying that activation of a single neuron could control bursting activity in the network (e.g., Menéndez de la Prida et al. 2006; Miles and Wong 1983), probably by indirect excitatory interconnections that increased or synapses that were silent and became active with the evolution of the epileptiform activity.

We also analyzed the effects of stimulating an excitatory synaptic input on burst synchronization and timing between simultaneously recorded CA3 pyramidal neurons (n = 7 pairs) under block of GABA_A inhibition with 40 µM PTX. At the onset of the abnormal activity the burst–sAHP sequence evoked by MF stimulation inhibited the epileptiform bursts during a somewhat variable interval in both cells that terminated by bursts that were not well synchronized (15 min + Stimulation, Fig. 11A). Later, when the epileptiform activity had stabilized, the burst–sAHP sequences evoked by MF stimulation were of similar duration and inhibited epileptiform bursts during a similar interval (4.5 ± 0.5 s, n = 6) that terminated by well-synchronized bursts when the sAHP ended in both neurons (30 min + Stimulation, Fig. 11A). The synchronization, estimated by measuring the reduction of the temporal dispersion between bursts in successive responses, increased by 208 ± 18.3% (P < 0.001, n = 6). Therefore the sAHP evoked by synaptic stimulation silenced the cells during a relatively fixed interval that synchronized and timed the subsequent rhythmic bursting activity. In addition, the response evoked by the MF stimulation was modified during the evolution of the epileptiform activity because in control conditions EPSPs were smaller, evoked bursts with fewer spikes, and were followed by a mAHP (Control, Fig. 11B), whereas later when the interictal-like activity had stabilized EPSPs were larger and evoked longer bursts with more spikes (the number of spikes increased by 167.3 ± 17.3%, P < 0.005, n = 7) and were followed by an ADP in all seven cells (30 min, Fig. 11B). Therefore both increases in synaptic efficacy and postsynaptic excitability contributed to the augmented network synchronization that developed with the evolution of the abnormal activity.

View larger version (26K): [in this window] [in a new window]   FIG. 11. Paired recordings: effects of MF stimulation on burst synchronization increase during epileptiform activity. A, left: superimposed records (n = 8) of bursting activity in two neurons (1 and 2) early after superfusion with 4-AP during mossy fiber (MF) stimulation. Stimulation (MF stim) generated a burst followed by sAHPs of slightly different durations in both neurons that inhibited bursting (dotted arrows); the subsequent bursts occurred at the end of the sAHPs (open arrow). A, right: later during the establishment of the abnormal activity, the sAHPs (dotted arrows) had similar durations in both cells and bursts occurred simultaneously at the termination of the sAHPs (i.e., the post sAHP activity was synchronized in both neurons, open arrows). B: superimposed averages (n = 10) showing the response evoked by the MF stimulation in control conditions (black trace) and 30 min after the establishment of the interictal-like activity in neuron 1 (gray trace). Note that mAHP recorded in the control (black trace) was substituted by an ADP (arrows) during the epileptiform activity (gray trace).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
We present evidence demonstrating that in the CA3 region, where epileptiform activity is initiated and from there spreads to other hippocampal regions (Colom and Saggau 1994; Luhmann et al. 2000; MacVicar and Dudek 1982; Miles and Wong 1983; Schwartzkroin and Prince 1978), the interictal-like activity in pyramidal neurons is precisely regulated by two Ca²⁺-activated K⁺-mediated currents with different kinetics. A downregulation of the mI_AHP/mAHP and the resulting increased excitability are central mechanisms contributing to the generation and regulation of the network bursts that characterize CA3 interictal-like activity. The results also point to a key role of the decreased excitability induced by the activation of the postburst sI_AHP/sAHP in controlling both the frequency and synchronization of the interictal-like network activity.

We show that there is a close relationship between the decline of the mAHP amplitude, the increased excitability, and the induction of interictal-like activity. Pharmacological manipulations that reduce the mAHP, as bath-applied apamin, that specifically inhibits SK Ca²⁺-activated K⁺-mediated channels, caused a marked increase of the bursting activity (McCown and Breese 1990). In contrast, EBIO that enhances Ca²⁺-activated K⁺ currents, and in CA3 pyramidal cells specifically augments the mI_AHP/mAHP, reduced bursting or even blocked the epileptiform activity (Garduño et al. 2005). The above-discussed results are consistent with a key involvement of the mI_AHP/mAHP acting as a negative feedback in the regulation of excitability and in the genesis of CA3 epileptogenesis.

Participation of SK channels in the genesis of the mI_AHP/mAHP has recently been questioned (Gu et al. 2005) because, although a clear SK-mediated apamin-sensitive component was evoked by brief depolarization under voltage-clamp mode, these authors could not evoke the Ca²⁺-activated K⁺-mediated current with similar depolarizations under current-clamp mode. Several ionic conductances may contribute to the mI_AHP/mAHP besides the apamin–EBIO-sensitive Ca²⁺-activated K⁺-mediated SK component. The Kv7/KCNQ M-current at depolarized (> –60 mV) and the H-current at hyperpolarized V_ms (< –70 mV) may add to the mI_AHP/mAHP (Bond et al.